02.taylor & Zappi 2004-Cacti of Eastern Brazil (LIVRO)

CACTI
OF E ASTERN B RAZIL
Facheiroa squamosa (Photo: Nigel Taylor)

Cereus jamacaru subsp. jamacaru (Mandacaru). The largest specimen ever seen, growing at Morro do Chapéu,
Bahia, 2002. Photograph by Marlon Machado.
CACTI
OF E ASTERN B RAZIL
Nigel Taylor and Daniela Zappi
Published by
The Royal Botanic Gardens, Kew
© The Board of Trustees of the Royal Botanic Gardens, Kew 2004
All rights reserved. No part of this publication may be reproduced,
stored in a retrieval system, or transmitted, in any form, or by any
means, electronic, mechanical, photocopying, recording or otherwise,
without written permission of the publisher unless in accordance with
the provisions of the Copyright Designs and Patents Act 1988.
First published in 2004 by

Royal Botanic Gardens, Kew
Richmond, Surrey, TW9 3AB, UK
www.kew.org
ISBN 1 84246 056 0
Production editor: Ruth Linklater

Typesetting and page layout: Christine Beard
Design by Media Resources,
Information Services Department,
Royal Botanic Gardens, Kew.
Printed in Italy by L.E.G.O.
For information or to purchase all Kew titles please visit

www.kewbooks.com or email publishing@kew.org
CACTI OF EASTERN BRAZIL iii
CONTENTS
CONTENTS
PREFACE by Prof. Dr Wilhelm Barthlott ................................................................ xi
1. INTRODUCTION........................................................................................ 1
1.1 Significance of the area delimited as ‘Eastern Brazil’ .................................... 1
1.2 Background to the Cacti of Eastern Brazil project ....................................... 2
1.3 Study methods and taxonomic concepts ...................................................... 3
2. HISTORY OF DISCOVERY, NAMING AND CLASSIFICATION .............. 13

2.1 Problems of the ‘Hortus Siccus’ ................................................................... 13
2.2 Count Johan Maurits in North-eastern Brazil: 1637–1644......................... 14
2.3 19th Century collectors .............................................................................. 14
2.4 The golden age of cactus discovery: 1900–1950 ........................................ 15
2.5 Modern collectors: post 1950 .................................................................... 19
2.6 Taxonomic history of the cacti from Eastern Brazil: 1890–1979 ................ 21
2.7 Cactaceae systematics 1980–2003 and the IOS consensus initiatives........... 25
3. PHYTOGEOGRAPHY................................................................................ 35
3.1 Overview .................................................................................................. 35
3.2 Definition of vegetation types.................................................................... 42
3.3 Analysis of distribution patterns ................................................................. 45
3.4 Phytogeographical links and palaeoclimates ............................................... 63
3.5 Reproductive and dispersal strategies ......................................................... 67
4. CONSERVATION ..................................................................................... 139

4.1 Why conserve the cacti of Eastern Brazil? ............................................... 139
4.2 Status of the environment and conservation of cacti in E Brazil............... 141
4.3 Criteria for the identification of priority taxa ........................................... 148
4.4 Conservation ‘hot-spots’.......................................................................... 155
4.5 Priority actions recommended ................................................................. 156
iv CACTI OF EASTERN BRAZIL
CONTENTS
5. TAXONOMIC INVENTORY ................................................................... 159

5.1 Introductory notes ................................................................................... 159
5.2 Artificial key to genera ............................................................................ 160
5.3 Pereskioideae: Pereskia..................................................................... 164
5.4 Opuntioideae........................................................................................... 173
Cylindropuntieae: Quiabentia ................................................................ 174
Opuntieae Tacinga ..................................................................... 175
Brasiliopuntia............................................................. 186
Nopalea .................................................................... 189
Opuntia .................................................................... 190
5.5 Cactoideae............................................................................................... 193
Hylocereeae Hylocereus ................................................................. 194
Selenicereus ................................................................ 198
Epiphyllum................................................................ 198
Echinocereeae: Pseudoacanthocereus .................................................... 200
Rhipsalideae Lepismium................................................................. 201
Rhipsalis ................................................................... 205
Hatiora ..................................................................... 228
Schlumbergera............................................................. 263
Cereeae Brasilicereus ............................................................... 266
Cereus....................................................................... 269
Cipocereus ................................................................. 282
Stephanocereus............................................................ 290
Arrojadoa .................................................................. 293
Pilosocereus ............................................................... 303
Micranthocereus ......................................................... 347
Coleocephalocereus ..................................................... 356
Melocactus ................................................................ 366
Trichocereeae Harrisia .................................................................... 425
Leocereus .................................................................. 428
Facheiroa .................................................................. 429
Espostoopsis .............................................................. 434
Arthrocereus .............................................................. 436
Discocactus ................................................................ 441
Uebelmannia ............................................................. 451
6. BIBLIOGRAPHY AND ACKNOWLEDGEMENTS .................................. 457
6.1 Literature cited or consulted .................................................................... 457
6.2 Acknowledgements and photo credit codes ............................................. 471
7. GLOSSARY & INDEXES........................................................................... 473
7.1 Glossary of specialized botanical terms (Cactaceae) .................................. 473
Glossário de termos botânicos (Cactaceae) ....................................................... 475
7.2 Index to botanical names and epithets...................................................... 477
7.3 Index to vernacular names ....................................................................... 489
7.4 Appendices I & II.................................................................................... 492
CACTI OF EASTERN BRAZIL v
CONTENTS
TEXT TABLES, MAPS, HALFTONE ILLUSTRATIONS

AND COLOUR PLATES
TEXT TABLES
1.1 Field excursions directly related to the project, involving the authors ............ 5
2.1 Status of genera of Cactaceae found in Eastern Brazil according
to the IOS consensus process ....................................................................... 28
2.2 Characters defining the subfamilies of Cactaceae represented in
Eastern Brazil............................................................................................... 32
2.3 Buxbaum’s tribal arrangement of Cactaceae-Cactoideae............................... 33
3.1 Distribution of species and subspecies of Cactaceae native in Eastern Brazil
by state ........................................................................................................ 37
3.2 Geographical subdivision of the caatingas dominion based on the
distribution of Cactaceae ............................................................................ 59
3.3 Cactaceae characteristic of the Northern and South-eastern campo rupestre
vegetation.................................................................................................... 60
3.4 Principal links between the E Brazilian Highlands and other areas (based
on Cactaceae) .............................................................................................. 62
3.5 Cactus evidence in support of historical migration routes to and from
the caatingas etc. ........................................................................................... 66
3.6 Seed dispersal vectors of species and subspecies of Cactaceae in
Eastern Brazil............................................................................................. 133
4.1 Brejo forests in North-eastern Brazil with notable Cactaceae ...................... 147
4.2 ‘League-table’ of prioritised Red List categories for all taxa assessed as
CR, EN or VU ......................................................................................... 150
4.3 ‘League-table’ of principal conservation area hot-spots............................... 155
7.1 Data on the scientific discovery and description of the native Cactaceae
of Eastern Brazil ....................................................................................... 492
MAPS
Map 1. The Caribbean, South America and Brazil with its states and the
core area studied ................................................................................... 69
Map 2. Distribution records of all taxa restricted to Mata atlântica, sens. lat.,
in Eastern Brazil.................................................................................... 70
Map 3. Distribution records of taxa typical of caatinga and caatinga-agreste............71
Map 4. Distribution records of taxa largely restricted to cerrado (including that
within campo rupestre) in Eastern Brazil .................................................. 72
vi CACTI OF EASTERN BRAZIL
CONTENTS
Map 5. Distribution records of all taxa restricted to campo rupestre (excluding

pure cerrado elements) in Eastern Brazil.................................................. 73
Map 6. Eastern Brazil, showing the principal phytogeographic Area
categories, highlands, ‘brejos’ and areas of potential interest that
await thorough investigation ................................................................. 74
Map 7. Distribution of Praecereus and Brasilicereus (South America) .................... 75
Map 8. Distribution of Cereus subg. Ebneria and C. subg. Mirabella
(South America)........................................................................................ 76
Map 9. Distribution of Espostoa, sens. str. and Espostoopsis (South America)........ 77
Map 10. Distribution of Harrisia: subg. Eriocereus, H. adscendens and subg.
Harrisia (Caribbean & South America) .................................................. 78
Map 11. Vicariant species-pairs: Pereskia guamacho and P. aureiflora; and
Pseudoacanthocereus: P. sicariguensis and P. brasiliensis (South America)...... 79
Map 12. Distribution of the P ERESKIA GRANDIFOLIA Group (South America) .... 80
Map 13. Distribution of Arthrocereus and Quiabentia (South America)................... 81
Map 14. Vicariant species-pairs in Lepismium (South America) ............................ 82
Map 15. Protected areas including threatened cactus taxa compared with sites
of rare and threatened taxa without protected areas in Eastern Brazil .... 83
Map 16. Second order Area subdivisions prioritized by importance for the
conservation of their cactus endemism .................................................. 84
Map 17. Distribution of Pereskia.......................................................................... 85
Map 18. Distribution of Quiabentia, Brasiliopuntia and Opuntia monacantha ........... 86
Map 19. Distribution of Tacinga .......................................................................... 87
Map 20. Distribution of Hylocereus setaceus and Pseudoacanthocereus ....................... 88
Map 21. Distribution of Lepismium...................................................................... 89
Map 22. Distribution of Rhipsalis subg. Phyllarthrorhipsalis.................................... 90
Map 23. Distribution of Rhipsalis subg. Epallagogonium ........................................ 91
Map 24. Distribution of Rhipsalis subg. Rhipsalis ................................................. 92
Map 25. Distribution of Rhipsalis subg. Erythrorhipsalis and R. subg.
Calamorhipsalis ....................................................................................... 93
Map 26. Distribution of Hatiora and Schlumbergera ............................................... 94
Map 27. Distribution of Brasilicereus..................................................................... 95
Map 28. Distribution of Cereus subg. Mirabella .................................................... 96
Map 29. Distribution of Cereus subg. Cereus ........................................................ 97
Map 30. Distribution of Cipocereus ...................................................................... 98
Map 31. Distribution of Stephanocereus................................................................. 99
Map 32. Distribution of Arrojadoa bahiensis, A. dinae and A. sp. nov. (Goiás) ...... 100
Map 33. Distribution of Arrojadoa penicillata and A. rhodantha ............................ 101
CACTI OF EASTERN BRAZIL vii
CONTENTS
Map 34. Distribution of Pilosocereus subg. Gounellea ........................................... 102

Map 35. Distribution of PILOSOCEREUS ARRABIDAE Species Group ................. 103
Map 36. Distribution of PILOSOCEREUS PENTAEDROPHORUS Species Group ..... 104
Map 37. Distribution of PILOSOCEREUS ULEI Species Group ............................ 105
Map 38. Distribution of PILOSOCEREUS AURISETUS Species Group .................. 106
Map 39. Distribution of PILOSOCEREUS PIAUHYENSIS Species Group ............... 107
Map 40. Distribution of Micranthocereus.............................................................. 108
Map 41. Distribution of Coleocephalocereus.......................................................... 109
Map 42. Distribution of MELOCACTUS OREAS Species Group (I): M. oreas
and M. ernestii ..................................................................................... 110
Map 43. Distribution of MELOCACTUS OREAS Species Group (II): M. bahiensis
and M. conoideus .................................................................................. 111
Map 44. Distribution of Melocactus deinacanthus, M. levitestatus and the
M. AZUREUS Species Group ............................................................... 112
Map 45. Distribution of MELOCACTUS VIOLACEUS Species Group (I):
M. salvadorensis, M. zehntneri and M. lanssensianus................................. 113
Map 46. Distribution of MELOCACTUS VIOLACEUS Species Group (II):
M. glaucescens, M. concinnus, M. paucispinus and M. violaceus.................. 114
Map 47. Distribution of Harrisia and Leocereus ................................................... 115
Map 48. Distribution of Facheiroa and Espostoopsis.............................................. 116
Map 49. Distribution of Arthrocereus .................................................................. 117
Map 50. Distribution of Discocactus .................................................................... 118
Map 51. Distribution of Uebelmannia ................................................................. 119
HALFTONE ILLUSTRATIONS
Example of herbarium specimen ......................................................................... 12
Some key individuals in the history of discovery and description of
the cacti of Eastern Brazil ..................................................................................... 18
COLOUR PLATES
1. Examples of vegetation types including cacti: Mata atlântica (agreste),
caatinga, cerrado (‘cerrado de altitude’), campo rupestre ......................................... 120
2. Pereskia aculeata, P. grandifolia subsp. grandifolia, P. grandifolia subsp. violacea........ 121
3. Pereskia grandifolia subsp. violacea, P. bahiensis, P. stenantha ............................. 122
4. Pereskia aureiflora, Quiabentia zehntneri, Tacinga funalis .................................... 123
5. Tacinga funalis, T. braunii............................................................................... 124
6. Tacinga werneri, T. palmadora, T. saxatilis subsp. estevesii ................................. 125
viii CACTI OF EASTERN BRAZIL
CONTENTS
7. Tacinga saxatilis subsp. saxatilis, T. inamoena ................................................. 126

8. Tacinga werneri T. inamoena, Brasiliopuntia brasiliensis .................................. 127
9. Brasiliopuntia brasiliensis ................................................................................ 128
10. Brasiliopuntia brasiliensis, Nopalea cochenillifera, Opuntia monacantha................. 129
11. Opuntia dillenii, O. ficus-indica, Hylocereus setaceus ......................................... 130
12. Hylocereus undatus, Epiphyllum phyllanthus, Pseudoacanthocereus brasiliensis ....... 131
13. Pseudoacanthocereus brasiliensis, Lepismium houlletianum, L. warmingianum,
L. cruciforme ................................................................................................. 132
14. Lepismium cruciforme, Rhipsalis russellii, R. elliptica ......................................... 229
15. Rhipsalis elliptica, R. oblonga, R. crispata, R. floccosa subsp. floccosa .................. 230
16. Rhipsalis floccosa subsp. oreophila, R. floccosa subsp. pulvinigera, R. paradoxa
subsp. septentrionalis ..................................................................................... 231
17. Rhipsalis pacheco-leonis subsp. catenulata, R. cereoides ...................................... 232
18. Rhipsalis cereoides, R. sulcata, R. lindbergiana.................................................. 233
19. Rhipsalis lindbergiana, R. teres f. capilliformis, R. baccifera subsp. baccifera,
R. baccifera subsp. hileiabaiana ....................................................................... 234
20. Rhipsalis pulchra, R. burchellii, R. juengeri, R. cereuscula.................................. 235
21. Rhipsalis clavata, R. pilocarpa......................................................................... 236
22. Rhipsalis hoelleri ........................................................................................... 237
23. Hatiora salicornioides, H. cylindrica, Schlumbergera kautskyi, S. opuntioides ......... 238
24. Schlumbergera microsphaerica, Brasilicereus phaeacanthus ..................................... 239
25. Brasilicereus phaeacanthus, B. markgrafii .......................................................... 240
26. Cereus mirabella, C. albicaulis ........................................................................ 241
27. Cereus fernambucensis subsp. fernambucensis, C. fernambucensis subsp. sericifer,
C. insularis................................................................................................... 242
28. Cereus jamacaru subsp. jamacaru, C. jamacaru subsp. calcirupicola, C. sp. nov. ..... 243
29. Cereus sp. nov., C. hildmannianus, Cipocereus laniflorus ................................... 244
30. Cipocereus crassisepalus, C. pusilliflorus, C. bradei ............................................ 245
31. Cipocereus minensis subsp. leiocarpus, C. minensis subsp. minensis..................... 246
32. Stephanocereus leucostele ................................................................................. 247
33. Stephanocereus leucostele, S. luetzelburgii.......................................................... 248
34. Arrojadoa bahiensis ........................................................................................ 249
35. Arrojadoa dinae subsp. dinae, A. dinae subsp. eriocaulis, A. dinae subsp.
dinae A. rhodantha.................................................................................... 250
36. Arrojadoa penicillata, A. rhodantha.................................................................. 251
37. Arrojadoa rhodantha, A. sp. nov. (‘A. marylanae’ nom. prov.)............................ 252
CACTI OF EASTERN BRAZIL ix
CONTENTS
38. Pilosocereus tuberculatus, P. gounellei subsp. gounellei ........................................ 253

39. Pilosocereus gounellei subsp. zehntneri, P. catingicola subsp. catingicola,
P. catingicola subsp. salvadorensis .................................................................... 254
40. Pilosocereus azulensis, P. floccosus subsp. quadricostatus P. multicostatus,
P. arrabidae..................................................................................................... 255
41. Pilosocereus brasiliensis subsp. brasiliensis, P. brasiliensis subsp. ruschianus,
P. flavipulvinatus........................................................................................... 256
42. Pilosocereus pentaedrophorus subsp. pentaedrophorus, P. pentaedrophorus subsp.
robustus, P. glaucochrous................................................................................. 257
43. Pilosocereus floccosus subsp. floccosus, P. floccosus subsp. cf. quadricostatus &
subsp. quadricostatus, P. subsimilis ............................................................... 258
44. Pilosocereus fulvilanatus subsp. fulvilanatus, P. fulvilanatus subsp. rosae,
P. pachycladus subsp. pachycladus ................................................................... 259
45. Pilosocereus pachycladus subsp. pachycladus, P. pachycladus subsp.
pernambucoensis, P. magnificus........................................................................ 260
46. Pilosocereus magnificus, P. pachycladus subsp. pachycladus, P. machrisii,
P. aurisetus subsp. aurisetus............................................................................ 389
47. Pilosocereus aurisetus subsp. aurilanatus, P. aurisetus subsp. aurisetus, P. aureispinus,
P. multicostatus ............................................................................................. 390
48. Pilosocereus multicostatus, P. piauhyensis, P. chrysostele...................................... 391
49. Pilosocereus densiareolatus, P. parvus, P. bohlei, Micranthocereus violaciflorus......... 392
50. Micranthocereus albicephalus, M. purpureus, M. auriazureus ............................... 393
51. Micranthocereus auriazureus, M. streckeri, M. polyanthus ................................... 394
52. Micranthocereus flaviflorus, M. dolichospermaticus .............................................. 395
53. Coleocephalocereus buxbaumianus subsp. buxbaumianus, C. buxbaumianus
subsp. flavisetus, C. fluminensis subsp. decumbens............................................ 396
54. Coleocephalocereus fluminensis subsp. decumbens, C. fluminensis subsp.
fluminensis, C. pluricostatus, C. goebelianus..................................................... 397
55. Coleocephalocereus aureus, C. purpureus........................................................... 398
56. Melocactus oreas subsp. oreas, M. oreas subsp. cremnophilus, M. ernestii subsp.
ernestii, M. ernestii subsp. longicarpus ............................................................. 399
57. Melocactus bahiensis subsp. bahiensis, M. bahiensis subsp. amethystinus,
M. conoideus, M. bahiensis (M. inconcinnus) .................................................... 400
58. Melocactus deinacanthus.................................................................................. 401
59. Melocactus levitestatus, M. azureus, M. ferreophilus........................................... 402
60. Melocactus pachyacanthus subsp. pachyacanthus, M. pachyacanthus subsp. viridis,
M. salvadorensis ............................................................................................ 403
61. Melocactus zehntneri, M. cf. lanssensianus ....................................................... 404
x CACTI OF EASTERN BRAZIL
CONTENTS
62. Melocactus glaucescens..................................................................................... 405

63. Melocactus concinnus, M. paucispinus............................................................... 406
64. Melocactus concinnus M. paucispinus, M. violaceus subsp. violaceus,
M. violaceus subsp. ritteri, M. violaceus subsp. margaritaceus ............................. 407
65. Harrisia adscendens ........................................................................................ 408
66. Leocereus bahiensis, Facheiroa ulei.................................................................... 409
67. Facheiroa cephaliomelana subsp. cephaliomelana, F. cephaliomelana subsp. estevesii,
F. squamosa.................................................................................................. 410
68. Espostoopsis dybowskii, Arthrocereus melanurus subsp. melanurus, A. melanurus
subsp. magnus .............................................................................................. 411
69. Arthrocereus melanurus subsp. odorus, A. rondonianus....................................... 412
70. Arthrocereus glaziovii ..................................................................................... 413
71. Discocactus zehntneri subsp. zehntneri, D. zehntneri subsp. boomianus,
D. bahiensis ................................................................................................. 414
72. Discocactus heptacanthus subsp. catingicola, D. placentiformis .............................. 415
73. Discocactus pseudoinsignis, D. horstii ............................................................... 416
74. Uebelmannia buiningii, U. gummifera.............................................................. 417
75. Uebelmannia gummifera ................................................................................. 418
76. Uebelmannia pectinifera subsp. pectinifera, U. pectinifera subsp. flavispina............ 419
77. Uebelmannia pectinifera subsp. horrida ............................................................. 420
CACTI OF EASTERN BRAZIL xi
PREFACE
PREFACE
‘The day has passed delightfully. Delight itself, however, is a weak term to express the feelings of a
naturalist who, for the first time, has wandered himself in a Brazilian forest. The novelty of the
plants, the beauty of flowers, but above all the general luxuriance of the vegetation, filled me with
admiration.’ It was not by accident that Charles Darwin wrote this remark in his journal
of researches on February 23rd in 1823, when he entered a tropical forest in Eastern
Brazil. And it was not by accident that one of the most significant international
agreements, the Convention on Biological Diversity (CBD), was initiated in Rio de
Janeiro in 1992.
It is generally accepted that Eastern Brazil is one of the global centres of biodiversity.
The region comprises some 2 million square kilometres of tropical Rain Forest (Mata
atlântica), dry woodlands (Caatinga), savannas (Campos cerrados) and highland vegetation
(Campos rupestres) with altogether at least 20,000 species of flowering plants, a high
percentage of these being endemics.
With cacti we associate the dry regions of the Sonoran, Chihuahuan or Atacama
Deserts, but, surprisingly, the species-rich forests and highlands of Eastern Brazil harbour
around 10% of the species of the Cactus Family. Remarkable and fantastic plants of
completely differing life forms: large foliage trees (Pereskia), lianas (Hylocereus) and a
wealth of epiphytes within the canopy, like the Christmas Cacti (Schlumbergera) and their
relatives (Rhipsalis). I think the most surprising fact is the occurrence of highly succulent,
columnar (eg. Coleocephalocereus) or even globular (Melocactus ernestii) cacti on rock
outcrops and inselbergs in the humid rain forests of the Mata atlântica. Even the Sugar-
loaf (Pão de Açucar) and adjacent inselbergs around Rio de Janeiro harbour endemic
species. The present work, however, focuses on the much less studied regions to the
north of Rio, which the ecotourist is only just beginning to discover.
Early explorers could document little of Eastern Brazil’s cactus riches due to the
difficulties these plants present for the preservation and description of living and museum
specimens, not to mention the inaccessibility of so many of the most biodiverse cactus
habitats prior to the mid-twentieth century. Most taxa now known were discovered in
the past 100 years, the majority being named and classified by horticulturists, whose
activities have rather distorted the number of botanically justifiable entities and their
geographical range. Even so, many genuinely distinct taxa do have rather restricted
ranges and in such a large area numerous, targeted field excursions are essential for a
proper understanding and characterisation of species and adequate documentation of
distribution and ecological data. Such activities have an overriding purpose in acquiring
the evidence and arguments for the conservation of these remarkable examples of our
xii CACTI OF EASTERN BRAZIL
PREFACE
natural inheritance — a richness which is slipping away day-by-day as man’s reckless

progress continues unabated.
Nigel Taylor is the Curator of the living collections at Kew, and his wife, Daniela
Zappi, Assistant Keeper of the Kew Herbarium. I admired Nigel’s broad knowledge in
various plant groups when I first met him at the Herbarium of the Royal Botanic Gardens
in the late 1970s. This led to a very fruitful co-operation on one of our mutual favourite
groups, epiphytic cacti, and to wonderful field trips in North and South America. The
same broad interests were later shared by Daniela — both are truly not typical ‘herbarium’
botanists. They have experience of cultivating the plants they have chosen to study and
through this have built up extensive personal networks in the amateur and academic
communities that serve the cactus hobby. They have also devoted considerable time to
teaching and enthusing a significant number of Brazilian university students at
undergraduate and postgraduate levels. Such contacts are essential not only for promoting
awareness and conservation, but also for filling the inevitable gaps in knowledge of the
plants in the field — gaps that exist even after a study of 15 years and many more
excursions in the wilds of Brazil. Their work has been one of critical inventory,
attempting to bring order to a complex group that, prior to the 1990s, was poorly
understood in terms of taxonomic relationships, phytogeography and conservation status.
The ‘Cacti of Eastern Brazil’ is a monographic treatment of more than 130 species,
with keys, maps and marvellous coloured photographic illustrations of the living plants,
mostly in habitat, and all carefully documented. It includes a wealth of ecological data and
chapters on their biogeography and conservation. A long-existing gap in our knowledge
of one of the most fascinating biomes of the world has been filled.
Wilhelm Barthlott
Director, Botanisches Institut
und Botanischer Garten der
Universität Bonn, Germany
CACTI OF EASTERN BRAZIL 1
INTRODUCTION
1. INTRODUCTION
1.1 SIGNIFICANCE OF THE AREA DELIMITED AS
‘EASTERN BRAZIL’
In the biodiverse neotropics the cactus family, with more than 1300 species (Hunt 1999a:
160), represents the second in order of size amongst higher plant groups that are endemic or
almost so (Bromeliaceae are in first place). In the Americas the Cactaceae have 4 major
geographical centres of diversity (Taylor in Oldfield 1997: 18–19), of which the most
significant is Mexico and the south-western USA. Baseline floristic-monographic treatments
of cacti from this North American region have been published in the last 25 years (cf. Hunt
1992b). The second centre of diversity is within the Andean chain, Peru and Bolivia being
especially rich. Here the taxonomic complexities of the family are too great and currently
too little understood (cf. Taylor et al. in Oldfield 1997: 111). The third centre in order of
importance is Eastern Brazil — a large but discrete region cut off from other areas of cactus
diversity by broad environmental zones that are ecologically unsuitable for most members of
this primarily dryland family. The last monographic-floristic treatment of the Cactaceae in
Brazil was published 114 years ago (Schumann 1890) and accounted for less than 10% of the
130+ native species now recorded from Eastern Brazil, most of which are endemic, making
this area a priority for taxonomic inventory and conservation.
As defined here ‘Eastern Brazil’ includes the habitats of all cacti native to the
vegetation known as caatinga, its ecotones with the Atlantic Forest (agreste) and all but 9
species endemic to campo rupestre and other, geographically associated rupicolous
formations, such as mata calcária, within the contiguous cerrados (for vegetation details, see
Chapter 3.2). The 9 taxa excluded are from the adjacent parts of the Brazilian states of
Goiás, Tocantins and Minas Gerais*, but expanding the study area westwards would
mean including elements of two more Brazilian Regions (‘Grandes Regiões’), ie.
Central-western and Northern Brazil, and for relatively little gain, since these are without
significant cactus floras. Expanding further southwards would increase the number of
epiphytic Rhipsalideae to be treated, but as noted below, these are now better understood
and less deserving of study than the cacti of the Brazilian drylands. The taxonomic survey
presented here deals with c. 1000 botanical names applied to cactus taxa found in Eastern
Brazil, more than 800 of these being treated as synonyms.
* Namely, Pilosocereus albisummus, P. flexibilispinus, P. vilaboensis, P. diersianus (Zappi 1994), Cereus bicolor, C. sp. nov. (Tocantins),
C. pierre-braunianus (Goiás), Arrojadoa sp. nov. (Goiás) and Micranthocereus estevesii (see Appendix II).
2 CACTI OF EASTERN BRAZIL
BACKGROUND
1.2 BACKGROUND TO THE CACTI OF EASTERN

BRAZIL PROJECT
The present study has its origins in a field excursion to Eastern Brazil led by Dr Ray
Harley (Herbarium, Royal Botanic Gardens, Kew), which departed the U.K. at the
close of September 1988. This excursion, in which the authors were invited to
participate as Cactaceae specialists with interests in Brazilian taxa (Taylor 1980, 1981,
1982, Zappi 1989), was part of a long-term collaborative project to study the flora of the
East Brazilian Highlands and especially the campos rupestres of the Chapada Diamantina,
Bahia (cf. Stannard 1995). This involved RBG Kew, the Instituto de Biociências of the
Universidade de São Paulo (São Paulo state) and the herbarium of the Centro de
Pesquisas de Cacau (CEPEC), Ilhéus (Bahia state). It was carried out under the terms of
a convênio (agreement) and collecting permit issued by the Brazilian authorities
(Conselho Nacional de Pesquisas Científicas e Tecnológicas, CNPq and Instituto
Brasileiro do Meio Ambiente, IBAMA). Whilst the excursion focused on the highland
areas, including visits to various localities in the state of Minas Gerais as well as the Pico
das Almas in Bahia, participating specialists were allowed time and resources to make
field collections of their chosen plant families in other vegetation types. Thus, the
authors were able to study and sample some of the Cactaceae characteristic of the caatinga
and cerrado biomes (Zappi had begun field studies on cacti in the campos rupestres the
previous year for her Master’s thesis, Table 1.1). Prior to Taylor’s return to the U.K., in
February 1989, there was also an opportunity to study members of the family in the
Atlantic Forest (Mata atlântica) of São Paulo state.
Having gained an insight into the ecology and variation patterns of Brazilian cacti in
habitat, the authors collaborated in an initial field excursion towards a taxonomic revision
of the genus Pilosocereus Byles & G. D. Rowley, the subject selected for Zappi’s doctoral
thesis (Zappi 1992, 1994). This also greatly furthered progress with a long-running
project to revise the genus Melocactus Link & Otto in Central and South America (Taylor
1991a), whose centre of diversity is in Bahia, which was extensively travelled by the
authors during 3 weeks in July 1989. Following this rewarding experience, approval was
gained from the Keeper of the Herbarium at Kew, Prof. Gren Lucas, to begin a regional
taxonomic monograph of the family, to be entitled ‘Cacti of Eastern Brazil’. The decision
to take this project forward was based on three factors: (1) the perceived high levels of
endemism of Cactaceae in this area, (2) the apparent lack of knowledge about their
taxonomy and distribution, and (3) the established series of collaborations and floristic
treatments already realized by RBG Kew in the North-eastern region of Brazil (eg.
Harley & Mayo 1980, Renvoize 1983, Harley & Simmons 1986, Lewis 1987).
However, whereas the above-cited treatments had focused on Bahia state, or
particular sites in its mountainous interior, the ‘Cacti of Eastern Brazil’ project, with a
more modest number of taxa, could afford to be more ambitious and cover a wider area.
Initial field studies suggested this should be the most species-rich parts of North-eastern
Brazil and the adjacent state of Minas Gerais from the South-eastern Region, which holds
many endemic taxa besides being the southern limit of range of many others. Ultimately
STUDY METHODS & CONCEPTS
the area chosen expanded further to encompass NE Brazil in its entirety and the northern
half of South-eastern Brazil, limited at c. 22°S and 46°W. Hence the convenient term,
‘Eastern Brazil’, an area amounting to some 2 million km2 (Map 1, page 69).
From 1990 further field excursions were planned and executed and opportunities taken
to collaborate with students and their professors, besides giving papers and mini-courses
at Brazilian botanical congresses and universities. In the interests of gaining a better
knowledge of the largely epiphytic cactus flora of the Atlantic Forest — a vegetation type
that has been all but comprehensively modified within Eastern Brazil — additional field
excursions within the contiguous South-eastern region of Brazil, in the states of Rio de
Janeiro and São Paulo, were also carried out. These contributed to a precursory treatment
of the tribe Rhipsalideae (Barthlott & Taylor 1995), whose species found in Eastern Brazil
would otherwise have been difficult to interpret, since their centre of greatest diversity is
outside the area dealt with here.
Subsequently the authors began developing ideas on the phytogeography and
conservation status of the cacti of Eastern Brazil (Taylor 1991c, Taylor & Zappi 1992a,
Taylor et al. in Oldfield 1997: 111–124, 143–144), which have become particular foci in
the work presented here, backed up by the detailed taxonomic inventory in Chapter 5.
Finally, a significant part of this book is based on studies made for our respective doctoral
theses (Zappi 1992, Taylor 2000).
1.3 STUDY METHODS AND TAXONOMIC CONCEPTS

In order to build a comprehensive understanding and documentation of morphology,
ecology and geographic distribution, field and herbarium studies and observations of
living materials in cultivation were made. The methods employed have already been
described in Taylor (1991a: 2–4) and Zappi (1994: 11) and have encompassed surveys of
seed-morphology using Scanning Electron Microscope facilities at RBG Kew. Field
excursions are detailed in Table 1.1, below. Field studies are especially important for
Cactaceae, because the difficulties that face the collector in their preparation for the
herbarium and for subsequent identification have meant that the family has tended to be
ignored or avoided by professional botanists and has in general remained poorly
represented in museum collections (see Chapter 2). Nevertheless, specimens at the
following 58 herbaria were studied and augmented in some cases with materials collected
for the project * (institutions are listed by their standard Index Herbariorum codes, those
indicated in bold typeface being the more important for Cactaceae in Eastern Brazil,
some of the remainder being significant at the state level): ALCB, ASE, B†, BAH, BHCB,
BHMG, BM, BONN†, BR†, [C — list by D. R. Hunt consulted], CEN, CEPEC,
CESJ, E†, EAN, ESA†, F, GUA, HB, HNT, HRB, HRCB†, HUEFS, IAN, IPA, JPB,
K†, L, M†, MAC, MBM, MBML, MEXU, MG, MO, NY†, P (loan only), PEUFR, R,
RB†, RSA, S, SI, SP, SPF, SPSF, U†, UB, UEC, UFG, UFMT†, UFP, UPS, US, W,
*Duplicates of specimens have been deposited in one local herbarium in each state of collection and at SPF or HRCB.
WAG, ZSS†, Herb. Fundação Zoo-Botânica de Belo Horizonte† and Herb. Univ.
Federal de Espírito Santo (the dagger symbol, †, indicates that associated living collections
were also examined). Most of the above-cited institutions have been visited in person, so
that all of their materials could be studied, this in view of the need to have as
comprehensive a view of taxon distribution as possible. Another major source of
distributional data has been the authors’ many contacts, both in Brazil and amongst the
cactus hobbyist community in the U.K., Germany and elsewhere. The extensive
literature on Cactaceae has also been consulted and published records and details of type
localities were particularly helpful in completing the discussion and distribution maps for
taxa from beyond Eastern Brazil, as detailed in Chapter 3 (Maps 7–14). Much of this
bibliography is difficult to access for those in Brazil who would like to study the family
and thus it is hoped that the majority of key literature references are cited in Chapter 6,
which should serve as a guide to what is available in European libraries. However, it
should be noted that recent publications by Russian botanist Prof. A. Doweld, in which
he has ‘atomised’ the Cactaceae, have not been considered. They add nothing to our
understanding of the family and further increase its overburdened nomenclature.
Distribution maps and calculations of ‘extent of occurrence’ for phytogeographic and
conservation purposes have been produced digitally with the assistance of Kew’s
Geographical Information Systems Unit (GIS Unit, Herbarium), employing ArcView
software and a convexhull algorithm run on a networked desktop computer with a 21
inch, high resolution screen. During the production of maps it became obvious that over-
reliance on latitude and longitude coordinates determined by collectors and their assistants
prior to the advent of global positioning systems equipment (GPS) can introduce
disturbingly great errors. Such records, therefore, have been localized using other label
data aided by various GIS overlays, eg. road and river systems. Obviously erroneous
coordinate label data have been excised from the records cited in Chapter 5. GIS overlays
have also been checked against a diverse range of available printed maps acquired in Brazil
during the course of this study (besides those held in the Library at RBG Kew) and also
compared with actual routes travelled on the ground, revealing significant discrepancies
in some areas.
Generic concepts. Unlike specific concepts, where many taxonomists can agree that it is
possible to recognize well-defined entities in nature whose circumscription is often
relatively unambiguous, concepts that group species into genera are much more liable to
differ in breadth. Cactaceae systematics is a classic exemplar of this, ‘with the archetypal
lumpers and splitters traditionally engaged in regular and unwearying conflict’
(Hunt 1999b: 3). The key point, however, is that for the purpose of communicating
evolutionary relationship the group defined should have a single common ancestor and
include all taxa derived from that ancestor, ie. it is monophyletic. Thus, in an earlier,
morphologically-based, cladistic treatment of the genera of Cereeae (the largest cactus
tribe in Eastern Brazil) we stated that ‘the operational taxonomic units to be employed .
. . as far as our knowledge permits . . . are monophyletic’ (Taylor & Zappi 1989: 15). This
is still very much the principle embodied here, so that the genera Melocactus Link & Otto
Table 1.1
Field excursions directly related to the project, involving the authors
(1987–2003).
Dates Areas visited Main purpose of Collaborator(s)

excursion (Index Herbariorum Acronyms)
1987–88 Minas Gerais (Serra do Master’s thesis research J. Prado et al. (SPF)
Espinhaço), various (Zappi only)
excursions/localities
Oct. 1988– Minas Gerais, Bahia, Collecting material for A.M. Giulietti &
Feb. 1989 coastal São Paulo above, plus ‘Flora of the N.L. de Menezes (SPF),
Pico das Almas’ and R.M. Harley et al. (K)
related projects
July 1989 Bahia, N Minas Gerais Collecting material for

Zappi’s PhD revision
of Pilosocereus
Jan.–Feb. Cent. & SW Ceará, As above

1990 SE Piauí, E Maranhão,
N Bahia, NW &
cent. Pernambuco
(by public bus)
May 1990 Coast of E Rio de Janeiro Cactaceae course at Univ. O.J. Pereira, Univ. Fed. ES
& S Espírito Santo Federal, Vitória, ES J.R. Pirani et al. (SPF)
Sep. 1990 Cent.-N Minas Gerais Collecting material for

‘Cacti of E Brazil’ project
(Zappi only)
Dec. 1990 NE & SE Minas Gerais, As above (Taylor/Zappi)

W Espírito Santo
Jan.–Feb. Bahia, E Goiás, N Minas As above U. Eggli (ZSS)

1991 Gerais, Sergipe, Alagoas,
Pernambuco
July 1991 S Minas Gerais As above (Zappi only) Univ. Fed. Juiz de Fora
April 1992 N & E Bahia Noting/photographing

additional distribution
records (Taylor/Zappi)
Feb. 1995 E Bahia, E Alagoas, As above BA: A.M. Carvalho

cent.-N Pernambuco, (CEPEC)
E & S Paraíba, N Piauí, AL: R. Lyra-Lemos (MAC)
NW Ceará PE/PB: M.F. Agra &
E.A. Rocha (JPB)
PI/CE: J.B. da Silva (MG)
Aug. 1998 NE Bahia (by public bus), Taylor, to determine identity M.L. Santos, Aracaju
E Sergipe of Pilocereus rupicola Werderm.
(Serra da Itabaiana, Sergipe)
Nov.–Dec. N Rio de Janeiro, Espírito Zappi only, for Rubiaceae RJ/ES: L. Aona (UEC)
1999 Santo, E Pernambuco (RJ/ES) and Cactaceae PE: E.A. Rocha (UFPE)
Course, Univ. Federal
Pernambuco
March– W Pernambuco, N & W Noting/photographing E.A. Rocha, Univ. Estadual

April 2000 Paraíba, S Rio Grande do additional distribution records Santa Cruz, Bahia &
Norte, S Ceará, SE Piauí, and testing phytogeographical P. Griffiths (K)
N Bahia hypotheses for Taylor’s PhD
thesis (Taylor only)
Aug. 2001 coastal Paraíba to SE Bahia Conservation assessments and PB–BA: J. Jardim (CEPEC)
(Zappi), cent.-S Minas additional geographical MG: M. Pimentel M. & J.
Gerais (Taylor) records Ordones Rego, Fundação
Zoo-Bot., BH, MG
Nov. 2001 E Bahia Cactaceae Course, Univ. J. Jardim et al. (HUEFS)

Federal Feira de Santana
(Zappi only)
Aug. 2002 N & E Bahia Conservation assessments and M. Machado plus staff &
additional records & students from HUEFS;
photographs (Taylor only) R. Augusto, Cruz das Almas
April 2003 SE Bahia Additional records and photos M. Machado (HUEFS) and
of new taxa (Taylor/Zappi) friends from Vitória da
Conquista, incl. UESB
Aug. 2003 S Minas Gerais Serra do Ibitipoca etc. for M. Hjertson (UPS)
photos/records
and Coleocephalocereus Backeb. are each regarded as monophyletic (they are held to be
sister taxa on the basis of the synapomorphy of fruits expressed from the cephalium).
Melocactus has autapomorphies of a terminal cephalium, seeds with relatively few testa-
cells and pollen with simple tectal perforations (Taylor & Zappi 1989, Taylor 1991a:
17–18), while its potentially paraphyletic sister taxon is now believed to be monophyletic
on the basis of having fruits dehiscent by means of a small basal pore. Recently, the
monophyly of both genera has been supported in a preliminary analysis of the chloroplast
rpl16 gene sequence (Soffiatti unpubl.).
Independent cladistic analyses of Rhipsalideae DC., based on morphological
characters (Taylor unpubl. [1996]) and molecular data (Wallace unpubl. [1996]), when
taken together, suggest that in comparison to the plesiomorphic, mesotonically branched
Lepismium Pfeiff. (Barthlott 1987), the more derived E Brazilian genera of the tribe are
each potentially monophyletic, as follows: (1) based on acrotonic branching, freely
disarticulating old/diseased stem-segments and expanded flowers with ± colourless

perianth-segments (Rhipsalis Gaertn.); (2) these same vegetative states combined with
strictly determinate stem-segments (ie. complete absence of indeterminate extension
shoots) and ± erect (vs pendent) habit, perianth highly coloured (Hatiora Britton & Rose,
sens. str. [3 spp.]); and (3) determinate stem-segments combined with the presence of a
well-developed flower-tube, highly coloured perianth and stamens inserted in 2 series
(Schlumbergera Lem.). The floral characters of Hatiora and Schlumbergera are here interpreted
as independent adaptations towards pollination by hummingbirds. However, in Wallace’s
molecular analysis, Hatiora is paired with Rhipsalis pachyptera, indicating Rhipsalis as
paraphyletic, but the morphological evidence clearly places this species (and all other
members of R. subg. Phyllarthrorhipsalis, see Chapter 5) together with the remainder of
Rhipsalis. Defined on this basis Rhipsalis, Hatiora (sens. str.) and Schlumbergera, are easily
recognized and may well be monophyletic, but there are also theoretically less clear cut,
yet practical considerations that flavour some of the generic circumscriptions presently
adopted in other tribes of the family.
Thus, various genera are recognised here on the basis of ‘obvious morphological
character[s]’ (Judd et al. 1999: 29) that are assumed to be autapomorphies, but cannot be
proved to be such, at least partly because of the lack of molecular phylogenies that would
help determine the polarities of the character states employed, eg. characters defining taxa
in the major tribes Cereeae and Trichocereeae (see Chapter 2.7 for explanations). In other
cases it may not even be wise to state that such obvious morphological characters can be
assumed to be autapomorphies, but their significance is that the taxa they define cannot be
allied with any other group and seem isolated taxonomically. This latter situation describes,
and is used to justify, the recognition of Pseudoacanthocereus F. Ritter, Leocereus Britton &
Rose (as a monotype) and Uebelmannia Buining (see Table 2.1), each of which possesses a
unique suite of characters and whose inclusion in any other genus — and it is hard to
decide which this might be (!) — would greatly increase the risk of creating at least a
paraphyletic, if not a polyphyletic assemblage. These uncertainties stem from a lack of
sufficient characters that can be safely employed in phylogenetic analysis (Taylor & Zappi
1989: 14), since the highly specialized and reduced nature of cactus morphology effectively
limits the features that can be used, many being strongly linked to environmental factors
and homoplasious. This said, a comprehensive investigation into anatomy (as begun by
Mauseth 1996) and stem chemistry might well add valuable data sets for Brazilian and other
South American taxa, as it has for the North American Pachycereeae (Gibson & Horak
1978). In the short term, molecular data look to be those most likely to contribute to a
better understanding of generic relationships in Brazilian cacti and various teams of
researchers, including Brazilians (eg. Dr P. Soffiatti, USP, São Paulo), are currently seeking
new sources of informative gene sequences.
In a couple of other cases it is not so much a lack of knowledge, as the needs of
practical convenience, that have shaped decisions on the circumscriptions adopted, it being
arguably better to be aware of a problem than to use ignorance as one’s excuse — the
situation described in the previous paragraph. For example, based on molecular data
(Wallace 1995: 9, Butterworth unpubl.) the species comprising Pereskia Mill. fall into 4
distinct clades, whose relationship to one another at present cannot be further resolved.
And, a more serious problem is that the genus is potentially paraphyletic in respect of the
Cactoideae (represented by Leptocereus in Wallace’s analysis, l.c.). It could, in theory, be
broken up and in that case the Brazilian taxa would end up being distributed between 3
different genera (see Chapter 5), not that this would be either helpful for identification
purposes or more informative than their recognition as subgenera or sections. However,
although the genus is implicitly circumscribed here to include all its named species, the
element that makes it paraphyletic in respect of the Cactoideae is P. lychnidiflora DC., a
native of Central America, not Brazil. A not dissimilar situation is presented by the
treatment of Cereus Mill. adopted here. Two sympatric subgenera are recognized, Subg.
Cereus and Subg. Mirabella (F. Ritter) N. P. Taylor (1991b, 1992a), but the latter lacks the
potential autapomorphy for the genus, which is the early-deciduous spent perianth
(inclusion of Subg. Mirabella has required Cereus to be keyed out twice in the ‘Key to
genera’, Chapter 5). However, on other characters there is no reason for its exclusion from
Cereus and it seems particularly close in habit, flowers and fruit to another allopatric-
vicariant element known as C. subg. Ebneria (see Chapter 3 and Map 8), which has the
early-deciduous perianth. In these circumstances the authors prefer to await molecular
evidence that may help determine whether the persistent perianth in Subg. Mirabella
represents the plesiomorphic state or an apomorphic reversal. The same remark must apply
to another case where, in contrast, two taxa have been kept apart. This concerns
Pilosocereus Byles & G. D. Rowley and Micranthocereus Backeb. (see Chapter 5 for details),
where there is an unclear relationship and fragile distinctions between two taxa that have
never been combined hitherto. The much larger and more widespread genus is Pilosocereus,
which has the younger of the two generic names involved, and in the interests of
nomenclatural stability would need to be conserved, if they were lumped as a single genus.
Besides the already-mentioned Leocereus, the status of two other monotypic genera
treated below merits particular justification. It could be argued that a monotype is an
admission of failure on the part of the taxonomist/systematist, since the otherwise
conveniently informative binomial system of nomenclature in this case may not indicate
the relationship of the species involved. Thus, the names of Espostoopsis dybowskii and
Brasiliopuntia brasiliensis do not give an immediate clue as to their relationship, although
the well-informed will get a steer from the generic etymologies. In fact, their names do
indeed refer to the putative relationship of Espostoopsis and Brasiliopuntia, which are
potentially close to Espostoa Britton & Rose and Opuntia Mill., respectively. In the case
of Espostoopsis its inclusion in Espostoa (tribe Trichocereeae) would be premature on
current evidence and require an expansion of the latter’s circumscription, which would
make for difficulties in writing a simple key to South American columnar genera.
Alternatively, the various unusual features or combinations of characters that Espostoopsis
dybowskii displays argue against its inclusion in any of the genera of tribe Cereeae (Taylor
& Zappi 1989), where its similarities could be due to convergence. On the basis of robust
evidence from gene sequences and seed anatomy (Wallace & Dickie 2002, Stuppy 2002)
Brasiliopuntia belongs in the same clade as Opuntia Mill. and shares some obvious
morphological similarities, eg. combination of flattened stem-segments and flowers with
spreading yellow perianth-segments. However, the previously ill-defined ‘dustbin genus’

Opuntia can now be most conveniently recognized in a restricted sense on the basis of the
autapomorphic character state of pollen with a reticulate exine, a unique feature within
the centrosperms, according to Wilhelm Barthlott (pers. comm.). This circumscription is
further supported in more practical terms, since all species to be included in Opuntia sens.
str. can be readily identified by flowers possessing sensitive stamens, which excludes
Brasiliopuntia and most other genera in the Opuntia clade (Taylor et al. 2002). In relation
to most of these the habit, dimorphic stems and pollen characters of Brasiliopuntia are
clearly strong autapomorphies, while its inclusion in any of them would likely result in
the creation of a paraphyletic group or at least one which was difficult to key out and
comprehend. (The somewhat similar Caribbean genus, Consolea Lem., differs in having
seeds with very distinct funicular envelope anatomy and flowers with numerous small
ovules, and is regarded as convergent.)
Specific concepts and use of subspecies. Judd et al. (1999: 128–130, Table 6.4) describe
more than seven different species concepts and admit that ‘there is no consensus about
species concepts in plants’. Earlier they note that ‘The ascendency of phylogeny as an
organising principle in systematics motivated a phylogenetic species concept’. This has
various potential interpretations, but we like the definition advanced by Nixon & Wheeler
(1990) of a phylogenetic species as the ‘smallest aggregation of populations (sexual) or
lineages (asexual) diagnosable by a unique combination of character states in comparable
individuals’. If such combinations of characters are fixed, diagnosable and testable by
phylogenetic analysis, then we have a better method of recognizing species than that which
has, in practice, been applied here — ie. essentially the phenetic species concept, based on
‘the overall similarity of members of a species, which are separated from other species by
a gap in variation’ (Judd et al., l.c.). For reasons explained in the preceding paragraphs it is
difficult to employ phylogenetic methods for testing species concepts in Brazilian cacti and
where it has been attempted (eg. Zappi 1994: 29) the robustness of the clades identified
could easily be questioned. Such techniques were also attempted in the investigative stages
of a revision of Melocactus (Taylor 1991a), but the phylogenies deduced for the species level
treatments were not published because of their inherent weaknesses. A sufficient number
of characters that can be confidently polarized into plesiomorphic and apomorphic states
is necessary and this has proved difficult enough at the generic level (Taylor & Zappi
1989), let alone at the rank of species, without access to molecular data. The size of the
treatment realized here (130 native species) has unfortunately prevented the application of
molecular techniques and phylogenetic methodology in a uniform way at this level.
Furthermore, this will only become a feasible and meaningful activity when the generic
relationships and, therefore, the significance of the characters and their states, can be
clarified in the tribes Cereeae and Trichocereeae.
In employing a phenetic species concept this treatment has, however, taken into
account, wherever possible, the reproductive strategy and inferred breeding system of the
plants, as observed in nature and in cultivation (where the authors have studied a
considerable diversity of taxa over many years). Understanding reproductive strategy is
important if similarities or differences and suites of linked characters that relate to

pollination and dispersal syndromes are not to be over-valued. For example, it seems that
some Brazilian cacti are probably able to take advantage of different pollinators by quite
minor changes to floral morphology and timing of anthesis, eg. Micranthocereus
(Austrocephalocereus) purpureus (Taylor & Zappi 1989: 22) and Pilosocereus glaucochrous
(Zappi 1994: 78). Little is known at the cytological level about the breeding systems of
cacti, but circumstantial evidence strongly implies that the majority of Brazilian taxa (as
is true of cacti in general) are self-incompatible and outbreeders (cf. Ross 1981). The
actual mechanism has recently been investigated in two genera, namely Schlumbergera
(O’Leary & Boyle 1998) and Echinopsis (Boyle & Idnurm 2001), but its operation in the
rest of the family is presently unknown. However, exceptions to the obligate outbreeder
status of most cacti are known or suspected. Many, but not all, Melocactus species appear
to be self-compatible (Taylor 1991a: 16, Nassar et al. 2002), and there is evidence that a
couple of Brazilian taxa are cleistogamous (see M. lanssensianus and other geographically
disjunct look-alikes of uncertain status discussed in Chapter 5). The species definition that
has been maintained in this latter circumstance is somewhat different, since there can be
no gene exchange between such lineages and thus it is possible that a series of very
narrowly defined taxa could be recognized if this phenomenon should prove to be more
widespread than current field knowledge suggests. Going down that road suggests
abandoning a pragmatic phenetic species concept in favour of the largely discredited
biological species concept (Judd et al., l.c.) and for the time being the temptation to
recognize additional narrowly defined taxa has been resisted.
The very few chromosome counts so far published for E Brazilian cacti, eg. in Pereskia
(Leuenberger 1986) and Rhipsalideae (Barthlott 1976), indicate that nearly all are
diploids, although the authors are aware of a recently published study documenting a
predominance of tetraploids in Melocactus (Assis et al. 2003). Clearly, there is much work
to be done here and studies of North American members of the family have painted a
more interesting picture of polyploidy in members of the Cactoideae (especially the ‘HPE
clade’) and Opuntioideae (Pinkava 2002).
Field studies of cacti conducted in Brazil and elsewhere indicate that related taxa, and
especially sister taxa, are only rarely sympatric and suggest that speciation has probably
occurred by allopatric means in a majority of cases. When species belonging to the same
genus occur together they can hybridize (eg. Tacinga, Cipocereus, Arrojadoa and Melocactus),
but this is not the norm and fewer hybrids have been recorded in the large genera
Pilosocereus and Rhipsalis, whose species are frequently sympatric (cf. Zappi 1994: 35,
Barthlott & Taylor 1995, Taylor 1999). However, even if inability to interbreed is a good
criterion for defining some species, its obvious limitation is the converse situation, since
more distantly related taxa hybridize on occasion and can produce viable offspring, eg.
Pilosocereus pentaedrophorus Micranthocereus purpureus, known from two sites at the eastern
edge of the Chapada Diamantina, Bahia. Likewise, no one would seriously wish to
suggest that Tacinga inamoena and T. palmadora, or for that matter Arrojadoa rhodantha and
A. penicillata, should be lumped together as more broadly defined species because they
hybridize at some (and certainly not all) sites of sympatry. In these cases the individual
species here recognized can each be separated on suites of mutually exclusive characters
and the Tacinga species would deserve being classified in different sections were the genus
not so small as to make this of limited value.
Leaving aside theoretical concepts it is worth recording the authors’ experience that
it has been a relatively straightforward task to delimit species amongst the cacti of Eastern
Brazil. Straightforward, at least once the necessary fieldwork has been completed, because
most initial uncertainties have revolved around regional or local variation, in the case of
a number of widespread taxa, and the geographical area to be surveyed has been
substantial. The initial difficulties have been created through the use of typological species
concepts by amateur ‘cactus hunters’, whose desire to discover and publish something
new has been greater than that to understand the overall patterns of variation (see Chapter
2.5). Put another way, their approach to defining and naming taxa has not been synthetic,
but driven only by a search for differences. While species have often been more broadly
defined here than in previous treatments, their regional variation is now better
understood and many of the more widespread taxa have been subdivided into subspecies.
In nature, many such cactus species, eg. Arrojadoa rhodantha and Pilosocereus pachycladus,
are comprised of numerous geographically sequential races, none of which is sufficiently
distinct to be worthy of being named (though many have). In some cases, however, these
races can grouped on a geographical basis into subspecific taxa sharing mutually exclusive
similarities. Thus, the races of P. pachycladus are presently grouped into two subspecies
defined on numbers of stem ribs correlated with amount of areolar wool. At their points
of contact in northern Bahia they are scarcely distinguishable, whereas towards their
margins it is likely that additional subspecies will be required once further field studies
have been made. A somewhat different situation where the rank of subspecies has been
employed is exemplified by P. fulvilanatus. This is not as wide-ranging as its
aforementioned relative and much less variable, but it includes a rather disjunct
population that has begun to evolve some constant differences suggestive of incipient
speciation (P. fulvilanatus subsp. rosae). Here similarities greatly outweigh differences, but
geographical separation makes its recognition as a subspecies appropriate and informative.
The most problematical taxa encountered in this study have been in the 3 largest
genera, two of which have already been monographed (ie. Melocactus and Pilosocereus).
Here, there remain doubts or controversies about the circumscriptions of the M. oreas /
M. bahiensis, P. pachycladus, P. machrisii / P. aurisetus and Rhipsalis burchellii complexes, but
it is anticipated that these can be resolved by more focused field studies.
Example of an herbarium specimen from the 1988 expedition to Bahia (Coleocephalocereus goebelianus) (K).
HISTORY OF DISCOVERY, NAMING AND CLASSIFICATION
2. HISTORY OF DISCOVERY, NAMING

AND CLASSIFICATION
2.1 PROBLEMS OF THE ‘HORTUS SICCUS’
The history of the discovery and naming of Cactaceae from Eastern Brazil, as defined here,
is detailed for each taxon in Appendix I and can be resolved into 4 major periods. First,
however, it is necessary to make some remarks about this subject in more general terms.
Being for the most part highly succulent, normally unpleasantly spiny, yet unusual to
look at, cacti have always been worthy of study and comment, but never easy to preserve
by conventional methods, nor to draw accurately. Early botanical expeditions to Brazil
undoubtedly encountered many cacti, but while even the most illustrious authorities, such
as VON MARTIUS, frequently refer to them in their field notebooks and scientific
publications (Martius 1832, 1841), remarkably few actually succeeded in bringing significant
numbers of preserved specimens back to their herbaria and museums. Many, however,
attempted and were evidently successful in sending live material back for cultivation in
Europe, eg. Melocactus violaceus subsp. violaceus (M. depressus) *, sent back to Britain from
Pernambuco by George GARDNER (Hooker 1838). Once introduced to the stove house,
their naming and description soon followed, but frequently the living specimens were never
subsequently preserved, either because they were considered too precious as items of
horticultural interest, or as simply ‘incapable of an hortus siccus’ (Bradley 1716–1727). Thus,
until relatively recent times detailed knowledge of the majority of species has been
hampered by a profound lack of both museum specimens and reliable literature, and this to
an extent much greater than with other, more ‘conventional’ plants.
As noted already, the last and only comprehensive study that focused on the Cactaceae
from this vast region of Eastern Brazil was made by Karl Moritz SCHUMANN, as part of
Martius’s monumental Flora brasiliensis (Schumann 1890). For the reasons just explained
his treatment is rather limited in its coverage, since many conspicuous species that were
accessible to, and very probably noticed by 17th and 19th Century collectors in Brazil are
either missing or not distinguished, for lack of adequate material, eg. the caatinga’s
ubiquitous Xique-xique (Pilosocereus gounellei). Unluckily for Schumann, the 20th
Century’s golden age of cactus discovery in Brazil was still to come, as the following notes
and Appendix I hopefully make clear. In fact, Schumann (1890) accounted for only 13 of
the 162 native taxa at the ranks of species and subspecies (heterotypic) now accepted from
the area covered here.
* Botanical authorities for E Brazilian cactus names are not cited here but can be found in Chapter 5.
HISTORY OF DISCOVERY
2.2 COUNT JOHAN MAURITS IN NORTH-EASTERN

BRAZIL: 1637–1644
Our knowledge of Brazilian cacti begins with the Dutch occupation of North-eastern
Brazil, during the years 1630–1654. This included a seven-year period, commencing in
1637, under the command of Governor General Johan MAURITS of Nassau-Siegen,
whose company included two accomplished scientists, Willem Pies (latinized to the more
familiar PISO) and Georg MARCGRAF (also written Marcgrave, Markgraf etc.), and the
talented artists, Albert ECKHOUT and Frans POST. Whitehead & Boeseman (1989) have
brought together a substantial body of information on the work of the Maurits team,
amongst which six cactus species can be confidently identified from the illustrations they
cite and reproduce: Tacinga palmadora, Brasiliopuntia brasiliensis, Cereus fernambucensis, C.
jamacaru, Pilosocereus gounellei and Melocactus violaceus (see Appendix I). They also illustrate
what may be Harrisia adscendens (l.c., tt. 89b & 99b) and refer to a Rhipsalis species that
remains to be identified (l.c., 84). Although the Dutch were mostly restricted to the
coastal regions of the Nordeste, ranging from Salvador (Bahia), to São Luis (Maranhão),
maps drawn by Marcgraf indicate that they travelled far inland, via the Rio São Francisco.
This explains how they encountered caatinga species, such as Cereus jamacaru, Pilosocereus
gounellei, Tacinga palmadora and, perhaps, Harrisia adscendens.
While other kinds of plants illustrated during the Dutch occupation were described in
the 18th Century by Linnaeus and his contemporaries, no cacti appear to have been
named from this source until 1814, when Willdenow published Cactus brasiliensis (=
Brasiliopuntia brasiliensis), followed in 1828 by Cereus jamacaru DC., both of these based on
illustrations in Piso (1648). The latter is now a conserved name, since the illustration
specifically cited by De Candolle is identifiable as C. fernambucensis, but subsequent
authors consistently employed the name for the much larger caatinga species we know
today (Taylor & Zappi 1992c).
2.3 19 TH CENTURY COLLECTORS

After the Dutch natural historians left North-eastern Brazil about 170 years passed before
records of additional cactus species were made from the region covered here. With the
Brazilian capital and principal port of entry now situated at Rio de Janeiro, which is
outside the geographical limits of this study, many of the humid forest and coastal sand-
dune species were first recorded from around that city long before they turned up further
to the north. Certainly, parts of Rio de Janeiro were intensively botanized in the 19th
Century, as was the southern half of Minas Gerais (Urban 1906). However, during the
period 1815–1860, only 6 additional species and 2 heterotypic subspecies of cacti were
collected in the area treated here and 4 of these — Pereskia grandifolia, Rhipsalis lindbergiana,
Arthrocereus glaziovii, Discocactus placentiformis — were found in northern Rio de Janeiro,
Minas Gerais and Espírito Santo, in areas relatively close to the then capital. The others —
Rhipsalis baccifera subsp. hileiabaiana, Pilosocereus pentaedrophorus, Melocactus oreas, M. violaceus
subsp. margaritaceus — were discovered in the state of Bahia, close to, or within a few days’
journey of the state capital and principal sea port of Salvador. The Pilosocereus was obtained
by a French collector, MOREL, who could also have been the discoverer and source of
Melocactus violaceus subsp. margaritaceus, which was originally described as M. ellemeetii in
1857–58. The collectors of Discocactus placentiformis and Melocactus oreas are unknown,
although the type of one of the contemporary synonyms of the former was said to have
been obtained by RIEDEL (see Chapter 5, for details of synonyms). Four of the remaining
taxa mentioned above were gathered as herbarium specimens by other famous European
collectors from this period, including Prince Maximilian of WIED-NEUWIED, SAINT-
HILAIRE, SELLO and BLANCHET (see Urban 1906, for details).
Two collectors, who were active during the remaining period up to the end of the
century, are worthy of particular mention: the Dane, WARMING, who spent 6 years at
Lagoa Santa in Minas Gerais during the 1860s, and the Frenchman, GLAZIOU, who sent
collections from the same state to Schumann prior to 1890. These individuals are notable
for reasons besides their relative productivity. Warming’s collections from the gallery
forests, mountains and Bambuí limestone outcrops around Lagoa Santa were mostly lost
or spoilt (Zappi 1994: 80–82), but he subsequently discussed and illustrated the plants
(Warming 1908), so that it is possible to be sure about their identity (here using modern
nomenclature): Hylocereus setaceus, Lepismium warmingianum, Cereus jamacaru subsp.
calcirupicola and Pilosocereus floccosus. Glaziou probably can be credited with the discovery
of Arthrocereus melanurus and Cereus hildmannianus, but his claims to have personally
collected Tacinga (Opuntia) inamoena and T. braunii (the latter described and named more
than a century later) are almost certainly not genuine. The Tacinga species are indeed
found in Minas Gerais, but not in the southern parts of the state that Glaziou is known
to have visited and thus it is fairly clear that here we have a further instance of his mis-
appropriation of material from another and as yet unidentified collector (cf. Wurdack
1970). There is also uncertainty arising between differences in the stated origins of taxa
‘collected’ by Glaziou and described by Schumann (1890), who sometimes gave Rio de
Janeiro as the source of plants that were subsequently stated by the collector to have come
from Minas Gerais (Glaziou 1909).
2.4 THE GOLDEN AGE OF CACTUS DISCOVERY:

1900–1950
From the beginning of the 20th Century until around 1930 six field collectors made major
contributions to our knowledge of the Cactaceae of Eastern Brazil. Between 1906 and
1907 the German, Ernst ULE, made a series of long expeditions into the interior of Piauí
and Bahia, partly in search of plants that would provide a source of rubber (Zappi 1994:
135). This was the first time that any serious attempt to collect the cacti of the caatinga
had been made and so nearly everything that Ule found was new to science. Most of the
novelties were named by Gürke at Berlin-Dahlem from 1906 onwards, the great
cactologist-monographer, Schumann, having died in 1903. The following key caatinga
species are amongst the many significant discoveries made by Ule: Pereskia bahiensis,
Stephanocereus leucostele, Arrojadoa penicillata, A. rhodantha, Pilosocereus catingicola, P.
pachycladus (the earliest definite record, but named much later), P. piauhyensis, Melocactus
ernestii, Harrisia adscendens (see 2.2 above), Facheiroa ulei and F. squamosa. He also made the
first collection of the most widespread of the Chapada Diamantina’s campo rupestre cacti,
Micranthocereus purpureus.
The Swiss-german emigré, Leo ZEHNTNER, who is known to have established a
cactus garden in the Horto Florestal (forestry station) at Juazeiro in northernmost Bahia,
remains a little-researched figure in the history of Brazilian plant collectors, yet there is
no doubting his importance in relation to Cactaceae. As a collector he was active
during the period 1912–1920 and seems to have complemented the activities of Ule,
visiting a number of Bahian sites the latter did not reach. He is justly commemorated
in the generic names of the cereoids, Zehntnerella (= Facheiroa) and Leocereus, both East
Brazilian endemics, and in the epithets of many other Brazilian cacti named by Britton
& Rose (1919–1923). This eponymy reflected the considerable assistance he gave to the
American collectors, Rose & Russell, as discussed below, who made field studies in
1915 in connection with the great monograph of the family by the above-cited authors.
At this period Zehntner must have relied extensively on access to habitats in Bahia by
means of the fluvial highway of the Rio São Francisco and via the few railway systems
that had already been constructed. Amongst the taxa he is believed to have collected
for the first time are: Pereskia stenantha, Quiabentia zehntneri, Tacinga funalis,
Pseudoacanthocereus brasiliensis, Pilosocereus gounellei subsp. zehntneri, Micranthocereus
flaviflorus, Coleocephalocereus goebelianus, Leocereus bahiensis, Discocactus zehntneri and D.
bahiensis. Another little-known collector from the earliest part of this century was
DYBOWSKI, the discoverer of the remarkable Espostoopsis dybowskii, Plate 68.
Ule and Zehntner between them are the most important discoverers of cacti from the
North-east Region of Brazil until quite recent times. Others amongst their
contemporaries, however, improved our knowledge of distribution, while also finding a
few novelties. Phillip von LUETZELBURG, another phytographer of Germanic origin,
made extensive journeys over much of North-eastern Brazil and also through Espírito
Santo state during the period 1913–1933, while in the employ of the Brazilian
government. However, his 3-volume work, Estudo Botânico do Nordéste, published by the
Inspetoria Federal de Obras Contra as Secas (Luetzelburg 1925–1926), is more important
as a source of information and illustrations than are the few cactus collections that
survived these expeditions. Unfortunately, as Werdermann (1933) noted, much of
Luetzelburg’s cactus material was lost, and the present study has revealed that the part
which survived suffers from many confusions of labelling, rendering his data unreliable
and his published lists of species ± unverifiable. His certain discoveries include Cereus
fernambucensis subsp. sericifer, Stephanocereus luetzelburgii (the remarkable bottle-cactus of
Bahia, Plate 33), Pilosocereus brasiliensis subsp. ruschianus, P. chrysostele and Coleocephalocereus
pluricostatus, but only two of these were described from his material.
More important as collectors were the two aforementioned North Americans, J. N.
ROSE and his assistant P. G. RUSSELL, who spent a couple of months collecting in Bahia,
via the railway system, in 1915. Apart from their obviously successful collaboration with
Leo Zehntner and the use made of his living collections at Juazeiro, they obtained
numerous gatherings in northern Bahia and made a significant contribution at the limits
of the caatinga with the Mata atlântica, in eastern Bahia, utilizing railway lines that
formerly connected the cities of Juazeiro and Jequié with Salvador. Amongst their
discoveries were Tacinga werneri, Rhipsalis russellii, Cereus albicaulis, Melocactus bahiensis
and M. zehntneri, but their extensive herbarium records, preserved at US, NY, K and
elsewhere, are perhaps more important for documenting the distribution of already
known taxa in parts of the region which have since suffered much habitat destruction
(eg. at Jaguaquara [‘Toca da Onça’]).
During the period 1930–1950 a greater number of collectors made mostly small but
worthy additions to our knowledge of the cacti of Eastern Brazil. By far the most
important of these was the German cactus specialist, Erich WERDERMANN, from the
Berlin-Dahlem Botanical Garden & Museum. He planned and executed a very
challenging expedition during 1932, which was reported the following year in his
entertaining botanical travelogue entitled Brasilien und seine Säulenkakteen, an English
translation appearing 9 years later (Werdermann 1933, 1942). With colleagues from
Berlin his journey began in Pernambuco, at the coast, then working inland via the
railroad, subsequently spending a considerable period in Bahia (Salvador, São Félix on
the Rio Paraguaçu, Morro do Chapéu, Tremedal, Caitité), before passing through
Minas Gerais (Diamantina, Serras do Cipó, Caraça and Curral) en route to São Paulo.
Much of the journey was made by Ford Zeppel car, but the condition of roads in the
Brazilian interior left much to be desired and his account tells of the many stops they
made to effect repairs. Werdermann made numerous collections, including living
specimens and some in spirit, but sadly much of his material was subsequently destroyed
at Berlin during the second World War (cf. Leuenberger 1978). Amongst the taxa he
discovered are Cipocereus minensis, Pilosocereus tuberculatus, P. glaucochrous, Micranthocereus
polyanthus and Melocactus salvadorensis. He also collected and described a significant
number of entities that have passed into synonymy, besides creating some confusions of
identity that survive even to this day (eg. he used the name Melocactus bahiensis for the
much more widespread M. zehntneri).
Other collectors who made notable contributions contemporary with that of
Werdermann included resident Brazilian botanists as well as foreign explorers, eg.
Bento P ICKEL (especially at the easternmost limits of the caatinga-agreste in
Pernambuco*), MARKGRAF et al. (Minas Gerais, discovered Brasilicereus markgrafii),
MELLO-BARRETO (Minas Gerais, Uebelmannia gummifera), HERINGER (Minas Gerais),
CUTLER (Ceará), DROUET (Ceará, Pilosocereus flavipulvinatus), BRADE (Minas Gerais and
Espírito Santo, Cipocereus bradei, C. minensis subsp. leiocarpus, Arthrocereus rondonianus),
DUARTE (Minas Gerais, Pilosocereus densiareolatus) and PINTO (Bahia, Pereskia aureiflora).
* Pickel collected many specimens at the adjacent localities of ‘Tapera’ (Mun. Moreno) and ‘São Bento’ (Mun. São Laurenço
da Mata), which are situated in an area where dry vegetation penetrates the otherwise broad band of humid Atlantic Forest in
this state.
1 2
3 4
5 6
Some key individuals in the history of discovery and description of the cacti of Eastern Brazil: 1. J.E.B.
Warming (K), 2. A.F.M. Glaziou (K), 3. E. Ule (B), 4. E. Werdermann (B), 5. W. Uebelmann (UE).
6. K.M. Schumann (B), author of the last monograph of Brazilian cacti in 1890.
2.5 MODERN COLLECTORS: POST 1950

In modern times, cactus discovery in Eastern Brazil began with the arrival of the German,
Friedrich RITTER, in 1959 (Eggli et al. 1995), and was enhanced in the 1960s by the
German-Brazilian, Leopoldo HORST, the latter’s Swiss nurseryman sponsor, Werner
UEBELMANN, and their Dutch collaborator, Albert BUINING (Uebelmann 1996). These
collectors were largely motivated by the horticultural trade and cactus hobbyist interests,
especially in Europe, where new discoveries were eagerly sought. However, although
many so-called ‘spec. nov.’ were collected, introduced to cultivation and swiftly named,
the number of genuine first discoveries of taxonomically ‘good’ species was far less than
it seems, for two reasons. First, many of the taxa that these and subsequent cactus plant
hunters claimed as new discoveries had already been collected much earlier by
professional botanists and deposited in herbaria, which such ‘cactophiles’ did not usually
consult (see Appendix I). Secondly, the taxonomic concepts they employed ignored
regional variation and resulted in a plethora of weakly defined microspecies, many of
which have since been dumped into synonymy or down-graded in rank by the studies of
the present authors (Taylor 1980, 1981, 1982, 1991a; Taylor & Zappi 1990, 1991, 1997;
Taylor in Hunt 1992b, 1999; Zappi 1994). Nevertheless, Ritter can lay claim to having
discovered at least 21 distinct taxa between 1959 and 1971, although he was not always
the first person to publish these (Ritter 1979): Tacinga saxatilis, Rhipsalis floccosa subsp.
oreophila (R. monteazulensis), Cereus mirabella (Mirabella minensis), Cipocereus crassisepalus, C.
pusilliflorus, Arrojadoa dinae, Pilosocereus floccosus subsp. quadricostatus, P. magnificus, P.
aurisetus subsp. aurilanatus, P. multicostatus, Micranthocereus albicephalus, Coleocephalocereus
buxbaumianus subsp. flavisetus, C. fluminensis subsp. decumbens, C. aureus, Melocactus ernestii
subsp. longicarpus, M. bahiensis subsp. amethystinus, M. levitestatus, M. concinnus, M. violaceus
subsp. ritteri (M. macrodiscus var. minor), Facheiroa cephaliomelana, Arthrocereus melanurus
subsp. odorus.
Horst and his associates, Uebelmann and Buining, were even more successful than
Ritter, since they were evidently able to devote more time to exploration of the remote
Brazilian interior (the ‘sertão’) and this often by means of suitable vehicles, on roads that
were steadily improving. Over a period of some 30 years, starting in the mid-1960s, a
steady stream of novelties was reported in journals and catalogues back in Europe and the
USA, and plant material was distributed to avid cactophiles via Uebelmann’s nursery. No
less than 27 distinct new taxa can be attributed to the efforts of their explorations,
although, as already indicated, the number they actually claimed as new was probably in
terms of hundreds (a measure of this can be gained from the many names listed for
relevant Brazilian genera in Eggli & Taylor 1991). The following were discovered
between the years 1966 and 1982:– Tacinga inamoena subsp. HU 336 (new — awaiting
description), Arrojadoa dinae subsp. eriocaulis, Pilosocereus pentaedrophorus subsp. robustus, P.
fulvilanatus (2 subspp.), P. aureispinus, P. parvus, Micranthocereus violaciflorus, M. auriazureus,
M. dolichospermaticus, Coleocephalocereus buxbaumianus subsp. buxbaumianus, C. purpureus,
Melocactus conoideus, M. deinacanthus, M. azureus, M. ferreophilus, M. pachyacanthus, M.
lanssensianus, M. glaucescens, Discocactus zehntneri subsp. boomianus, D. heptacanthus subsp.
catingicola, D. pseudoinsignis, D. horstii, Uebelmannia buiningii, U. pectinifera (3 subspp.). The

achievement and significance of these collectors is reflected in the fact that many of the
above novelties are extremely localized in habitat and now figure amongst the threatened
taxa discussed in Chapter 4.
Other collectors that have been active during this period include professional
botanists, such as our Kew colleague David HUNT, who collected cacti in Pernambuco
in 1966. The Argentinian cactus specialist, Alberto CASTELLANOS, made frequent visits
to Eastern Brazil while resident in Rio de Janeiro and collected cacti on a number of
occasions. In 1968 he collected material of a Pilosocereus, P. azulensis, which was only
recognised as new quite recently, Plate 40 (Taylor & Zappi 1997). He is known to have
made other herbarium collections from the core area, but it is believed that an important
part of his material was destroyed in a fire (M. Vianna [GUA], pers. comm.). Around the
mid 1960s Paulo MARTINS collected Rhipsalideae at the behest of Prof. F. Brieger
(Universidade Federal de Brasília) and discovered Rhipsalis paradoxa subsp. septentrionalis
in Bahia, which was not described until 1995, Plate 16 (Barthlott & Taylor 1995).
Dárdano de ANDRADE-LIMA, Raymond HARLEY (discoverer of Melocactus oreas subsp.
cremnophilus), Leopoldo KRIEGER (discoverer of Arthrocereus melanurus subsp. magnus) and
Gustavo MARTINELLI are amongst a handful of mostly Brazilian botanists that included
cacti in their general field collections for the herbarium during the 1970s. In particular
Andrade-Lima deserves praise for his many collections in areas that have since undergone
considerable habitat modification, and for his various helpful publications on Brazilian
cacti (cf. Prance & Mori 1982).
During the 1980s and 90s there has been much more field activity, both by Brazilian
botanists and their European collaborators, as well as hobbyist cactophiles, but this has
resulted in fewer discoveries of genuinely new taxa. Some of those that have been found
are extreme rarities, narrow endemics or plants from more or less inaccessible habitats.
Nonetheless, a steady stream of novel plants has been reported in the specialist cactus &
succulent literature. In 1981, a team of botanists from universities in São Paulo state,
including Antônio FURLAN and Inês CORDEIRO, discovered a peculiar new Arrojadoa,
A. bahiensis, restricted to nearly vertical cliffs in the Chapada Diamantina, Bahia (Taylor
& Zappi 1996). In the same year another remarkable cactus, Melocactus paucispinus, which
imitates the unrelated genus Discocactus, was discovered in the same region by the German
cactophiles, HEIMEN et al., Plate 63. In 1984, Brazilian Eddie ESTEVES PEREIRA found
two new subspecies at a locality in southern Bahia (Tacinga saxatilis subsp. estevesii and
Facheiroa cephaliomelana subsp. estevesii), while the following year VAN HEEK and VAN
CRIEKINGE discovered the very rare Micranthocereus streckeri in central Bahia. In 1986,
Prof. Werner RAUH and Roberto KAUTSKY collected a series of little-known or
undescribed Rhipsalideae from the region of Domingos Martins in Espírito Santo state:
Rhipsalis pacheco-leonis subsp. catenulata, R. cereoides, R. sulcata, R. clavata, R. pilocarpa and
Schlumbergera kautskyi. Their work was soon after complemented by field studies
conducted by Prof. Wilhelm BARTHLOTT and Dieter SUPTHUT, and thereafter by the
former’s student, Stefan POREMBSKI, from the University of Bonn, with surveys of
inselberg vegetation. Meanwhile, in 1987, Countess Beatrix ORSSICH obtained the
HISTORY OF DISCOVERY, NAMING AND CLASSIFICATION
extraordinary, red-flowered Rhipsalis hoelleri from Espírito Santo, Plate 22 (Barthlott &
Taylor 1995). Also in 1987, one of the present authors, Daniela ZAPPI, and Brazilian
colleague, Vera SCATENA, found a most unusual new cereoid cactus, Cipocereus laniflorus,
on exposed rocks of the Serra do Caraça, in central-southern Minas Gerais, Plate 29.
Then, the following year the present authors discovered the ‘Critically Endangered’
Melocactus pachyacanthus subsp. viridis in central-northern Bahia (Taylor 1991a).
Since the last-mentioned there have been no significant discoveries until very recently
(apart from presumed hybrids and other plants of dubious status). However, in the period
2000–2002 three quite remarkable ‘good’ new taxa have come to light. The first is
Pilosocereus bohlei Hofacker — see Chapter 5, under Pilosocereus ‘Insufficiently known taxa’.
The second and third await further study prior to their description, but are briefly
mentioned in Chapter 5 under Cereus jamacaru and Arrojadoa rhodantha. Both are
discoveries made by staff from the Universidade Estadual do Sudoeste da Bahia, at Vitória
da Conquista, and communicated to the authors through the ever enthusiastic Marlon
MACHADO. Besides these, during the past 15 years, extensive field studies focused on
Cactaceae have been carried out for this book and have involved numerous collaborators
(see Chapter 6.2), especially Urs EGGLI, Emerson ROCHA and, latterly, Marlon Machado.
Others, from the University of Feira de Santana, including Luciano Paganucci de
QUEIROZ and Jomar JARDIM, have made significant herbarium collections in the Nordeste
and, unlike many botanists, have not been afraid of gathering substantial numbers of cacti,
greatly extending our knowledge of geographical distribution and improving our
estimations of current levels of threats to habitats and taxa.
As this text was being finalized a review focusing on the last 25 years of Brazilian
cactus discovery was published (Braun & Esteves Pereira 2002). This review contains
large amounts of literature references, colour photographs and opinion, but is generally
short on useful data, especially precise localities, although its authors emphasize their
many years of exploration in the field. In particular, students of the family should be wary
of the distributional information offered (eg. l.c. 79), which is mostly not backed up by
cited herbarium vouchers or photographs, and in part contradicts the distribution maps
presented. In the introduction (l.c. 8) it is noted that there is now a greater amount of
material in herbaria, presumably for study, but little evidence is presented of this resource
having been utilized except, of course, for the deposition of obligatory nomenclatural
types. In addition, the application of Brazilian vegetation terminology is sometimes quite
bizarre, especially with reference to carrasco [‘charasco’], cerradão, agreste and caatinga — a
tradition established much earlier by Ritter and Buining.
2.6 TAXONOMIC HISTORY OF THE CACTACEAE

FROM EASTERN BRAZIL: 1890–1979
In terms of monographic and synoptic taxonomic treatments, the following names are
particularly relevant in relation to the cacti of Brazil, amongst botanists and field collectors
that published prior to the 1980s: Schumann, Britton & Rose, Berger, Werdermann,
Backeberg, Buxbaum, Hunt and Ritter. The first treatment that merits mention is that by
HISTORY OF CLASSIFICATION
SCHUMANN (1897–98), the Gesamtbeschreibung der Kakteen, which followed closely on the
heels of his accounts for Martius’s Flora brasiliensis and Engler & Prantl’s Das
Pflanzenfamilien (Schumann 1890, 1894). Schumann’s classification was, by present
standards, extremely conservative at generic level, where he recognized the following
‘hold-all’ genera: Cereus Mill. (including a variety of Brazilian columnar-cereoid and
scandent species, plus some rhipsaloids now referred to Schlumbergera Lem., but excluding
part of the modern Pilosocereus Byles & G. D. Rowley as Pilocereus Lem.), Rhipsalis Gaertn.
(for a diversity of epiphytic taxa belonging to tribe Rhipsalideae DC., but also including
some now referred to tribe Hylocereeae Buxb.), Echinocactus Link & Otto (for all the low-
growing globular forms like Discocactus Pfeiff., but not Melocactus) and Opuntia Mill., this
last being used in the traditional broad sense that has persisted until quite recent times. In
addition he accepted Melocactus Link & Otto and Pereskia Mill. in their current senses,
Hariota DC. (nom. illeg. = Hatiora Britton & Rose), Pilocereus Lem. (in a sense excluding
its type), Phyllocactus Link (correctly Epiphyllum Haw.), Epiphyllum (for part of what is
now Schlumbergera) and Zygocactus K. Schum. (= Schlumbergera). Schumann classified the
genera of Cactaceae into 3 subfamilies, Pereskioideae, Opuntioideae and Cereoideae
(correctly Cactoideae), an arrangement which has remained little changed until very
recently (Maihuenia (Weber) K. Schum., which he placed in Pereskioideae, is now placed
in its own subfamily).
In their great 4-volume work, The Cactaceae, BRITTON & ROSE (1919–1923)
recognized Schumann’s 3 subfamilies as tribes and otherwise radically changed the
classification of the family at generic level, liberally dividing his hold-all genera and
describing many new ones to account for the numerous discoveries made in the early
years of the 20th Century. In total they recognised 124 genera for the family. A good part
of the changes they made to the abundant cereoid species was considerably influenced by
Alwin BERGER’s detailed subgeneric rearrangement of Cereus (Berger 1905), some of
whose subgeneric names and/or taxa they upgraded to generic status. Thus, Schumann’s
concept of Cereus became restricted to the group immediately close to the type species,
C. hexagonus (L.) Mill. However, Pilocereus was included in Cephalocereus Pfeiff., whose
type (Cactus senilis Haw.) and Brazilian taxa (ie. Pilosocereus) are nowadays placed in
different tribes (Pachycereeae and Cereeae, respectively). The scandent cereoids
including Brazilian species were separated into the new genera, Hylocereus and Mediocactus.
The then known E Brazilian globular cacti were placed in the reinstated Discocactus, but
allied with Melocactus, for which Britton & Rose dug up the abandoned name Cactus L.
These were the only genera included in their subtribe Cactinae (Melocactus is now seen as
the most derived element in tribe Cereeae, while Discocactus is referred to Trichocereeae,
Taylor & Zappi 1989). Britton & Rose also split up the E Brazilian elements in
Schumann’s concept of Rhipsalis, recognizing Lepismium Pfeiff., Erythrorhipsalis A. Berger
(1920) and Epiphyllanthus A. Berger (1905), as well as Hatiora Britton & Rose. New
genera were created for various taxa that had been described in Cereus and Cephalocereus
shortly after Schumann’s death, namely, Arrojadoa, Leocereus, Facheiroa and Zehntnerella.
New opuntioid genera were also created for recently discovered taxa: Quiabentia (for a
leafy species initially described as a Pereskia) and Tacinga (for a curious, scandent species).
While some commentators welcomed these new generic names, others, especially in
Germany, where Schumann’s view still held sway, rejected some or all, or even hedged
their bets. For example, Berger (1929) ambiguously treated many of Britton & Rose’s
genera simultaneously as subgenera and genera in his handbook, Kakteen.
WERDERMANN (1933, 1942), in dealing with the cereoid cacti of Eastern Brazil,
adopted a compromise position, recognising some of the segregates employed by Britton
& Rose, but sinking others. For example, Trichocereus (Berger) Riccob. he accepted for
some of the Brazilian taxa subsequently realigned by others in Arthrocereus A. Berger
(1929), but Britton & Rose’s Arrojadoa went into the synonymy of Cephalocereus, from
which he separated Pilocereus auctt. non Lemaire. In the latter genus he placed the curious
Cereus luetzelburgii, and also C. leucostele, for which Berger (1926) had recently created the
then monotypic Stephanocereus. Another of Berger’s splits he did not accept was
Brasiliopuntia A. Berger (1926), which he referred back to Opuntia, although he seems to
have been prepared to accept Tacinga Britton & Rose.
If Britton & Rose started a trend towards the splitting of genera, then the German
cactus nurseryman, Curt BACKEBERG (1938, 1958–62), went many stages further, his
6-volume monograph promulgating a more than 10-fold increase in genera over
Schumann’s treatment (220 vs 21)! His poorly researched innovations, typological
species concepts, disregard for taxonomic and nomenclatural conventions and
unsatisfactory suprageneric classification (Barthlott 1988, Hunt 1991: 152) resulted in
little less than a state of taxonomic and nomenclatural chaos, which probably
frightened most professional botanists away from serious study of the family until the
mid-1960s. In 1938 he established 3 very poorly defined genera, which have
fortuitously turned out to be worthy of recognition, now that the family is better
understood in Eastern Brazil: Brasilicereus, Micranthocereus and Coleocephalocereus. At the
same time he published Austrocephalocereus, which modern authors now agree should
be subsumed in Micranthocereus, according to the type species he cited for the former
(Cephalocereus purpureus Gürke), although it is very doubtful whether he was using that
binomial in its correct sense! He also accepted nearly all previous generic splits
involving E Brazilian species. His great adversary and critic was the cactus
evolutionary morphologist, Franz BUXBAUM, who supported the description of
perhaps the most distinct of all the newer genera to emanate from Eastern Brazil —
Uebelmannia Buining (1967). On reflection, it is surprising that Backeberg made no
attempt to name this group himself, having previously described its earliest-known
species (Parodia gummifera Backeb. & Voll) in 1950.
David HUNT (1967), in John Hutchinson’s The Genera of the Flowering Plants (vol. 2),
provided a comprehensive botanical account of the Cactaceae down to generic level. He
adopted a more conservative attitude than both Britton & Rose and Backeberg and
classified genera within a modified version of the former monographers’ tribal/subtribal
scheme. In total 84 genera were recognised for the family and their arrangement was partly
influenced by the morphological-phylogenetical studies of Buxbaum (see below). Notable
in Hunt’s treatment was his use of Cephalocereus in a very broad sense to include both
North American and Brazilian columnar taxa bearing cephalioid structures, ie. Arrojadoa,
Stephanocereus, Pilosocereus, Micranthocereus, Austrocephalocereus and Coleocephalocereus. The

only Brazilian exception to this was Facheiroa, which was subsumed into the Andean genus
Espostoa Britton & Rose, a move earlier proposed by Buxbaum (1959). Similarly, Rhipsalis
was used in a more inclusive sense than either Schumann or Britton & Rose, but
Schlumbergera was circumscribed in the manner accepted today. Both Leocereus and
Arthrocereus were recognised, but the latter, besides Brazilian taxa, included a divergent
Argentinian species now placed in Echinopsis (E. mirabilis Speg.).
The aforementioned Friedrich RITTER based his classification scheme on very careful
observation of the plants in the field, which knowledge he had gained during many years
of exploration (Eggli et al. 1995). In 1959 he was the first to recognize the generic status
of what later was named Uebelmannia Buining (1967), for which his manuscript name was
‘Gummocactus’ (cf. Ritter 1979). Later, in a paper on Brazilian cephalium-bearing cacti
(Ritter 1968) he independently recognised the distinctness of the rare Bahian endemic,
Cereus dybowskii, which he named Gerocephalus dybowskii, this, unfortunately, a few weeks
after Franz Buxbaum had published the priorable generic name, Espostoopsis, for the same
plant (Buxbaum 1968). He was also the first among modern authors to recognize that the
globular-stemmed genus, Melocactus, was closely related to the cereoid Coleocephalocereus
(Ritter, l.c., Taylor 1991a: 17). Subsequently, when writing up his many years’ results of
cactus study in South America, he published a series of new genera reflecting his excellent
knowledge of the Cactaceae (Ritter 1979). Of these, Pseudoacanthocereus and Cipocereus are
recognized today, but Floribunda and Mirabella have passed into the synonymy of
Cipocereus and Cereus, respectively (Zappi & Taylor 1991, Taylor 1991b, 1992a).
Although Ritter lacked a proper training in plant systematics, his generic concepts were
certainly more realistic than those of either Backeberg or Buxbaum. This is evidenced by
his treatment of Pilosocereus as including Pseudopilocereus Buxb., of Micranthocereus including
Austrocephalocereus, and of Coleocephalocereus including Buiningia Buxb., all of which are
circumscriptions followed in standard treatments today (eg. Hunt 1992b, 1999a, Barthlott
& Hunt 1993, Zappi 1994).
While Buxbaum described and named a new genus and a few new subgenera
amongst the cacti of Eastern Brazil, as well as rearranging some others, the major
contribution he made to the study of the Cactaceae as a whole was the evolution of a
tribal classification based on a comprehensive understanding of morphology (Buxbaum
1950–1954). His system was first presented in 1958 with further notes and adjustments
appearing over the next 17 years (Buxbaum 1958, 1959, 1962, 1968a & b, 1975; Endler
& Buxbaum 1974). His studies, terminology and tribal nomenclature still form the basis
for current schemes of classification, as described in the following section (Chap. 2.7).
Towards the close of the period under discussion a new tool for investigating micro-
morphology and especially that of pollen and seeds, was becoming important for the
study of cacti — the Scanning Electron Microscope (Leuenberger 1976, Barthlott &
Voit 1979). Influenced by data derived from these sources, the last treatment of the
Cactaceae to appear prior to the 1980s recognised a little over 100 genera (Barthlott
1977, 1979).
MODERN CLASSIFICATION
2.7 CACTACEAE SYSTEMATICS 1980–2003 AND THE

IOS CONSENSUS INITIATIVES
An unusual feature of Cactaceae systematics over the past two decades is the degree to
which its proponents have been organised. Since 1950, the International Organization
for Succulent Plant Study (IOS) has encompassed a significant group of cactus specialists
(as well as those with wider interests in succulent plants). A notable exception to this was
the wayward Curt Backeberg, whose major detractors were mostly IOS members! A
study of the names of those in attendance at successive biennial IOS congresses highlights
the opportunities that have existed for those interested in ‘cactology’ to exchange ideas
(Supthut 1999). Thus, it was perhaps not surprising that a proposal by IOS Secretary,
David Hunt, to establish a Cactaceae Working Party, which would aim towards a
‘consensus classification’, was made at the Frankfurt-am-Main congress of the
organization in 1984 (Anderson 1999). This consensus approach was partly driven by the
need of two of its major protagonists, David Hunt and Wilhelm Barthlott, to complete
treatments of the family for major reference works on vascular plants (ie. Walters et al.
1989: 202–301, Barthlott & Hunt 1993). Such treatments were at risk of being
hampered by the singular lack of orthodox taxonomic revisions of individual genera and
the widely contrasting approaches adopted in modern ‘standard works’ on the family,
such as those by the ‘splitters’, Curt Backeberg and Friedrich Ritter (see Chapter 2.6)
and ‘lumper’, Lyman Benson. The latter’s circumscription of Cereus Mill. was as broad
as that of its 18th Century author, Phillip Miller, taking in elements placed in at least 4
of Buxbaum’s tribes (Benson 1982).
In these circumstances, and recognizing that a considerable body of knowledge
existed amongst the IOS membership, but was fragmentary in nature, the drawing
together of a Working Party at annual meetings could bring benefits. This process was
agreed in 1984 and meetings have been held at least annually since then, facilitated by the
decision to hold ‘inter-congress’ meetings of IOS between its biennial congresses, starting
in 1985. Following a second meeting of the Working Party a draft list of genera was
published in 1986 (Hunt & Taylor 1986). This recognized a total of 86 genera in 3
categories: ‘unanimously accepted’ by the majority of Working Party participants (51
genera), ‘less than unanimously accepted’ by at least a third of participants (26 genera) and
the remainder (9 genera) to be retained out of nomenclatural expediency or because they
were considered ‘incertae sedis’. This first ‘consensus’ list was relatively conservative, but
as the Working Party grew, and with it the sources of useful information, there was a
gradual inflation of genera, so that the next list published (Hunt & Taylor 1990)
recognized a total of 93, once again in 3 categories: 55 accepted by at least 80% of
participants, 21 accepted by c. 60% and 17 for nomenclatural reasons or as incertae sedis
and accepted by 35–60%. The genera accepted in the second list were determined by a
postal ballot, which attracted responses from more than 20 specialists. Proposed changes
to the previous list were justified in a series of short printed notes and accompanied by
the sometimes contrasting views of particular participants, the present authors included
(Hunt & Taylor 1990: 98–104). Further thoughts and views were subsequently aired by
consensus participants in a paper printed the following year (Hunt & Taylor 1991) and
opportunities were taken to publish new names for use in the first edition of the CITES
Cactaceae Checklist (Hunt 1992b), which closely followed the IOS consensus list, but
accepted a further 8 genera in the third category, bringing the total for the family to 101.
This Checklist went down to the level of species and consequently drew attention to taxa
whose generic placement was particularly controversial. A series of short papers by
members of the Working Party, as well as by others not necessarily connected with the
IOS, has continued to appear (eg. Hunt & Taylor 1992 and in the newsletters, Cactaceae
Consensus Initiatives nos. 1–8, 1996–99, Cactaceae Systematics Initiatives nos. 9–15,
2000–03). Besides these shorter commentaries, more substantial monographic, synoptic
and cladistic treatments of larger or complex groups by members of the IOS Working
Party have begun to appear, those with relevance to Brazil including Pereskia
(Leuenberger 1986), Facheiroa (Braun & Esteves Pereira 1986–89), tribe Cereeae (Taylor
& Zappi 1989), Melocactus (Taylor 1991a), Pilosocereus (Zappi 1994), tribe Rhipsalideae
(Barthlott & Taylor 1995) and Uebelmannia (Nyffeler 1998). More recently, a second
edition of the CITES Checklist has appeared (Hunt 1999a), drawing heavily on the
above-cited treatments and accepting 108 genera, with their included taxa listed down to
the level of subspecies.
Since 1992, as with other groups of organisms, our understanding of Cactaceae
systematics has become heavily influenced by molecular data, in the form of phylogenies
derived from analysis of gene sequence variation (DNA/RNA), and this has been further
supported by anatomical, phytochemical, pollen and seed micro-morphological
information, which the molecular data are beginning to help interpret (Stuppy & Huber
1991, Wallace 1995, Wallace & Cota 1996, Maffei et al. 1997, Santos et al. 1997, Barthlott
& Hunt 2000, Stuppy 2002, Taylor et al. 2002, Wallace & Dickie 2002, Wallace &
Gibson 2002). The ways in which the IOS consensus process and subsequent studies have
contributed to the classification of the 30 cactus genera here recognized for Eastern Brazil
are listed in Table 2.1.
Relationships are clearly defined on molecular and other evidence at the subfamily
level (ie. Pereskioideae, Opuntioideae and Cactoideae, see Table 2.2; Barthlott & Hunt
1993, Wallace 1995, Wallace & Cota 1996, Wallace & Dickie 2002, Wallace & Gibson
2002, Stuppy 2002, Nyffeler 2002), but in the complex Cactoideae, comprising > 100
genera, the situation is less clear, though there appear to be c. 7 major clades of uneven size.
Unfortunately, some of the molecular research already conducted is yet to be formally
published, although regular presentations and updates have been provided by Robert
Wallace and his students (Iowa State Univ., USA) at IOS and other meetings, and most
recently by him and also Reto Nyffeler (Harvard Univ. Herbaria, USA; cf. Nyffeler 1999,
2000a–c) at the 26th IOS congress in Zürich, March 2000.
These investigations have demonstrated that the Brazilian Rhipsalideae DC. are a
monophyletic group (Nyffeler 2000a, 2002) and in Eastern Brazil recognition of 4 genera
based on morphological characters is well supported by the trnL-F intergenic spacer and
rpl16 intron molecular markers (fide Wallace). As already argued by Barthlott (1988), this
strongly contradicts Buxbaum’s tribal arrangement (Buxbaum 1958 etc., cf. Leuenberger
1976), where the slender-stemmed, small-flowered South American rhipsaloid epiphytes

were included within the Hylocereeae, which are larger-growing epiphytes and climbers
with a centre of diversity in Central America.
In-so-far as they have been sampled, the remaining E Brazilian Cactoideae divide
into two groups of Buxbaumian tribes (based on the rpl16 intron and rbcL markers,
Wallace & Gibson 2002, and trnK intron, including matK, Nyffeler 2000a, 2001, 2002):
the Hylocereeae-Pachycereeae-Echinocereeae clade (HPE, also includes Leptocereeae
pro parte and Corryocactus Britton & Rose) and the Browningieae-Cereeae-Trichocereeae
(BCT) clade. In the Taxonomic Inventory in Chapter 5 the Rhipsalideae are positioned
after the HPE clade and before the BCT clade, since more recently announced
molecular evidence suggests that the Rhipsalideae (excluding Pfeiffera, sens. lat.) are basal
to the latter (Nyffeler 2000c). The HPE clade includes the large-flowered, robust
climbers and epiphytes belonging to the Hylocereeae (Hylocereus, Epiphyllum etc.) to
which should probably be added the similar if enigmatic Pseudoacanthocereus, whose
position awaits confirmation with molecular evidence. In an earlier molecular analysis
the Hylocereeae appeared as the most basal element in Cactoideae (Wallace 1995: Fig.
11), but that position has now been taken by Calymmanthium F. Ritter, an aberrant
Peruvian cereoid.
The BCT clade is robustly monophyletic, characterised by a substantial deletion of
some 300 nucleotide base-pairs from the rpl16 intron, but its further resolution into tribes
awaits the completion of molecular systematics projects currently in progress. The clade
represents 16 out the total of 30 cactus genera of Eastern Brazil. However, preliminary
molecular studies by Soffiatti (unpubl.) have raised doubts about the current
circumscription of tribe Cereeae (8–9 genera). In contrast, evidence from a
phytochemical source suggests that the Cereeae could be a monophyletic component
within BCT, since the cuticular n-alkanes of Cereus, Pilosocereus and Melocactus,
representatives of markedly different parts of the tribe (Taylor & Zappi 1989: Figs 2–5),
are reported to be much more similar to each other than to those of other tribes (Maffei
et al. 1997). Brasilicereus is provisionally included in Cereeae here, but has aberrant, scaly
flowers and potentially might represent the Browningieae component of BCT. The
presumed Trichocereeae representatives are 7 genera, of which Leocereus, Espostoopsis and
Uebelmannia are each aberrant in different ways and their inclusion in the tribe is
provisional at present. Indeed, the last-named appears closer to Brazilian endemic genera
of Cereeae than to Trichocereeae, according to recent gene sequence studies (Soffiatti
unpubl.). Unlike the naked-flowered Cereeae, the Trichocereeae are characterised by
flowers ± clothed in hair-spines and/or sometimes spines or bract-scales, although
Espostoopsis has almost naked flowers (and may be convergent with the Cereeae).
The remaining major clades of Cactoideae recognizable on molecular evidence, that
are absent from Eastern Brazil, are the Cacteae (N Hemisphere), Notocacteae sens. str. (S
South America), Calymmanthium (Peru) and Copiapoa Britton & Rose (Chile), the latter
two lacking tribal names at present. A comparison of Buxbaum’s higher level classification
of Cactaceae with that implied by modern molecular phylogenies is offered in Table 2.3,
together with brief notes on the circumscription of his groupings.
Table 2.1
Status of genera of Cactaceae found in Eastern Brazil according to the
IOS consensus process (Hunt & Taylor 1986, 1990), CITES Checklist
editions (Hunt 1992b, 1999a) and their present treatment, with brief
justifications and notes.
(See further discussion under individual generic treatments in Chapter 5.3–5.5)
E Brazilian genus IOS Consensus status Comments/justification

(* = introduced)
P ERESKIOIDEAE K. Schum.:
1. Pereskia Mill. 1754 Unanimously accepted since 1986 Earliest generic name in subfam.,
but genus potentially paraphyletic
as circumscribed here
O PUNTIOIDEAE K. Schum.:
2. Quiabentia Britton Sunk into Pereskiopsis 1986 & 1990; Seed anatomy (Stuppy 2002)
& Rose 1923 recognised by Hunt (1992b, 1999a) supports separation from Pereskiopsis
3. Tacinga Britton & Unanimously accepted 1986 & Recognized in an expanded sense
Rose 1919 1990, but only in original here for 6 spp., based on seed
circumscription (sens. str.): 2 spp. anatomy and floral similarities
(Taylor et al. 2002)
4. Brasiliopuntia Sunk into Opuntia /1986 & 1990 Recognized here based on seed-
(K. Schum.) and pollen-morphology (lacks
A. Berger 1926 reticulate exine of Opuntia sens. str.)
and unique, autapomorphic habit
and behaviour (Taylor et al. 2002)
*5. Nopalea Salm-Dyck Sunk into Opuntia / 1986 & 1990 Recognized here based on pollen-
1850 morphology (lacks reticulate exine
of Opuntia sens. str.) and floral
differences (Taylor et al. 2002)
6. Opuntia Mill. 1754 Unanimously accepted since 1986 Earliest generic name in subfamily,
with sens. lat. circumscription here recognized sens. str. for taxa with
unique autapomorphic reticulate
pollen exine and sensitive stamens
C ACTOIDEAE (HPE clade) — Hylocereeae:

7. Hylocereus Britton Unanimously accepted 1986 & Circumscription expanded here to
& Rose 1909 1990, employing circumscription include Selenicereus sect. Salmdyckia
adopted by Hunt (1967) D. R. Hunt, based on trigonous
stem-morphology and taxa with
transitional floral/fruit characters,
supported by gene sequence data
(Wallace unpubl.)
*8. Selenicereus Britton Unanimously accepted 1986 & 1990 Circumscription adjusted here, as
& Rose 1909 with its circumscription adjusted in indicated above
1990 in line with Hunt (1989)
9. Epiphyllum Haw. Unanimously accepted since 1986 Earliest generic name in tribe
1812
C ACTOIDEAE (HPE clade?) — Echinocereeae(?):
10. Pseudoacanthocereus Included in Acanthocereus in 1986 & Fruit- and seed-morphology (Taylor
F. Ritter 1979 1990, and by Hunt (1992b) with a et al. 1992) unique within the HPE
note about its probable distinctness. clade and sister group as yet
Recognised by Taylor et al. (1992) unidentified
and by Hunt (1999a)
C ACTOIDEAE — Rhipsalideae:
11. Lepismium Pfeiff. Accepted to include only the Brazilian Circumscription employed here
1835 type species for reasons of defined on the basis of plesiomorphic
nomenclatural expediency in 1986 stem- and fruit-morphology
then in an expanded sense by a character states, as detailed in
minority of IOS WP members in Barthlott (1987) and Barthlott &
1990; accepted by Hunt (1992b, Taylor (1995). Brazilian elements
1999a) shown to be a monophyletic group
on the basis of gene sequence data
(Wallace unpubl.), but genus becomes
paraphyletic if some Andean taxa
are included (Nyffeler 2000c)
12. Rhipsalis Gaertn. Unanimously accepted 1986–90 Earliest generic name in the tribe;
present generic circumscription is
supported by phylogenetic analyses
13. Hatiora Britton Unanimously accepted 1986, then by Circumscription employed here
& Rose 1915 only a minority in 1990; subsequently defined on the basis of stem- and
accepted by Hunt (1992b, 1999a) floral-morphology, as described by
Barthlott (1987) and Barthlott &
Taylor (1995). E Brazilian type
species shown to be part of a
monophyletic group (Hatiora subg.
Hatiora) on the basis of phylogenetic
analyses, but genus becomes
paraphyletic if taxa referable to
Rhipsalidopsis B. & R. are included
14. Schlumbergera Lem. Unanimously accepted since 1986 Recognition of this genus is supported
1858 following expanded circumscription by gene sequence data (Wallace
employed in revision by Hunt (1969) unpubl.) and floral morphology:
presence of true perianth tube
C ACTOIDEAE (BCT clade) — Cereeae:

15. Brasilicereus Included in Monvillea Britton & Rose Defined on the basis of a unique
Backeb. 1938 1920 in 1986, but accepted by a combination of vegetative and floral
minority of IOS WP members in features within the BCT clade
1990, following recommendations (cereoid habit, short-tubed flowers
in Cereeae paper by Taylor & Zappi with glabrous but scaly pericarpel,
(1989); accepted by Hunt (1992b, stamens inserted in 2 series and
1999a) indehiscent fruits bearing brownish,
not blackish, perianth remains).
Strongly supported as distinct in a
recent gene sequence analysis
(Soffiatti unpubl.)
16. Cereus Mill. 1754 Unanimously accepted since 1986 Earliest generic name in the tribe
17. Cipocereus F. Ritter Lumped with Pilosocereus in 1986, then Distinguished from other Cereeae by
1979 accepted (incl. Floribunda F. Ritter) by its blue-waxy, globose, indehiscent
a minority of IOS WP members in fruits with translucent funicular pulp
1990, following recommendations in and persistent, erect, blackened
Cereeae paper by Taylor & Zappi perianth remains in combination
(1989); accepted by Hunt (1992b, with small seeds. Shown to be
1999a) monophyletic in a recent gene
sequence analysis (Soffiatti unpubl.)
18. Stephanocereus Accepted to include only the type Present circumscription reflects
A. Berger 1926 species for reasons of nomenclatural presumed homology between the
expediency in 1986, then in an two included species in respect of
expanded sense by a minority of IOS juvenile to adult developmental
WP members in 1990; accepted by stages and reproductive structures
Hunt (1992b, 1999a) (cephalia, flowers & fruit). It differs
from other Cereeae in these same
characteristics (Taylor & Zappi 1989)
19. Arrojadoa Britton Unanimously accepted since 1986 Earlier generic name than its sister
& Rose 1920 group, Stephanocereus, having derived
pollen characters (Taylor & Zappi
1989)
20. Pilosocereus Byles Unanimously accepted 1986 (incl. Differentiated from other Cereeae by
& G. D. Rowley Cipocereus) and then with present its depressed-globose, dehiscent fruits
1957 [Pilocereus circumscription in 1990 (Taylor & Zappi 1989, Zappi 1994)
K. Schum. 1894
nom. illeg.]
21. Micranthocereus Sunk into Arrojadoa in 1986, but Since 1990 its circumscription has
Backeb. 1938 accepted by a majority of IOS WP followed that of Ritter (1968, 1979)
members in 1990, following papers by and Taylor & Zappi (1989), including
Taylor & Zappi (1989) and Braun & Austrocephalocereus Backeberg 1938,
Esteves Pereira (1990) which the 1986 IOS report accepted
as distinct. Probably very close to
Pilosocereus, but with small,
hummingbird-syndrome flowers and
indehiscent fruit
22. Coleocephalocereus Accepted by a majority of IOS WP The presence of fruits dehiscent by

Backeb. 1938 members in 1986, then by fewer votes a small basal pore is supported by
in 1990, following the suggestion that molecular evidence (Cowan &
it might be paraphyletic in respect of Soffiatti unpubl.) as a sound
Melocactus by Taylor & Zappi (1989). autapomorphy, distinguishing the
Accepted by Hunt (1992b, 1999a) genus from both its potential sister
group, Melocactus, and other Cereeae
(cf. Taylor 1991a: 18)
23. Melocactus Link Unanimously accepted since 1986 Highly derived within Cereeae, with
& Otto 1828 autapomorphic non-chlorophyllous
terminal cephalium and distinctive
fruit, pollen and seeds (Taylor &
Zappi 1989, Taylor 1991a)
C ACTOIDEAE (BCT clade) — Trichocereeae:

24. Harrisia Britton Unanimously accepted since 1986 in Distinctive for its seed-anatomy
& Rose 1908 the broad sense including Eriocereus (Barthlott & Hunt 2000: 27,
Riccob. (1909) pl. 23–24); circumscription
supported by gene sequence data
(Wallace 1997)
25. Leocereus Britton Accepted less than unanimously by Fruit- and seed-morphology unique
& Rose 1920 IOS WP members in 1986 (when it within the BCT clade and sister
was hinted that Arthrocereus A. Berger group as yet unidentified. Position
might be included), then fully accepted awaits confirmation based on
as a monotypic entity in 1990, molecular data
following commentary by Taylor
(Hunt & Taylor 1990: 100) and data
in Taylor & Zappi (1990)
26. Facheiroa Britton Accepted less than unanimously by Sister group within tribe at present
& Rose 1920 IOS WP members in 1986 & 1990; uncertain and its position remains
accepted by Hunt (1992b, 1999a) to be elucidated with molecular data
27. Espostoopsis Buxb. Maintained as a synonym of In stem and habit characters this plant
1968 Austrocephalocereus in 1986 (which strongly resembles the Andean genus,
had been Buxbaum’s later view, Espostoa Britton & Rose, but has
cf. Leuenberger 1976), then naked flowers and polycolpate
recognized by a minority of IOS pollen. Its placement in Trichocereeae,
WP members in 1990, following rather than Cereeae, is provisional
recommendations by Taylor & and based on the strong suspicion
Zappi (1989); accepted by Hunt that its floral features are merely
(1992b, 1999a) convergent with Cereeae, where it
appears to lack any obvious relatives
28. Arthrocereus Tentatively referred to Leocereus in Close to the very large and complex
A. Berger 1929 1986, then accepted by a minority of genus, Echinopsis Zucc., but differing
IOS WP members in 1990, following in its pollen (Leuenberger 1976) and
debate reported in Hunt & Taylor and indehiscent fruits. Molecular
1990: 99–100; accepted by Hunt evidence is needed to confirm its
(1992b, 1999a) generic status or otherwise
29. Discocactus Pfeiff. Unanimously accepted since 1986 Assumed to be the sister group of the
1837 much larger Gymnocalycium Pfeiff. ex
Mittler 1844, but distinguished by
having 12–15-colpate (vs 3-colpate)
pollen (Leuenberger 1976) and the
presence of a terminal cephalium —
a parallelism with Melocactus
30. Uebelmannia Accepted less than unanimously by Strongly supported as monophyletic

Buining 1967 IOS WP members in 1986, then in a recent gene sequence analysis
unanimously in 1990; accepted by (Soffiatti unpubl.), which places it
Hunt (1992b, 1999a) close to Cipocereus (cf. Cereeae above)
Table 2.2
Characters defining the subfamilies of Cactaceae represented in Eastern
Brazil.
Subfamily Characters / molecular markers
P ERESKIOIDEAE K. Schum. (Pereskia only) Woody, with essentially non-succulent stems bearing
broad, functional, scarcely succulent leaves and
un-barbed spines; seeds of ‘common ancestral
centrospermous type’ (Barthlott & Hunt 1993);
pollen 3–15-colpate. Has no shared cpDNA restriction
site changes with the putatively basal Maihuenioideae
(Maihuenia only), but may be paraphyletic in respect of
Cactoideae (Wallace 1995: 9)
O PUNTIOIDEAE K. Schum. Stems succulent, at least at first, bearing cylindrical to

awl-shaped or rarely flattened, succulent and often
caducous leaves and two distinct kinds of barbed
spines; seeds covered in a bony aril formed from the
funicle (unique in the centrosperms); pollen mostly
12–18-porate. Has autapomorphic deletion in the
plastid gene accD, ORF 512, from the large single
copy portion of cpDNA (Wallace & Gibson 2002)
C ACTOIDEAE Stems succulent, leaves reduced to minute ephemeral

scales or usually entirely absent; spines un-barbed; seeds
various, lacking an aril or this a corky appendage
at the hilum only; pollen mostly 3(–12)-colpate.
Has autapomorphic c. 700 base-pair deletion from the
chloroplast encoded gene rpoC1 intron (Wallace &
Cota 1996)
Table 2.3
Buxbaum’s tribal arrangement of Cactaceae-Cactoideae (after
Leuenberger 1976) evaluated by modern molecular data
(Wallace, pers. comm. etc.). For terminology, see Chapter 7.1.
Tribes/subtribes of Buxbaum, Status based on molecular Comments

principal included genera [* = native data (gene sequence
in E Brazil] and their characteristics variation)
L EPTOCEREEAE Buxb.
Leptocereus, Calymmanthium, Armatocereus, Buxbaum’s circumscription is not Buxbaum was
Neoraimondia, Samaipaticereus etc. supported by molecular data and grouping these taxa
Erect columnar-cereoid taxa with few these genera should be dispersed on the basis of what
high stem ribs and shortly tubular flowers amongst the tribes that follow, are now seen to be
bearing areoles and spines except for Calymmanthium, which symplesiomorphies
probably merits its own tribe
H YLOCEREEAE Buxb.: 5 subtribes

Nyctocereus, Peniocereus, Acanthocereus, Molecular evidence requires the As circumscribed
Harrisia*, Aporocactus, Selenicereus, Deamia, removal of the small-flowered, by Buxbaum, the
Hylocereus*, Epiphyllum*, Disocactus, S Hemisphere epiphytes as the Hylocereeae should
Pseudorhipsalis, Pfeiffera, Hatiora*, Rhipsalideae DC., sens. str. (excl. have been called
Schlumbergera*, Rhipsalis*, Lepismium* etc. Pseudorhipsalis, and also Pfeiffera sens. the Rhipsalideae
Suberect shrubs, lianas and epiphytes lat.; Nyffeler 2000c). Acanthocereus, DC., which is an
(both erect and pendent) with few-ribbed Leptocereus (see above), Deamia older tribal name
cylindric or flattened stems, very diverse and the Hylocereeae sens. str. are
flowers and ± smooth seedcoats (except basal to the Echinocereeae
Harrisia) (Pachycereeae) elements in the
HPE clade. Harrisia belongs in
Trichocereeae
PACHYCEREEAE Buxb.: 4 subtribes

Pterocereus, Escontria, Pachycereus, Stenocereus, These form a significant part of the Buxbaum included
Carnegiea, Neobuxbaumia, Cephalocereus, HPE clade, to which must be added Pilosocereus p.p.
Myrtillocactus etc. the relevant parts of Buxbaum’s (tribe Cereeae)
Erect columnar-cereoid taxa with woody Leptocereeae and Hylocereeae, and in Cephalocereus
tissues organized into discrete rods and the Echinocereeae (q.v.)
diverse flowers with spiny, bristly, scaly or
woolly pericarpels
B ROWNINGIEAE Buxb.
Browningia, Castellanosia etc. Part of Browningia sens. lat. belongs
Erect columnar-cereoid taxa with densely to the BCT clade, but recent
scaly tubular flowers reports suggest that Castellanosia
is not related (Nyffeler 2002)
C EREEAE Salm-Dyck
Jasminocereus, Stetsonia, Praecereus, Cereus*, With the exception of Jasminocereus, Cereus, Praecereus
Pseudopilocereus [Pilosocereus p.p.]*, which belongs with Armatocereus and Brasilicereus
Stephanocereus*, Coleocephalocereus*, in the HPE clade, the remainder may not be close to
Brasilicereus* included here by Buxbaum, where the E Brazilian taxa
Erect columnar-cereoid taxa, bearing sampled, are BCT clade members included here
± naked, tubular, mostly nocturnal flowers and those having a naked pericarpel according to rpl16
and ± naked fruit are presently referred to Cereeae gene sequences
(Soffiatti unpubl.)
T RICHOCEREEAE Buxb.: 4 subtribes

Echinopsis, Haageocereus, Espostoa, Facheiroa*, All elements included here by The position of
Austrocephalocereus (incl. Espostoopsis)*, Buxbaum belong in the BCT Leocereus remains to
Leocereus*, Arthrocereus*, Cleistocactus, clade, but some taxa are here be clarified on the
Oreocereus, Matucana, Micranthocereus*, referred to the Cereeae on the basis basis of molecular
Arrojadoa*, Rebutia etc. of floral morphology, and other markers
Erect columnar-cereoid to low-growing BCT-Trichocereeae were
taxa, bearing long-tubed flowers with misplaced in the Notocacteae.
bract-scales and/or spines/hair-spines Samaipaticereus (Leptocereeae) and
Harrisia (Hylocereeae sensu Buxb.)
belong here (Wallace 1997)
NOTOCACTEAE Buxb.: 5 subtribes

Corryocactus, Austrocactus, Eriosyce, Notocacteae in its strictest sense The Notocacteae as
Eulychnia, Copiapoa, Parodia, Frailea, (cf. Nyffeler 2001; ie. Parodia previously defined
Uebelmannia*, Astrophytum, Gymnocalycium, sens. lat. and Eriosyce) is strongly were recognizable
Sulcorebutia [= Rebutia p.p., see above], supported by molecular data, on the basis of
Discocactus*, Melocactus* etc. but the tribe sensu Buxbaum floral characters
Mostly globular-stemmed, ribbed or is otherwise a complete mixture, (woolly pericarpel
tuberculate cacti, bearing diurnal short- including elements of at least 4 with bristles above)
tubed flowers with mostly scaly and/or other major lines, eg. the BCT in combination with
bristly and woolly pericarpels; many clade, the HPE clade (Eulychnia, unique seedcoat
unique seedcoat characters Austrocactus, Corryocactus), Copiapoa morphology
& Cacteae (Astrophytum)
E CHINOCEREEAE Buxb.
Bergerocactus, Echinocereus. Not recognizable as a discrete entity Echinocereeae will
Low-growing, shortly cereoid cacti on molecular data, being an integral likely prove to be
bearing diurnal flowers with areolate part of the large HPE clade and the oldest name for
-spiny pericarpels close to Stenocereus (Pachycereeae) the HPE clade
C ACTEAE : 4 subtribes
Echinocactus, Sclerocactus, Thelocactus, Well-defined major clade strongly The majority of
Turbinicarpus, Lophophora, Strombocactus, supported by molecular and other taxa are
Ariocarpus, Ferocactus, Escobaria, data when the North American characterized by a
Mammillaria, Coryphantha etc. Astrophytum is included from suite of apomorphic
Mostly globular-stemmed, ribbed or Buxbaum’s Notocacteae seedcoat characters
tuberculate cacti, bearing diurnal that sets them apart
short-tubed flowers with naked, from all other cacti
scaly and/or woolly pericarpels
PHYTOGEOGRAPHY
3. PHYTOGEOGRAPHY
3.1 OVERVIEW
Following detailed study of the systematics, habitats and range of the Cactaceae native in
Eastern Brazil it becomes obvious that there are some more or less well-defined
distribution patterns into which the majority of taxa can be classified. These patterns
generally correspond with geographical, climatic, edaphic and other ecological
phenomena and thus may offer help in defining phytogeographical regions and give clues
about past vegetational history (Harley 1988, Prado & Gibbs 1993). Such studies are also
important from the standpoint of conservation, since they indicate the minimum number,
range and diversity of areas and habitats that need to be considered for protection (for
priorities, see Chapter 4: Table 4.3).
Perhaps the most interesting and potentially informative aspect of this study is to use
the areas defined below as a model to compare with other floristically important plant
groups found in Eastern Brazil, especially those with ecological preferences similar to
Cactaceae, such as Araceae (eg. Philodendron, Anthurium) and Bromeliaceae (eg. Dyckia,
Encholirium), both of which are frequently lithophytic. Another group including caatinga
species, which is sometimes compared in the discussion below, is Caesalpinia (Lewis
1998), while the Hyptidinae (Lamiaceae) include a number of campo rupestre taxa with
ranges similar to those of Cactaceae (cf. Harley 1988). However, at present, opportunities
for making such comparisons are few, because, if they are to be meaningful, it is necessary
to have a rather detailed data set on the group concerned. This means monographic
studies, to enable the careful determination of species limits, relationships, geographical
range and reproductive-dispersal strategies, all requiring extensive field knowledge. This
is perhaps the first study of its kind to focus on a taxonomic group of reasonable size and
ecological complexity. Hopefully, there will be future opportunities to evaluate its
phytogeographical significance in relation to other families found in Eastern Brazil,
especially those mentioned above.
This chapter considers all named, fully accepted taxa of Cactaceae that are native of
the North-eastern region (‘Nordeste’) of Brazil, and of its South-eastern region
(‘Sudeste’) north of c. 22°S and east of c. 46°W, as indicated on Map 1. As defined, the
area chosen includes all Brazilian Cactaceae endemic to the great caatingas dominion
(Andrade-Lima 1981) and all but two of those endemic to the campos rupestres, the
vegetation type that characterizes the East Brazilian Highlands (Harley in Stannard 1995:
25), whose western part, with few cacti (W Minas Gerais & Goiás), is excluded (see Map
1). Three-quarters or 76% (123) of the 162 native species and heterotypic subspecies of
cacti found in Eastern Brazil are endemic and we believe this justifies its choice for study.
PHYTOGEOGRAPHY
While the whole of this geographical area has been considered, initial field and
herbarium investigations suggested focusing on a somewhat smaller core area delimited at
c. 7°S and c. 46°W, ie. including south-eastern Piauí, the southernmost tip of Ceará, the
southern half of Paraíba, then southwards to 22°S, in which there is a much greater
diversity of taxa. Cereus insularis (Fernando de Noronha), is the only endemic from the
Nordeste that does not enter this core area. Table 3.1 indicates taxa ranging beyond the
core area and those endemic to it, the latter being two-thirds (107 taxa) of the overall
total. A significant part of the territories of the northernmost states of the Brazilian
Nordeste, comprising Maranhão, Piauí, Ceará, Rio Grande do Norte and Paraíba, has a
relatively poor cactus flora, none having more than 16 native taxa. Only Minas Gerais and
Bahia have significant numbers of state endemic taxa. Possible explanations for these
differing levels of diversity will be discussed below.
As stated above, the core geographical area defined within Eastern Brazil and
particularly focused on in this study has been recognized primarily for its remarkable
endemism of Cactaceae genera and species, especially those of the caatinga of the North-
eastern states and associated highland campos rupestres. Both vegetation types extend into
the northern part of the SE Brazilian state of Minas Gerais, where the highest
concentration of Cactaceae taxa is found. Habitats including a significant representation
of cacti also reach into the Atlantic Forest (Mata atlântica) of eastern Minas Gerais and the
neighbouring state of Espírito Santo, where a substantial area drained by the Rio Doce
and adjacent river systems receives less than 1000–1250 mm of rain per year (see map
‘isoietas anuais 1914–1938’ in Azevedo 1972; Nimer 1973: 40, fig. 18). In this extensive
core area representatives of the family appear to have evolved for long periods in isolation,
so that many distinctive taxa without parallel elsewhere have arisen. It is also true that
various major cactus genera, widespread in the neotropics, are absent from this region.
These include Selenicereus (NB. S. subg. Salmdyckia is here referred to Hylocereus),
Disocactus, Pseudorhipsalis, Echinopsis and Cleistocactus, besides others mentioned below,
which, unlike the above, are replaced in Eastern Brazil by obvious, closely related
vicariants. Genera from the widespread, species-rich, South American Notocacteae are
likewise absent. Other significant neotropical genera are represented by only a single
native species in each case, eg. Opuntia (sens. str.), Hylocereus, Epiphyllum and Harrisia. In
the case of Opuntia, the single species, O. monacantha, is only marginally represented,
being a rarity at the northern limits of its range.
It is perhaps not difficult to see why this should be so, since today the region is
effectively cut off from other centres of cactus diversity by very broad zones of habitat
unfavourable to most members of the family. To the north-west of the dry region of
North-eastern Brazil is the nearly constantly humid Amazonian region, with its
rainforests and included savanna formations, where only few specialized cacti are able
to compete with other plants (see Appendix II). Amongst these is the epiphyte,
Rhipsalis baccifera subsp. baccifera, whose range southwards into Eastern Brazil stops at c.
8°S in the coastal forests of Pernambuco (around Recife), indicating the southern limit
of Amazonian floristic influence in terms of Cactaceae. To the west, in Central-western
Brazil, are the extensive fire-swept cerrados, which are avoided by most cactus genera
PHYTOGEOGRAPHY
Table 3.1
Distribution of species and subspecies of Cactaceae native in Eastern
Brazil by state.
Key: † = taxon endemic to the total area included in this table; ‡ = taxon endemic to core area (northern
limit 7°S); § = ‘single-site endemic’ (definition in Chap. 3.3); + = definitely recorded as native; ± =
record requiring confirmation or of dubious identity; +? = possibly native or cultivated; c = cultivated;
FN = Fernando de Noronha (PE) only; MG1/RJ2 = these states included only east of 46ºW and north
of c. 22ºS. For locations of states/key to 2-letter codes, see caption to Map 1.
TAXON / STATE MA PI CE RN PB PE AL SE BA MG1 ES RJ2
Pereskia aculeata + + + + + + + +
P. grandifolia subsp. grandifolia +? + + + + + +
P. grandifolia subsp. violacea ‡ c + +
P. bahiensis ‡ +
P. stenantha ‡ + +
P. aureiflora ‡ + +
Quiabentia zehntneri ‡ + +
Tacinga funalis ‡ ± + +
T. braunii ‡ +
T. werneri ‡ + +
T. palmadora † + + + + + +
T. saxatilis subsp. saxatilis + +
T. saxatilis subsp. estevesii ‡§ +
T. inamoena † + + + + + + + + + +
Brasiliopuntia brasiliensis + + + + + + + +
Opuntia monacantha c c + +
Hylocereus setaceus ± + + + + + + + + +
Epiphyllum phyllanthus + + + + + + + + + + +
Pseudoacanthocereus brasil. ‡ + +
Lepismium houlletianum +
L. warmingianum + +
L. cruciforme + + + +
Rhipsalis russellii + + +
R. elliptica +
R. oblonga + +
R. crispata + +
R. floccosa subsp. floccosa + + + +
R. floccosa subsp. oreophila ‡ + +
R. floccosa subsp. pulvinigera + +
R. paradoxa subsp. septentrion. ‡ + + + +
R. pacheco-leonis subsp. catenul. +
R. cereoides +
R. sulcata ‡§ +
R. lindbergiana + + + + +
R. teres + +
R. baccifera subsp. baccifera + + + +
R. bacc. subsp. hileiabaiana ‡ +
PHYTOGEOGRAPHY
R. pulchra +
R. burchellii ± +
R. juengeri +
R. clavata ± + +
R. cereuscula + + +
R. pilocarpa + +
R. hoelleri ‡§ +
Hatiora salicornioides + + + +
H. cylindrica + ±
Schlumbergera kautskyi ‡ +
S. microsphaerica + ±
S. opuntioides +
Brasilicereus phaeacanthus ‡ + +
B. markgrafii ‡ +
Cereus mirabella + + +
C. albicaulis † + + + ± + ±
C. fernambucensis subsp. fern. + + + + + + + +
C. fernambuc. subsp. sericifer + + +
C. insularis † FN
C. jamacaru subsp. jamacaru †? + + + + + + + + + + c c
C. jamacaru subsp. calcirupicola‡ + +
C. hildmannianus +
Cipocereus laniflorus ‡§ +
Cipocereus crassisepalus ‡ +
C. bradei ‡ +
C. minensis subsp. leiocarpus ‡ +
C. minensis subsp. minensis ‡ +
C. pusilliflorus ‡§ +
Stephanocereus leucostele ‡ +
S. luetzelburgii ‡ +
Arrojadoa bahiensis ‡ +
A. dinae subsp. dinae ‡ + +
A. dinae subsp. eriocaulis ‡ + +
A. penicillata ‡ + +
A. rhodantha ‡ + + + +
Pilosocereus tuberculatus ‡ + + +
P. gounellei subsp. gounellei † + + + + + + + + +
P. gounellei subsp. zehntneri ‡ + +
P. catingicola subsp. catingicola ‡ +
P. catingic. subsp. salvadorensis † ± + + + + + +
P. azulensis ‡§? +
P. arrabidae + + +
P. brasiliensis subsp. brasiliensis +
P. brasiliensis subsp. ruschianus ‡ + + +
P. flavipulvinatus † + + +
P. pentaedrophorus subsp. pent. ‡ + + +
P. pentaedroph. subsp. robustus ‡ + +
P. glaucochrous ‡ +
P. floccosus subsp. floccosus ‡ +
PHYTOGEOGRAPHY
P. flocc. subsp. quadricostatus ‡ +

P. fulvilanatus subsp. fulvil. ‡ +
P. fulvilanatus subsp. rosae ‡§ +
P. pachycladus subsp. pachycl. ‡ + +
P. pachycl. subsp. pernambuco. † + + + + + + +
P. magnificus ‡ +
P. machrisii + +
P. aurisetus subsp. aurisetus ‡ +
P. aurisetus subsp. aurilanatus ‡ +
P. aureispinus ‡ +
P. multicostatus ‡ +
P. piauhyensis † + + +
P. chrysostele † + + + +
P. densiareolatus ‡ + +
Micranthocer. violaciflorus ‡ +
M. albicephalus ‡ + +
M. purpureus ‡ +
M. auriazureus ‡§ +
M. streckeri ‡§ +
M. polyanthus ‡ +
M. flaviflorus ‡ +
M. dolichospermaticus ‡ + +
Coleocephalocer. buxb. subsp. b ‡ + +
C. buxbaum. subsp. flavisetus ‡ +
C. fluminensis subsp. fluminensis + + +
C. fluminensis subsp. decumbens ‡ +
C. pluricostatus ‡ + +
C. goebelianus ‡ + +
C. aureus ‡ +
C. purpureus ‡§ +
Melocactus oreas subsp. oreas ‡ +
M. oreas subsp. cremnophilus ‡ +
M. ernestii subsp. ernestii † ± ± + + ± + + +
M. ernestii subsp. longicarpus ‡ + +
M. bahiensis subsp. bahiensis ‡ + +
M. bahiensis subsp. amethystinus ‡ + +
M. conoideus ‡§ +
M. deinacanthus ‡§ +
M. levitestatus ‡ + +
M. azureus ‡ +
M. ferreophilus ‡ +
M. pachyacanthus subsp. pachy. ‡ +
M. pachyacanthus subsp. viridis ‡ +
M. salvadorensis ‡ +
M. zehntneri † + + + + + + + +
M. lanssensianus ‡§? ± +
M. glaucescens ‡ +
M. concinnus ‡ + +
M. paucispinus ‡ +
PHYTOGEOGRAPHY
M. violaceus subsp. violaceus + + + + + + +

M. violaceus subsp. ritteri ‡ +
M. violac. subsp. margaritaceus ‡ + + +
Harrisia adscendens ‡ ± + + + + + +
Leocereus bahiensis ‡ ± + + +
Facheiroa ulei ‡ +
F. cephaliomelana subsp. c. ‡ + +
F. cephaliom. subsp. estevesii ‡§ +
F. squamosa ‡ + + +
Espostoopsis dybowskii ‡ +
Arthrocereus melanurus subsp. me. +
A. melanurus subsp. magnus ‡§ +
A. melanurus subsp. odorus ‡ +
A. rondonianus ‡ +
A. glaziovii ‡ +
Discocactus zehntneri subsp. z. ‡ +
D. zehntneri subsp. boomianus ‡ +
D. bahiensis † + + ± +
D. heptacanthus subsp. catingic. + + +
D. placentiformis ‡ +
D. pseudoinsignis ‡ +
D. horstii ‡§ +
Uebelmannia buiningii ‡ +
U. gummifera ‡ +
U. pectinifera subsp. pectinifera ‡ +
U. pectinifera subsp. flavispina ‡ +
U. pectinifera subsp. horrida ‡§ +
STATE MA PI CE RN PB PE AL SE BA MG1 ES RJ2

STATE ENDEMICS 0 0 0 0 0 1? 0 0 30 35 3 0
TAXA SCORED + 8 16 15 11 15 33 14 19 95 97 34 11
SPECIES SCORED + 8 16 15 11 15 33 14 19 79 80 30 11
(except Discocactus in gravelly areas), although included rock outcrops and gallery forests
do provide some suitable habitats (eg. for Rhipsalis russellii, Pilosocereus machrisii and
allies, Cereus spp. and Arthrocereus spinosissimus) and have probably permitted a limited
amount of migration to and from the caatingas, Mata atlântica and campos rupestres in the
past (cf. Prado & Gibbs 1993). To the south the diversity of non-epiphytic cacti
abruptly decreases as humidity increases, and these are replaced by numerous epiphytic
species from tribe Rhipsalideae (Barthlott & Taylor 1995) until the grasslands (or
campos) of Rio Grande do Sul are reached. Here the diversity of non-epiphytic cacti
suddenly increases again; these, however, are globular species from tribe Notocacteae.
Such ecological barriers in recent times must have severely limited migration of cactus
species and genera to and from neighbouring regions of high diversity, such as the
Caribbean, the Andes and the part of south-eastern South America where the
Notocacteae are most abundant.
PHYTOGEOGRAPHY
While Table 3.1 indicates which taxa are endemic to Eastern Brazil and the core area,
it is also worthwhile to summarize endemism in terms of the major taxonomic groups
represented:
Pereskioideae. Pereskia, nowadays the only genus included in this subfamily, is

comprehensively represented in Eastern Brazil in terms of diversity, since elements of the
3 major groups of species recognized by Leuenberger (1986) are present and give a total
of 5 species, with 3 endemic to the core area. The endemic species (P. bahiensis, P.
stenantha and P. aureiflora) are restricted to the caatinga-agreste, while P. aculeata and P.
grandifolia (and its endemic subsp. violacea, which may be specifically distinct) are Mata
atlântica species.
Opuntioideae. In terms of species, the 4 rather distinct genera of Opuntioideae native

of Eastern Brazil are very small, with only 9 species in total (6 belonging to Tacinga, sens.
lat., an endemic of the NE and SE regions of Brazil). At first sight this seems hard to
explain, since elsewhere this subfamily is generally species-rich in the dry zones of the
Americas, eg. in the Andes, Mexico and south-western United States, where they
invariably constitute a major component of cactus floras. So why have so few taxa
evolved in Eastern Brazil? The answer may lie in the origin of the subfamily. It includes
taxa resembling the genus Maihuenia (Maihuenioideae, cf. Wallace 1995), a strong
candidate for the plesiomorphic sister group of Opuntioideae and a clear parallel to some
members of the Portulacaceae, a part of which is the sister group of the Cactaceae
(Hershkovitz & Zimmer 1997, Savolainen et al. 2000: Fig. 2). Maihuenia is a genus from
temperate South America (Patagonia etc.) and the Opuntioideae in major part are also
represented by plants of cooler mountainous regions (Andes) or higher plateau lands
(northern Mexico etc.), perhaps suggesting that they are generally less well adapted to life
in the constantly warm, lowland parts of the tropics. In the drier parts of such tropical
regions the Opuntioideae have, however, evolved some highly specialized and derived
forms (derived when compared to a supposed Maihuenia-like ancestor), such as Tacinga
(including 2 spp. of cactus lianas, E Brazil), Brasiliopuntia (1 treelike sp., neotropics south
of the River Amazon), Consolea (c. 8 treelike spp., E Caribbean) and Pereskiopsis (6 semi-
scandent or treelike spp., trop. Mexico & Cent. America). Quiabentia (2 treelike spp.,
South American Chaco region & E Brazil) is not included in this list, since at least its
western species comes from a region that experiences frost in winter (Map 13) and its
Brazilian counterpart is confined to the more seasonal south-western caatingas. Tacinga,
sens. lat. (6 spp.), extends considerably beyond the E Brazilian core area into the
northernmost caatingas (T. inamoena), and slightly beyond it to the west, in Minas Gerais,
on edaphically dry limestone outcrops (T. saxatilis). It completely replaces Opuntia, sens.
str., in most of Eastern Brazil, but is lacking at the southern margins of the core area where
O. monacantha reaches its northern limit.
Cactoideae. Out of the 8 tribes and 3 or more other major clades of Cactoideae
currently recognised (Wallace 1995, Wallace & Gibson 2002, Nyffeler 2000a–c; see
Chapter 2.7) nearly all E Brazilian cactus genera of this relationship fall into tribes Cereeae
PHYTOGEOGRAPHY
(9 genera), Trichocereeae (7 genera) and Rhipsalideae (4 genera), with only Hylocereus (1

sp.) and Epiphyllum (1 sp.) in the Hylocereeae and Pseudoacanthocereus (1 sp.) doubtfully in
Echinocereeae, sens. lat. (incl. Leptocereeae & Pachycereeae). Therefore, Eastern Brazil
is apparently rather lacking in diversity at the tribal level or equivalent. However, at
generic level the situation is different. Cereeae has its major centre of diversity in Eastern
Brazil (Taylor & Zappi 1989) and many of its genera and subgenera are endemic:
Cipocereus (5 spp.), Brasilicereus (2 spp.), Stephanocereus (2 spp.), Micranthocereus subg.
Micranthocereus (4 spp.), M. subg. Austrocephalocereus (3 spp.), Coleocephalocereus subg.
Buiningia (2 spp.) and C. subg. Simplex (1 sp.). Micranthocereus (9 spp.), Arrojadoa (4–5 spp.)
and Coleocephalocereus (6 spp.) are almost endemic, the first with only M. estevesii (Subg.
Siccobaccatus) located outside the core area on limestone outcrops between southernmost
Tocantins and NW Minas Gerais, the second with a recently discovered taxon in NE
Goiás. The Trichocereeae, while represented by relatively few species, are also diverse at
generic level, with 4 out of 7 genera being endemic: Uebelmannia (3 spp.), Facheiroa (3
spp.), Espostoopsis (1 sp.) and Leocereus (1 sp.). The stronghold of this tribe in Eastern Brazil
at species/subspecies level is in the South-eastern campos rupestres and adjacent cerrados,
where they total nearly half of the Cactaceae. All 4 genera of Rhipsalideae sens. str. are
found in Eastern Brazil, but none is endemic and few species and subspecies are endemic,
since the Mata atlântica they mostly inhabit extends well beyond the limits of the core area
to the south and south-west. Excepting Rhipsalis baccifera, their range northward
terminates at the northern border of Pernambuco.
3.2 DEFINITION OF VEGETATION TYPES

In most cases, the terminology used in this book to identify types of vegetation in Eastern
Brazil follows that employed by Prado (1991) and Harley in Stannard (1995: 7–11, 13–27,
37–40, 47–50, 52–66), which contain more detailed treatments including numerous
valuable references. Azevedo (1972), Cavalcanti Bernardes (1951), Domingues (1973),
King (1956) and Nimer (1973) have been consulted as sources of edaphic, geological and
climatic data.
Mata atlântica (Atlantic Forest, Plate 1.1, Map 2). In its broadest sense this term covers
a wide range of evergreen and semi-evergreen forest formations, beginning at the coast,
from the high tide mark, and stretching far inland until, in NE Brazil, the deciduous
caatinga is reached. It also reappears, in fragmented form, at the base of the eastern side of
the Chapada Diamantina in Bahia and as riverine forest along the Rio São Francisco
(Brandão 2000: 75). In SE Brazil such vegetation continues much further inland, on to
the planalto central, and there once formed extensive tracts of semi-humid ‘mata de
planalto’, which is considered a part of the Atlantic Forest system by many Brazilian
botanists. In some areas the forest includes extensive outcrops of smooth gneissic rocks or
larger inselbergs and these represent an important habitat for rupicolous cacti in such
areas, especially from south-eastern Bahia (Floresta Azul) southwards and south-
westwards to Rio de Janeiro, where, for example, Pilosocereus brasiliensis and
PHYTOGEOGRAPHY
Coleocephalocereus subg. Coleocephalocereus are characteristic rupicolous taxa (Maps 36 &

41). Although normally thought of as a very humid environment, this is by no means
uniformly the case, since rainfall can vary from more than 2000 mm down to less than
800 mm per annum and often includes a more or less well-marked dry season. The cactus
flora varies considerably with levels and frequency of precipitation and, as a consequence,
many taxa show markedly disjunct distributions. Minimum temperatures vary in line with
latitude and altitude, but are probably never less than 0°C anywhere in the regions
considered in this study, and are never <12°C, from Rio Grande do Norte to coastal S
Bahia (Nimer 1973: 42).
In the drier or rocky and less dense phases of the Mata atlântica large, terrestrial (ie.
non-epiphytic) cacti are able to compete with other forest species, such as in the coastal
forests around the border regions of Alagoas, Sergipe and Bahia (eg. Cereus jamacaru —
typically a caatinga species), in eastern Minas Gerais and central Espírito Santo (eg. Pereskia
grandifolia subsp. violacea) and in the lowlands of southern Espírito Santo and north-eastern
Rio de Janeiro (eg. Brasiliopuntia brasiliensis). At or near the coast the forest is termed
restinga and even in more humid areas frequently includes edaphically dry sand-dunes in
which an open or denser scrub, or taller forest supports a variety of non-epiphytic cacti.
However, the dense, multilayered forest in areas where there is rain every month of the
year includes only cactus epiphytes, but even these can be absent or rare in some phases
of this ecosystem, eg. the ‘mata de tabuleiro’ found on elevated Tertiary Barreiras sediments
in many coastal areas of NE Brazil. Further landwards the ‘brejo’ forests (mata de brejo)
represent isolated patches of more humid, evergreen vegetation on mountain ridges
surrounded by seasonally dry Atlantic Forest or even caatinga (Map 6), eg. the Pico do
Jabre (Mun. Maturéia, Paraíba; Rocha & Agra 2002). These benefit from lower average
temperatures, greater cloud cover and overnight dews that compensate for rain during dry
periods (Rodal et al. 1998). Besides Melocactus ernestii, which is common on exposed
rocks, they are often rich in epiphytes, with a greater variety of Rhipsalideae than is found
generally in Mata atlântica. Finally, in eastern parts of the Chapada Diamantina (Bahia) and
Serra do Espinhaço (N Minas Gerais) there occur isolated montane cloud forests (‘mata de
neblina’ at 1200–1800 m), which, at the highest elevations, are extremely humid. These
are treated as part of the campo rupestre mosaic here (see below), because their flora lacks
various of the more widespread cactus epiphytes characteristic of forests further east and
even includes an endemic subspecies (Rhipsalis floccosa subsp. oreophila).
Connecting the extremes of perhumid coastal rainforest and dry inland caatinga is the
ecotonal type known as agreste (here used to include Mata de cipó). Like much of the Mata
atlântica (Myers et al. 2000), most of this has been destroyed eliminating an important
habitat for certain cacti, but the remaining fragments in SE Bahia contain some of the
world’s largest cacti, including 20–25 metre tall Brasiliopuntia and the massive Cereus sp.
(5c) — see Chapter 5.
Caatinga (Plate 1.2, Map 3). The low forest or semi-open thorny thicket vegetation
most prevalent in NE Brazil, in which the great majority of species is drought-deciduous.
There is a pronounced dry season, which can be of unpredictable duration (2–12+
PHYTOGEOGRAPHY
months), the total annual rainfall being less than 1000 mm and sometimes less than 500
mm. Soils vary from very shallow and stony to deep and sandy, but many of the cacti are
restricted to rock outcrops of various kinds, including gneiss/granite, sandstone and
limestone (± flat expanses of rock are called ‘lajedos’). The minimum temperature
throughout the area never drops below 8°C and the maximum can reach 40°C (Nimer
1973: 49). This vegetation is normally encountered at less than 750 metres altitude,
occurring at sea level at its northern limits in coastal Ceará, but occasionally from
750–1100 metres (‘caatinga de altitude’), such as around Morro do Chapéu, Bahia, c. 50
km west of Seabra (Mun. Ibitiara, BA) and on the west slope of the Serra Geral, east of
Monte Azul (northern Minas Gerais). The southern limit of true caatinga is uncertain
(although cacti may help in defining it, see below), but in the Rio São Francisco valley
islands of taller, dry forest (deciduous or semi-deciduous mata seca) mostly on limestone
outcrops (‘mata calcária’) and associated deep soils extend far to the south of the main
caatinga area. At its southern and western limits the caatinga merges into floristically
somewhat different formations termed cerradão, which are ecotonal with the cerrado.
Andrade-Lima (1981) divides caatinga vegetation into 12 different types, only some
having a diversity of Cactaceae, while others are quite devoid of the family.
Cerrado (Map 4). This is the Brazilian savanna, which occupies a huge area in the
Central-Western and South-eastern Regions, occurring in only a few, mostly small
patches in the North-east Region where precipitation and soil conditions dictate. It is a
kind of grassland, but usually includes an open, ± evergreen layer of trees with curious
contorted trunks and a continuous cover of mixed, herbaceous and woody ground flora.
The plants show many adaptations against high insolation and fire: sclerophyllous foliage,
xylopodia, thick bark etc. Usually around 1500 mm or more of rain falls each year, but
there is a regular dry season during winter when fires sometimes sweep through. The soils
are very deep and strongly leached of nutrients, often with a very low pH. Cacti are rare
or quite absent, except in open areas on some deep sandy or gravelly substrates, or
restricted to rock outcrops protected from the fire, such as those on limestone supporting
dry forest (‘mata calcária’). In Eastern Brazil cerrado vegetation, as described here, is usually
found below 750 metres altitude. When occurring above this elevation it is generally
included with the following type, although some localities at higher altitudes of certain
specialized taxa are marked on Map 4.
Campo rupestre (Plates 1.3 & 1.4, Map 5). This is often described as a mosaic of different
vegetation types, which reflects the intimate mixture of different topographies, substrates
and microclimates, resulting from the juxtaposition of mountain ridges, grassy or marshy
valleys, plateaux with ± open vegetation (‘gerais’), bare rocks, sand and gravel deposits and
sharp diurnal fluctuations in temperature, humidity and light, all combined with elements
of the 3 preceding vegetation types (hence the terms ‘mata de neblina’, ‘caatinga de altitude’
and ‘cerrado de altitude’; also ‘carrasco’ — sometimes used for the caatinga/ campo
rupestre/cerrado ecotone). However, campo rupestre is most commonly and intricately
associated with cerrado, so that the distinction between the two is often difficult to make
and, therefore, in Area No. 4, below, the two are treated together. Rainfall is rather
PHYTOGEOGRAPHY
variable, but frequently well in excess of 1000 mm per annum and accompanied by heavy
dews and mist. Maximum temperatures are lower than in the other vegetation types
mentioned above and winter minima can descend to between 0°C and -4°C at the
highest elevations (Nimer 1973: 46). Cacti are mainly restricted to the rock outcrops and
areas of deep sands and gravels, from c. 750–1950 m, examples near the upper altitudinal
limits being the Pico das Almas, BA (Arrojadoa bahiensis) and Pico do Itambé, MG
(Cipocereus minensis), although vegetation of this nature can descend to c. 400 metres (eg.
near Andaraí, Bahia: Micranthocereus purpureus). Such vegetation is commonest along the
mountain backbone of Eastern Brazil, formed by the Serra do Espinhaço and Chapada
Diamantina (Serra da Moeda, MG, northwards to Jacobina, BA), but also occurs in
smaller pockets further to the east on more isolated ridges, such as at Rui Barbosa and
Monte Santo (BA), and further south in the Serra da Mantiqueira (Serras do Lenheiro and
Ibitipoca, MG).
3.3 ANALYSIS OF DISTRIBUTION PATTERNS

The phytogeographical Areas and their subdivisions elaborated below are largely defined
by the distribution patterns of cacti indigenous to Eastern Brazil. The lines delimiting
Areas 2–4 on Map 6 reflect the recorded native occurrences of the taxa described and
documented in Chapter 5. Gaps in the authors’ field knowledge, and where herbarium
records are also lacking, are indicated on Map 6*, so that any extrapolated lines and other
assumptions can be identified. No less than 95% of the taxa (including c. 9% thought to
be ‘single-site’ endemics, ie. those known from only a single locality or group of closely
adjacent localities) have distributions that show a strong correlation with the major
vegetation types and topographic features, falling into one of the 3 major categories
described below (Area Nos 2–4).
Leaving aside the first distribution category (No. 1, below), the other major categories
recognised, ie. Area Nos 2, 3 & 4, are as delimited on Map 6, with minor exceptions, ie.
where disjunct occurrences of one vegetation type are found inside the area of another
or as rare disjunct records of particular taxa, these being noted in the discussion under
each Area heading below. In terms of vegetation types, Area No. 2 broadly corresponds
with the Mata atlântica (Atlantic Forest) and its constituent sub-types, such as restinga (at
the coast) and mata de brejo (on higher ground well inland). Further west, Area No. 3
corresponds to the caatinga in its entirety, which also surrounds and includes the campos
rupestres/cerrados of the northern part of the East Brazilian Highlands and disjunct islands
*By state, from north to south, areas deserving of further investigation in the field are as follows. Maranhão: northernmost
(coastal dunes) and southernmost limits (towards Piauí); SW Piauí; Ceará (most of state except NW & S); northern Rio Grande
do Norte; SE Pernambuco; N & W Alagoas (especially around Mata Grande); southernmost coastal Sergipe; Bahia: NE (Raso
da Catarina), cent.–W (the entire region N & S of road BR 242, including land draining into the Rio São Francisco from its
E side, to the borders with Minas Gerais, Goiás and Piauí) & SE inselbergs (BA/MG/ES border region); Minas Gerais: NW,
NE (drained by Rios Jequitinhonha & Mucuri), E (Caratinga eastwards) & SW (towards 46°W); and Espírito Santo: NW
(inselbergs) & S (at > 600 m for Rhipsalideae).
PHYTOGEOGRAPHY
of dry forest on limestone outcrops beyond its vaguely defined southern limits. Area No.
4 comprises the western cerrados and the campos rupestres of the south-eastern part of the
East Brazilian Highlands, which are bounded by the Mata atlântica (Area No. 2) on their
eastern and southern slopes. At the borders between these adjacent Areas species assigned
to one or other may sometimes be sympatric over a small distance, eg. upon gneissic rocks
south of Padre Paraíso, Minas Gerais (Nos 2 & 3) and in the municípios of Grão Mogol
and Bocaiúva, MG (Nos 3 & 4). Apart from these, the zones of overlap, sometimes
ecotonal in character, are generally not very extensive or do not appear to be so today
due to their widespread destruction, as is the case, for example, with the agrestes, which
lie between the caatinga and Mata atlântica. However, in a few places such overlaps can be
detected, eg. in the region south and east of Vitória da Conquista, Bahia, where there are
occasional records of caatinga-agreste taxa (Brasilicereus phaeacanthus, Pereskia bahiensis, Maps
27, 17) to the east of sites of typical Mata atlântica species (Lepismium cruciforme, Rhipsalis
baccifera subsp. hileiabaiana, Maps 21, 24). At their southern limits Areas 2–4 can no longer
be distinguished geographically, differing only as ecological concepts (Map 6). This means
that species referred to No. 2 or to No. 3 may sometimes grow in very close proximity
to campo rupestre taxa from No. 4. From the above it should be noted that while each of
the Areas 2, 3 and 4 broadly corresponds to a major vegetation type, namely the Mata
atlântica, caatinga and cerrado, the campos rupestres are divided between Areas 3 & 4. This
reflects the strong divergence in cactus genera and species between the northern and
south-eastern parts of the East Brazilian Highlands, a feature already noted in relation to
studies of Eriope by Harley (1988). These points are discussed under the four Areas,
below.
As will shortly become clear, within each Area it is possible to define further patterns
or distribution categories. These are summarized below with lists of their constituent taxa
and they are also employed in short-hand form at the beginning of the statement
describing the range and ecology of each taxon in the taxonomic inventory (Chapter 5).
These patterns/subcategories are not all mutually exclusive, ie. they overlap, especially in
the areas of highest diversity, eg. in the Southern caatingas: (3a:iii:c), (3b:iii), (3d:ii).
AREA 1. This is Eastern Brazil in its entirety, ie. for taxa that are widespread and/or non-
specific, or disjunct between different vegetation/Area categories (9 taxa; 5.5% of total).
(1a). Very widespread/non-specific in Areas 2–4 (2 taxa): Hylocereus setaceus, Epiphyllum
phyllanthus.
(1b). Ranging between Area No. 2 and 3 or 4, in restinga/Mata atlântica/mata de brejo/mata de
neblina/campo rupestre/caatinga/agreste etc. (7 taxa): Brasiliopuntia brasiliensis, Rhipsalis
lindbergiana, R. baccifera subsp. hileiabaiana, R. russellii, Hatiora salicornioides,
Pilosocereus catingicola subsp. salvadorensis, Melocactus violaceus subsp. violaceus.
Under this Area heading are listed the taxa which cannot be fitted into any of the more
specific vegetation/area categories or patterns outlined below (Nos 2–4). Only one is
endemic to Eastern Brazil, and the remainder includes a few taxa very widely distributed
PHYTOGEOGRAPHY
in the neotropics as well as those ranging into other parts of Brazil but not beyond. That
most of these are epiphytes confirms the assessment made in Ibisch et al. (1996), in
relation to Peru, that epiphyte taxa are generally more wide-ranging than terrestrial taxa
(see also Area No. 2).
The two taxa in Category (1a) are wide-ranging neotropical species of broad
ecological tolerance, being strongly xerophytic epiphytes and lithophytes (Epiphyllum
phyllanthus can even be found in the cactus-poor cerrado). They belong to tribe
Hylocereeae, which phylogenetic research based on gene sequences (Wallace 1995 &
pers. comm.) has shown to be amongst the most basal branches in the Cactoideae or its
HPE clade. It is also possible that these species themselves are rather old, as is suggested
by their very extensive distributions in tropical America, which in the case of Hylocereus
setaceus could represent the combined trans-Amazonian and Andean historical migration
route hypothesized by Prado (1991).
In Category (1b) Rhipsalis baccifera subsp. hileiabaiana and Hatiora salicornioides are
taxa mostly occurring in areas of relatively high rainfall in the Mata atlântica/mata de brejo
of Area No. 2 and in campos rupestres and montane cloud forests of the East Brazilian
Highlands further west (Chapada Diamantina, Bahia, Area No. 3). Strictly speaking, H.
salicornioides occurs in Areas 2–4, but in Area No. 2 is restricted to its southern part, so
hardly qualifies for inclusion in the previous Category, (1a). The disjunctions of such
plants, restricted to habitats with very high humidity, is striking and is unlikely to be an
artefact of poor collecting or habitat destruction. In Eastern Brazil Rhipsalis russellii is
known from central-E/SE Bahia and two apparently disjunct collections from
central/southern Minas Gerais (as an epiphyte) and southern Espírito Santo. From there
it ranges into Paraná, Goiás and Mato Grosso, where its habitat details are unknown at
present. R. lindbergiana has most of its distribution within Eastern Brazil, ranging from
Pernambuco to the border region between eastern São Paulo and Rio de Janeiro. It is
found in the restinga/Mata atlântica on the coast, in mata de brejo on the Serra Negra
(Pernambuco), and more rarely upon rocks or in semi-humid forests within or
surrounding the East Brazilian Highlands, as well as in patches of more humid forest
within the caatinga area east of the Chapada Diamantina (Bahia) and in mata seca west
of the Serra do Espinhaço, Minas Gerais.
Melocactus violaceus subsp. violaceus (Map 46) is commonly found at the coast (restinga)
and up to 35 km inland, in open or low-shrubby vegetation on sands, at c. 0–150 metres
altitude, from Rio de Janeiro to Rio Grande do Norte. However, it has also been
collected in north-eastern Minas Gerais, in a very sandy phase of ‘cerrado de altitude’ at
1100 metres elevation (Taylor 1991a: 56). This is an example of the links between the
montane and coastal floras of Eastern Brazil, such as have been reported for other plant
groups (Giulietti & Pirani 1988: 47, 53, 60; Harley 1988: 100).
Brasiliopuntia (Map 18) is a most unusual cactus, inhabiting semi-humid, high forest
where it forms a tree with a well-developed trunk. It ranges from dry restinga forest at the
coast into the agrestes and caatinga (especially along seasonal watercourses), from north-
western Paraíba to NE Minas Gerais, and also occurs in brejo forest (central Pernambuco,
western Alagoas and Bahia), reappearing in north-western Bahia (at Formosa do Rio Preto)
PHYTOGEOGRAPHY
and then on limestone outcrops amongst mata seca in central Minas Gerais (Area No. 3).
Pilosocereus catingicola subsp. salvadorensis appears to be the only cactus that is
characteristic of both restinga and true caatinga (Map 35). While its range is primarily littoral,
from Salvador (Bahia) northwards along the coast erratically to Rio Grande do Norte (and
perhaps Ceará), it is also a significant component in caatinga and similar vegetation from
the southern foot of the Serra Negra (PE) and the Raso da Catarina (BA) eastwards,
through northern Sergipe and southern Alagoas. The clue to this unusual distribution
pattern is the corridor provided by the rather dry climate characteristic of the lower parts
of the Rio São Francisco’s drainage, which links the inland caatinga with the coastal restinga.
AREA 2. Humid/subhumid, evergreen/semi-evergreen forest, including Mata atlântica,

restinga, mata de neblina, mata de brejo, capão de mato, agreste in part etc. (Map 6, east of line
C–D; summarized distribution in Map 3) (42 taxa; 26% of total).
(2a) Widespread taxa and those from SE Brazil or beyond with disjunct occurrences in brejo forests
of NE Brazil (9 taxa): Pereskia aculeata, P. grandifolia subsp. grandifolia, Lepismium
cruciforme, Rhipsalis crispata, R. floccosa subsp. floccosa, R. paradoxa subsp.
septentrionalis, R. cereuscula, Hatiora cylindrica, Cereus fernambucensis subsp.
fernambucensis.
(2b) Ranging mainly to the north of the core area (2 taxa): Rhipsalis baccifera subsp.
baccifera, Cereus insularis.
(2c) Restingas etc. between Alagoas and Salvador, Bahia: Melocactus violaceus subsp.
margaritaceus.
(2d) Southern subhumid and perhumid forest (SE Bahia southwards) (30 taxa):
(i) Taxa characteristic of less humid, seasonally dry habitats, especially gneiss/granite
inselbergs below 900 metres altitude (8 taxa): Pereskia grandifolia subsp. violacea, Cereus
fernambucensis subsp. sericifer, Pilosocereus brasiliensis subsp. ruschianus,
Coleocephalocereus subg. Coleocephalocereus (C. fluminensis [2 subspecies], C.
buxbaumianus [2 subspp.], C. pluricostatus).
(ii) Taxa from more humid habitats (22 taxa): Opuntia monacantha, Lepismium
houlletianum, L. warmingianum, Rhipsalis elliptica, R. oblonga, R. floccosa subsp.
pulvinigera, R. pacheco-leonis subsp. catenulata, R. cereoides, R. sulcata, R. teres, R.
pulchra, R. burchellii, R. juengeri, R. clavata, R. pilocarpa, R. hoelleri,
Schlumbergera kautskyi, S. microsphaerica, S. opuntioides, Cereus hildmannianus,
Pilosocereus arrabidae, P. brasiliensis subsp. brasiliensis.
As might be expected, the high proportion of epiphytic taxa in this category means that
there are fewer endemics in this part of the core area of Eastern Brazil in comparison with
Areas 3 & 4 (caatingas/campos rupestres). Nevertheless, in Eastern Brazil this vegetation type
is characterised by some distinctive and widespread taxa (eg. the endemic Rhipsalis
paradoxa subsp. septentrionalis, Map 23) and there are some interesting disjunctions of non-
endemic taxa. In its northern parts, north of the Rio Paraguaçu (BA), the Mata atlântica
PHYTOGEOGRAPHY
becomes a generally rather narrow coastal band, of which the more humid phases
sometimes reappear further inland on higher ground as isolated stands of mata de brejo.
These are located at or near the limit with Area No. 3 (caatingas), or actually within it,
west of the line on Map 6, but are assigned to Area No. 2.
It should also be re-emphasized that Area No. 2 includes an extensive zone receiving
less than 1000–1250 mm of rain/year (see map ‘isoietas anuais 1914–1938’ in Azevedo
1972; Nimer 1973: 40, fig. 18), located between eastern Minas Gerais, central & western
Espírito Santo (Rio Doce drainage) and northern Rio de Janeiro, at less than 900 metres
altitude. Although this south-east region includes all 5 representatives of Coleocephalocereus
subg. Coleocephalocereus, whose other two subgenera are restricted to the caatingas (Area
No. 3), its referal to the Mata atlântica is supported, in terms of Cactaceae at least, by the
widespread presence of Pereskia aculeata (Map 2), a species characteristic of this vegetation
and its ecotone with the caatinga (the agreste), but one rarely found in the caatinga proper.
This SE area also lacks any of the taxa typical of Area No. 3, such as Tacinga,
Pseudoacanthocereus, Cereus jamacaru etc., and instead is characterized by vicariant
subspecies of widespread Mata atlântica species, such as Pereskia grandifolia subsp. violacea
and Cereus fernambucensis subsp. sericifer. However, its lower rainfall appears to interrupt
the ranges of various species characteristic of more humid phases of the Mata atlântica (eg.
Lepismium cruciforme, Rhipsalis floccosa, R. oblonga, R. russellii) and separates closely related
taxa such as R. baccifera subsp. hileiabaiana and R. teres (Map 24). An area of high diversity
and endemism of Rhipsalideae is thus restricted to much wetter habitats in southern
Espírito Santo (ie. Serra do Caparaó, Domingos Martins, Santa Teresa), which is
effectively isolated from the stronghold of this group further southwest, in the mountains
of Rio de Janeiro etc. Such habitat disjunction can explain the morphological divergence
of the northern and southern taxa in the widespread species, Rhipsalis paradoxa, the
southernmost record for subsp. septentrionalis (at Domingos Martins, ES) being separated
by some 250 km from the nearest of subsp. paradoxa (Macaé de Cima, Rio de Janeiro).
Lepismium cruciforme, Rhipsalis crispata and R. cereuscula (Category 2a) are Rhipsalideae
whose range in NE Brazil is mainly restricted to montane brejo forest and agreste (Maps
21, 22 & 25), where all three reach their north-eastern limits of distribution in
Pernambuco. They are mostly of more frequent occurrence in SE Brazil and two range
much further in a south-westerly direction, R. cereuscula attaining Argentina and central
Bolivia (La Paz). Of these, R. crispata is an interesting example of disjunction between
subhumid restinga of eastern Rio de Janeiro (from near the state capital to Cabo Frio),
forests of the planalto of inner São Paulo and brejo forest rising out of the caatinga-agreste in
eastern Bahia and Pernambuco (Caruaru). It is also recorded from Santa Catarina in
Southern Brazil.
As noted already, in terms of Cactaceae, Rhipsalis baccifera subsp. baccifera (Category 2b)
potentially marks the limits of Amazonian floristic influence, or past expansions of its flora
(cf. Oliveira-Filho & Ratter 1995: 144). Significantly, it is the only member of the large
tribe Rhipsalideae recorded from the northern third of Eastern Brazil. Cereus insularis is
an endemic of the Atlantic archipelago of Fernando de Noronha and seems taxonomically
close to C. fernambucensis subsp. fernambucensis from (2a).
PHYTOGEOGRAPHY
Melocactus violaceus subsp. margaritaceus, Category (2c), is the cactus representative

amongst other endemics known from the extensive coastal sand-dunes that begin north
of Salvador, Bahia (cf. Harley 1988: 100–101, Taylor 1991a). It also occurs in restinga-like
vegetation at up to 400 metres elevation on the coastal slope of the Serra da Itabaiana,
Sergipe (Map 46).
In Category (2d:i) the 5 taxa of Coleocephalocereus subg. Coleocephalocereus and Cereus
fernambucensis subsp. sericifer have extensive distributions in the drainage basins of the Rios
Mucuri (MG/BA), Doce (MG/ES) and Paraíba do Sul (MG/RJ), with Co. buxbaumianus
subsp. flavisetus reaching westwards to the watershed between the Rio Grande and Rio São
Francisco (MG): Maps 29 & 41. These six, plus the more northerly-ranging Pilosocereus
brasiliensis subsp. ruschianus (Map 36), are characteristic inhabitants of this region’s abundant,
smooth, gneiss/granite inselbergs, which do not retain much water or permit the
accumulation of soil and are otherwise home mainly to certain bromeliads and specialized
bulbs, besides annual herbs and more wide-ranging cacti (eg. Pereskia aculeata, Hylocereus
setaceus). Pereskia grandifolia subsp. violacea is an endemic plant with a similar distribution area,
but not always associated with inselbergs and occasionally recorded above 900 metres
altitude at the western limits of the Area (Map 17). It may be specifically distinct and
represents the basal element of the P. GRANDIFOLIA Group within Eastern Brazil (Wallace
1995), from which the caatinga taxa belonging to this Group have been derived.
Of those included in Category (2d:ii), 10 out of the total of 18 Rhipsalideae are taxa
with the major part of their range outside of Eastern Brazil. Only 3 in this category are
endemic and appear to have very restricted ranges, in southern Espírito Santo. This tribe
has a considerable number of species endemic to the region comprising E São Paulo, SE
Minas Gerais and Rio de Janeiro, where its centre of diversity is located. The non-
endemic Rhipsalis oblonga (Map 22) appears to be restricted to the ‘Hiléia Baiana’ or
Bahian phase of the Atlantic forest, but this could reflect lack of collecting, eg. in
Espírito Santo (where there is one doubtful record from Luetzelburg). Its occurrence
seems to be linked to regions of very high rainfall (ie. 1750 mm/year or more). It
reappears to the west of the city of Rio de Janeiro, reaching coastal São Paulo (Ilha São
Sebastião) and represents an ecological vicariant of R. crispata, see (2a). Pilosocereus
arrabidae and P. brasiliensis subsp. brasiliensis (Maps 35 & 36) are restinga taxa almost
endemic to the area, reaching southwards to the environs of Rio de Janeiro. Opuntia
monacantha, a widely distributed species from SE South America and at its northern limit
in Eastern Brazil, has a mainly littoral range in SE & S Brazil (from S Espírito Santo
south-westwards), but has also been recorded from dunes near Diamantina and Pedra
Menina (Mun. Rio Vermelho), central-eastern Minas Gerais, at c. 1000 metres altitude
(Map 18). Its occurrence in the latter is comparable with records from inland sites in São
Paulo state (eg. near Piracicaba, and historically at Lorena and, even, Congonhas —
nowadays engulfed by the city of São Paulo).
The most diverse tribe of Cactaceae in this major area are the epiphytic Rhipsalideae,
but all of these appear to have a southern or south-western origin. Thus, the Cactaceae
of the Mata atlântica can be said to show little direct floristic affinity with those of the
Amazonian rainforest.
PHYTOGEOGRAPHY
AREA 3. Caatingas & Northern campos rupestres (+ included cerrados)/caatinga-agreste/mata

seca (Map 6, north of line A–D; summarized distribution of caatinga elements in Map 3)
(83 taxa; c. 51% of total).
(3a) Widespread from east to west etc. (23 taxa):
(i) Widespread east to west and north to south (5 taxa): Tacinga inamoena, Cereus
albicaulis, C. jamacaru subsp. jamacaru, Pilosocereus gounellei subsp. gounellei,
Melocactus zehntneri.
(ii) Northern caatingas (S Piauí, N Bahia, Alagoas & Pernambuco northwards) (5 taxa):
Pilosocereus flavipulvinatus, P. pachycladus subsp. pernambucoensis, P. piauhyensis,
P. chrysostele, Discocactus bahiensis.
(iii) Central-Southern caatingas, from S Piauí, S Ceará, S Paraíba southwards (13 taxa):
(a) Wide-ranging in the central-southern caatingas (7 taxa): Tacinga funalis,
Arrojadoa penicillata, A. rhodantha, Pilosocereus tuberculatus, P. pachycladus subsp.
pachycladus, Harrisia adscendens, Leocereus bahiensis.
(b) Bahian caatingas, including and surrounding the Chapada Diamantina (2 taxa):
Pereskia bahiensis, Stephanocereus leucostele.
(c) Southern caatingas (Cent./S Bahia & N Minas Gerais) (4 taxa): Pereskia
aureiflora, Brasilicereus phaeacanthus, Coleocephalocereus goebelianus, Melocactus
ernestii subsp. longicarpus.
(3b) Caatingas (and mata seca on southern limestone outcrops) in the middle and upper drainage of
the Rio São Francisco (cent. Minas Gerais to W & cent.-N Bahia/SW Pernambuco) and adjacent
SE Piauí (20 taxa):
(i) Widespread in the caatingas of the middle part of the Rio São Francisco drainage (2 taxa):
Pilosocereus gounellei subsp. zehntneri, Facheiroa squamosa.
(ii) Caatingas of central-northern Bahia: Melocactus azureus, M. ferreophilus, M.
pachyacanthus (2 subspp.), Facheiroa ulei, Discocactus zehntneri subsp. zehntneri.
(iii) Southern caatingas and other seasonally dry forests (on islands of limestone) of the Rio São
Francisco/Rio das Velhas drainage (western & central-southern Bahia to central Minas Gerais) (12
taxa): Pereskia stenantha, Quiabentia zehntneri, Tacinga saxatilis (2 subspp.), Cereus
jamacaru subsp. calcirupicola, Pilosocereus floccosus subsp. floccosus, P.
densiareolatus, Micranthocereus dolichospermaticus, Melocactus deinacanthus, M.
levitestatus, Facheiroa cephaliomelana (2 subspp.).
(3c) Northern campos rupestres etc., East Brazilian Highlands (Chapada Diamantina and northern
part of Serra do Espinhaço, BA/MG) (17 taxa):
(i) Chapada Diamantina & Northern Serra do Espinhaço (BA/MG) (3 taxa): Rhipsalis
floccosa subsp. oreophila, Arrojadoa dinae subsp. eriocaulis, Melocactus paucispinus.
(ii) Chapada Diamantina (BA) (9 taxa): Stephanocereus luetzelburgii, Arrojadoa
bahiensis, Pilosocereus glaucochrous, Micranthocereus purpureus, M. flaviflorus, M.
streckeri, Melocactus oreas subsp. cremnophilus, M. glaucescens, Discocactus
zehntneri subsp. boomianus.
PHYTOGEOGRAPHY
(iii) Northern Serra do Espinhaço (Bocaiúva MG – Caetité BA) (5 taxa): Arrojadoa dinae
subsp. dinae, Cipocereus pusilliflorus, Micranthocereus violaciflorus, M. albicephalus,
M. polyanthus.
(3d) Eastern caatingas-agrestes/campos rupestres (Minas Gerais & Bahia, from the Serra do
Espinhaço & Chapada Diamantina, eastwards & north-eastwards to Pernambuco and sometimes
beyond) (23 taxa):
(i:a) Widespread eastern taxa and those restricted to E/NE Bahia and Paraíba/Pernambuco
(10 taxa): Tacinga palmadora, Pseudoacanthocereus brasiliensis, Pilosocereus
catingicola subsp. catingicola, P. pentaedrophorus subsp. pentaedrophorus, Melocactus
ernestii subsp. ernestii, M. bahiensis subsp. bahiensis, M. lanssensianus, M.
concinnus, M. violaceus subsp. ritteri, Espostoopsis dybowskii.
(i:b) Restricted to the E & SE caatingas-agrestes/and associated campos rupestres from the Rio
Paraguaçu drainage (BA) southwards (6 taxa): Tacinga werneri, Pilosocereus
pentaedrophorus subsp. robustus, Melocactus oreas subsp. oreas, M. bahiensis subsp.
amethystinus, M. conoideus, M. salvadorensis.
(ii) SE caatingas-agrestes of NE Minas Gerais (Rio Jequitinhonha drainage and watersheds
with Rio Pardo and Rio Mucuri) (7 taxa): Tacinga braunii, Pilosocereus azulensis, P.
floccosus subsp. quadricostatus, P. magnificus, P. multicostatus, Coleocephalocereus
subg. Buiningia (C. aureus & C. purpureus).
This major area category includes all that considered here as part of the caatingas
dominion in Eastern Brazil (cf. Andrade-Lima 1981), plus the following:– those areas
referred to caatinga to the west of the Rio São Francisco (Andrade-Lima 1975); the areas
of campo rupestre and cerrado from the northern part of the East Brazilian Highlands (ie.
the northern parts of the Serra do Espinhaço and Chapada Diamantina, and their
extensions eastwards); ecotones at the southern limits of the caatinga and mata seca in
northern Minas Gerais; and part of the agreste or ecotonal vegetation between the caatinga
and Mata atlântica in the east. (Much of the agreste has now been so severely altered or
destroyed that its cactus flora is mainly to be inferred from a few herbarium records and
scrappy extant remnants of what was once an extensive vegetation zone.) In addition, to
the west and south of the caatinga proper, in western Bahia, south-western Piauí and
central Minas Gerais (Rio São Francisco drainage), a few taxa characteristic of this area
occasionally appear in disjunct dry habitats, eg. Melocactus zehntneri, Leocereus bahiensis,
Pilosocereus flavipulvinatus, P. floccosus and Cereus jamacaru subsp. calcirupicola. These are
mainly rock outcrops located inside Area No. 4, but represent disjunct extensions from
Area No. 3. The same applies to occurrences of Cereus jamacaru and Pilosocereus
pentaedrophorus in the Mata atlântica to the east (Area No. 2). A cactus-based
approximation to the caatingas area is offered in Map 3.
Area No. 3, as delimited here, makes good sense on the basis of endemism of
Cactaceae, as the following list hopefully makes clear. At generic level the endemics are:
Facheiroa (3 spp.), Stephanocereus (2 spp.), Espostoopsis (1 sp.) and Leocereus (1 sp.), with
Arrojadoa (4–5 spp.) almost endemic. There are also significant infrageneric taxa in
Pilosocereus, Micranthocereus, Coleocephalocereus and Melocactus. Facheiroa and Espostoopsis are
PHYTOGEOGRAPHY
typical caatinga elements. However, the diverse species of Arrojadoa and Stephanocereus are
equally divided between the caatingas and Northern campos rupestres and this pair of genera
is believed to have common ancestry with Cipocereus, a more southerly ranging, endemic
genus from Minas Gerais, which is associated with campo rupestre and appears to be relictual
(Taylor & Zappi 1996). This suggests an autochthonous E Brazilian source for some of the
most characteristic elements of the caatinga’s cactus flora, such as A. penicillata, A. rhodantha
and S. leucostele, and this may also be assumed to be the case with the majority of species
in the largest caatinga genera, Pilosocereus and Melocactus (Zappi 1994, Taylor 1991a).
Micranthocereus subg. Austrocephalocereus is a Northern campo rupestre endemic.
The southernmost records of cactus species that can be considered as typical elements
of the Caatinga/Northern campo rupestre flora are, west to east:– Pilosocereus pachycladus
(near Francisco Dumont, MG), Melocactus concinnus (Peixe Cru, Mun. Turmalina, MG —
Map 46), M. bahiensis (Mercês, Diamantina, MG — Map 43) and Brasilicereus phaeacanthus
(near Padre Paraíso, MG — Map 27). However, this is not intended to suggest that any
of these species are actually found in caatinga vegetation at these southern sites. The
southern limit of caatinga vegetation proper is uncertain and depends on one’s definition,
but, in the Rio São Francisco valley, deciduous thorn forest including cacti, such as
Pereskia stenantha (Map 17), Pilosocereus pachycladus and Arrojadoa rhodantha (Map 33),
extends southwards at least as far as the municípios of Varzelândia, Janaúba and
Porteirinha, MG (c. 15°40'S). South of there cacti interpreted as belonging to the caatingas
are found only on edaphically dry, exposed rock outcrops, mostly of Precambrian Bambuí
limestone and gneiss/granite, and are represented by taxa from categories (3a) & (3b). The
southernmost occurrences on such outcrops are those of Tacinga saxatilis, Melocactus
levitestatus, Cereus jamacaru subsp. calcirupicola and Pilosocereus floccosus subsp. floccosus, the
latter two (Maps 29 & 36) inhabiting mata seca as far south as the region of Lagoa Santa
(c. 19°40'S). Arrojadoa dinae, from the Northern campo rupestre flora, has its southern limit
at c. 17°30'S. Further east, crossing the Serra do Espinhaço, the south-easternmost
caatingas are isolated in the middle section of the Rio Jequitinhonha valley between Mun.
Jacinto (16°10'S) and Mun. Araçuaí (16°50'S), as determined by the ranges of species
representative of Categories (3a) and (3d:ii). The Cactaceae characteristic of campo rupestre
vegetation and endemic to Area No. 3 are about 45% fewer than those endemic to the
South-eastern campos rupestres (Area No. 4c below).
In Category (3a) all the taxa listed are primarily caatinga elements. They represent less
than a third (28%) of the taxa included in Area 3, the remainder falling into three parallel
geographical and/or ecological categories: (3b), (3c), (3d).
The widespread taxa included in (3a:i) are amongst the cacti that are most
characteristic of the caatinga and agreste of North-eastern Brazil, with Tacinga inamoena also
ascending into the campos rupestres. Cereus jamacaru ranges further than any of the others
included here, reaching western Maranhão, entering drier phases of the Atlantic forest in
NE Brazil and penetrating the area of the SE campos rupestres as far as the region of
Diamantina, MG. The distribution of the endemic and much less frequent, C. albicaulis,
is linked to sandy habitats. It occurs mainly in caatinga associated with the Serra do
Espinhaço and Chapada Diamantina in the southern half of its range, but expands
PHYTOGEOGRAPHY
northwards into SE Piauí, W & NE Bahia, S Pernambuco etc., finally reaching the Serra
da Ibiapaba (N Piauí & NW Ceará). Its sister species is C. mirabella, with a complementary
range west and south (Map 28), and is one of the few cacti restricted to the cerrado, being
a rarity limited to eastern parts of this biome.
The five taxa characteristic of the Northern caatingas (Category 3a:ii) are amongst a
small minority with a significant part of their ranges north of the core area of Eastern
Brazil (ie. north of 7°S). This area corresponds to some degree with the eco-region
[‘ecorregião’] termed as ‘Depressão Sertaneja Setentrional’ in Velloso et al. (2002). The
most characteristic species of this area, Pilosocereus chrysostele and P. piauhyensis, are
relatives of the southern endemic, P. multicostatus, see (3d:ii) below. P. flavipulvinatus,
which reaches beyond the limit of the caatingas as far as Carolina (W Maranhão), is the
sister taxon of the P. pentaedrophorus/P. glaucochrous species pair from (3c:ii) & (3d:i). A
legume whose range corresponds to Category (3a:ii) is Caesalpinia gardneriana Benth.
(Lewis 1998: 129).
The Central-Southern caatingas Category, (3a:iii:a), corresponds with the eco-region
termed as ‘Depressão Sertaneja Meridional’ (Velloso et al., l.c.). Excepting Leocereus
bahiensis and Pilosocereus pachycladus subsp. pachycladus, which have more western
distributions beginning in the Chapada Diamantina (Maps 47 & 37), included taxa are
restricted to the caatinga proper and, together with those listed for Category (3a:i/ii), are
the most important cactus species of the caatingas generally. Although falling reasonably
within the widespread category, in detail the distribution of Leocereus is unlike that of any
other and is hard to characterize in terms of its prefered vegetation type, being a plant of
ecotones between mostly higher altitude caatinga, cerrado and the margins of campo rupestre,
but is generally associated with arenitic rocks and sands. Pilosocereus tuberculatus is perhaps
an example of an erratically distributed relict species (see Chapter 5), whose range (Map
34) fits neatly inside Köppen’s semi-arid climatic zone (Cavalcanti Bernardes 1951). Like
Cereus albicaulis, it is usually found on dunes or light sandy substrates (referred to the
‘Cipó’ soil series) and both species appear to be characteristic of a distinct type of caatinga
identified as ‘No. 5’ by Andrade-Lima (1981: 159). A legume with a distribution pattern
not dissimilar to that of Pilosocereus tuberculatus is Caesalpinia microphylla G. Don, while the
more widespread C. laxiflora Tul. matches that of Tacinga funalis (Lewis 1998: 140, 155).
Both species categorized as (3a:iii:b) are widespread and typical elements of the Bahian
caatingas located within and around all sides of the East Brazilian Highlands, including the
Chapada de Maracás. On present knowledge, they do not range southwards into Minas
Gerais or, apparently, northwards into Pernambuco or Piauí. They also avoid the drier
NW and NE parts of Bahia.
The four species from the Southern caatingas, (3a:iii:c), are typical elements on, or
associated with, gneiss/granite outcrops (inselbergs) or derived substrates in the
southernmost caatingas, outside of the semi-arid climatic zone as defined by Köppen
(Cavalcanti Bernardes 1951). Coleocephalocereus goebelianus and Melocactus ernestii subsp.
longicarpus are absent from NE Minas Gerais (east of the Serra do Espinhaço), where they
are replaced by sister taxa, see (3d:ii/iii). Brasilicereus phaeacanthus has its much rarer sister
species in the South-eastern campos rupestres (4c:ii:a).
PHYTOGEOGRAPHY
All bar one of the taxa listed for the Rio São Francisco drainage, (3b), are endemic to
this region (11 restricted to Bahia), giving it a very characteristic cactus flora and
suggesting that it may have been a refugium for such drought-adapted plants during past
periods of greater humidity and/or cooler conditions. It should be noted that all 3 species
of Facheiroa are endemic here, with one in each of the subdivisions (3b:i–iii), F. squamosa
straying slightly outside of the Rio São Francisco drainage basin in southern Piauí and
central Bahia (Mun. Rio de Contas). Pereskia stenantha and Pilosocereus gounellei subsp.
zehntneri replace their respective allies, Pe. bahiensis and Pil. gounellei subsp. gounellei,
which have extensive ranges to the north and east, Category (3a). Both these endemics
are relatively widespread, but many of the others are restricted either to central-northern
or S Bahia (south or east of the curving course of the Rio São Francisco), or to western
Bahia and northern Minas Gerais on both sides of the São Francisco on Bambuí limestone,
eg. Pilosocereus densiareolatus and Melocactus levitestatus (cf. Andrade-Lima 1977). Most of
the endemics are found only on one of various rock types, eg. the widespread limestone
(numerous taxa, including two endemic species-groups in Melocactus) or gneiss/granite
(Melocactus deinacanthus, a taxonomically isolated species), but some, eg. Quiabentia
zehntneri and Pilosocereus gounellei subsp. zehntneri, are less specific. West of the river on
soils derived from limestone the cactus flora is rather different, since most of the
widespread caatinga species (from 3a:i/ii/iii) are lacking, the chief exceptions being Cereus
jamacaru, Pilosocereus pachycladus and Arrojadoa rhodantha, none of which is common. A
legume whose distribution conforms to (3b) is Caesalpinia pluviosa DC. var. sanfranciscana
G. P. Lewis (1998: 150).
Within (3b) 3 subcategories are recognized, (ii/iii) conforming to the limits between
Köppen’s semi-arid and hot/humid climates (Cavalcanti Bernardes 1951). The range of
both widespread taxa (3b:i) is somewhat disjunct between their northern and southern
occurrences (Maps 34 & 48), which probably in part reflects a lack of suitable habitat,
there being extensive intervening areas of cerrado and marshy sand-dunes (Tricart 1985:
209–211). Facheiroa squamosa (3b:i) is restricted to crystalline rocks or sandstones, whereas
its south-western congener, F. cephaliomelana (3b:iii), occurs only on limestone, and a
third, more poorly understood species, F. ulei, is found on non-calcareous rocks in
central-northern Bahia (3b:ii). The Melocactus taxa included in (3b:ii) represent the M.
AZUREUS Group, an endemic of this area, and are restricted to outcrops of limestone in
valleys draining northwards (Rios Verde, Jacaré & Salitre) and dissecting the Chapada
Diamantina area (cf. Category 3c), but they clearly belong to (3b), being absent from
limestone further to the east. These and other taxa included here are threatened with
extinction due to extensive modification of their habitat by agriculture and formerly by
the great Represa de Sobradinho dam-lake.
Most of the species included in (3b:iii) are rock-dwellers. The two Melocactus each
represent a monotypic species-group endemic to this area — M. deinacanthus being
restricted to a few outcrops of gneiss, M. levitestatus to Bambuí limestone, but more
widespread. The most widespread taxon is Cereus jamacaru subsp. calcirupicola, which is
found in the middle and upper drainage of the Rio São Francisco, in both the south-
western caatingas and on comparable limestone islands, in cerrado, further south, in central
PHYTOGEOGRAPHY
Minas Gerais (Map 29). It occasionally occurs on substrates other than limestone, straying
into the South-eastern campos rupestres, in the region of the Serra do Cabral. In the
southern part of its range it is often found in association with Pilosocereus floccosus subsp.
floccosus (Map 36), which occurs in proximity to campo rupestre vegetation (Zappi 1989),
but is actually a plant of mata seca.
In relation to Category (3c) it should be noted that the term Chapada Diamantina as
employed here extends this mountain area somewhat farther north than the definition
implied in Bandeira (1995), including parts of the municípios of Sento Sé, Umburanas,
Campo Formoso and Jacobina, whose highlands are ± continuous with those to the
south. All of the taxa included are endemic to Eastern Brazil and none has a close link
with an extra-Brazilian cactus flora. Represented are some morphologically rather
unusual cacti, such as Stephanocereus luetzelburgii, Arrojadoa dinae and Melocactus
paucispinus, which have evolved specialized habit forms in keeping with their
environment and/or pollinators. Despite the availability of many suitable habitats, it is
curious that the genus Discocactus is absent from all but the northernmost part of this area
(having the strongest caatinga influence), which is in stark contrast to the ecologically
similar Category (4c) below, where the genus is well-represented. The taxa included
here display the full spectrum from widespread within either one of the two subdivisions
recognized (eg. Stephanocereus luetzelburgii, Micranthocereus purpureus, Arrojadoa dinae
subsp. dinae), to those known from single localities (eg. Micranthocereus streckeri, Cipocereus
pusilliflorus), but only 3 are known from both subdivisions (3c:i) and each of these is
infrequent or disjunct. Notable species are Pilosocereus glaucochrous, restricted to dense
‘caatinga/cerrado de altitude’, Rhipsalis floccosa subsp. oreophila, found on rocks and as an
epiphyte in pockets of perhumid cloud forest (mata de neblina), and Melocactus glaucescens
and Discocactus zehntneri subsp. boomianus, plants of northern habitats associated with
sparse caatinga on sand and gravel. The remainder comprise typical campo rupestre
elements (including ‘cerrado de altitude’). Arrojadoa dinae has a recently discovered,
probable sister-species in NE Goiás (Map 32).
As can be seen from the above list for Category (3c:ii), the Chapada Diamantina has a
significant number of endemic Cactaceae, although it should be noted that most of these
are restricted to its eastern segment. They are nearly all allopatric or ecological-vicariant
sister taxa of species from the adjacent caatingas at lower elevations, such as Pilosocereus
pentaedrophorus and Melocactus oreas subsp. oreas (3d:i:a/b), or from the campos rupestres
further south (Serra do Espinhaço, S Bahia & Minas Gerais). The southern portion of this
mountain region, (3c:iii), is separated from the broader ranges to its north and south by
areas of only moderate elevation and is itself generally lower, and as a consequence
probably drier, being surrounded by caatinga on all sides. The flora exists as a complex
patchwork of small areas of campo rupestre and cerrado, and their ecotones with caatinga, the
latter ascending high up the west-facing slopes. Of the five taxa listed here Arrojadoa dinae
is the most widespread and ecologically most tolerant, being found on very sandy cerrado
as well as on more stony campo rupestre and caatinga ecotone substrates (Map 32). Cipocereus
pusilliflorus is the only member of its genus found outside the South-eastern campos rupestres
of Area category (4c) and is related to C. minensis and C. bradei from that area (Map 30).
PHYTOGEOGRAPHY
While the distribution patterns and phytogeographical affinities of campo rupestre cacti
are closely matched by those described for Eriope in Harley (1988), there is one key
difference that concerns Category (3c:iii). Harley (l.c. 83, 93), on the basis of one of his
species, E. salviifolia (Benth.) Harley (l.c. 87), would include the highlands between
Caetité (BA), Vitória da Conquista (BA) and Grão Mogol (MG) within the South-eastern
campos rupestres of Minas Gerais, whereas on many cactus distributions and even generic
affinities they are considered to be part of the Northern highlands here (eg. see Melocactus
concinnus, Map 46). Nevertheless, it is clear from both cactus and Eriope distributions that
the Chapada Diamantina (BA) and South-eastern campos rupestres (4c, below) are discrete
entities, the latter having their northern limit near Grão Mogol.
All except two of the taxa listed for Category (3d) are currently thought to be
endemic to Eastern Brazil and comparison with Category (3b) suggests that the East
Brazilian Highlands have been an important barrier isolating the cactus flora of the Rio
São Francisco drainage from that further east. Thus, east of the Chapada Diamantina we
have Pseudoacanthocereus and Espostoopsis, while the genera Facheiroa and Discocactus,
characteristic of the Rio São Francisco valley, are lacking. Some of the species in (3d)
are associated with the now much depleted part of the caatinga biome that forms the
Mata atlântica ecotone (the agreste), especially Pseudoacanthocereus brasiliensis and Pilosocereus
pentaedrophorus, while others, also found partly associated with such transitional
vegetation, are characteristic of the granitic/gneissic outcrops found in many parts of this
region, eg. Tacinga werneri and Melocactus ernestii. Four of the taxa included in (3d:i) —
Tacinga palmadora, Pilosocereus pentaedrophorus subsp. pentaedrophorus, Melocactus ernestii and
M. bahiensis subsp. bahiensis — have extensive distributions, extending northwards as far
as Pernambuco or beyond. In contrast, Melocactus violaceus subsp. ritteri is found on
isolated areas of quartz sand/gravel on the eastern slope of the Chapada Diamantina at
Jacobina and further south at Rui Barbosa (BA). M. lanssensianus is a local, cleistogamous
endemic of uncertain status from the region of Garanhuns (PE), which can be associated
with other, very similar plants from elsewhere in that state and from neighbouring
Paraíba. The distribution of the Bahian endemic, Espostoopsis, is markedly disjunct (2
areas c. 400 km apart) and is presumably indicative of its relict status (Map 48).
Pilosocereus pentaedrophorus ranges only within Köppen’s hot/humid zone, avoiding the
semi-arid climate (cf. Cavalcanti Bernardes 1951). It has recently been reported from the
coastal restinga vegetation in NE Bahia (N of Salvador) and also occurs in Mata atlântica
clothing the eastern slopes of the Chapada Diamantina (Map 36). A non-cactus with an
eastern distribution matching Category (3d) is Caesalpinia pyramidalis Tul. var. pyramidalis
(Lewis 1998: 129).
Tacinga werneri, Pilosocereus pentaedrophorus subsp. robustus and Melocactus bahiensis subsp.
amethystinus, (3d:i:b), are southern relatives of taxa in Category (3d:i:a). M. salvadorensis
replaces the more widespread M. zehntneri (3a:i) in southern and eastern Bahia, except for
a small area to the east of Brumado where they are sympatric (Map 45).
The cactus flora of the relatively small region categorized as (3d:ii) is extremely
interesting, not only because of the variety of unusual cactus endemics it has, but also for
the absence (or only marginal presence) of certain widespread caatinga-agreste Cactaceae,
PHYTOGEOGRAPHY
while other ‘indicator’ species, with which they are normally associated, are present and
common, especially those from Categories (3a) & (3d:i:a/b). In many cases these absent
or marginally present cacti are replaced by vicariant sister taxa or species from the same
infrageneric group, eg. Tacinga funalis by T. braunii, Pilosocereus pachycladus by P. magnificus,
P. catingicola by P. azulensis, and Coleocephalocereus subg. Simplex (C. goebelianus) by C.
subg. Buiningia. An instance of a more disjunct distribution pattern is that of the species
group to which Pilosocereus multicostatus belongs (Map 39), its nearest relatives being P.
piauhyensis and P. chrysostele from the distant Northern caatingas. While there can be no
doubt that parts of the middle Rio Jequitinhonha valley have typical caatinga vegetation,
the aforementioned species composition and endemism of Cactaceae also suggest that this
region has been somewhat isolated from the main caatingas area further north and west
during a substantial period in its history. However, there is a small zone of contact
between the two caatinga areas, near the borders of the Municípios of Taiobeiras and
Águas Vermelhas (see the ‘Biomas’ map for Minas Gerais in Costa et al. 1998: 21), and,
interestingly, it is here that the otherwise allopatric Pilosocereus pentaedrophorus (3d:i:a/b)
and its relative P. floccosus subsp. quadricostatus (Map 36) occur together. The Rio
Jequitinhonha valley deserves to be a key area for the attention of conservationists.
In summary, the ranges of most (60 out of 83, ie. c. 72%) of the cactus taxa from Area
No. 3 can be characterized in terms of three topographical/ecological subdivisions,
running in parallel from SSW to NNE, and broadly corresponding to the major river and
mountain systems. These are: (3b) the middle and upper parts of the Rio São Francisco
drainage, with 20 taxa, the majority caatinga elements; (3c) the campos rupestres etc. on
primarily crystalline rocks and sandstones of the East Brazilian Highlands (Chapada
Diamantina & northern Serra do Espinhaço), with 17 taxa; and (3d) the complex of
caatingas-agrestes, campos rupestres etc. for taxa ranging eastwards and north-eastwards from
within or to the east of the latter mountain system, with 23 taxa (Table 3.2).
The southern part of the Caatingas/Northern campos rupestres area is particularly rich
in species and there are instances of high levels of sympatry. For example, climbing
eastwards into the Serra Geral east of Monte Azul (MG) it is possible, including epiphytes,
to find at least 16 cactus species over a distance of less than 2 kilometres, and in southern
Bahia the caatinga may often have 10 or more sympatric cactus species. However, from
the limits of Bahia northwards the diversity of Cactaceae diminishes rapidly and it is
significant that none of the 3 caatinga species of the widespread neotropical genus Pereskia
appears to range outside of Bahia and northern Minas Gerais. There are no cacti endemic
to the extensive northern caatingas in Paraíba, Rio Grande do Norte, Ceará and northern
Piauí and only 4 species can be said to have the majority of their ranges north of 7°S
(3a:ii). This perhaps suggests that either the northern caatingas are younger than the dry
areas further south, or that they have experienced stronger forces of extinction in the past,
or that they lack the diversity of habitats and refugia created by the combination of the
Rio São Francisco valley, the East Brazilian Highlands and more humid ecotonal areas
connected with the Atlantic Forest. It is certainly curious that some rather widespread
caatinga species, in Bahia and northern Minas Gerais, with distribution patterns around or
between the East Brazilian Highlands and the agrestes, have not spread further north,
PHYTOGEOGRAPHY
Table 3.2
Geographical subdivision of the caatingas dominion based on the
distribution of Cactaceae
Caatinga area Endemic/characteristic Cactaceae
(3a:ii). Northern caatingas: MA, PI, Pilosocereus flavipulvinatus, P. pachycladus subsp. pernambucoensis,
CE, RN, PB, PE, AL, northern BA P. piauhyensis, P. chrysostele, Discocactus bahiensis
(3a:iii). Central-southern caatingas: Pereskia aureiflora, P. bahiensis, Tacinga funalis, Brasilicereus

southern PI & PB, southernmost CE, phaeacanthus, Stephanocereus leucostele, Arrojadoa penicillata,
PE, AL, BA, SE, MG A. rhodantha, Pilosocereus tuberculatus, Coleocephalocereus
goebelianus, Melocactus ernestii subsp. longicarpus, Harrisia adscendens
(3b). Caatingas of the Rio São Francisco Pereskia stenantha, Quiabentia zehntneri, Tacinga saxatilis,
drainage (MG, BA, PE) and adjacent Pilosocereus gounellei subsp. zehntneri, P. densiareolatus,
south-eastern PI Micranthocereus dolichospermaticus, Melocactus deinacanthus,
M. levitestatus, M. azureus, M. ferreophilus, M. pachyacanthus,
Discocactus zehntneri subsp. zehntneri, Facheiroa (3 spp.)
(3d:i). Eastern caatingas-agrestes: CE Tacinga werneri, T. palmadora, Pseudoacanthocereus brasiliensis,

[part] & RN southwards to central Pilosocereus pentaedrophorus, P. catingicola subsp. catingicola,
BA & north-eastern MG (from the Melocactus oreas subsp. oreas, M. ernestii subsp. ernestii,
E Brazilian Highlands eastwards) M. salvadorensis, M. concinnus, Espostoopsis dybowskii
(3d:ii). South-eastern caatingas-agrestes: Tacinga braunii, Pilosocereus azulensis, P. floccosus subsp.

endemics of north-eastern Minas Gerais quadricostatus, P. magnificus, P. multicostatus, Coleocephalocereus
subg. Buiningia (2 spp.)
examples being Pereskia bahiensis and Pseudoacanthocereus brasiliensis. Has the drier zone
along the Rio São Francisco valley (and elsewhere in NE Bahia), or the great river itself,
halted their expansion northwards?
AREA 4. Cerrados and South-eastern campos rupestres (Map 6, west of line A–B & south of
line B–C) (28 taxa; 17% of the total).
(4a) Widespread (cerrados): Cereus mirabella.
(4b) Western cerrados (including those immediately east of the Rio São Francisco in central-southern
Bahia) (3 taxa): Pilosocereus machrisii, P. aureispinus, Discocactus heptacanthus
subsp. catingicola.
(4c) South-eastern campos rupestres and associated sandy/gravelly cerrados, Minas Gerais (Map 6,
south of line B–C) (24 taxa):
(i) Widespread taxa (3 taxa): Cipocereus minensis subsp. leiocarpus, Pilosocereus
aurisetus subsp. aurisetus, Discocactus placentiformis.
PHYTOGEOGRAPHY
(ii) Northern part of area (15 taxa):

(a) Municípios of Grão Mogol, Botumirim & Cristália (Rio Jequitinhonha drainage) (5
taxa): Pilosocereus fulvilanatus subsp. fulvilanatus, Discocactus pseudoinsignis, D.
horstii, Micranthocereus auriazureus, Brasilicereus markgrafii.
(b) Serra do Cabral and western slopes of Serra do Espinhaço (Rio São Francisco drainage)
(5 taxa): Cipocereus bradei, Pilosocereus fulvilanatus subsp. rosae, P. aurisetus subsp.
aurilanatus, Arthrocereus rondonianus, Uebelmannia pectinifera subsp. horrida.
(c) Município Diamantina & E to the Serra Negra / Serra do Ambrósio (5 taxa):
Cipocereus crassisepalus, Uebelmannia (3 spp. + 1 heterotypic subsp.).
(iii) Southern part: Município Diamantina south- and south-westwards (6 taxa): Cipocereus
laniflorus, C. minensis subsp. minensis, Arthrocereus glaziovii, A. melanurus (3
subspp.).
As in the case of the Caatingas and Northern campos rupestres (Area No. 3), this region is
well-defined in terms of endemic cactus taxa, including the genus Uebelmannia (3 spp. &
2 heterotypic subspp.), an ecologically highly specialized group with no close relatives and
apparently relictual. Also notable is Cipocereus, a peculiar member of tribe Cereeae, which
is represented by 4 endemic species plus one heterotypic subspecies. A characteristic non-
endemic genus found only in Area No. 4 within Eastern Brazil is Arthrocereus (3 spp. + 2
subspp.). Excluding taxa from the widespread Area Category No. 1, the genera of cacti
that occur in common with Area No. 3 are all represented by different species, although
in some cases the southern species are actually sisters or probable sister taxa of those from
further north or east, eg. Brasilicereus phaeacanthus (3a:iii:c) & B. markgrafii (4c:ii:a) and
Cereus albicaulis (3a:iii:a) & C. mirabella (4a). Table 3.3 summarizes the principal
differences between the Northern and South-eastern campos rupestres in terms of
Cactaceae (excluding a few very narrow endemics).
Table 3.3
Cactaceae characteristic of the Northern and South-eastern
campo rupestre vegetation
(† = taxon also found in other included vegetation types — especially caatinga — of Area No. 3).
Campo rupestre areas within E Brazil Characteristic taxa of Cactaceae
Area 3. Northern campos rupestres (and Stephanocereus subg. Lagenopsis, Arrojadoa p.p. (2 spp.),
included cerrados): Bahia & northern Pilosocereus glaucochrous†, P. pachycladus subsp.
Minas Gerais pachycladus†, Melocactus paucispinus, M. oreas subsp.
cremnophilus, M. violaceus subsp. ritteri, Micranthocereus
subg. Austrocephalocereus (3 spp.), M. polyanthus,
M. flaviflorus, Discocactus zehntneri subsp. boomianus
Area 4. South-eastern campos rupestres Uebelmannia, Cipocereus p.p. (4 spp.), Arthrocereus p.p.
(and included cerrados): Minas Gerais (3 spp.), Brasilicereus markgrafii, Pilosocereus aurisetus,
P. fulvilanatus, Micranthocereus auriazureus, Discocactus
placentiformis, D. pseudoinsignis
PHYTOGEOGRAPHY
The geography of cactus taxa within this southern area is very far from being random
and a series of well-defined patterns of distribution and endemism can be recognized.
Two exceptions to the line delimiting this Area on Map 6 should be noted, both in (4b).
The first is provided by a narrow zone of cerrado located on the east side of the Rio São
Francisco in central-southern Bahia, where Pilosocereus aureispinus is endemic (its relatives
in the P. AURISETUS Group are all Cerrado/South-eastern campo rupestre taxa). The second
is a similarly located disjunct record of Discocactus heptacanthus subsp. catingicola.
The only widespread cerrado taxon (4a), Cereus mirabella, on present knowledge has a
markedly disjunct distribution, including the Rio Doce/Rio Jequitinhonha watershed
(W of Água Boa, MG — the locus classicus), western & northern Minas Gerais, western
Bahia and western Maranhão (near Carolina). It is assumed to be a relict species, the major
part of whose range appears to be within Eastern Brazil, mostly in sandy phases of the
cerrado (Map 28). C. albicaulis, its sister species, has a parallel range eastwards in caatinga.
Turning to Category (4b), Pilosocereus aureispinus, from east of the Rio São Francisco,
is the sister species of P. vilaboensis from Goiás (Zappi 1994: 126–129), while the other
two species are wide-ranging in Central-western Brazil and eastern Paraguay. The
variable P. machrisii, which is found only on rock outcrops, is the sister species of P.
aurisetus, Category (4c:i). In spite of it having received the epithet catingicola, the so-called
subspecies of Discocactus heptacanthus is usually found in the cerrados and ecotonal areas.
Nearly all of the taxa listed for Category (4c) are endemic to the core area of Eastern
Brazil and none of the species restricted to this type of habitat is common to that of the
comparable (3c), Chapada Diamantina and N Serra do Espinhaço, where there are only
17 taxa, although some genera are shared. The absence of various non-endemic genera
and species groups characteristic of adjacent areas, including Tacinga, Harrisia, Leocereus,
Coleocephalocereus and Facheiroa, is particularly worthy of note. From the conservation
standpoint it is also important to point out that (4c) has a significant number of ‘single-
site’ endemics (6 out of its 24 taxa).
Of those taxa that are categorized as widely distributed in the South-eastern campos
rupestres (4c:i), Discocactus placentiformis and Pilosocereus aurisetus subsp. aurisetus have somewhat
differentiated regional forms in each of the included subdivisions (4c:ii:a–c/4c:iii). These
taxa have been formally recognised by Braun & Esteves Pereira (see Chapter 5), but are not
as clearly distinct as other regional variants given such recognition in this treatment.
Subdivision (4c:ii:a) is at the border with Area No. 3, and the above taxa can be found
sympatric with, or in close proximity to Tacinga inamoena, Arrojadoa dinae, Melocactus
bahiensis and M. concinnus, which are characteristic members of the Caatingas/Northern
campos rupestres flora. Pilosocereus fulvilanatus, whose northern relatives are all caatinga taxa,
is the sister species of P. ulei, from the dry coastal forest at Cabo Frio, eastern Rio de
Janeiro (Zappi 1994). Brasilicereus markgrafii is considered to be a plesiomorphic relict. All
the taxa listed for (4c:ii:b) have sister taxa elsewhere in (4c). Cipocereus crassisepalus and the
species from Uebelmannia subg. Uebelmannia endemic to (4c:ii:c) are restricted to the
abundant deposits of quartz sands and gravels found eastwards from Diamantina. In
relation to the taxa listed for (4c:ii:c) & (4c:iii) it is worth noting the statement by
Giulietti & Pirani (1988: 65), that Hensold regards the most primitive forms of
PHYTOGEOGRAPHY
Paepalanthus subg. Xeractis (Eriocaulaceae) as being from the southern and eastern Serra do
Espinhaço. This observation can also be applied to Cipocereus, whose taxa with most
plesiomorphies are C. laniflorus (Serra do Caraça), C. minensis subsp. minensis (Serra do
Cipó southwards) and C. crassisepalus (E of Diamantina).
The relationships of the East Brazilian Highlands’ cactus flora (including the campos
rupestres, sensu stricto, plus ‘cerrado de altitude’ and montane cloudforest) with other
vegetation zones are quite varied, although most do not extend much beyond the core
area. They include significant links with the caatinga flora, contrary to what is stated in
Table 3.4
Principal links between the E Brazilian Highlands and other areas
(based on Cactaceae).
Highland area/substrate or vegetation Other area Linking taxon/taxa
Widespread/humid montane forest & rocks Mata atlântica/ Rhipsalis russellii, R. floccosa, sens. lat.
mata de planalto
E drainage of highlands from Itamarandiba (MG) Eastern caatingas- Pilosocereus catingicola subsp.
to Jacobina (BA)/rocks, sands & gravels agrestes catingicola, Melocactus oreas, sens. lat.,
M. ernestii subsp. ernestii,
M. bahiensis, M. concinnus
E edge of Highlands and disjunct serras: coastal restinga Melocactus violaceus, sens. lat.
Serra da Areia (MG), Rui Barbosa & Jacobina from RN to RJ
(BA), Serra da Itabaiana (SE)/quartz sand
& gravel
From Grão Mogol (MG) northwards/arenitic SW caatingas & Micranthocereus

rocks, quartz sands & gravels planalto central,
on limestone
N Chapada Diamantina, region of Morro do caatingas of Rio Discocactus zehntneri, sens. lat.
Chapéu northwards/arenitic rocks, sands & São Francisco
gravels
Chapada Diamantina & N Serra do Central-southern Arrojadoa

Espinhaço/sands & gravels caatingas
Chapada Diamantina/perhumid montane forest Mata atlântica Rhipsalis baccifera subsp. hileiabaiana,
(BA, ES, MG) Hatiora salicornioides
SE campos rupestres, between Grão Mogol and coastal rocks, Pilosocereus fulvilanatus/P. ulei
Augusto de Lima/arenitic rocks Cabo Frio, RJ species pair
SE campos rupestres, Diamantina & Serra do Cent.-western PILOSOCEREUS AURISETUS

Cabral to Serra do Cipó/rocks & gravels cerrados, on Group
diverse rocks
SE campos rupestres, Serra do Cabral (MG) Chapada dos Arthrocereus

southwards/crystalline rocks Guimarães, MT
PHYTOGEOGRAPHY
general terms by Giulietti et al. in Davis et al. (1997: 400), but cacti may represent an
exception because of their common preference for rupicolous habitats, which exist in
both ecosystems. These links are summarized in Table 3.4.
Taking together all the species of Cactaceae confined to campo rupestre and associated
vegetation (eg. ‘cerrado de altitude’, ‘mata de neblina’), ie. most taxa from categories (3c) and
(4c), plus those from (3d:i), there is a total of 42 taxa, or more than one quarter (26%) of
the cacti of Eastern Brazil, all but one of these being endemic to the core area. These
include the endemic genera Cipocereus (5 spp.) and Uebelmannia (3 spp.), and the
subgenera Micranthocereus subg. Micranthocereus (4 spp.), M. subg. Austrocephalocereus (3
spp.) and Stephanocereus subg. Lagenopsis (1 sp.). Making the same analysis for species found
mainly in the caatinga-agreste, we have a total of 58 taxa or 36%, of which 50 (c. 31%) are
endemic to the core area of Eastern Brazil. Thus, the number of endemic Cactaceae
restricted to either campo rupestre (26%) or caatinga-agreste (31%) in Eastern Brazil is not
remarkably different and accounts for c. 57% of the total. However, it should not be
forgotten that there are some important taxa that occur in both vegetation types (eg.
Tacinga inamoena, Melocactus bahiensis, Pilosocereus glaucochrous), as well as a few that have an
even wider ecological tolerance. The taxa found in neither campo rupestre nor caatinga-
agreste, are from either the more humid forests, their included rock outcrops and coastal
sand-dunes (Area No. 2), or the cerrados, and amount to c. 26% of the total, only 7% being
endemic to the core area.
To summarize, these figures indicate that, although one might have expected the great
majority of Cactaceae of Eastern Brazil to be from the extensive caatingas, in point of fact
less than half occur there, and the representation of cactus taxa in the campos rupestres
appears to be nearly as important, yet the area they occupy is relatively much smaller. Taxa
that can be regarded as the best overall markers of each vegetation type include, for the
caatinga-agrestes: Brasilicereus phaeacanthus, Pilosocereus gounellei subsp. gounellei and P.
pachycladus, sens. lat. (see Map 3, noting that the last-named occasionally strays into the
campos rupestres); for the campos rupestres: Arthrocereus, Cipocereus and Micranthocereus subg.
Micranthocereus & subg. Austrocephalocereus (Map 5); for the Atlantic Forest, sensu lato,
including restinga, riverine vegetation and eastern slopes of the Chapada Diamantina etc.:
Pereskia aculeata and P. grandifolia (Map 2); and for the cerrados (including ‘cerrado de altitude’):
Cereus mirabella, Melocactus paucispinus, Discocactus heptacanthus and D. placentiformis (Map 4).
3.4 PHYTOGEOGRAPHICAL LINKS AND

PALAEOCLIMATES
Amongst the endemic cactus genera and subgenera there are some which have obvious
sister taxa or close relatives beyond the bounds of Eastern Brazil. For example, the genus
Brasilicereus Backeb., with a pair of allopatric species (see Area Nos 3 & 4, above), is
judged to be the sister group of the widespread Praecereus Buxb. (N South America to N
Argentina; Hunt 1999a), which it replaces in Eastern Brazil (Taylor 1992a, 1997a & b):
Map 7. Praecereus ranges westwards and northwards from São Paulo and Paraná, Brazil,
through NE and NW Argentina, Paraguay, Bolivia, Peru, Ecuador, Colombia and
PHYTOGEOGRAPHY
Venezuela. This range is complemented by that of Brasilicereus in Eastern Brazil (Minas

Gerais & Bahia) and, when these are taken together, they conform closely to the extended
Pleistocene Arc distribution pattern described by Prado & Gibbs (1993) for dry seasonal
woodlands in South America. Another vicariant pair of taxa comparable with the
Praecereus/Brasilicereus case is that of Cereus subg. Mirabella (2 spp.), which replaces C.
subg. Ebneria in Eastern Brazil (the latter comprises 6–7 species, ranging through Central-
western Brazil, Bolivia, Paraguay and Argentina; Taylor 1992a): Map 8. A different and
much more disjunct example of vicariance is that of Espostoopsis and its presumed relative,
the central Andean genus, Espostoa — Map 9. Interestingly, this is matched by the almost
identical example of Hyptis sect. Leucocephala, discussed by Harley (1988: 110). The sister
relationships proposed for the three Cactaceae cases identified above await confirmation
with molecular data.
Below generic rank the high level of endemism that characterizes the cacti of Eastern
Brazil means that most have no close relatives outside of this region. However, some
notable exceptions provide clear phytogeographical links with other, more or less remote
cactus floras. The most important of these involve the Caribbean, northern South America
and Amazonia. Thus, the PILOSOCEREUS PENTAEDROPHORUS Group (8 spp., 5 in E Brazil,
3 endemic to the core area) includes one markedly disjunct species in Roraima and the
Guianas (P. oligolepis; Zappi 1994), while the MELOCACTUS VIOLACEUS Group (10 spp., 7
in E Brazil, 5 endemic to the core area) includes the geographically isolated M. smithii and
M. neryi in the Amazonas-Orinoco drainage region and M. matanzanus in Cuba (Taylor
1991a). A gene sequence phylogeny (Wallace 1997: 11) suggests that the geographically
isolated caatinga endemic, Harrisia adscendens, links members of the basal H. subg. Eriocereus
(E Bolivia, Paraguay, Argentina & Uruguay) with the Caribbean taxa of Subg. Harrisia,
indicating their path of radiation: Map 10. This would seem to have been via an eastern
route, rather than a western Andean route, as is suggested by the Praecereus/Brasilicereus
example given in the previous paragraph. Equally interesting links are those between the
caatinga-agreste of Eastern Brazil and similar dry habitats in northern South America
(Colombia & Venezuela): Pereskia aureiflora and P. guamacho (Wallace 1995: 9);
Pseudoacanthocereus brasiliensis and P. sicariguensis (Taylor et al. 1992). These vicariant species-
pairs are disjunct across the currently humid Amazonas-Orinoco region: Map 11. A
geographically similar case, but not so disjunct, involving a pair of species, or perhaps a
single widespread species with regional forms, is that of the E Brazilian Cereus jamacaru and
the Venezuelan/Amazonian C. hexagonus. Links with the cactus flora of SE Bolivia,
Argentina, Paraguay and immediately adjacent parts of Central-western and Southern
Brazil are provided by an E Brazilian group of 3 Pereskia spp. (P. grandifolia, P. bahiensis, P.
stenantha) and Quiabentia zehntneri, which have south-western counterparts in the taxa
comprising P. nemorosa/P. sacharosa and Q. verticillata (a species ranging in and around the
western part of the Chaco). Quiabentia (Map 13) is the only clear link between Chaco and
caatinga (cf. Prado 1991), while the Pereskia species-group is another which conforms to
the Pleistocene Arc distribution pattern described by Prado & Gibbs (1993): Map 12.
Likewise, the monotypic Brasiliopuntia, a wide-ranging species, distributed through Eastern
Brazil, eastern Paraguay, the Misiones nucleus (Argentina) and the eastern Andes of Peru
PHYTOGEOGRAPHY
and Bolivia, is another clear example of this Pleistocene pattern. Arthrocereus is a genus of
four species, with 3 native to Eastern Brazil (South-eastern campos rupestres) and one
disjunct in the Chapada dos Guimarães, Mato Grosso (A. spinosissimus): Map 13. Further
examples of such ‘western’ connections include Pilosocereus machrisii and Discocactus
heptacanthus, sens. lat., from eastern Paraguay and Central-western & Northern Brazil etc.,
both ranging into the western part of Eastern Brazil and having endemic sister taxa in
Minas Gerais and/or Bahia: P. aurisetus and D. placentiformis/D. bahiensis etc. Other such
links involve, for example, Rhipsalis cereuscula, which ranges from the Mata atlântica of NE
& SE Brazil via S Brazil and Paraguay to the west as far as the east Andean Yungas of
Bolivia. The widespread and almost endemic R. lindbergiana is morphologically most
similar to R. baccifera subsp. shaferi, which ranges between São Paulo state (Campinas) and
the eastern Andes. Two Lepismium species, L. cruciforme (NE to S Brazil etc.) and L.
warmingianum (SE & S Brazil etc.), have sister taxa in the eastern Andes of Bolivia (in the
Yungas) and north-western Argentina — L. incachacanum and L. lorentzianum, respectively:
Map 14. Pereskia aculeata, from dry habitats in the Mata atlântica in Eastern Brazil, has a
bimodal distribution north and south of the Amazon basin (Leuenberger 1986: 62), mainly
on the eastern side of the neotropics. Its allies are all Andean (Wallace 1995: Fig. 10).
Cactus phytogeography lends support to hypotheses on historical plant migration routes
involving dry seasonal forest formations in South America (Pennington et al. 2000).
Amongst the origins and routes discussed by Prado (1991: Fig. 8.3) for the flora of the
caatingas, the following are supported by examples from the Cactaceae of the caatinga-agrestes
and adjacent drier phases of the Mata atlântica of Eastern Brazil: (i) the Caribbean islands and
coastal regions of northern South America (‘Guajira province’), via a trans-Amazonian
route; (ii) the same region and/or dry valleys of the central Andes, via a western route
including the ‘Pleistocene Arc’; (iii) N South America or beyond, via a ‘pincers movement’
(ie. via both trans-Amazonian and Andean routes); (iv) foothills of the eastern Andes
(including the ‘Piedmont’ and W Chaco) and Misiones nucleus, via the ‘Pleistocene Arc’;
(v) the Atlantic Forests of Brazil; and (vi) the cerrados of Central-western Brazil. Some of the
clearest examples supporting routes (i)–(vi) are summarized below in Table 3.5. It should
be noted that here it is not necessarily the origin of the caatinga flora that is being elucidated,
rather that the caatinga is one of the nodes for migration routes, which may either have
brought taxa to it, or from it, to other dry areas. Indeed, molecular phylogenies for the
PERESKIA GRANDIFOLIA Group and Coleocephalocereus indicate radiations in opposite
directions along the same route — the former from the Mata atlântica (2d:i) into the
Southern caatingas (3a:iii/3b:iii), the latter in reverse (Wallace 1995, Cowan & Soffiatti
unpubl.). And, as noted above, another molecular phylogeny suggests that Harrisia both
entered the caatinga (from the eastern Andes etc.) and radiated from it (into the drier parts
of the eastern Caribbean). A further point to note is that Prado’s routes should not been
seen as mutually exclusive, since, for example, (i), (ii) and (iv) are essentially subsets of (iii).
Nevertheless, for ease of comparison his list has been followed in the table below.
In addition to his discussion of migration routes for species of dry seasonal forests in
South America, Prado (1991: 232–239) considered palaeoclimates and particularly the
Pleistocene fluctuations, when the caatingas and similar dry forests appear to have
PHYTOGEOGRAPHY
Table 3.5
Cactus evidence in support of historical migration routes to and from the
caatingas etc., as proposed by Prado (1991).
Region/migration Vicariant taxa/ranges (habitat/country codes Notes

route (Map refs) [+ Brazilian state codes]: E Brazilian
vegetation types)
(i) Caribbean & (1) Pereskia guamacho (CO,VE) & allies (Caribbean): (1) & (3) were
‘Guajira province’, P. aureiflora (S caatingas-agrestes) examples given in
via Amazônia (see (2) Pseudoacanthocereus sicariguensis (CO,VE): Prado, l.c.; (1) & (4)
Maps 10 & 11) P. brasiliensis (E caatingas-agrestes) are now confirmed by
(3) Cereus hexagonus (CO,VE,GY,GF,SR,BR molecular evidence.
[RR,PA]): C. jamacaru (caatingas-agrestes) NB. Pseudoacanthocereus
(4) Harrisia subg. Harrisia (Caribbean): H. adscendens has only 2 species
(caatingas)
(ii) Ibid., via dry cent. (1) Praecereus (CO,VE,EC,PE,BO,PY,AR,BR Relationship awaits
Andean valleys and [MS,SP,PR]): Brasilicereus (S caatingas-agrestes confirmation by
the ‘Pleistocene Arc’ & adjacent campos rupestres) molecular data
(see Map 7)
(iii) N South America (1) Hylocereus setaceus complex (MX,BE,CO,EC, See Chapter 5.5 for
and beyond, via both PE,BO,PY,AR,GY,SR,BR[RR,PA,MS,RJ, notes on this complex
(i) and (ii) above SP,PR]/caatingas & Mata atlântica)
(iv) Via the (1) Pereskia sacharosa & P. nemorosa (BO,AR,PY,BR Molecular evidence
‘Pleistocene Arc’, [MT,MS,RS]: remainder of P. GRANDIFOLIA has the western
from the floristic Group (Mata atlântica & S caatingas) Harrisia spp. and the
nodes it connects (see (2) Quiabentia verticillata (BO,PY,AR): Q. zehntneri andean pereskias
Maps 8, 10, 12 & 13) (SW caatingas) as basal or potentially
(3) Brasiliopuntia brasiliensis (PE,BO,AR,PY/caatingas- basal within their
agrestes & dry phases of Mata atlântica) respective groups
(4) Cereus subg. Ebneria (BO,AR,PY,BR[MT,MS, (Wallace 1995, 1997).
RS]): C. subg. Mirabella (caatingas & E Brazilian NB. Quiabentia has
cerrados) only 2 species
(5) Harrisia subg. Eriocereus (BO,AR,PY,UY,BR
[MS]): H. adscendens (caatingas)
(v) Brazilian Atlantic (1) Pereskia grandifolia (Mata atlântica, BR[CE,PE, Molecular evidence
Forest (see Maps 17, BA,MG,ES,RJ,SP,SC]): P. bahiensis & P. stenantha indicates Pereskia
35 & 41) (Bahian & SW caatingas) grandifolia subsp.
(2) Pilosocereus catingicola subsp. salvadorensis (restinga/ violacea and
caatinga BR[RN,PB,PE,AL,SE,BA]): P. catingicola C. goebelianus as basal
subsp. catingicola (Bahian caatingas-agrestes) taxa in each case
(3) Coleocephalocereus subg. Simplex (S caatingas [BA, (Wallace 1995, Cowan
MG]): C. subg. Coleocephalocereus (Mata atlântica) & Soffiatti unpubl.)
(vi) Central-western (1) Discocactus heptacanthus, sens. lat. (PY,BO,BR[MS, See Chapter 5.5
Brazilian cerrados MT,GO,TO,PI,BA,MG]): D. bahiensis &
(Map 50) D. zehntneri (N caatingas & adjacent N campos
rupestres)
PHYTOGEOGRAPHY
expanded (12–18,000 years BP). He presents evidence that the northern caatinga area
expanded westwards into Maranhão and thence northward, forming a corridor along a
route including the Monte Alegre/Faro area (region of Santarém, Pará) and north of the
Amazon River to the Roraima/Guyana area, potentially linking ultimately with the dry
areas in present day northern South America (‘Guajira province’). Such a route seems
plausible on the basis of the known distribution of the as yet inadequately resolved Cereus
jamacaru/C. hexagonus taxonomic complex, C. jamacaru subsp. jamacaru ranging north-
westwards at least as far as the limits of Eastern Brazil (see Map 29). Equally significant is
Pilosocereus flavipulvinatus, which also ranges to the western border of Maranhão, reaching
Carolina (see Map 36). Furthermore, Zappi (1994: 69) suggests that this species provides
a link to the related P. oligolepis, known only from the Roraima/Guyana region and the
northernmost member of the diverse and largely endemic E Brazilian P.
PENTAEDROPHORUS Species Group. Do the present ranges of P. flavipulvinatus and P.
oligolepis represent remnants of a Pleistocene expansion of the caatingas? If so, then it is
possible that the presence of Tacinga inamoena, Melocactus zehntneri and Leocereus bahiensis
on rock outcrops in the cerrados of westernmost Bahia, of Pilosocereus pentaedrophorus in
restinga north of Salvador, Bahia and of P. ulei at Cabo Frio, Rio de Janeiro (Ab’Sáber
1974, Araújo in Davis et al. 1997: 373) are likewise remnants representative of former
caatinga expansions (Maps 19, 45, 47 & 36). Similarly, the disjunct occurrences of Tacinga
funalis, Arrojadoa penicillata and Harrisia adscendens in sand-dunes west of the Rio São
Francisco (NW Bahia) could represent historical range extensions during the Pleistocene
fluctuations, when the barrier presented by the river was removed as it dried up in its
middle section (Tricart 1985: 210). These are taxa whose current range is mainly in the
eastern sector of the caatingas and are not likely to have been long-distance dispersed into
NW Bahia, to judge from their presumed dispersal strategies (see below).
3.5 REPRODUCTIVE AND DISPERSAL STRATEGIES

While distribution patterns may be heavily influenced by geographic, climatic, edaphic
and temporal factors, the range of a taxon will also be dependent, to a greater or lesser
extent, on its reproductive ability and dispersal strategies. As far as can be determined, all
of the cactus taxa native of Eastern Brazil reproduce by means of seeds, only very few also
employing vegetative means. The chief examples of the latter are in the Opuntioideae,
where the jointed stem-segments are frequently capable of being detached, transported
and then forming roots upon contact with the ground. A few of the epiphytic and
scrambling taxa in tribes Rhipsalideae and Hylocereeae may also indulge in a limited
amount of vegetative propagation, but this is unlikely to spread the plant much beyond
adjacent branches of the tree in which it originally established itself. Table 3.6, page 133,
lists taxa in systematic order noting, where possible, the observed, reported or presumed
dispersal vectors, principal habitat type(s) and a categorization of geographical range
within specified habitat type(s) in Eastern Brazil into either ‘widespread’, ‘restricted’ or
‘single site’. Chapter 5 should be consulted for data on size, morphology and dehiscence
of fruit and seed.
PHYTOGEOGRAPHY
Amongst the range of observed or probable dispersal strategies demonstrated by the cacti
of Eastern Brazil, the rarest appear to be those involving wind or water — Micranthocereus
dolichospermaticus (W. Barthlott, pers. comm.), Pilosocereus gounellei (Zappi 1994) and
Discocactus bahiensis. However, it is likely that in each case zoochory is still part of the initial
stages of dispersal, enabling the seeds to escape from the fruit and/or wool of the cephalium
or stem apex. Wind dispersal of the seeds of M. dolichospermaticus from its elevated cephalium
is a strategy matched by associated species of the bromeliaceous genus Encholirium, both taxa
needing to disperse their seeds across deeply fissured, karstic limestone outcrops. However,
it may be doubted whether this strategy alone is capable of achieving dispersal over longer
distances. Pilosocereus gounellei is very widely distributed, perhaps because its fruits are
attractive to bats and other vectors (large wasps have been observed flying away bearing
funicular pulp with seeds attached), but its buoyant seeds may be adapted for local dispersal
during flash-floods, which affect the impermeable, flat rocky substrates the species tends to
frequent. The same strategy for local dispersal may apply to the much rarer Discocactus
bahiensis, which has been observed on a low-lying river flood plain subject to occasional
inundation. It could also play a role in other species of this genus.
Bats, birds, lizards and ants are almost certainly the commonest seed vectors for E
Brazilian cacti, with non-flying mammals being of lesser significance and linked to species
with larger fruits and seeds. The most interesting amongst the latter vectors is that
suggested by the behaviour of 3 species of Pereskia (the P. GRANDIFOLIA Group),
Brasiliopuntia brasiliensis, Opuntia monacantha and Pseudoacanthocereus brasiliensis. In these,
the fruits seem to ripen only once they have fallen to the ground, turning yellow or
reddish and then smelling strongly of pineapple, just like the ripe infructescences of
ground-dwelling species of Bromelia, with which they are often associated. Here it is
hypothesized that such taxa are (or were formerly) dispersed by peccaries (‘caitité’,
‘caititu’, ‘queixada’ etc.), within whose historic range they are included. This could
explain the wide distribution achieved by Brasiliopuntia, Pseudoacanthocereus and the 5
species of the PERESKIA GRANDIFOLIA Group across South America, via the Amazon
basin, cerrado and/or dry seasonal forest environments. In the much altered caatingas of the
present day, where the peccary is either extinct or very rare, the fallen fruits of P. bahiensis
and P. stenantha are eaten by man’s cattle and good crops of pereskia seedlings can be
observed germinating amongst recently deposited cow-pats. Two climbing or epiphytic
taxa of Hylocereeae (Hylocereus & Epiphyllum), wide-ranging in the neotropics, produce
large succulent fruit that may be of interest to monkeys, which have been observed
Map 1. The Caribbean and South America, with Brazil, indicating its states, the core study area (box
delimited at 7 & 22°S and 46°W), Rio São Francisco, East Brazilian Highlands [ ] and Atlantic
Archipelago of Fernando de Noronha, the home of Cereus insularis [ ]. The overall area treated in this
study includes the states of Maranhão (MA), Piauí (PI), Ceará (CE), Rio Grande do Norte (RN), Paraíba
(PB), Pernambuco (PE), Alagoas (AL), Sergipe (SE) and Bahia (BA), which comprise the NE Brazil
region, and parts of both Minas Gerais (MG) & Rio de Janeiro (RJ) and Espírito Santo (ES) from SE
Brazil. Note that in the following maps (Nos 2–51) a symbol may sometimes represent two or more
closely situated localities, as cited in the text of Chapter 5, and the position of some symbols has been
slightly adjusted where different taxa are coincident at a given locality.
MAP 1
MAP 2
Map 2. Distribution records in Eastern Brazil of all cactus taxa restricted to Mata atlântica, sens. lat.,
distinguishing the marker species Pereskia aculeata [ ] and P. grandifolia, sens. lat. [ ]. NB. Only records
within the study area, ie. north of 22°S are displayed (this vegetation extends much further south).
MAP 3
Map 3. Distribution records of taxa typical of caatinga and caatinga-agreste vegetation, distinguishing
Brasilicereus phaeacanthus [ ], Pilosocereus gounellei subsp. gounellei [ ] and P. pachycladus, sens. lat. [ ]
(omitting a southern campo rupestre site) and indicating the approximate limits of the caatinga biome
(excluding south-western mata seca).
MAP 4
Map 4. Distribution records in Eastern Brazil of taxa restricted to, or with specialised adaptations for
cerrado (including that within campo rupestre), distinguishing the marker species Cereus mirabella [ ],
Melocactus paucispinus [ ], Discocactus heptacanthus [ ] and D. placentiformis [ ].
MAP 5
Map 5. Distribution records in Eastern Brazil of all taxa restricted to campo rupestre (excluding pure cerrado
elements), distinguishing Arthrocereus [ ], Cipocereus [ ] and Micranthocereus [ ] (subgenera Micranthocereus
& Austrocephalocereus).
MAP 6
Map 6. Eastern Brazil, showing the principal phytogeographic Area categories (2 = Mata atlântica, 3 =
Caatingas & Northern campos rupestres, 4 = Cerrados & South-eastern campos rupestres), ‘brejo’ forests [ ], campo
rupestre highlands [ ] and areas of potential interest that await thorough investigation [ ? ]. Taxa defining lines
A-B-C-D: Pereskia stenantha, Quiabentia zehntneri, Tacinga braunii, T. inamoena, Rhipsalis paradoxa subsp.
septentrionalis, Brasilicereus phaeacanthus, Cereus jamacaru subsp. calcirupicola, Cipocereus laniflorus, Pilosocereus
gounellei subsp. gounellei, P. floccosus, P. pachycladus subsp. pernambucoensis, P. aurisetus subsp. aurisetus, Melocactus
bahiensis subsp. amethystinus, M. concinnus, Harrisia adscendens, Uebelmannia gummifera.
MAP 7
Map 7. Distribution of Praecereus [ ] and Brasilicereus [ ]. (South America)

MAP 8
Map 8. Distribution of Cereus subg. Ebneria [ ] and C. subg. Mirabella [ ]. (South America)
MAP 9
Map 9. Distribution of Espostoa, sens. str. [ ] and Espostoopsis [ ]. (South America)

MAP 10
Map 10. Distribution of Harrisia: subg. Eriocereus [ ], H. adscendens [ ] and subg. Harrisia [ ]. (Caribbean
& South America)
MAP 11
Map 11. Vicariant species-pairs: Pereskia guamacho [ ] and P. aureiflora [ ]; and Pseudoacanthocereus
sicariguensis [!] and P. brasiliensis [%]. (South America)
MAP 12
Map 12. Distribution of the PERESKIA GRANDIFOLIA Group: P. sacharosa [ ], P. nemorosa [ ], P.
grandifolia, sens. lat. [ ], P. bahiensis [ ] and P. stenantha [ ]. (South America)
MAP 13
Map 13. Distribution of the genus Arthrocereus: A. spinosissimus [ ] and E Brazilian taxa [ ], see Map
49; and of the genus Quiabentia: Q. verticillata [j] and Q. zehntneri [%]. (South America)
MAP 14
Map 14. Vicariant species-pairs in Lepismium: L. lorentzianum [ ] and L. warmingianum [ ]; and L.

incachacanum [ ] and L. cruciforme [ ]. (South America)
MAP 15
Map 15. Locations of protected areas [!] including threatened cactus taxa compared with sites of rare
and threatened taxa [ ] lacking protected areas in Eastern Brazil (endemic taxa only).
MAP 16
Map 16. Second order Area subdivisions prioritized by importance for the conservation of their cactus
endemics, according to the summation of short-list values of taxa scoring 8 or more (see Table 4.3). Key.
1 (1st) = SE Campos Rupestres (SECR) – Southern part, 2 (2nd) = Northern Campos Rupestres (NCR) –
N Serra do Espinhaço, 3 (3rd) = SECR – Diamantina (MG) eastwards, 4 = SECR – Serra do Cabral
etc., 5 = SECR – Grão Mogol etc., 6 = NCR – Chapada Diamantina, Bahia, 7 = Rio São Francisco
Caatingas (RSFC) – southern caatingas, 8 = RSFC – cent.-northern Bahia, 9 = Eastern Caatingas-agrestes
/ Campos Rupestres, 10 = SE Caatingas-agrestes, Minas Gerais.
MAP 17
Map 17. Distribution of Pereskia: P. aculeata [ ], P. grandifolia subsp. grandifolia [ ] (both non-endemic)
& subsp. violacea [ ], P. bahiensis [ ], P. stenantha [ ] and P. aureiflora [ ] (all endemic).
MAP 18
Map 18. Distribution of Quiabentia zehntneri [ ] (endemic), Brasiliopuntia brasiliensis [ ] and Opuntia
monacantha [j] (Eastern Brazil only).
MAP 19
Map 19. Distribution of Tacinga: T. funalis [ ], T. braunii [ ], T. werneri [ ], T. palmadora [ ], T.

saxatilis subsp. saxatilis [%] & subsp. estevesii [!]; and T. inamoena [j] — eastern-, southern- and
westernmost records only.
MAP 20
Map 20. Distribution of Hylocereus setaceus in Eastern Brazil [ ; ? = identity of record to be confirmed]
and the endemic Pseudoacanthocereus brasiliensis [ ].
MAP 21
Map 21. Distribution of Lepismium in Eastern Brazil: L. houlletianum [ ], L. warmingianum [ ] and L.

cruciforme [ ].
MAP 22
Map 22. Distribution of Rhipsalis subg. Phyllarthrorhipsalis in Eastern Brazil: R. russellii [ ], R. elliptica
[ ], R. crispata [ ] and R. oblonga [!].
MAP 23
Map 23. Distribution of Rhipsalis subg. Epallagogonium in Eastern Brazil: R. floccosa subsp. floccosa [ ], the
endemic subsp. oreophila [ ] & subsp. pulvinigera [ ], the endemic R. paradoxa subsp. septentrionalis [ ]
and R. pacheco-leonis subsp. catenulata, R. sulcata & R. cereoides [all indicated by !].
MAP 24
Map 24. Distribution of Rhipsalis subg. Rhipsalis in Eastern Brazil: R. lindbergiana [ ], R. baccifera subsp.
baccifera [ ] & the endemic subsp. hileiabaiana [ ] and R. teres [ ].
MAP 25
Map 25. Distribution of Rhipsalis subg. Erythrorhipsalis and R. subg. Calamorhipsalis in Eastern Brazil: R.
pulchra [!], R. juengeri [ ], R. clavata [ ], R. indet. cf. R. juengeri / R. clavata [?], R. cereuscula [ ], R.
pilocarpa [ ] and R. burchellii & the endemic R. hoelleri [both indicated by ].
MAP 26
Map 26. Distribution of Hatiora and Schlumbergera in Eastern Brazil: H. salicornioides [ ], H. cylindrica
[ ; ? = identity of record to be confirmed], the endemic S. kautskyi [ ], S. microsphaerica [V] and S.
opuntioides [ ].
MAP 27
Map 27. Distribution of Brasilicereus: B. phaeacanthus [ ] and B. markgrafii [ ].

MAP 28
Map 28. Distribution of Cereus subg. Mirabella in Eastern Brazil: C. mirabella [ ;? = record to be confirmed]
and the endemic C. albicaulis [ ].
MAP 29
Map 29. Distribution of Cereus subg. Cereus in Eastern Brazil (excluding C. insularis, see Map 1): C.
fernambucensis subsp. fernambucensis [ ] & subsp. sericifer [ ], C. jamacaru subsp. jamacaru [j] — eastern,
southern- and westernmost records only & subsp. calcirupicola [ ] and C. hildmannianus [!].
MAP 30
Map 30. Distribution of Cipocereus: C. laniflorus [ ], C. crassisepalus [ ], C. bradei [ ], C. minensis subsp.
leiocarpus [ ] & subsp. minensis [ ] and C. pusilliflorus [!]. (Minas Gerais)
MAP 31
Map 31. Distribution of Stephanocereus: S. leucostele [ ] and S. luetzelburgii [ ].

MAP 32
Map 32. Distribution of Arrojadoa bahiensis [ ], A. dinae subsp. dinae [ ] & subsp. eriocaulis [ ] and A.
cf. dinae (Goiás) [ ].
MAP 33
Map 33. Distribution of Arrojadoa penicillata [ ] and A. rhodantha [ ].

MAP 34
Map 34. Distribution of Pilosocereus subg. Gounellea: P. tuberculatus [ ] and P. gounellei subsp. gounellei
[ ] & subsp. zehntneri [ ].
MAP 35
Map 35. Distribution of the PILOSOCEREUS ARRABIDAE Species Group: P. catingicola subsp. catingicola [ ]
& subsp. salvadorensis [ ], P. azulensis [%] and P. arrabidae [ ].
MAP 36
Map 36. Distribution of the PILOSOCEREUS PENTAEDROPHORUS Species Group: P. brasiliensis subsp.
brasiliensis [ ] & subsp. ruschianus [ ], P. flavipulvinatus [ ], P. pentaedrophorus subsp. pentaedrophorus [ ]
& subsp. robustus [ ], P. glaucochrous [V] and P. floccosus subsp. floccosus [ ] & subsp. quadricostatus [ ].
MAP 37
Map 37. Distribution of the PILOSOCEREUS ULEI Species Group: P. fulvilanatus subsp. fulvilanatus [ ] &
subsp. rosae [ ], P. pachycladus subsp. pachycladus [ ] & subsp. pernambucoensis [ ] and P. magnificus [ ].
MAP 38
Map 38. Distribution of the PILOSOCEREUS AURISETUS Species Group in Eastern Brazil: P. machrisii [ ],
P. parvus [%] and the endemics P. aurisetus subsp. aurisetus [ ] & subsp. aurilanatus [ ], P. bohlei [!] and P.
aureispinus [ ].
MAP 39
Map 39. Distribution of the PILOSOCEREUS PIAUHYENSIS Species Group: P. multicostatus [ ], P. piauhyensis
[ ], P. chrysostele [ ] and P. densiareolatus [%].
MAP 40
Map 40. Distribution of Micranthocereus in Eastern Brazil (all taxa treated here are endemic): M. violaciflorus
[j], M. albicephalus [!], M. purpureus [%], M. auriazureus [ ], M. streckeri [ ], M. polyanthus [ ], M.
flaviflorus [ ] and M. dolichospermaticus [ ].
MAP 41
Map 41. Distribution of Coleocephalocereus in Eastern Brazil: C. buxbaumianus subsp. buxbaumianus [%] &
subsp. flavisetus [V], C. fluminensis subsp. fluminensis [ ] & subsp. decumbens [ ], C. pluricostatus [!], C.
goebelianus [ ], C. aureus [ ] and C. purpureus [ ].
MAP 42
Map 42. Distribution of the MELOCACTUS OREAS Species Group (I): M. oreas subsp. oreas [ ] & subsp.
cremnophilus [ ] and M. ernestii subsp. ernestii [ ; ? = sites to be confirmed] & subsp. longicarpus [ ].
MAP 43
Map 43. Distribution of the MELOCACTUS OREAS Species Group (II): M. bahiensis subsp. bahiensis [ ] &
subsp. amethystinus [ ] and M. conoideus [ ].
MAP 44
Map 44. Distribution of Melocactus deinacanthus [ ], M. levitestatus [ ] and the M. AZUREUS Species Group:
M. azureus [ ], M. ferreophilus [ ] and M. pachyacanthus subsp. pachyacanthus [ ] & subsp. viridis [ ].
MAP 45
Map 45. Distribution of the MELOCACTUS VIOLACEUS Species Group (I): M. salvadorensis [V], M. zehntneri
[ ] and M. lanssensianus [ ; = records of comparable taxa].
MAP 46
Map 46. Distribution of the MELOCACTUS VIOLACEUS Species Group (II): M. glaucescens [!], M. concinnus
[ ], M. paucispinus [ ] and M. violaceus subsp. violaceus [ ], subsp. ritteri [ ] & subsp. margaritaceus [ ].
MAP 47
Map 47. Distribution of Harrisia adscendens [ ] and Leocereus bahiensis [!].

MAP 48
Map 48. Distribution of Espostoopsis dybowskii [!] and Facheiroa: F. ulei [ ], F. cephaliomelana subsp.
cephaliomelana [ ] & subsp. estevesii [ ] and F. squamosa [ ]. Notable is the marked disjunction of both
Espostoopsis and Facheiroa squamosa between northern and southern Bahia.
MAP 49
Map 49. Distribution of Arthrocereus: A. melanurus subsp. melanurus [ ], subsp. magnus [ ] & subsp. odorus
[ ], A. rondonianus [ ] and A. glaziovii [ ; ? = unconfirmed record]. (Minas Gerais)
MAP 50
Map 50. Distribution of Discocactus: D. zehntneri subsp. zehntneri [ ] & subsp. boomianus [ ], D. bahiensis
[ ; ? = records to be confirmed or localized], D. heptacanthus subsp. catingicola [%], D. placentiformis [ ],
D. pseudoinsignis [ ] and D. horstii [!].
MAP 51
Map 51. Distribution of Uebelmannia: U. buiningii [ ], U. gummifera [ ] and U. pectinifera subsp. pectinifera
[ ], subsp. flavispina [ ] & subsp. horrida [ ]. Minas Gerais
PLATE 1
1 2
3 4
Plate 1: examples of vegetation types including cacti. 1.1 Mata atlântica. Lepismium cruciforme and Rhipsalis russellii,
in agreste (mata de cipó), SE Bahia, 2002 (MM). 1.2 Caatinga. Melocactus pachyacanthus on a limestone lajedo, caatinga
forest behind, N Bahia, 2002 (NT). 1.3 Cerrado. Partly scorched Melocactus conoideus in ‘cerrado de altitude’, SE Bahia,
2002 (MM). 1.4 Campo rupestre. Young Pilosocereus pachycladus, nr Brejinho das Ametistas, S Bahia, 1991 (NT).
PLATE 2
1 2
3 4
Plate 2. 2.1 Pereskia aculeata. Cult., Univ. Heidelberg (WR). 2.2 Ibid. Cult. Les Cèdres (WR). 2.3 P. grandifolia
subsp. grandifolia. SW Minas Gerais, Harley et al. 24802 (NT). 2.4 P. grandifolia subsp. violacea, with Daniela
Zappi. E Minas Gerais, 2 km N of Penha do Norte, 1990 (NT).
PLATE 3
1 2
3 4
Plate 3. 3.1 Pereskia grandifolia subsp. violacea. E Minas Gerais, Taylor & Zappi 773 (NT). 3.2 Ibid. Ibid. (NT).
3.3 P. bahiensis. Bahia, Harley et al. 25542 (NT). 3.4 P. stenantha. N Minas Gerais, Harley et al. 25523 (NT).
PLATE 4
1 2
3 4
Plate 4. 4.1 Pereskia aureiflora. NE Minas Gerais, Harley et al. 25525 (DZ). 4.2 Quiabentia zehntneri. N Minas
Gerais, Capitão Enéas, 2002 (GC). 4.3 Tacinga funalis. N Bahia, NE of Gruta dos Brejões, 2002 (NT). 4.4 Ibid.
Bahia, W of Morro do Chapéu, 2002 (MM).
PLATE 5
1 2
Plate 5. 5.1 Tacinga funalis, in fruit. Bahia, E of Seabra, 1989 (WB). 5.2 T. braunii. NE Minas Gerais, W of
Itaobim, Harley et al. 25530 (NT). 5.3 Ibid. Loc. cit. (GC).
PLATE 6
1 2
3 4
Plate 6. 6.1 Tacinga werneri, with open flower and buds. Bahia, S of Morro do Chapéu, 2002 (MM). 6.2 Ibid.,
in fruit. NE Minas Gerais, Taylor et al. 1520 (NT). 6.3 T. palmadora. N Bahia, N of Brejão, 2002 (NT). 6.4 T.
saxatilis subsp. estevesii. Type locality, Horst & Uebelmann HU 761, cult. A. Hofacker (AH).
PLATE 7
2 3
Plate 7. 7.1 Tacinga saxatilis subsp. saxatilis. N Minas Gerais, Capitão Enéas, 2002 (GC). 7.2 T. inamoena.
N Bahia, N of Brejão, 2002 (NT). 7.3 Ibid., in fruit. Bahia, 1989 (NT).
PLATE 8
2 3
Plate 8. 8.1 Tacinga werneri T. inamoena. Bahia, W of Morro do Chapéu, 2002 (NT). 8.2 Brasiliopuntia brasiliensis.
E Pernambuco, Caruaru, E.A. Rocha et al. (DZ). 8.3 Ibid., seedling. Rio de Janeiro, Ilha Grande, cult. Kew,
1992 (K).
PLATE 9
Plate 9. Brasiliopuntia brasiliensis, in fruit. Bahia, S of Vitória da Conquista, 2003 (NT).

PLATE 10
1 2
3 4
Plate 10. 10.1 Brasiliopuntia brasiliensis. Rio de Janeiro, Ilha Grande, cult. Kew, flower, 1999 (K). 10.2 Nopalea
cochenillifera. Pernambuco, Igarassu, 1999 (DZ). 10.3 Opuntia monacantha. Espírito Santo, Zappi 470 (DZ). 10.4
Ibid. Ibid. (DZ).
PLATE 11
1 2
3 4
Plate 11. 11.1 Opuntia dillenii. Bahia, Taylor et al. 1350 (NT). 11.2 O. ficus-indica. N Bahia, Morro do Chapéu,
Brejão, 2002 (NT). 11.3 Hylocereus setaceus, in mata de encosta. Piauí, J.B. Fernandes da Silva 365 (NT).
11.4 Ibid., fruit. Espírito Santo, Taylor & Zappi 779 (NT).
PLATE 12
1 2
3 4
Plate 12. 12.1 Hylocereus undatus. Bahia, Morro do Chapéu, 2002 (NT). 12.2 Epiphyllum phyllanthus. Brazil,
Horst, cult. Univ. Heidelberg (WR). 12.3 Pseudoacanthocereus brasiliensis. Bahia, Taylor et al. 1579 (NT).
12.4 Ibid., in fruit. Ibid. (NT).
PLATE 13
1 2
3 4
Plate 13. 13.1 Pseudoacanthocereus brasiliensis. NE Minas Gerais, Taylor & Zappi 768 (NT). 13.2 Lepismium
houlletianum f. regnellii (left) & f. houlletianum. Cult., Univ. Bonn (WB). 13.3 L. warmingianum, in fruit. Ibid.
(WB). 13.4 L. cruciforme, 2 different forms. Ibid. (WB).
PHYTOGEOGRAPHY
Table 3.6
Seed dispersal vectors of species and subspecies of Cactaceae in Eastern
Brazil, with details of habitat and extent of range within that habitat type.
Actual observations of presumed vectors at work are indicated by (). Some infraspecific taxa are not
distinguished from their respective species, where their entries would otherwise be identical and
repetitive (" = as above)
TAXON DISPERSAL VECTOR HABITAT RANGE
Pereskia aculeata mammal Mata atlântica, widespread

sens. lat.
P. grandifolia subsp. grandifolia ground-feeding mammal (peccary?) " "
P. grandifolia subsp. violacea " " restricted
P. bahiensis " caatinga "
P. stenantha " " "
P. aureiflora mammal " "
Quiabentia zehntneri ? " "
Tacinga funalis ? " "
T. braunii ? " "
T. (Opuntia) werneri mammal " "
T. (Opuntia) palmadora " " widespread
T. (Opuntia) saxatilis subsp. saxatilis " " restricted
T. (Opuntia) saxatilis subsp. estevesii " " single site
T. (Opuntia) inamoena mammal (peccary?) " widespread
Brasiliopuntia (Opuntia) brasiliensis ground-feeding mammal (peccary?) diverse forests "
Opuntia monacantha " Mata atlântica: restricted
dunes
Hylocereus setaceus mammal (bat/primate?) & bird diverse forest widespread
Epiphyllum phyllanthus mammal: monkey () & bat (?); bird? " "
Pseudoacanthocereus brasiliensis ground-feeding mammal (peccary?) caatinga restricted
Lepismium houlletianum bird & small monkeys Mata atlântica "
L. warmingianum " " "
L. cruciforme " " widespread
Rhipsalis russellii " Mata atlântica & "
campo rupestre
R. elliptica " Mata atlântica restricted
R. oblonga " " "
R. crispata " " widespread
R. floccosa subsp. floccosa " " "
R. floccosa subsp. oreophila " " "
R. floccosa subsp. pulvinigera " " restricted
R. paradoxa subsp. septentrionalis " " widespread
R. pacheco-leonis subsp. catenulata " " single site
R. cereoides " " "
R. sulcata " " "
R. lindbergiana " " widespread
R. teres " " restricted
R. baccifera subsp. baccifera " " "
R. baccifera subsp. hileiabaiana " Mata atlântica/ "
de neblina
PHYTOGEOGRAPHY
R. pulchra " " "

R. burchellii " Mata atlântica "
R. juengeri " " "
R. clavata " " "
R. cereuscula " " widespread
R. pilocarpa " " restricted
R. hoelleri " " single site
Hatiora salicornioides " Mata atlântica & widespread
campo rupestre
H. cylindrica bird and (?) small lizards Mata atlântica
Schlumbergera kautskyi bird & small monkeys " restricted
S. microsphaerica " mata de neblina "
S. opuntioides " " "
Brasilicereus phaeacanthus bat? caatinga-agreste restricted
B. markgrafii " campo rupestre "
Cereus mirabella mammal cerrado widespread
C. albicaulis " caatinga "
C. fernambucensis subsp. fernambuc. bat, other mammal & bird restinga "
C. fernambucensis subsp. sericifer " Mata atlântica restricted
C. insularis ? restinga "
C. jamacaru subsp. jamacaru bird () caatinga-agreste widespread
C. jamacaru subsp. calcirupicola " " restricted
C. hildmannianus bat & bird Mata atlântica "
Cipocereus laniflorus " campo rupestre single site
C. crassisepalus " campo rupestre/ restricted
cerrado
C. bradei " campo rupestre "
C. minensis subsp. leiocarpus " " "
C. minensis subsp. minensis " " "
C. pusilliflorus " " single site
Stephanocereus leucostele bat & ants caatinga restricted
S. luetzelburgii bat campo rupestre "
Arrojadoa bahiensis bird? " "
A. dinae subsp. dinae bird & mammal? campo rupestre/ "
cerrado
A. dinae subsp. eriocaulis " " "
A. penicillata " caatinga "
A. rhodantha " " widespread
Pilosocereus tuberculatus bat " restricted
P. gounellei subsp. gounellei bat, insect (wasp ) & water " widespread
P. gounellei subsp. zehntneri " " restricted
P. catingicola subsp. catingicola bird () & bat " restricted
P. catingicola subsp. salvadorensis bat (Locatelli et al. 1997) caatinga-agreste restricted
& restinga
P. azulensis bat caatinga-agreste single site?
P. arrabidae lizard (), bat & bird? restinga restricted
P. brasiliensis subsp. brasiliensis bat & bird? " "
P. brasiliensis subsp. ruschianus " Mata atlântica "
P. flavipulvinatus " caatinga restricted
P. pentaedrophorus subsp. " caatinga-agreste widespread
pentaedrophorus
P. pentaedrophorus subsp. robustus " " restricted
PHYTOGEOGRAPHY
P. glaucochrous " caatinga de widespread

altitude
P. floccosus subsp. floccosus " mata seca restricted
P. floccosus subsp. quadricostatus " caatinga-agreste "
P. fulvilanatus subsp. fulvilanatus " campo rupestre "
P. fulvilanatus subsp. rosae " " single site
P. pachycladus subsp. pachycladus " caatinga & campo widespread
rupestre
P. pachycl. subsp. pernambucoensis " caatinga-agreste "
P. magnificus " " restricted
P. machrisii " rocks in cerrado "
P. aurisetus subsp. aurisetus " campo rupestre "
P. aurisetus subsp. aurilanatus " " "
P. aureispinus bat, bird (?) & ant rocks in cerrado "
P. multicostatus bat & bird? caatinga-agreste "
P. piauhyensis " caatinga "
P. chrysostele " " "
P. densiareolatus " " "
Micranthocereus violaciflorus ant? campo rupestre "
M. albicephalus " " "
M. purpureus bat & bird () " "
M. auriazureus bird " single site
M. streckeri " " "
M. polyanthus " " restricted
M. flaviflorus " " "
M. dolichospermaticus wind caatinga "
Coleocephalocereus buxbaumianus, ant () (also bird/bat?) Mata atlântica "
sens. lat.
C. fluminensis subsp. fluminensis (ant ) " "
C. fluminensis subsp. decumbens ant (also bird/bat?) " "
C. pluricostatus " " "
C. goebelianus (ant ) " caatinga "
C. aureus (ant ) " " "
C. purpureus ant (also bird/bat?) " single site
Melocactus oreas subsp. oreas lizard () & bird " restricted
M. oreas subsp. cremnophilus " campo rupestre "
M. ernestii subsp. ernestii " (lizard ) caatinga widespread
M. ernestii subsp. longicarpus " " restricted
M. bahiensis subsp. bahiensis " caatinga & campo restricted
rupestre
M. bahiensis subsp. amethystinus " " "
M. conoideus " cerrado de altitude single site
M. deinacanthus " caatinga "
M. levitestatus " " restricted
M. azureus " " "
M. ferreophilus " " "
M. pachyacanthus, sens. lat. " " "
M. salvadorensis " " "
M. zehntneri " (lizard ) " widespread
M. lanssensianus, sens. lat. " " restricted
M. glaucescens " campo rupestre "
PHYTOGEOGRAPHY
M. concinnus " caatinga & campo "

rupestre
M. paucispinus " cerrado de altitude "
M. violaceus subsp. violaceus " (lizard ) restinga widespread
M. violaceus subsp. ritteri " campo rupestre restricted
M. violaceus subsp. margaritaceus " restinga "
Harrisia adscendens bat, other mammal & bird caatinga widespread
Leocereus bahiensis " campo rupestre "
& caatinga
Facheiroa ulei bat caatinga restricted
F. cephaliomelana subsp. " " "
cephaliomelana
F. cephaliomelana subsp. estevesii " " single site
F. squamosa " " restricted
Espostoopsis dybowskii bat & bird " "
Arthrocereus melanurus subsp. bat, other mammal & bird campo rupestre "
melanurus
A. melanurus subsp. magnus " " single site
A. melanurus subsp. odorus " " restricted
A. rondonianus " " "
A. glaziovii " " "
Discocactus zehntneri subsp. zehntneri ant caatinga restricted
D. zehntneri subsp. boomianus " campo rupestre "
D. bahiensis ant () & water caatinga "
D. heptacanthus subsp. catingicola ant cerrado "
D. placentiformis " () campo rupestre/ "
cerrado
D. pseudoinsignis " cerrado "
D. horstii " campo rupestre/ single site
cerrado
Uebelmannia buiningii ant? " restricted
U. gummifera " " "
U. pectinifera subsp. pectinifera " campo rupestre "
U. pectinifera subsp. flavispina " " "
U. pectinifera subsp. horrida " " single site
PHYTOGEOGRAPHY
feeding in forest where E. phyllanthus was fruiting (bats are also possible vectors here).
Bats are assumed to be important seed vectors for the great majority of Brazilian
columnar cacti with fleshy fruit (Taylor & Zappi 1989: 22; Locatelli et al. 1997) and this
role is certainly well known in relation to the famous North American Sahuaro cactus,
Carnegiea gigantea (Engelm.) Britton & Rose, and other cereoid cacti from the Northern
Hemisphere (Dobat & Peikert-Holle 1985). It seems reasonable to assume that the
distances fruit-eating bats regularly travel are not likely to be a limiting factor in the
expansion of range for such columnar species, and the substantial distributions attained by
genera such as Pilosocereus in the Americas presumably reflect this. Birds may play a similar
role (eg. Cereus jamacaru, Melocactus) and are likely also to be important vectors for the
many smaller-fruited, epiphytic Rhipsalideae (Lepismium, Rhipsalis, Hatiora &
Schlumbergera), whose ranges vary from ‘widespread’ to ‘single site’ (Table 3.6). Some of
these taxa have very sticky fruit pulp and are assumed to be dispersed by birds in a manner
similar to mistletoes (Viscum spp.), but small monkeys (‘sagüis’) have also been reported
eating their fruits in South-Eastern Brazil (herbarium label data).
Lizards have been frequently observed upon Melocactus plants, eating the watery fruit,
and are assumed to be effective local dispersal agents (Taylor 1991a, Figueira et al. 1993,
1994). However, it is improbable that this genus has achieved its currently extensive, yet
discontinuous, neotropical range by this means alone and dispersal over longer distances
by birds is presumed to have occurred (Taylor 1991a). Lizards may also consume the fruits
of other cacti that grow close to the ground or which are not so densely covered in sharp
spines as to render them inaccessible (however, dead lizards impaled on the spines of
Melocactus have been observed on more than one occasion!). These vectors were also
observed eating Pilosocereus arrabidae fruits.
Ants are frequently attracted to the funicular pulp in which nearly all cactus seeds are
embedded, but generally do not appear to penetrate the often leathery pericarp of the fruit
on their own and are probably less effective with or unable to transport larger seeds.
However, the small-seeded genera, Coleocephalocereus and Discocactus, have dehiscent fruits,
the former opening by means of a basal pore, the latter by lateral fissures, and each type has
been observed in the process of being raided by ants, which were carrying seeds away.
Furthermore, Discocactus bahiensis plants have been observed growing in the centre of old
ants’ nests. Both genera are characterized by erratic occurrence, being abundant in some
suitable habitats yet curiously absent from others, suggesting the possibility that longer
distance dispersal is less effective and may rely on birds. It is interesting to note that while
their fruits at first sight appear to differ markedly, Coleocephalocereus goebelianus and
Stephanocereus leucostele have seeds of remarkably similar shape, yet are not closely related taxa.
However, it turns out that both have fruits which open by means of a basal pore and are
strongly attractive to ants (M. Machado, unpubl. observations for S. leucostele), this suggesting
that the coincidence in seed morphology may be connected with dispersal vectors.
While it has been possible to make observations in habitat of a few taxa, and speculate
on rather more, in relation to their dispersal vectors, there remain some species for which
it is difficult to be sure how dispersal might be achieved. These include the opuntioids,
Quiabentia zehntneri, Tacinga funalis and T. braunii, whose fruits are scarcely either
PHYTOGEOGRAPHY
conspicuously coloured, fragrant or juicy/fleshy at maturity. While Quiabentia may

indulge in vegetative propagation on a local scale, this can hardly explain the present
range of either the genus or its Brazilian species, unless this was once much more
continuous than it is today, or was achieved over a long time scale. The same remark
applies, to a lesser degree, to the above-mentioned Tacinga species, but careful studies of
these, and of Brazilian cacti in general, are clearly desirable.
Study of Table 3.6 suggests that dispersal strategies do have a significance for the
interpretation of distribution patterns, but that this is probably much less important than
other, even if interrelated, climatic and edaphic factors. While, as noted occasionally above,
the range and frequency of certain taxa may well be explained by such strategies, many
closely related taxa, with presumably similar or identical dispersal vectors, differ markedly
in the range they have attained, suggesting that other habitat factors, climatic history or the
age of the taxon itself have come into play. This is assumed to be the case with the disjunct
occurrences of Harrisia adscendens, Arrojadoa penicillata and Tacinga funalis to the west/north-
west of the Rio São Francisco, none of these displaying dispersal strategies, such as very
sticky fruit pulp enabling epizoochory, which could facilitate the extension of range over
longer distances. The points made in this paragraph can also be applied to pollination
vectors, which are briefly discussed under the heading of Conservation (Chapter 4).
Pollinators are clearly of considerable importance, since extant knowledge suggests that the
family is largely comprised of self-incompatible taxa (Ross 1981).
CONSERVATION
4. CONSERVATION
4.1 WHY CONSERVE THE CACTI OF EASTERN BRAZIL?

There are various arguments to be advanced in support of the conservation of the
Cactaceae of Eastern Brazil. Perhaps the strongest, however, is the degree of biological
uniqueness these plants represent in terms of endemic genera and species. This can be
expressed both in terms of the family in Brazil and in the Americas as a whole, to which
the Cactaceae is all but endemic (save only for Rhipsalis baccifera, which ranges from the
neotropics into the palaeotropics). Brazil has a total of 37 native genera of Cactaceae (ie.
c. 30% out of a New World total of c. 120), of which 28 or 75% are found in Eastern
Brazil, 12 (32%) of these being Brazilian endemics native in Eastern Brazil. Of the 28
genera from this area, 8 (29%) are endemic to it and the remaining 4 Brazilian endemics
have the major parts of their ranges and nearly all of their biological diversity within the
area. The 28 native cactus genera of Eastern Brazil comprise 130 named, fully accepted
species, of which 117 (90%) are endemic to Brazil and 88 (68%) are endemic to the area
covered here. If heterotypic subspecies are added to the total for Eastern Brazil, then we
have 162, of which 123 (76%) are endemic. Taking the family as a whole, 3 out of the 4
subfamilies occur in Eastern Brazil, and the largest of these, the Cactoideae, includes
about 9 tribal groups. One such is Cereeae, whose 10 genera all occur in Brazil, 3 being
endemic to the Eastern region (Brasilicereus, Cipocereus & Stephanocereus) and a further 5
having most of their diversity in the region. Another large tribe are the Trichocereeae,
with 7 genera in Eastern Brazil, 4 of these being endemic to the region (Leocereus,
Facheiroa, Espostoopsis & Uebelmannia). The taxonomic isolation, remarkable morphology
and ecology of Uebelmannia make the strongest of biological arguments for its
conservation and the preservation of its habitats.
While it may be legitimate to analyse the above cactus diversity in isolation, it is
more environmentally relevant to consider the ways in which its loss might impact the
overall ecology of the habitats the plants occupy and sometimes co-dominate. We know
too little of the overall ecology of cacti in Eastern Brazil to be able to cite specific cases,
but at least in parts of the caatinga, campo rupestre, drier phases of the Mata atlântica and
coastal restinga there exist communities where the loss of cacti could certainly affect the
survival of their pollinators and seed vectors, and vice versa. Perhaps the most vulnerable
of higher animal groups in this respect are bats, which visit columnar cacti for nectar,
pollen and fruit (Zappi 1994, Ruiz et al. 1997, Locatelli et al. 1997, Petit 1999). In some
of these communities the cacti flower and fruit for much of the year, providing an
ongoing food resource even when other vegetation is seasonally dormant or suffering
WHY CONSERVE?
from drought. Another group reliant on cactus nectar as an energy source, at least in
some of the driest areas, are the hummingbirds, and in particular those which locate their
territories amongst populations of Melocactus, Arrojadoa and Tacinga, which like some
columnar cacti tend to flower for a significant part of the year (Taylor 1991a, Taylor &
Zappi 1996, Raw 1996, Locatelli & Machado 1999a). In the case of Melocactus, the
subsequent and regular production of juicy fruits is an important water resource for
lizards (Figueira et al. 1993, 1994), which locally disperse the seeds. Besides these
organisms, there are other birds, lepidopterans (Taylor 1991a, Locatelli et al. 1997,
Locatelli & Machado 1999b), bees (Schlindwein & Wittmann 1997), ants and terrestrial
mammals that interact with Brazilian cacti in various ways (see also Chapter 3.5), not to
mention the important roles played by other plants. These last are many and various,
ranging, for example, from the carnaúba palm (Copernicia prunifera) in flooded forests in
northern Piauí and Ceará, upon which Pilosocereus gounellei is often epiphytic, to great
trees in the Atlantic Forest supporting many epiphytes belonging to the tribe
Rhipsalideae. And while birds may need cacti as a source of energy, they can also be a
source of nesting material, the authors having seen birds collecting the cephalium wool
of Micranthocereus purpureus for this purpose. The same is reported by Schulz & Machado
(2000: 63), involving Pilosocereus aurisetus and a hummingbird.
The numerous locality records for cacti in Eastern Brazil assembled for the present
study may in future enable assessments of the well-being of diverse vegetation types to be
made and guide those who have the power to create reserves, as environments inevitably
deteriorate further. Few, if any, cacti found in Eastern Brazil appear to increase
significantly when habitat disturbance occurs (Quiabentia zehntneri, Cereus jamacaru and
Pilosocereus pachycladus subsp. pernambucoensis being possible exceptions), but it is probable
that a more complete knowledge of the ecology of their diverse habitats will permit the
identification of cactus species assemblages indicative of primary vegetation. This would
seem a real possibility in the case of certain Rhipsalideae, whose presence, diversity and
abundance in the more humid parts of the Atlantic Forest should be a reliable indicator
of its primary or secondary status, since these epiphytes seem only to occur on mostly
large trees of considerable antiquity.
Another compelling argument in favour of cactus conservation in Eastern Brazil is
their economic potential, if used sustainably, both at the local level and for international
trade. They are already employed for a variety of purposes locally, whether as substitute
livestock fodder in times of drought, for their delicious fruit, for making ‘cactus candy’
or planted to form living, impenetrable fences (especially Pereskia bahiensis and Cereus
jamacaru). The frequent attributions of medicinal value made by locals in relation to cacti
mostly lack any kind of scientific insight at present (cf. Agra 1996: 32), but in other
countries these plants are frequently used in alternative medicine and some have been
cultivated commercially for extraction of drugs, eg. Selenicereus (see Backeberg 1966: plate
pp. 10–11). While they are already recognized for their potential in amenity horticulture
in Brazil, where an established nursery industry exports large volumes to Europe, the
USA and beyond, more could be made of native species from Eastern Brazil. Some are
especially attractive as seedlings and certainly marketable, provided their production does
CONSERVATION STATUS
not rely on a constant supply of seeds and plants from the wild, but employs artificial
propagation. A further area of commerce that native cacti have a role to play in is that of
the ever-increasing trend for ‘eco-tourism’. There are enough cactophiles in the world to
make it worthwhile for specialised tour companies to arrange bespoke holidays for
aficionados to visit the more spectacular plants and habitats, eg. around Morro do
Chapéu, BA, once again given that the object is to photograph, video and observe in
nature, but not collect! Such tours have operated in Mexico for some years now and, if
properly policed, can heighten public awareness of the value of nature conservation.
Cacti are also important to preserve for purely aesthetic and cultural reasons. The
Xique-xique, Mandacaru and Palmatória (Pilosocereus gounellei, Cereus jamacaru & Tacinga
palmadora, respectively) frequently appear on the signs of restaurants and bars in the
Brazilian Nordeste, where they form a part of the folklore, as well as providing names for
numerous villages and towns. Their vernacular nomenclature is extensive, interesting and
sometimes quite entertaining for those with knowledge of Brazilian Portuguese and, as
such, is listed in Chapters 5 & 7.3. Cacti are frequently planted as decorative symbols of
the region in village and town squares and sometimes allowed to remain in the fields when
other vegetation has been cleared during agricultural development. Last, but not least, it is
hard to forget the Brazilian passer-by in a remote part of Minas Gerais, who upon noticing
our interest in a large planted Cereus flowering by the roadside, warmly offered the
comment, “É Mandacaru da Bahia! — Quer uma muda?” [ ... Do you want a cutting?].
4.2 STATUS OF THE ENVIRONMENT AND

CONSERVATION OF CACTI IN E BRAZIL
The following background information has been updated from the relevant Brazilian
elements in the chapter on South America in the IUCN Species Survival Commission’s
Cactus & Succulent Specialist Group Action Plan, by Taylor et al. (see Oldfield 1997).
Since then an important official report recording the levels of habitat destruction in Brazil
has also been published (Brasil 1998: 49 etc.).
In terms of East Brazilian cactus species, special conservation concerns include endemic
taxa from the genera Discocactus (7 species, 5 endemic), Uebelmannia (endemic, 3 spp.) and
Melocactus (M. conoideus, M. deinacanthus, M. glaucescens, M. paucispinus), which are
‘Critically Endangered’, ‘Endangered’ or ‘Vulnerable’ (sensu IUCN 2001) and, since 1992,
have been placed in Appendix I of CITES (Convention on Trade in Endangered Species
of fauna & flora) to afford them protection from the export trade (Taylor 1991c). Most of
these ‘Appendix I’ plants and equally threatened taxa in the genera Cipocereus, Arthrocereus
and Espostoopsis are known from only one or very few localities, where the populations
number between less than ten to at most a few thousand individuals. Discocactus
placentiformis, D. horstii, D. pseudoinsignis, D. zehntneri subsp. boomianus, Melocactus glaucescens
(all except the first being known from between only 1 and 5 small areas each) and all
Uebelmannia species (U. buiningii being ‘Critically Endangered’, cf. Braun & Esteves Pereira
1988) are threatened in part by trade, via regular collection of plants for seed production,
or of seeds, for wholesale export in large quantities. Discocactus bahiensis and Melocactus
CONSERVATION STATUS
deinacanthus (the latter currently known from only a single very small area) are also seriously
threatened by agricultural development, and both the former and D. zehntneri subsp.
zehntneri have had their ranges and numbers significantly reduced by inundation from the
Represa de Sobradinho, a huge dam-lake created in the 1970s on the Rio São Francisco
(Bahia/Pernambuco). Repeated commercial collecting was only partly responsible for the
decline of Melocactus conoideus at its type locality above the expanding city of Vitória da
Conquista, southern Bahia, a species that remains ‘Critically Endangered’ due to the
extraction of the quartz gravel in which it grows (Taylor 1992b). On the positive side the
‘Critically Endangered’ Melocactus glaucescens has recently received increased protection by
the creation of the Parque Estadual do Morro do Chapéu (BA), whose area includes most
of the small remaining population at its locus classicus where, however, it still needs
protection from the burning of vegetation by local farmers.
Caatinga. The driest zone of Eastern Brazil is the caatinga (‘Area 3’, see Chapter 3) and
its ecotones with Atlantic Forest to the east (agreste), dry forests to the south (in Minas
Gerais & Espírito Santo) and savannas (cerrados) to its west. It represents a severely
disturbed ecosystem (Andrade-Lima 1981) that has been subject to forest clearance for
agriculture and fuelwood over more than two centuries (Lleras in Davis et al. 1997: 395).
Over half of the energy needs in North-eastern Brazil are met by plant biomass (Brasil
2000: 10). Desertification in its northern part has recently begun to accelerate at an
alarming rate, as reports in the popular press testify and have been stimulating government
action towards grand irrigation schemes (Gusmão 1999). Sadly, this seems to be driven
more by an understandable desire to further exploit the land than to conserve its
remaining biodiversity.
Nevertheless, many cacti have probably suffered less than most other plants as a
consequence of their frequent occurrence on rock outcrops unsuitable for cultivation or
livestock grazing. Thus, species of Coleocephalocereus, various Pilosocereus and some
Melocactus (eg. M. ernestii), have significant populations in places dominated by
gneiss/granite inselbergs, which are probably at less risk from habitat modification, unless
situated near expanding towns or selected for quarrying.
Of those cacti that are not mainly restricted to rock outcrops, the least threatened are
those which seem able to regenerate when their forest habitat is cut over. These include
Cereus jamacaru, Pereskia grandifolia (Plate 2.4), P. bahiensis and P. stenantha, and all are also
conserved by their use in the form of impenetrable livestock hedges and fences
surrounding homesteads, both within and sometimes outside their natural ranges. A few
very widely distributed endemic cacti, which inhabit little-utilised or sufficiently diverse
habitats, are probably not at risk, even though their numbers may have dropped
significantly, eg. Facheiroa squamosa, Harrisia adscendens, Leocereus bahiensis, Tacinga
inamoena, Pilosocereus gounellei subsp. gounellei and P. pachycladus, sens. lat. However, other,
mostly wide-ranging species that are mainly found growing in the soil of the caatinga-
agreste, or on exposed rocks more or less level with the floor of the surrounding thorn
forest, have suffered considerable reductions in their distributions and abundance through
forest clearance. Endemic species affected in this way, whose ranges now appear to some
CONSERVATION STATUS
degree fragmented, include Arrojadoa penicillata, A. rhodantha, Brasilicereus phaeacanthus,

Cereus albicaulis, Coleocephalocereus goebelianus, Melocactus salvadorensis, M. zehntneri, Pereskia
aureiflora, Pseudoacanthocereus brasiliensis, Pilosocereus catingicola, sens. lat., P. floccosus subsp.
quadricostatus, P. flavipulvinatus, P. glaucochrous, P. pentaedrophorus, sens. lat., Stephanocereus
leucostele, Tacinga palmadora, T. braunii and T. funalis. Although most of these are unlikely
to become seriously threatened in the immediate future (some are already ‘Vulnerable’),
regular monitoring and protection is essential if they are not to become ‘Endangered’ in
the longer term.
Of more urgent concern are taxa, such as Melocactus azureus and M. pachyacanthus, sens.
lat., which have smaller ranges and are restricted to local, low-lying outcrops of limestone
(Rio São Francisco basin, Plate 1.2) whose vegetation gets destroyed when the
surrounding caatinga forest is cleared for cultivation and grazing. These taxa are here
assessed as ‘Endangered’ or ‘Critically Endangered’ on the basis of their known
populations, but more field studies are needed in the remoter parts of northern Bahia,
where additional and less disturbed habitats could exist, although as time goes on this
seems less likely. Even if cacti found on raised rock outcrops within the caatinga are
generally at less risk from agricultural development etc., some, and particularly those close
to roads or human settlements, are at risk from the quarrying of stone for building
materials. Those found only on limestone or gneiss/granite outcrops and probably most
at risk are: Coleocephalocereus aureus, Facheiroa cephaliomelana, Melocactus ferreophilus, M.
levitestatus, Micranthocereus dolichospermaticus, Pilosocereus densiareolatus, P. floccosus, P.
multicostatus, P. gounellei subsp. zehntneri and Tacinga saxatilis.
Examples of the few and mostly relatively small protected areas (WCMC 1992, Brasil
1998: 66–91) within the vast ‘Caatinga dominion’ (Andrade-Lima 1981) are, as follows:
‘Parque Nacional Serra de Capivara’ (includes Pilosocereus piauhyensis) and ‘Parque Nacional
Sete Cidades’ (incl. P. flavipulvinatus; both parks are in Piauí state), ‘Parque Nacional de
Ubajara’ (Ceará), ‘Estação Ecológica de Seridó’ (Rio Grande do Norte), ‘Reserva
Ecológica do Raso da Catarina’ (NE Bahia), the ‘Áreas de proteção ambiental (APA)’
known as ‘Serra de Baturité’ (Ceará), ‘Lago de Pedra do Cavalo’ (Bahia — includes
Pseudoacanthocereus brasiliensis) and ‘Gruta dos Brejões/Vereda do Romão Gramacho’ (BA
— includes Melocactus pachyacanthus and M. ferreophilus), ‘Estação Ecológica Federal de
Aiuaba’ (Ceará) and the ‘Estação Experimental do IPA’, Caruaru (PE). These can offer
protection to only few species, since, unfortunately, there are currently no significant
protected areas in the southern part of the caatinga zone in cent.-S Bahia and N Minas
Gerais, where a very high species diversity and endemism is matched by a most disturbing
level of habitat destruction (mainly for agriculture and charcoal production). One of the
most important areas needing protection amongst the southern caatinga-agrestes is the
middle section of the Rio Jequitinhonha valley (between the towns of Araçuaí and Jacinto)
in north-eastern Minas Gerais, where a remarkably rich assortment of cacti exists,
including many endemic and potentially threatened cactus species (Taylor & Zappi 1992a).
Another promising site for protection, with a comprehensive range of southern caatinga
cacti, including the rare Espostoopsis dybowskii, is situated to the east of the village of Porto
Alegre, on the north bank of the Rio de Contas drainage, Mun. Maracás, Bahia. Other
CONSERVATION STATUS
sites need to be identified for the conservation of taxa characteristic of the deep soils and
‘Bambuí’ limestone outcrops in the middle section of the valley of the São Francisco River
(especially for columnar Cactaceae). One such would be the massive raised outcrop south
of the town of Iuiú on the east bank of the river (SW Bahia), which has two very local
endemics restricted to the rock itself (Facheiroa cephaliomelana subsp. estevesii and Tacinga
saxatilis subsp. estevesii). Other sites should be found to the west of the river, where further
endemics, such as the aforementioned Micranthocereus dolichospermaticus and Facheiroa
cephaliomelana subsp. cephaliomelana, are located.
Campo rupestre & cerrado. The East Brazilian Highlands, with their mosaic of campo
rupestre and cerrado vegetation (Giulietti & Pirani 1988, Zappi & Taylor 1994: 77),
represent the least modified of the environments in Eastern Brazil. However, they have
much greater concentrations of threatened species than the caatingas-agrestes just discussed,
and many are of extremely local occurrence and therefore potentially at considerable risk.
Widespread and mostly common, non-threatened exceptions include Cipocereus minensis
subsp. leiocarpus, Melocactus bahiensis, M. concinnus, Pilosocereus aurisetus subsp. aurisetus,
Micranthocereus purpureus and Stephanocereus luetzelburgii, the latter two endemic to the
extensive uplands of the Chapada Diamantina, Bahia, and also found within its national
park (‘Parque Nacional da Chapada Diamantina’, Mucugê–Lencóis).
Utilization of the campos rupestres is generally limited to cattle grazing, with associated
burning to induce re-growth of edible pasture, and local extraction of some plants, eg.
Eriocaulaceae (dried flower export trade — a serious conservation issue), orchids and
Vellozia spp. (Harley in Stannard 1995: 35–37, Giulietti et al. in Davis et al. 1997: 403), and
there is also limited disturbance caused by small-scale mining for gold and precious stones.
In certain areas tourism, including eco-tourism, is becoming popular due to the spectacular
scenery, and, as noted already, this can create either negative or positive environmental
pressures, depending on how sustainably it is managed. Some parts where cerrado
vegetation is more abundant are being cut over for the production of charcoal and later
converted into Eucalyptus plantations, especially in Minas Gerais, where this activity is one
of the factors threatening Uebelmannia species and Cipocereus crassisepalus. Burning the land
for grazing and trampling by cattle does affect some native populations of cacti, but the
regular collection of plants, and nowadays more especially of seed, of certain rare cacti may
be cause for greater concern. In addition to some of the CITES Appendix I taxa noted
above, the following campo rupestre/cerrado cacti are known from only one or two small
populations, or at best have a very localized range which does not include any kind of
designated protected area: Arrojadoa dinae (especially the rare variant, subsp. eriocaulis),
Arthrocereus rondonianus, Cipocereus bradei, C. crassisepalus, C. pusilliflorus, Melocactus violaceus
subsp. ritteri, Micranthocereus albicephalus, M. polyanthus, M. streckeri, Pilosocereus aurisetus
subsp. aurilanatus and P. fulvilanatus subsp. rosae.
Similarly restricted taxa located within protected areas are rather few: Arrojadoa
bahiensis (partly inside the ‘Parque Nacional Chapada Diamantina’ and the ‘APA da Serra
do Barbado’, Bahia), Brasilicereus markgrafii, Micranthocereus auriazureus, M. violaciflorus,
Discocactus pseudoinsignis & D. horstii (inside the ‘Parque Estadual da Serra do Barão’, Minas
CONSERVATION STATUS
Gerais), Uebelmannia gummifera (found within the ‘Parque Estadual da Serra Negra’, MG),
Cipocereus minensis subsp. minensis (partly within the ‘Parque Nacional da Serra do Cipó’,
MG), C. laniflorus (entirely within the ‘Reserva Particular de Patrimônio Natural do
Caraça’, MG) and Arthrocereus melanurus subsp. magnus (‘Parque Estadual de Ibitipoca’,
MG). If extended slightly to its west, the ‘Parque Nacional Chapada Diamantina’ would
include a further population of the remarkable Arrojadoa bahiensis. The Serra da Piedade
(Mun. Caeté, Minas Gerais) is not a designated protected area, but benefits from some
protection as a site of religious significance, which has a population of Arthrocereus
glaziovii, a specialized species generally restricted to rocks very rich in iron (known as
canga). Many of its former habitats have disappeared through ore extraction. A peculiar
and specialized cactus, found in the sandy cerrados bordering on the caatinga and campo
rupestre zones, from western Maranhão to central-eastern Minas Gerais, is Cereus mirabella.
It is widespread, but erratically occurring, and much of the habitat in the southern part
of its range is being destroyed by charcoal producers. Its status needs to be monitored.
Locations within the campos rupestres, where new protected areas have been suggested
to assist the conservation of the above-listed rarities (Taylor in Oldfield 1997, Costa et al.
1998), including the earlier discussed CITES Appendix I taxa, are: the southern end of
the Serra Chapada and associated cerrado (27–28 km W of Seabra, BA), the quartzitic
outcrops at Brejinho das Ametistas (S Bahia), the Serra Geral c. 12–15 km east of Monte
Azul (Minas Gerais), the Serra Geral with white sand cerrado 12 km east of Mato Verde
(MG), the Serra do Cabral (MG), the western slopes of the Serra de Minas east of Santa
Bárbara (Mun. Augusto de Lima, MG) and one or more sites for Uebelmannia pectinifera
in the vicinity of Diamantina (MG).
Mata atlântica. Of great concern, in general terms, is the Brazilian Atlantic Forest. This
comprises the coastal rainforest (Mata atlântica in its strictest sense) and sandy littoral dunes
(restingas) of North-eastern Brazil and their extensions southwards, where the former
broadens and merges with the planalto forests of South-eastern Brazil. This more or less
humid area, which has a high alpha-diversity of epiphytic cacti from the tribe
Rhipsalideae (cf. Ibisch et al. 1996), is represented by only a small fraction of the original
forest. Myers et al. (2000: Table 1) state that only 7.5% of the original primary forest
remains. Rising above the forest or its remnants are the frequently encountered
gneiss/granite inselbergs, which represent important habitats for rupicolous taxa. While
these may appear safe, they are nevertheless affected when the surrounding forest is cut,
because this tends to allow weedy species and especially alien grasses (eg. ‘capim gordura’)
to occupy the niches that would be taken by seedling cacti.
The Atlantic Forest zone is home to some horticulturally and economically important
genera, such as Schlumbergera (species of ‘Christmas cactus’) and Hatiora. A few very
widespread or regionally common taxa, such as the epiphytic H. salicornioides, Lepismium
cruciforme, L. houlletianum, L. warmingianum, Rhipsalis floccosa, R. teres, R. elliptica and R.
cereuscula, and the non-epiphytes Brasiliopuntia brasiliensis, Opuntia monacantha, Cereus
fernambucensis, Pilosocereus arrabidae and P. brasiliensis are here regarded as of ‘Least
Concern’ or ‘Near Threatened’. However, the remaining Brazilian endemic species are
CONSERVATION STATUS
of conservation concern to varying degrees and the not infrequent but mostly small
coastal protected areas may not be sufficient to halt their decline. For example, the wide-
ranging, but erratically occurring restinga taxa, Melocactus violaceus subspp. violaceus &
margaritaceus, and other coastal cacti, are threatened at many points in their ranges by ever-
expanding tourism, urban and agricultural developments (sugarcane, coconut and
pineapple). At least one extensive population of M. violaceus subsp. margaritaceus and
Pilosocereus catingicola subsp. salvadorensis are, however, currently offered some protection
in the ‘Estação Ecológica da Serra de Itabaiana’, Sergipe and ‘Reserva Ecológica Dunas
do Abaete’, Bahia.
The flora of southern Espírito Santo is poorly understood and its habitats severely
altered, but it can count a recently described species of Christmas Cactus, Schlumbergera
kautskyi (known from only 2 small sites), and the remarkable, red-flowered Rhipsalis
hoelleri, as yet not localized with certainty. Other species, represented by disjunct
populations, are Rhipsalis cereoides, R. pilocarpa and Schlumbergera microsphaerica (the last
within the boundaries of the ‘Parque Nacional do Caparaó’, which straddles the border
with the state of Minas Gerais). Another protected area in this state, which merits
investigation, is the ‘Reserva Biológica Federal de Nova Lombardia’ (Mun. Santa Teresa).
Destruction of the Atlantic Forest has been greatest in North-eastern Brazil, where
very little remains (Wayt Thomas in Davis et al. 1997: 364) and, therefore, our knowledge
of the flora is correspondingly fragmentary. To judge from the number of taxa known
from very few, disjunct records, it is quite possible that epiphytic Cactaceae from here
have become extinct before discovery and description. In Pernambuco remnants of this
forest include those termed as ‘brejos’, on higher land far away from the coast, where the
watersheds are an important resource for the human populations living below them. Such
forests have recently been studied and catalogued (Rodal et al. 1998, Sales et al. 1998) as
part of an Anglo-Brazilian project, supported by the U.K. Government’s Darwin
Initiative and the Royal Botanic Gardens, Kew (‘Plantas do Nordeste’ programme). Great
emphasis is being placed on the need to preserve these floristic refuges, which, inter alia,
include disjunct northern populations of cactus epiphytes, such as Lepismium cruciforme and
Rhipsalis crispata. A protected area including one of these brejos is the ‘Reserva Biológica
Federal da Serra Negra’, Pernambuco, which includes a population of the monotypic
Brasiliopuntia brasiliensis.
Further south, in coastal Bahia (up to 100 km inland), between the capital Salvador
and Teixeira de Freitas, where annual rainfall is generally in excess of 1750 mm, there are
occasional records of various species of Rhipsalideae, indicating a once rich centre of
diversity, including Hatiora cylindrica (‘Vulnerable’), Rhipsalis crispata, R. paradoxa subsp.
septentrionalis (‘Endangered’), R. baccifera subsp. hileiabaiana & R. russellii (both
‘Vulnerable’) and R. oblonga. With so little forest remaining, post 1971, when road BR
101 was completed, it seems reasonable to assume that all of these are threatened to a
significant extent. Hopefully, some may benefit from protection in local reserves, such as
the ‘Reserva Biológica Federal de Una’ (south of Ilhéus, BA) and ‘Parque Nacional de
Monte Pascoal’ (N of Itamaraju, BA) — other, smaller protected areas are described by
Wayt Thomas in Davis et al. (1997: 367). Table 4.1 lists brejo forests in NE Brazil (see Map
CONSERVATION STATUS
Table 4.1
Brejo forests in North-eastern Brazil with notable Cactaceae and their most
recent collection dates.
Precise localities are given under materials cited for the above in Chapter 5.
Locality of brejo (state) Cactaceae (date of most recent record), notes
Serra de Baturité (Ceará) Rhipsalis baccifera subsp. baccifera (1979)

N slope of Chapada do Araripe (Ceará) Pereskia grandifolia subsp. grandifolia (1971)
Areia (Paraíba) Rhipsalis baccifera subsp. baccifera (1998)
Pico do Jabre (Paraíba) Melocactus ernestii (1998) — see Rocha & Agra (2002)
Taquaritinga do Norte (Pernambuco) Lepismium cruciforme & Rhipsalis floccosa (2001), R. cereuscula
(1972)
Poção (Pernambuco) Pereskia aculeata (1991)
Brejo da Madre de Deus (Pernambuco) Rhipsalis floccosa (1980), R. cereuscula (1995), Melocactus ernestii
(1993)
Caruaru (Pernambuco) Rhipsalis crispata (1970), R. floccosa (1971) — these species
records urgently need re-confirmation in habitat
Serra Negra, Inajá (Pernambuco) Brasiliopuntia brasiliensis & Rhipsalis lindbergiana (1995)
Serra do Cumanati (Pernambuco) Rhipsalis floccosa (1969) — this brejo may have been destroyed,
since it does not figure in Rodal et al. (1998)
Mata Grande (Alagoas) Brasiliopuntia brasiliensis (1974)
Serra de São José, Feira de Santana Pereskia aculeata, Brasiliopuntia brasiliensis, Rhipsalis floccosa,
(Bahia) R. lindbergiana & Melocactus ernestii (all 2001)
Rui Barbosa (Bahia) Pereskia aculeata (1973), Brasiliopuntia brasiliensis (1978)
Monte Verde, Itaberaba (Bahia) Pereskia aculeata (1973)
Serra da Jibóia, Santa Teresinha (Bahia) Rhipsalis baccifera subsp. hileiabaiana, R. crispata & Melocactus
ernestii (2002)
Três Braços, Ubaíra (Bahia) Lepismium cruciforme & Rhipsalis floccosa (1994)
Jaguaquara (Bahia) Pereskia aculeata, Brasiliopuntia brasiliensis, Rhipsalis russellii
(all 1915) — brejo assumed to have been destroyed
Venceslau Guimarães (Bahia) Hatiora cylindrica (1993) — in protected area?
Almadina/Floresta Azul (Bahia) Rhipsalis russellii (before 1966), R. baccifera subsp. hileiabaiana
(before 1966), R. cereuscula (1972)
Barra do Choça (Bahia) Rhipsalis floccosa, R. paradoxa subsp. septentrionalis,
R. lindbergiana, Hatiora cylindrica (all 2003)
Juçari/Jussari (Bahia) Rhipsalis paradoxa subsp. septentrionalis (before 1966)
Camacã/Camacan (Bahia) Lepismium cruciforme (2003), Rhipsalis russellii (1971), R. oblonga
(1971), R. paradoxa subsp. septentrionalis (before 1979)
CONSERVATION STATUS
6) including notable cacti, with dates of most recent collections/observations.

Unfortunately, it is likely that many of these are now much altered or destroyed. Those
that survive merit regular monitoring (Rodal et al. 1998). Also part of the Brazilian
Nordeste, is the Archipelago of Fernando de Noronha, a Federal Environment Protection
area. These Atlantic islands are home to an endemic cactus, Cereus insularis (a close relative
of the Brazilian coastal C. fernambucensis), which seems adequately protected at present.
The locations of reserves that potentially assist the conservation of threatened cacti in
Eastern Brazil are indicated on Map 15, which also marks the sites of rare and threatened
or narrowly endemic taxa that are currently without protected areas.
4.3 CRITERIA FOR THE IDENTIFICATION OF

PRIORITY TAXA
The internationally agreed process for assessing the level of threat of extinction to living
organisms is that established by IUCN, whose system aims, as far as is possible, to provide
an objective means of determining Categories of Threat, or ‘Red List Categories’ (IUCN
2001). The principal categories employed in the IUCN system are ‘Extinct (EX)’,
‘Extinct in the Wild (EW)’, ‘Critically Endangered (CR)’, ‘Endangered (EN)’,
‘Vulnerable (VU)’ and ‘Least Concern (LC)’. The series of standard criteria that lead to
their determination should be stated when an assessment is published, in order that the
category can be easily verified or reassessed in the future, as circumstances change. The
Categories CR, EN & VU are defined by any one of 5 sets of criteria, which are of a
consistent type throughout, but differ in degree for each Category.
It is a relief to be able to report that the present study has not identified any Cactaceae
native of Eastern Brazil that belong in either the ‘Extinct’ or ‘Extinct in the Wild’
categories. However, there is no shortage of taxa whose assessments place them in the
remaining categories. Besides the categories for threatened taxa listed above, there is also
‘Near Threatened (NT)’ for taxa that are close to qualifying as VU and ‘Data Deficient
(DD)’ where there is inadequate information to make a direct or indirect assessment of
risk of extinction based on geographical distribution and/or population status. This last
subcategory applies to 10 out of the 162 taxa covered in the present study, while a further
3 present taxonomic difficulties and together with the few natural hybrids have not been
assessed, being referred to as ‘Not Evaluated (NE)’. The assessments made in this study
are recorded under each taxon in the taxonomic inventory (Chapter 5) and are based on
the total range of the taxon, whether or not it is endemic to Eastern Brazil as defined here.
They were communicated to IUCN, via the Species Survival Commission’s Cactus &
Succulent Specialist Group, in August 2001 (some have been revised here).
Farjon & Page criteria. While the standard methodology embodied in the published
IUCN system has been employed here, it has recently been noted by Farjon & Page
(1999) that its application can result in rather long lists of equally threatened taxa. This
suggests that further prioritisation is desirable if the limited resources currently applied to
nature conservation are not to be spread too thinly. Farjon & Page (1999: 28) have
devised a novel additional formula for achieving such prioritisation, which has proved
RED LISTING
eminently capable of being applied here. This formula calculates a score based on the
Category of Threat, where arbitrarily assigned values of CR=4, EN=3, VU=2 and
LC/NT=1, are multiplied by the sum of 3 other criteria, namely ‘Phylogenetic
Distinction’, ‘Ecological Importance’ and ‘Genetic Diversity’, whose ranges of values are
given below. The resulting total score effectively short-lists and prioritises between taxa
with the same IUCN rating, as can be seen for the cacti of Eastern Brazil in Table 4.2.
The Farjon & Page criteria are defined as follows:
Phylogenetic distinction (PD). This is a measure of the relative taxonomic isolation
of an organism and recognises that not all taxa of the same rank are equal. Thus, if the
taxon is representative (a) of a monotypic genus, it scores 4; (b) of a species or infraspecific
rank of a small genus (2–5 species) or monotypic infrageneric rank within any genus, it
scores 3; (c) of a species of a larger genus (>5 species*), it scores 2; and (d) of an
infraspecific rank of a species of a larger genus, it scores 1.
Ecological importance (EI). This was devised with forest conifers in mind, but can be
applied equally to cacti, which are also woody and frequently take the place, or are an
important component of arborescent vegetation in dryland ecosystems. Here, if a taxon is
co-dominant in a distinct vegetation type, it scores 2; if it is only a more minor constituent
of the vegetation, it scores 1. (Farjon & Page, l.c., give a score of 3 to what they call
‘keystone species of a biotic community’, but none of the cacti treated here seems to fall
into this category, although some of the arborescent cereoids arguably come close and
those found elsewhere would definitely qualify, eg. some Mexican Pachycereeae.)
Genetic diversity (GD). This criterion distinguishes between taxa known to be
unusually rich in regional diversity, often expressed in high levels of morphological
variation or varied ecological adaptation, from those whose genetic diversity is less. This
is the most subjective of the Farjon & Page criteria, at least if based only on morphological
and ecological assessments, because it is known from modern studies of conservation
genetics that taxa which appear to be variable may have low levels of measurable genetic
diversity and vice versa. However, since it will be many years before the population
genetics of the cacti of Eastern Brazil have been properly investigated in the laboratory,
use of this criterion remains justifiable as the best option available. Taxa displaying genetic
diversity within and/or between regional populations receive a score of 2 (eg.
Uebelmannia gummifera and Espostoopsis dybowskii), less diverse taxa scoring 1 (eg. Melocactus
violaceus subsp. ritteri) — see Chapter 5 for details in each case.
The short-list formula is, therefore, calculated as follows: IUCN category of threat
(score 1–4) × (PD [score 1–4] + EI [score 1–2] + GD [score 1–2]); see Table 4.2 (below).
More sophisticated uses of phylogenetics and population genetics for assessing
conservation priorities are discussed by Linder (1995), in relation to the Southern African
orchid genus, Herschelia, but such methods require a considerably more detailed
knowledge of relationships and biology/ecology than is currently available for most
Brazilian cacti.
* Farjon & Page, l.c., divide small and larger genera on the basis of ‘2–5’ versus ‘>6’ species, leaving genera with exactly 6 species
in limbo. The authors are grateful to Aljos Farjon (pers. comm., 30.11.99) for clarifying that the stated ‘>6’ should be read as
>5 species.
RED LISTING
Table 4.2
‘League-table’ of prioritised Red List categories for all taxa assessed as
CR, EN or VU.
These have short-list scores ranging from 24 to 6. The 65 taxa below represent 40% of the 162 Cactaceae
native to Eastern Brazil considered in this exercise (including 3 that were ‘Not Evaluated’). The full range
of scores was from 24 down to 3 and it is important to stress that many of the remaining 94 taxa rated as
‘Data Deficient’, ‘Near Threatened’ or ‘Least Concern’ also merit conservation action, including
taxonomic field studies and monitoring. Key. † = Convention on International Trade in Endangered
Species (CITES): trade in wild-collected CITES Appendix I taxa and their derivatives is prohibited, while
those in App. II and artificially propagated App. I taxa may be traded under export/import licences issued
by national management authorities.
Red List Categories prioritised by IUCN Red List Area References to existing and
conservation short-list scores (see Category [criteria]/ code proposed protected areas and
Chapter 5): priority order/taxon CITES App.† (Chapter recommendations
name/(Brazilian state code) 3.3)
CRITICALLY ENDANGERED (CR)
24
1. Melocactus deinacanthus (BA) CR [B1ab(iii) + (3b:iii) A cluster of small sites is
2ab(iii)]/App. I recommended for
protection here (see below).
20
2. Cipocereus pusilliflorus (MG) CR [D]/App. II (3c:iii) Taylor in Oldfield (1997): 144,
prop. 105(e).
2. Melocactus glaucescens (BA) CR [B1ab(iii) + (3c:ii) Offered protection by the Parque

2ab(iii)]/App. I Estadual do Morro do Chapéu.
Protected status for the new
populations discovered in
2000–03 in Mun. Morro do
Chapéu should be sought.
Needs monitoring!
2. Uebelmannia buiningii (MG) CR [C2a]/App. I (4c:ii:c) Costa et al. (1998): 62, prop.
C7; Taylor in Oldfield (1997):
144, prop. 105(l).
Monitoring essential.
16
3. Pilosocereus azulensis (MG) CR [D]/App. II (3d:ii) Taylor in Oldfield (1997): 143,
prop. 105(a).
3. Micranthocereus streckeri (BA) CR [B1ab(iii) + (3c:ii) Taylor in Oldfield (1997): 144,

2ab(iii); C2a(ii); prop. 105(c).
D]/App. II
3. Coleocephalocereus purpureus (MG) CR [C2b]/App. II (3d:ii) Taylor in Oldfield (1997): 143,

prop. 105(a).
RED LISTING
3. Melocactus conoideus (BA) CR [B1ab(iii,iv,v) + (3d:i:b) Adequate protection for a

2ab(iii,iv,v)]/App. I recently designated site is
needed (see below).
Re-introduction is planned.
12
4. Pilosocereus fulvilanatus subsp. CR [B1ab(iii) + (4c:ii:b) Taylor in Oldfield (1997): 144,
rosae (MG) 2ab(iii)]/App. II prop. 105(j).
4. Melocactus pachyacanthus subsp. CR [B2ab(i,ii,iii, (3b:ii) Protected status for the
viridis (BA) iv); C2a(ii); D]/ populations discovered in
App. II 2002–03 in Mun. Morro do
Chapéu should be sought
(see below).
4. M. violaceus subsp. ritteri (BA) CR [B2ab(iii,v)]/ (3d:i:a) Sites are recommended for
App. II protection here (see below).
ENDANGERED (EN)
24
1. Espostoopsis dybowskii (BA) EN [B2ab(iii)]/ (3d:i:a) Taylor in Oldfield (1997): 143,
App. II prop. 105(b). A second site is
recommended for
protection here (see below).
18
2. Cipocereus minensis subsp. EN [B2ab(iii)]/ (4c:iii) North-eastern part of range
minensis (MG) App. II included in the Parque Nacional
Serra do Cipó.
2. Arrojadoa dinae subsp. EN [B2ab(iii)]/ (3c:iii) Costa et al. (1998): 64, prop. N22;
eriocaulis (BA/MG) App. II Taylor in Oldfield (1997): 144,
props 105(f/h).
2. Arthrocereus glaziovii (MG) EN [B1ab(i,ii,iii, (4c:iii) Costa et al. (1998): 63, prop. C10.
iv,v) + 2ab(i,ii,iii,
iv,v)]/App. II
15
3. Brasilicereus markgrafii (MG) EN [B1ab(iii) + (4c:ii:a) Protected in the Parque Estadual
2ab(iii)]/App. II Serra do Barão, Grão Mogol.
3. Cipocereus laniflorus (MG) EN [D]/App. II (4c:iii) Adequately protected in the
RPPN do Caraça, Catas Altas.
3. C. bradei (MG) EN [B1ab(iii) + (4c:ii:b) Costa et al. (1998): 63, prop. C6;
2ab(iii)]/App. II Taylor in Oldfield (1997): 144,
prop. 105(i).
3. Melocactus azureus (BA) EN [B1ab(i,ii,iii, iv) (3b:ii) Conservation ex situ may be the
+ 2ab(i,ii,iii, iv)]/ only viable option unless
App. II population discovered in 2002
can be adequately protected.
3. Discocactus bahiensis (BA/PE?/ EN [B2ab(i,ii,iii, (3a:ii) A site is recommended for
PI/CE) iv,v)]/App. I protection here (see below).
RED LISTING
ENDANGERED (EN) (continued)
12
4. Schlumbergera kautskyi (ES) EN [B2ab(iii)]/ (2d:ii) Sites are recommended for
App. II protection here (see below).
4. Pilosocereus aurisetus subsp. EN [B1ab(iii) + (4c:ii:b) Costa et al. (1998): 63, prop. C6;
aurilanatus (MG) 2ab(iii)]/App. II Taylor in Oldfield (1997): 144,
prop. 105(i).
4. Micranthocereus auriazureus (MG) EN [B1ab(iii) + (4c:ii:a) Protected in the Parque Estadual

2ab(iii)]/App. II Serra do Barão, Grão Mogol.
4. M. polyanthus (BA) EN [B2ab(iii)]/ (3c:iii) Taylor in Oldfield (1997): 144,

App. II prop. 105(d).
4. Coleocephalocereus fluminensis EN [B1ab(iii) + (2d:i) Costa et al. (1998): 67, prop. E7?
subsp. decumbens (MG) 2ab(iii)]/App. II
4. Melocactus ferreophilus (BA) EN (B1/2c–e)/ (3b:ii) Protected in the APA Gruta dos
App. II Brejões.
4. Melocactus pachyacanthus subsp. EN [B1ab(i,ii,iii, (3b:ii) Protected in the APA Gruta dos
pachyacanthus (BA) iv)]/App. II Brejões, Plate 1.2.
4. M. paucispinus (BA) EN [B2ab(v)]/ (3c:i) Taylor in Oldfield (1997): 144,

App. I prop. 105(c). Now protected on
a private reserve at Morro do
Chapéu.
4. Discocactus pseudoinsignis (MG) EN [B1ab(iii) + (4c:ii:a) Protected in the Parque Estadual

2ab(iii)]/App. I Serra do Barão, Grão Mogol.
4. D. horstii (MG) EN [B1ab(iii) + (4c:ii:a) Protected in the Parque Estadual

2ab(iii)]/App. I Serra do Barão, Grão Mogol.
9
5. Rhipsalis paradoxa subsp. EN [B2ab(iii)]/ (2a) Costa et al. (1998): 67, prop. C12?
septentrionalis (PE/BA/MG/ES) App. II
VULNERABLE (VU)
14
1. Facheiroa cephaliomelana subsp. VU [B1ab(iii)]/ (3b:iii) Taylor in Oldfield (1997):
cephaliomelana (BA/MG) App. II 143–144, prop. 105(c)(ii).
12
2. Pseudoacanthocereus brasiliensis VU [C1 + 2a(i)]/ (3d:i:a) Taylor in Oldfield (1997): 143,
(BA/MG) App. II prop. 105(a). May be offered
some protection in the APA do
Lago de Pedra do Cavalo, Bahia.
2. Arrojadoa bahiensis (BA) VU [D2]/App. II (3c:ii) Probably not a high priority for
further protection.
2. A. dinae subsp. dinae (BA/MG) VU [B1ab(iii)]/ (3c:iii) Taylor in Oldfield (1997): 14

App. II 4, prop. 105(e).
RED LISTING
2. Uebelmannia gummifera (MG) VU [B1ab(iii,v)]/ (4c:ii:c) Some populations may be offered

App. I protection inside the Parque
Estadual da Serra Negra.
2. U. pectinifera subsp. pectinifera VU [B1ab(iii,v)]/ (4c:ii:c) Costa et al. (1998): 62, prop. C7;
(MG) App. I Taylor in Oldfield (1997): 144,
prop. 105(k).
2. U. pectinifera subsp. flavispina VU [B1ab(iii,v)]/ (4c:ii:c) Costa et al. (1998): 62, prop. C7;
(MG) App. I Taylor in Oldfield (1997): 144,
prop. 105(k).
2. U. pectinifera subsp. horrida VU [D2]/App. I (4c:ii:b) Costa et al. (1998): 62, prop. C7.
(MG) Monitoring essential.
10
3. Tacinga werneri (BA/MG) VU [B2ab(ii,iii,iv, (3d:i:b) Taylor in Oldfield (1997): 143,
v)]/App. II prop. 105(a/b).
3. Rhipsalis russellii (BA/MG/ES) VU [B2ab(iii,iv)]/ (1b) Part of range incl. within Parque
App. II Nacional Chapada Diamantina
(nr Mucugê, BA) and, possibly,
Parque Nac. de Monte Pascoal
etc. (S Bahia).
3. R. pilocarpa (MG/ES) VU [B2ab(i,ii,iii, (2d:ii) See site recommended for
iv)]/App. II Schlumbergera kautskyi here.
3. Cipocereus crassisepalus (MG) VU [B1ab(iii) + (4c:ii:c) Some populations may be offered
2ab(iii)]/App. II protection inside the Parque
Estadual da Serra Negra.
3. Facheiroa cephaliomelana subsp. VU [D2]/App. II (3b:iii) Taylor in Oldfield (1997): 143,
estevesii (BA) prop. 105(c): (i).
3. Arthrocereus melanurus subsp. VU [B1ab(iii)]/ (4c:iii) Costa et al. (1998): 64, prop. S9.
melanurus (MG) App. II
3. A. melanurus subsp. odorus (MG) VU [D2]/App. II (4c:iii) Part of range incl. in Parque
Nacional Serra do Cipó.
3. A. rondonianus (MG) VU [D2]/App. II (4c:ii:b) Costa et al. (1998): 63, prop. C6;
Taylor in Oldfield (1997): 144,
prop. 105(i).
3. Discocactus placentiformis (MG) VU [B2ab(iii,v)]/ (4c:i) Costa et al. (1998): 62, prop. C7;
App. I Taylor in Oldfield (1997): 144,
prop. 105(g, i–k).
8
4. Pereskia aureiflora (BA/MG) VU [A2c + 3c]/ (3a:iii:c) Taylor in Oldfield (1997): 143,
App. II prop. 105(a).
4. Tacinga braunii (MG) VU [B1ab(iii) + (3d:ii) Taylor in Oldfield (1997): 143,
2ab(iii)]/App. II prop. 105(a).
4. Rhipsalis crispata (PE/BA) VU [B2ab(iii)]/ (2a) Field studies are required to
App. II determine if it is still extant
in PE and monitoring in BA.
RED LISTING
VULNERABLE (VU) score 8 (continued)
4. R. cereoides (ES) VU [B2ab(iii)]/ (2d:ii) See site recommended for

App. II Schlumbergera kautskyi here.
4. Hatiora cylindrica (BA/ES?) [VU B2ab(iii)]/ (2a) In theory protected at some of

App. II its Bahian localities, but these are
likely to remain under threat.
4. Cereus mirabella (MA/BA/MG) VU [B2ab(iii)]/ (4a) Known populations require

App. II regular monitoring.
4. Pilosocereus brasiliensis subsp. VU [B2ab(iii)]/ (2d:ii) Sites are recommended for

brasiliensis (ES/RJ) App. II protection here (see below).
4. P. fulvilanatus subsp. fulvilanatus VU [B1ab(iii)]/ (4c:ii:a) Protected in the Parque Estadual

(MG) App. II Serra do Barão, Grão Mogol.
4. Micranthocereus violaciflorus (MG) VU [D2]/App. II (3c:iii) Protected in the Parque Estadual

Serra do Barão, Grão Mogol.
4. Coleocephalocereus buxbaumianus VU [B2ab(iii)]/ (2d:i) A site is recommended for

subsp. flavisetus (MG) App. II protection here (see below).
4. Melocactus violaceus subsp. violaceus VU [A3c]/App. II (1b) Recorded from within 3

(RN/PB/PE/SE/BA/MG/ES) small protected areas (PB/
ES/RJ), but others are needed
to maintain genetic diversity.
4. M. violaceus subsp. margaritaceus VU [A3c]/App. II (2c) Protected in the Serra de

(AL/SE/BA) Itabaiana reserve (IBAMA),
Sergipe and Res. Ecol. Dunas de
Abaete, Bahia.
Additional reserves needed.
4. Discocactus zehntneri subsp. VU [D2]/App. I (3c:ii) Offered protection by the Parque

boomianus (BA) Estadual do Morro do Chapéu.
4. D. heptacanthus subsp. catingicola VU [B2ab(iii)]/ (4b) Needs regular monitoring.

(PI/BA/MG) App. I
6
5. Tacinga saxatilis subsp. estevesii VU [D2]/App. II (3b:iii) Taylor in Oldfield (1997): 143,
(BA) prop. 105(c): (i).
5. Rhipsalis baccifera subsp. VU [B2ab(i,ii,iii, (1b) The possibility for its

hileiabaiana (BA) iv)]/App. II protection inside the area of
CEPLAC (Ilhéus, BA) should
be investigated.
5. Pilosocereus floccosus subsp. VU [B2ab(iii)]/ (3d:ii) Taylor in Oldfield (1997): 143,

quadricostatus (MG) App. II prop. 105(a).
CONSERVATION HOTSPOTS
4.4 CONSERVATION HOTSPOTS

The above short-listing process can also be used to identify geographical areas of high
priority for attention by conservationists, these nowadays commonly being referred to as
‘hotspots’ (Reid 1998, Myers et al. 2000). Here (Table 4.3), the short-list scores of the 65
threatened taxa in Table 4.2 have been summed according to the geographical areas they
are characteristic of (Chapter 3), giving a prioritised list of these area subdivisions, which
can be analysed at different levels. This indicates that of the First Order subdivisions, the
South-eastern campos rupestres (4c) are by far the most important for attention by
conservationists and land managers, having a score nearly twice that of any of the others.
This subdivision is followed by 3 with broadly similar scores: 2nd the Northern campos
rupestres (3c), 3rd the Eastern caatingas-agrestes/campos rupestres (3d) and 4th the Caatingas of
the Rio São Francisco (3b). In 5th and last place is the Southern subhumid/humid forests
and inselbergs (2d), which is part of the Atlantic Forest (Area 2), whose present state of
destruction is so great that all remaining areas merit immediate preservation.
Unfortunately, cactus diversity can only play a small role in the biological and other
arguments in support of this obvious need.
Table 4.3
‘League-table’ of principal conservation area hot-spots based on a
summation of the scores presented in Table 4.2. See also Map 16.
Priority hot-spots (sum of taxon scores from Table 4.2) Endemics: no. of taxa × score
A). Second order Area subdivisions

1 (71). SE campos rupestres (MG): Southern part (4c:iii) 2×18, 1×15, 2×10
2 (70). Northern campos rupestres: N Serra do Espinhaço (3c:iii) 1×20, 1×18, 2×12, 1×8
3 (66). SE campos rupestres (MG): Diamantina & E (4c:ii:c) 1×20, 3×12, 1×10
4 (61). SE campos rupestres (MG): Serra do Cabral etc. (4c:ii:b) 1×15, 3×12, 1×10
5 (59). SE campos rupestres (MG): Grão Mogol etc. (4c:ii:a) 1×15, 3×12, 1×8
6 (56). N campos rupestres: Chapada Diamantina, BA (3c:ii) 1×20, 1×16, 1×12, 1×8
7 (54). Rio São Francisco caatingas: southern caatingas (3b:iii) 1×24, 1×14, 1×10, 1×6
8 (51). Rio São Francisco caatingas: cent.-northern Bahia (3b:ii) 1×15, 3×12
9 (48). Eastern caatingas-agrestes/campos rupestres (3d:i:a) 1×24, 2×12
10 (46). South-eastern caatingas-agrestes (3d:ii) 2×16, 1×8, 1×6
B). First order Area subdivisions

1 (267). South-eastern campos rupestres, Minas Gerais (4c) 1×20, 2×18, 3×15, 9×12, 5×10, 1×8
2 (138). Northern campos rupestres, Bahia & Minas Gerais (3c) 2×20, 1×18, 1×16, 4×12, 2×8
3 (120). Eastern caatingas-agrestes/campos rupestres (3d) 1×24, 3×16, 2×12, 1×10, 1×8, 1×6
4 (105). Rio São Francisco caatingas, Bahia & Minas Gerais (3b) 1×24, 1×15, 1×14, 3×12, 1×10, 1×6
5 (58). Southern subhumid/humid forest/inselbergs, MG/ES (2d) 2×12, 1×10, 3×8
CONSERVATION HOTSPOTS
At the level of Second Order subdivisions (Table 4.3: A), the top 5 are all contiguous
campo rupestre areas, followed by the campos rupestres of the Chapada Diamantina (6th) and
central-southern caatingas (7th –10th). Thus, in terms of conservation importance for cacti,
the campos rupestres stand well in front of the caatingas, even though the threats to the
former environments are generally of lesser significance (see Map 16). Nevertheless, all
subdivisions include various taxa of key conservation importance.
4.5 PRIORITY ACTIONS RECOMMENDED

Since the late 1970s when eminent Brazilian botanists, such as Dárdano de Andrade-Lima
(1981) and Nanuza de Menezes (many verbal presentations), drew attention to the levels
of modification and threats being suffered by the caatingas and campos rupestres, only
relatively few and mostly very small protected areas have been created in the more than
one million square kilometres of land surface in which these major ecosystems are
represented in Eastern Brazil. The various federal and state reserves offering protection to
Cactaceae have been mentioned already, but many more are needed (eg. see Map 15), as
has been indicated by recent symposia and studies conducted by international,
governmental and non-governmental interests (Oldfield 1997, Costa et al. 1998). The
most encouraging development of recent times is the evolution of various legal
instruments that offer Brazilian municipalities and citizens opportunities for the setting
aside for conservation reasons of natural areas they oversee or own (cf. Costa et al. 1998).
In addition to these, Brazilian cactus enthusiasts, such as Marlon Machado, are persuading
municipal authorities and private land owners to establish local protected areas and take
pride in the rare taxa and unusual habitats that they were probably unaware of until very
recently. Such developments are to be greatly applauded and, none more so, than the
remarkable achievement represented by the publication of the comprehensive report,
entitled ‘Biodiversidade em Minas Gerais’ (Costa et al. 1998). This places the collective
knowledge of conservation priorities for the state of Minas Gerais on a level that all other
Brazilian states from the area covered in this study need to emulate. It is only to be hoped
that the recommendations in this ground breaking report will be actioned by the
authorities empowered to do so, and especially, since the present study identifies by far
the greatest concentration of conservation priority taxa as endemics of Minas Gerais (see
Tables 3.1, 4.2 & 4.3, Maps 15 & 16). The same aspirations apply to the need for
protected areas and associated actions detailed in the IUCN-SSC Cactus & Succulents
Action Plan (Oldfield 1997). The following additional actions are recommended here:
• Establishment of local reserves/protected areas (A1 = taxon in CITES App. I):

1. For the highest scoring taxon in the ‘Critically Endangered’ category in Table 4.2,
Melocactus deinacanthus (A1), known for certain from only a small cluster of sites within a
few kilometres north and east of the type location at ‘Morro da Barriguda, Juá’, Mun.
Bom Jesus da Lapa (at the border of Mun. Riacho de Santana), southern Bahia, to protect
it from agricultural activities including depredation by livestock and collecting.
CONSERVATION PRIORITY ACTIONS
2. Similarly, for the highest priority taxon in the ‘Endangered’ category in Table 4.2,
Espostoopsis dybowskii, in the Município of Jaguarari, northern Bahia, to complement that
recommended for its morphologically distinct southern populations (Oldfield 1997: 143)
and to protect it from urban expansion.
3. Designation as a reserve of the area known as ‘Lagedo Bordado’ and its surroundings
(course of the Rio Salitre, N of Brejão, W of Icó, Mun. Morro do Chapéu, Bahia) for
recently discovered, contiguous populations of the Critically Endangered Melocactus
pachyacanthus subsp. viridis and M. glaucescens (A1). Also, expansion of the APA Gruta dos
Brejões to include an adjacent population of the former taxon discovered in 2003.
4. A reserve for Melocactus conoideus (A1), above Mun. Vitória da Conquista, Bahia, has
recently been designated, but needs better enforcement, since habitat destruction caused
by gravel extraction and urban expansion continue. Establishment of a secure area will be
essential if the re-introduction programme planned for this species by staff from the Univ.
Estadual Sudoeste da Bahia is to succeed.
5. For Discocactus bahiensis (A1), in an area of former flood plain of the Rio São Francisco,
near Rodeadouro, WSW of Juazeiro, northern Bahia. More than half of this population
was recently destroyed by the construction of an embanked asphalt road, but healthy
plants remain at the western edge of the former site dominated by bushes of jurema preta,
Mimosa tenuiflora.
6. For Schlumbergera kautskyi, at mountain sites (inselbergs) in Mun. Domingos Martins
(Pico da Pedra Azul) and Mun. Alfredo Chaves (São Bento de Urânia), Espírito Santo,
from which region a range of other rare Rhipsalideae is also reported: Rhipsalis pacheco-
leonis subsp. catenulata, R. cereoides, R. sulcata, R. burchellii, R. pilocarpa and, perhaps, R.
hoelleri. This region is being developed as a high cost residential and tourism area,
alongside the appearance of plantations of the dreaded Eucalyptus.
7. For Melocactus paucispinus (A1), at its sites around the Pico das Almas, Mun. Rio de
Contas/Érico Cardoso (Água Quente), Chapada Diamantina, Bahia. (It is already receiving
protection at Morro do Chapéu.) Inclusion of the Pico das Almas itself would also protect
a population of the remarkable Arrojadoa bahiensis.
8. For Melocactus violaceus subsp. ritteri, at its two sites near to the towns of Jacobina and
Rui Barbosa, Bahia, to protect it from urban expansion and general habitat disturbance,
as well as from collectors.
9. For any surviving western populations of Coleocephalocereus buxbaumianus subsp.
flavisetus, such as that recently found at Carmo da Mata, south-western Minas Gerais (see
Chapter 5).
10. For Pilosocereus brasiliensis subsp. brasiliensis along the coast at Guarapari and
southwards (Espírito Santo), which could also protect one or more populations of the
Vulnerable Melocactus violaceus subsp. violaceus.
• Establishment of ex situ gene banks to enable future re-introductions, if Extinction in the Wild
should occur:
1. Specifically for Melocactus azureus, whose known habitats are in imminent peril of
destruction and hold wild populations that are highly fragmented or numbering only tens
CONSERVATION PRIORITY ACTIONS
of individuals. It is desirable that projects for the cold storage of seeds in one or more seed
banks (Dickie et al. 1990, Yang 1999), supplemented by ex situ gene banks of living plants
under controlled conditions of pollination (to avoid hybridization), are established.
However, this should not detract from any efforts to conserve wild populations, if
sufficiently intact habitats can be protected, such as that located by Aloísio Cardoso in 2002.
2. The above actions are equally applicable to many of the taxa discussed in this chapter,
as well as others in Chapter 5, where it is difficult to identify secure protected areas, or
where there is a severe risk of catastrophic reduction in genetic diversity. In the latter case,
seed banking may provide an effective insurance, if collecting programmes are carefully
designed.
• Monitoring of wild populations

Regular monitoring of the status in the wild of many of the taxa documented in Chapter
5 is essential, whether or not these are currently rated as threatened (eg. those listed in
Table 4.1, above). Many taxa presently assessed as ‘Least Concern’ or ‘Near Threatened’
are likely to become ‘Vulnerable’ in due course and could become yet more seriously
threatened shortly thereafter, if timely conservation actions are not triggered. In
particular, populations of Uebelmannia pectinifera and U. buiningii need regular investigation
to assess the ongoing impact of collecting for the horticultural trade. The latter species is
already ‘Critically Endangered’ and the former could easily pass from ‘Vulnerable’ to
‘Endangered’ if a few more of its sites are ransacked. In addition to monitoring known
populations there is also the need for field studies of those taxa that are apparently
restricted to only one or two small sites in the hope that new populations can be located
and protected before their habitats are irreversibly modified. Examples include a number
of the taxa listed in Table 4.2: Rhipsalis crispata, Melocactus deinacanthus, Cipocereus
pusilliflorus, Micranthocereus streckeri, Pilosocereus azulensis, P. fulvilanatus subsp. rosae,
Coleocephalocereus purpureus and C. fluminensis subsp. decumbens. Discovery of one or two
additional localities for any of these could make a considerable difference to our
perception of their conservation status. While some may be truly restricted in range, we
know that significant extensions are possible when similar habitats are located. Thus, only
recently the range of the very rare Micranthocereus violaciflorus was considerably extended,
from a single small site in the region of Grão Mogol to a new one at Serranópolis.
Locating a third site for this taxon might well remove it from the ‘Vulnerable’ to the
‘Least Concern’ category of risk. Likewise, the discovery between 2000 and 2003 of four
additional sites for the ‘Critically Endangered’ Melocactus glaucescens, some 40 and 80 km
from its previously known area, increases our optimism that it can be saved.
TAXONOMIC INVENTORY
5. TAXONOMIC INVENTORY
5.1 INTRODUCTORY NOTES

As its scope this account attempts to include all taxa of Cactaceae that are native,
naturalized or commonly cultivated outdoors in North-eastern Brazil, and in South-
eastern Brazil north of 22°S and east of 46°W (entries for non-native taxa are indicated
by an asterisk [*] below). However, the focus from the start has been on the terrestrial,
rather than epiphytic taxa, the former having a centre of high endemism in the dry
environments comprising south-eastern Piauí, southernmost Ceará, the southern half of
Paraíba, Pernambuco, Bahia, Alagoas, Sergipe, Espírito Santo, most of Minas Gerais and
the northern half of Rio de Janeiro (Map 1, box). While epiphytic taxa native within this
area are fully treated below, their main centre of diversity in SE Brazil, around 22–26°S,
has been intentionally excluded, since a preliminary account of the all-important tribe
Rhipsalideae was published quite recently (Barthlott & Taylor 1995).
Except where expressly limited, synonymy aims to be comprehensive for names based
on Brazilian types, but all relevant botanical nomenclature is included in the ‘Index to
botanical names and epithets’ (Chapter 7.2). Also included for some accepted names and
synonyms are references to good illustrations. Vernacular names are listed from various
sources, eg. Menezes (1949), and as reported to the authors by ‘sertanejos’. The
morphological descriptions do not aim in principle to account for all aspects of the plant,
but are meant to be primarily diagnostic, supplementing details in the dichotomous keys to
facilitate identification. In some cases they are comprehensive where the data have been
available or when the differences between taxa are particularly subtle. Details mentioned in
descriptions of genera, subgenera or other infrageneric categories, which should be assumed
to apply to all subordinate taxa, are not always repeated in descriptions of the included
species etc. Unqualified measurements refer to length (or height in the case of erect plants);
those connected by a multiplication sign (×) refer to length followed by width/diameter
and, occasionally, also thickness, in the case of laterally compressed structures. Dimensions
of the whole plant or its stem are always given exclusive of the spines; and distances between
structures that themselves have a size, eg. between areoles on a rib, are given from organ
centre to centre. Specialized terms are defined in the Glossary (Chapter 7.1).
Besides nomenclatural and diagnostic/descriptive details, bibliographical,
phytogeographical and ecological data (see Chapter 3) are provided, likewise distribution
records and conservation assessments (IUCN 2001 Categories of Threat and conservation
short-listing values, including Phylogenetic Distinction [PD], Ecological Importance [EI]
and Genetic Diversity [GD] — see Chapter 4 and cf. Golding & Smith 2001). Where
KEY TO GENERA
appropriate, commentary is also provided on some historical aspects, other taxonomic

treatments, relationships, uses and the need for further studies.
All materials cited have been seen unless indicated by ‘n.v.’ (non vidi). The cited records
are organized by state and município (abbreviated to ‘Mun.’), and mostly arranged from
north to south and west to east. A minimum of at least one record per município has been
included and, in the case of many taxa, all available records are cited, since Brazilian
Cactaceae remain in general very poorly collected. For three taxa that are widespread,
frequent and taxonomically well-defined within Eastern Brazil — Tacinga inamoena,
Epiphyllum phyllanthus and Cereus jamacaru subsp. jamacaru — only a few representative
collections/observations or records marking the limits of distribution etc. have been cited,
since their abundance and range are otherwise not in doubt. Apart from herbarium
exsiccata, whose locations are indicated by the standard Index Herbariorum codes, other field
observations by the authors (or from reliable sources, even if unsupported by vouchers) are
mentioned in some cases; likewise published and unpublished illustrations and photographs
of plants in localized habitats. A bracketed questionmark (?) preceding a record’s entry
indicates uncertainty about the collection’s identity, whereas after the name of the município
it indicates doubt about the correct municipal attribution of the locality cited.
5.2 ARTIFICIAL KEY TO GENERA

1. Actively growing stems bearing broad or awl-shaped leaves (Pereskia &
Opuntioideae) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
1. Actively growing stems leafless or with only minute scale-like leaf
primordia (Cactoideae) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
2. Spines not microscopically barbed, not becoming strongly attached if allowed
to penetrate the skin; glochids lacking; seed with blackish testa visible . 1. Pereskia
2. Spines microscopically barbed, very difficult to detach if allowed to penetrate
the skin, or true spines lacking and glochids present (at least on older stem-
segments or the trunk); seed encased in a pale and sometimes fibrous/hairy
funicular envelope (Opuntioideae) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
3. Leaves broad, like those of Pereskia (SW Bahia & Cent.-N Minas Gerais) . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2. Quiabentia
3. Leaves awl-shaped (widespread) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
4. Tree to 4 m or more with dimorphic stems, comprising a cylindric
indeterminate leader shoot and flattened determinate lateral segments,
the ultimate thin and leaf-like; seed 8–10 mm . . . . . . . . . . . . . 4. Brasiliopuntia
4. Trees to < 4 m, or shrubs, subshrubs or lianas; stems not as above;
seed to c. 5 mm . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
5. Stamens and perianth spreading, the former sensitive, closing around
the style when touched, the latter patent but never strongly reflexed,
at least partly yellow; pollen exine reticulate . . . . . . . . . . . . . . . . . . . 6. Opuntia
KEY TO GENERA
5. Stamens erect, clustered around the style, not sensitive; perianth erect,
spreading or strongly reflexed, greenish, deep pinkish, red or purplish,
or orange-yellow and plants not exceeding 50 cm; pollen exine not as above . . . 6
6. Stamens long-exserted and perianth-segments erect, not spreading; plants
spineless (introduced and cultivated) . . . . . . . . . . . . . . . . . . . . . . . .*5. Nopalea
6. Stamens not as above or perianth-segments spreading to strongly reflexed
and/or stems spiny . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3. Tacinga
7. Flower > 10 cm long, or 8–10 cm and tube bract-scales bearing hairs
(hair-spines) in their axils . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
7. Flowers < 8 cm long, or 8–10 cm and tube bract-scales naked in their axils
or lacking or minute . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
8. Fruit yellow, globose, surface weakly ribbed, > 5 cm diam., falling to the
ground, smelling of pineapple and spineless when ripe; seed light brown when
fresh, brown when old (N & E Bahia & NE Minas Gerais: caatinga-agreste) . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10. Pseudoacanthocereus
8. Fruit not as above; seed blackish . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
9. Stems flat, trigonous or 3-winged . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
9. Stems with > 3 wings/ribs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
10. Pericarpel and flower-tube bearing conspicuous spines/bristles or broad-based
bract-scales; perianth > 15 cm diam. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
10. Pericarpel and flower-tube with inconspicuous, narrow-based bract-scales,
minute spines and/or trichomes only, or expanded perianth < 15 cm diam. . . . 12
11. Stems trigonous or 3-winged . . . . . . . . . . . . . . . . . . . . . . . . . . . 7. Hylocereus
11. Stems flattened . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .*8. Selenicereus
12. Epiphyte or garden plant with mostly flattened stems . . . . . . . . . 9. Epiphyllum
12. Terrestrial with mostly trigonous or 3-winged stems . . . . . . . . . . 16. Cereus p.p.
13. Perianth remains cleanly abscissing from young fruit or persistent and
strongly blackened . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16. Cereus p.p.
13. Perianth remains not as above (blackening only if attacked by fungi) . . . . . . . . . 14
14. Fruit > 5 cm, dehiscent; seeds 3–4 mm (Bahia northwards: caatinga) . 24. Harrisia
14. Fruit 1–5 cm, indehiscent; seeds 1–2 mm (Minas Gerais: campo rupestre) . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28. Arthrocereus
15. Epiphytic, or epilithic on coastal rocks, in Mata atlântica or at altitudes of
>1500 m, with flattened or 3–5-winged/angled stem-segments or stems
<2 cm diam., often only slightly succulent; flowers < 3 cm long or if
larger then magenta and zygomorphic (tribe Rhipsalideae) . . . . . . . . . . . . . . . . 16
15. Terrestrial, or epilithic in caatinga/campo rupestre and/or stems and flowers not
as above . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
KEY TO GENERA
16. Flowers ± zygomorphic, tube to 8 mm long or more (S Espírito Santo &

S Minas Gerais) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14. Schlumbergera
16. Flowers actinomorphic, tube 0 or < 3 mm long . . . . . . . . . . . . . . . . . . . . . . . . 17
17. New stem-segments (excluding greatly elongated, usually basal, extension
shoots) arising mostly 2 or more together from the apices of older segments
(branching acrotonic), old and diseased stems separating from the plant at
the joints between segments, deciduous . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
17. New stem-segments arising only at base or singly from the sides of older
segments (branching basi-/mesotonic), not deciduous when old or diseased . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11. Lepismium
18. Flowers whitish or not strongly coloured, or developed laterally on stem-
segments of ± indeterminate pendent growth . . . . . . . . . . . . . . . . 12. Rhipsalis
18. Flowers bright yellow or orange, from composite areoles at the apex of
ultimate or penultimate segments of strictly determinate ± erect growth . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13. Hatiora
19. Plant unbranched, never segmented, shortly columnar, ± globose or
depressed-globose, with or without a non-chlorophyllous bristly-woolly
terminal cephalium . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
19. Plant branched (at least basally), slender cylindric to tall columnar, or stems
solitary with lateral cephalia, segmented with cephalia at the joints or bottle-
shaped and the juvenile stem tapering into an elongate chlorophyllous
terminal cephalium . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
20. Flowers yellowish, rarely greenish, diurnal (central Minas Gerais) . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30. Uebelmannia
20. Flowers not as above (widespread) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
21. Flowers to 4 cm long, tubular, deep magenta-pink to red at least without,
diurnal to crepuscular . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
21. Flowers > 4 cm long, salverform, whitish, nocturnal . . . . . . . . 29. Discocactus
22. Flowers from a cephalium . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23. Melocactus
22. Flowers from the stem apex, cephalium lacking (Chapada Diamantina,
Bahia: A. bahiensis) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19. Arrojadoa p.p.
23. Pericarpel and tube clothed in conspicuous bract-scales and/or wool and
bristle-spines . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
23. Pericarpel and tube ± naked, bract-scales lacking, minute or very widely spaced . . 27
24. Fruit with red or purplish pulp when ripe, spiny at first; stems to 2.5 cm
diam. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25. Leocereus
24. Fruit with white or translucent pulp, spineless, or if spiny then stem
>2.5 cm diam. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
25. Pericarpel and tube bearing areoles and bristle-spines; fruit covered in an
intensely blue waxy bloom (Cent.-S Minas Gerais) . . . . . . . 17. Cipocereus p.p.
KEY TO GENERA
25. Pericarpel and tube with bract-scales but lacking bristle-spines; fruit not
as above . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26
26. Flowers shortly funnelform, at least 4 cm diam. at full anthesis; seed 2–3 mm . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15. Brasilicereus
26. Flowers tubular, to 2.5 cm diam. at full anthesis; seed to c. 1.5 mm . . 26. Facheiroa
27. Fruit depressed-globose (rarely globose), 2–6 cm diam., bursting open
laterally or apically due to pressure from the expanding funicular pulp;
stems never regularly segmented (widespread) . . . . . . . . . . . . . 20. Pilosocereus
27. Fruit not as above, opening via a basal pore or < 2 cm diam.; stems various . . . . 28
28. Stems segmented, with bristly/woolly ring cephalia at the joints and apex
of distal stem-segments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
28. Stems not as above; cephalium, if present, lateral and ± continuous or
terminal and chlorophyllous . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
29. Flowers green without, 8–10 cm long; fruit >3 cm diam. (Bahia, caatinga) . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . 18. Stephanocereus subg. Stephanocereus
29. Flowers deep pink to bright red without, < 4 cm long; fruit < 3 cm
diam. (widespread) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19. Arrojadoa p.p.
30. Mature plant bottle-shaped, the upper part narrowed into a terminal
chlorophyllous cephalium (Bahia, Chapada Diamantina) . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18. Stephanocereus subg. Lagenopsis
30. Mature plant not as above . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
31. Flower-bearing areoles not differing markedly from those on purely
vegetative stems and/or fruit covered in blue wax (Minas Gerais) . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17. Cipocereus p.p.
31. Flower-bearing areoles ± modified or comprising a lateral cephalium;
fruit not as above . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
32. Fruit clavate, sometimes laterally compressed, > 11 mm diam., deep pink to
red, with a small basal pore allowing ants to enter (plants of naked gneiss/
granite inselbergs of SE Brazil; C. goebelianus also on other rocks and in
stony soil of the caatinga in cent.-E to S Bahia) . . . . . . 22. Coleocephalocereus
32. Fruit depressed-globose, globose or very shortly clavate, variously coloured
or whitish, or < 11 mm diam., not opening at base . . . . . . . . . . . . . . . . . . . . . 33
33. Stem tissues almost lacking mucilage; shrub branched at base and above,
not glaucous; perianth-segments white inside and out (caatinga to the
N & E of the Chapada Diamantina, Bahia) . . . . . . . . . . . . . . . 27. Espostoopsis
33. Stem tissues highly mucilaginous and at least the outermost perianth-
segments coloured (Minas Gerais & Bahia: campo rupestre), or plant a
glaucous column to > 3 m (W of Rio São Francisco, N Minas
Gerais & W Bahia: Bambuí limestone) . . . . . . . . . . . . . . . 21. Micranthocereus
PERESKIA
5.3 PERESKIOIDEAE K. Schum.

This subfamily comprises only one genus following the removal of Maihuenia (S Chile &
S Argentina: Patagonia) to the Maihuenioideae Fearn (cf. Wallace 1995). The
Pereskioideae are distinguished from the 2 broad-leaved genera of Opuntioideae by their
unbarbed spines and unspecialized seeds lacking a funicular envelope (aril).
1. PERESKIA Mill.
Gard. dict. abr. ed. 4 (1754). Type: Pereskia aculeata Mill.

Including Rhodocactus Backeb. & F. M. Knuth (1936).
Literature: Leuenberger (1986), Wallace (1995: 9, Fig. 10).
Leafy and spiny trees, shrubs or scramblers, 1–20 m or more. Areoles in the axils of deciduous leaves, sparsely
to densely tomentose, always with felt and sometimes with long hairs, producing spines and sometimes
brachyblast leaves. Leaves alternate, broad, exstipulate, somewhat fleshy, deciduous. Spines solitary or
clustered, or paired, claw-like and recurved (see species No. 1), sometimes lacking on the flowering stems
but increasing in number on the trunk. Flowers solitary or in paniculate to cymose inflorescences,
sometimes developing by proliferation from the flower receptacle; flowers 2–7 cm diam., perigynous or
epigynous, receptacle smooth or with prominent podaria and areoles, bract-scales leafy, fleshy, green or
sometimes tinged like the sepaloids; perianth multiseriate, segments free, sepaloids lacking axillary areoles
and partly tinged like the petaloids, these spreading, rotate or forming a campanulate-urceolate corolla;
stamens numerous; filaments shorter than the perianth; ovary superior to inferior, unilocular, with septal
ridges at the ovary roof only, or rarely septate; stigma-lobes 3–20, erect to spreading. Fruit solitary or
clustered-proliferous, pear-shaped, turbinate or globular, terete or angled, with a narrow to broad umbilicus
and persistent or deciduous flower remnants or leafy bract-scales; pericarp mucilaginous; locule without
funicular pulp (in the Brazilian species). Seeds few to numerous, obovate to lenticular-reniform, 1.8–7.5
mm, black, smooth, shiny; hilum whitish.
A genus widespread in the neotropics with 17 species, of which 5 are native to Eastern
Brazil (3 species and one heterotypic subspecies are endemic). Two further species are
reported from western and southern Brazil, P. sacharosa (Mato Grosso & Mato Grosso do
Sul) and P. nemorosa (Rio Grande do Sul), and are related to Nos 2–4 from Eastern Brazil.
The E Brazilian taxa are restricted to various phases of the Mata atlântica, agreste and
Southern caatingas. Plates 2–4.1.
1. Scrambling or climbing plant, spines on vigorous shoots paired, recurved; perianth white
or cream, flower strongly scented (diosmin) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1. aculeata
1. Erect shrubs or trees, spines always straight, spreading, never paired-recurved; perianth not
as above . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
2. Flower bright yellow; fruit globose, 1–3-seeded . . . . . . . . . . . . . . . . . . . . . . . . . . . 5. aureiflora
2. Flower pink, magenta, orange or red; fruit turbinate, with > 3 seeds . . . . . . . . . . . . . . . . . . . . . . 3
3. Leaves narrowly elliptic to obovate-lanceolate; lateral veins (7–)10–13; seeds (5–)6–7 mm
(Mata atlântica, sens. lat., including brejo forest) . . . . . . . . . . . . . . . . . . . . . . . . . . . 2. grandifolia
3. Leaves ovate to broadly elliptic-obovate; lateral veins 5–7; seeds 4–5.5 mm (caatinga) . . . . . . . . . . 4
4. Flowers with campanulate to urceolate perianth; flower-buds orange; perianth-segments
reddish pink, erect, recurving at apex only, stamens and style enclosed . . . . . . . . . . 4. stenantha
PERESKIA
4. Flowers with rotate, widely opening perianth; flower-buds greenish or

pinkish; perianth-segments pink or magenta-pink, spreading; stamens and
style not enclosed . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3. bahiensis
Three types of sclereids are found in the genus (Leuenberger 1986: fig. 22) and each is
represented amongst the species treated here, dividing them into three groups as
indicated below:
PERESKIA ACULEATA Group (No. 1): sclereids fusiform-simple; stomata present

on stem; periderm formation early; brachyblast leaves 0.
1. Pereskia aculeata Mill., Gard. dict. ed. 8 (1768). Lectotype (Leuenberger 1986):
Dillenius, Hort. eltham.: tab. 227, fig. 294 (1732).
Cactus pereskia L., Sp. pl. 1: 467 (1753). Pereskia pereskia (L.) H. Karst., Deut. Fl.: 888 (1882), nom. inval.
(Arts 23.4 & 32.1(b)). Type: as above.
vernacular names. Ora-pro-nobis, Azedinha, Lobolôbô, Espinho-de-Santo-Antônio, Espinho-preto,

Surucucú, Cipó-Santo.
Climbing shrub or liana, stems 3–10 m long; extension shoots scandent, sometimes ascending up trees
to heights of 20–30 metres and trunk becoming 10 cm thick near the ground. Areoles initially 2 mm
diam., with white long hairs, later accrescent, cushion-like, to 15 mm diam., producing claw-like,
geminate primary spines and secondary straight spines, but not brachyblast leaves. Leaves lanceolate to
oblong or ovate, 4.5–7.0(–11.0) × 1.5–5.0 cm, shortly petiolate, base cuneate, attenuate or rounded;
blade green, concolorous or purplish below; lateral veins 4–7, often inconspicuous. Inflorescence
terminal and lateral on long shoots, racemose to profusely paniculate, with up to 70 flowers of more.
Flowers 2.5–5.0 cm diam., white or creamy-white, strongly fragrant (diosmin); pedicels 5–15 mm;
receptacle cup-shaped to turbinate, with 6–15 hairy areoles on inconspicuous to prominent podaria;
bracts leaf-like, fleshy, lower bracts 9–12, spreading to recurved, upper bracts 3–8, erect or adpressed
in bud; outer perianth-segments 2–5, pale green to whitish; inner segments 6–11, obovate to
spathulate, to 2.5 cm, delicate, white; stamens 5–10 mm, filaments white, yellow or orange-red; ovary
not distinctly delimited from the style, which is 10–12(–19) × 1–2 mm, thickening above, stigma-lobes
4–7, erect or suberect. Infructescences large or fruits solitary, pedicellate, 1.5–2.5 cm, globular, yellow
to orange, pericarp fleshy, usually bracteate and spiny when immature, mostly naked when ripe; locule
with gelatinous tissue enclosing the seeds. Seeds 2–5, lenticular, 4.5–5.0 mm diam., laterally
compressed. Chromosome number: 2n = 22 (Leuenberger 1986: 43).
Humid/subhumid evergreen forest element: scrambling over vegetation and inselbergs of gneiss/granite
etc., Mata atlântica, including restinga, mata de brejo and agreste (mata de cipó), rarely in ‘carrasco’ (near the
Rio São Francisco, cent.-N Minas Gerais) and caatinga (SE Bahia), Maranhão, eastern Pernambuco to
central-eastern and south-eastern Bahia, eastern and southern Minas Gerais and Espírito Santo, from sea
level to > 1000 m; Goiás, South-eastern and Southern Brazil; eastern Paraguay and Argentina, Mexico,
Central America, Caribbean and northern South America (once recorded from Peru). Maps 2 & 17.
maranhão: Mun. Bom Jardim [‘Santa Luzia’], ‘Faz. Agripec. da VARIG, margem esquerda do Rio
Pindaré’, 2 Apr. 1983, M.G.A. Lobo et al. 337 (MO).
pernambuco: Mun. Paulista [‘Olinda’], ‘restinga do Rio Doce’, 5 Jan. 1948, Leal 41 (RB); Mun. São
Lourenço da Mata, Engenho São Bento, Mata do Camocim (fide Andrade-Lima 1966: 1454); Mun.
Recife, Bonji, 3 Oct. 1968, I. Pontual 68/878 (PEUFR); Mun. Poção, road to Jataúba, 21 Oct. 1991,
F.A.R. Santos 2 (HUEFS, PEUFR); Mun. Bonito, 1 km before limit with Camocim de São Félix, 10
Feb. 1967, Andrade-Lima 67-4944 (SP, IPA).
PERESKIA
alagoas: Mun. União dos Palmares, ‘Faz. Santo Antônio, mata ao N da sede’, 27 Nov. 1966, Andrade-
Lima 66-4769 (IPA).
sergipe: Mun. Frei Paulo, 8 Apr. 1986, G. Viana 1386 (ASE 4249); Mun. Simão Dias, 5 Feb. 1975,
A.C. Barreto s.n. (ASE 295); Mun. São Cristóvão, ‘Faz. Rio Comprido, 5 km após entrada para Rita
Cacete – S. Cristo’, 25 Mar. 1982, G. Viana 384 (IPA, ASE).
bahia: Mun. Feira de Santana, Faz. Boa Vista, Serra de São José, 12°15'S, 38°58'W, 10 May 1984,
Noblick 3182 (CEPEC, K, HUEFS); Barragem de Bananeiras, Rio Paraguaçu and Rio Jacuípe, 12°32'S,
39°5'W, Apr. 1980, Iscardino et al. in Grupo Pedra do Cavalo 4 (ALCB, HUEFS); Mun. Cruz das Almas,
30 May 1981, G.C.P. Pinto 83 (HRB); Mun. Salvador, 26 May 1915, Rose & Russell 19639 (NY, US);
Mun. Rui Barbosa, Serra do Orobó, 1 July 1973, G.C.P. Pinto s.n. (ALCB); Mun. Itaberaba, Faz.
Monte Verde, Nov. 1973, A.L. Costa (ALCB); Mun. Jaguaquara [Toca da Onça], 27–29 June 1915,
Rose & Russell 20069 (NY, US); Mun. Aiquara, 2 km N of town, 13 May 1969, Jesus 375 & T.S. dos
Santos 424 (CEPEC, K); Mun. Aurelino Leal, between Itajuípe and Ubaitaba, 12 km N of Banco
Central, 24 Apr. 1965, Belém & Magalhães 877 (CEPEC); Mun. Uruçuca, 20 May 1994, W.W. Thomas
et al. 10407 (CEPEC, MO, HUEFS); Mun. Ilhéus, 14°48'S, 39°10'W, 27 July 1995, L.A. Mattos-Silva
3153 (HUEFS); l.c., Km 22 from Ilhéus on road to Itabuna, CEPLAC, 25 Apr. 1983, Leuenberger & Brito
3103 (CEPEC, K, MO); l.c., Zehntner s.n. (US); Mun. Vitória da Conquista, 7 km NE on road BR 116,
then 3.5 km W, caatinga, 16 April 2003, Taylor & Zappi (obs.); l.c. 12 km from town centre towards
Itambé, 17 April 2003, Taylor (K, photos); Mun. (?), between Prado and Alcobaça, 16 Apr. 1967, Lanna
1390 & Castellanos 26413 (CEPEC, K, HB); Mun. Nova Viçosa, 10 Apr. 1984, Hatschbach 47785
(CEPEC, HRB, MBM).
minas gerais: Mun. Itacarambi, Vale do Peruaçu, near the entrance to Fazenda Terra Brava, 15°6'45"S,
44°15'36''W, carrasco, 14 Feb. 1998, A. Salino & A. Gotschalg 3997 (K, BHCB); Mun. Caraí, 12 km N of
Catuji, 13 Dec. 1990, Taylor & Zappi 757 (K, HRCB, ZSS, BHCB); Mun. Teófilo Otoni, Pedro
Versiani, Aug. 1958, Mendes Magalhães 17397 (IAN); Mun. Galiléia, 23 km SE from São Vítor, 15 Dec.
1990, Taylor & Zappi 772 (K, HRCB, ZSS, BHCB); Mun. Aimorés, Terreno do Varejão, 7 July 1997,
M.F. de Vasconcelos s.n. (K, BHCB); Mun. Santana do Pirapama, 18 Feb. 1971, Krieger 10145 (CESJ);
Mun. Pedro Leopoldo, 11 Feb. 1995, M.L. Guedes 5042 (ALCB); Mun. Caratinga, road BR 458,
between Ipatinga and Iapu, 12 km from the BR 381, 13 Dec. 1990, Taylor & Zappi 753 (K, HRCB,
ZSS, BHCB); Mun. Itabira, Cauê, 11 Feb. 1934, Sampaio 7069 (R); Mun. Arcos, Serra do Minério, July
1983, E. Esteves Pereira 170 (UFG); Mun. Viçosa, 19 Dec. 1929 & 27 June 1930, Mexia 4129, 4804 (K,
NY, U, US, MO); Mun. São João del Rei, Apr. 1896, Silveira 943 (R).
espírito santo: Mun. Nova Venécia, 3 km de Todos os Santos, 18°37'S, 40°43'W, 8 Sep. 1989,
H.Q.B. Fernandes 2823 (MBML, HRCB); Mun. São Gabriel da Palha, 30 km from Barra de São
Francisco on road to Águia Branca, 16 Dec. 1990, Taylor & Zappi 785 (K, HRCB, ZSS, MBML); Mun.
Linhares, Reserva Florestal da CVRD, Est. Flamengo, 5 Apr. 1986, Folli 589 (Univ. Fed. ES); l.c.,
Reserva Biológica de Comboios, IBDF, Regência, near Fazenda Cata-Vento, 22 May 1989, Folli 916
(Univ. Fed. ES); l.c., road BR 101 20 km S of town, 7 Apr. 1984, Hatschbach 47706 (MBM); Mun.
Colatina, 30 Apr. 1934, Kuhlmann 290 (RB, MO); Mun. Santa Teresa, Rio Cinco de Novembro, 22
Sep. 1988, H.Q.B. Fernandes et al. 2550 (MBML, HRCB), 13 Mar. 1980, Weinberg 502 (MBML); Mun.
Fundão, Três Barras, 1 Aug. 1984, Pizziolo 220 (MBML, HRCB); Mun. Vila Velha, between Lagoa do
Milho and Barra do Jucu, 20 July 1973, Araújo 369 (RB); l.c., Barra do Jucu, 15 Mar. 1981, Weinberg 499
(MBML); l.c., Interlagos II, 27 July 1982, Weinberg 262 (R); Mun. Guarapari, Setiba, 25 Apr. 1989, O.J.
Pereira 1959 (Univ. Fed. ES), 1 May 1981, Krieger & Souza 19564 (CESJ); Mun. Itapemirim, Bello 472
(R); l.c., road ES 060, between Marataízes and Pontal, 10 May 1987, H.C. de Lima et al. 2917 (K, MO).
rio de janeiro: NE Rio de Janeiro, Mun. Campos, Apr. 1939, Sampaio s.n. (R).
conservation status. Least Concern [LC] (1); PD=2, EI=1, GD=2. Short-list score (1×5) = 5.
This taxon is variable in leaf size and shape and in its inflorescence morphology, both
within and between populations. Its leaves and those of the following (No. 2) are eaten
as a potherb.
PERESKIA
PERESKIA GRANDIFOLIA Group (Nos 2–4): sclereids fusiform-aggregated;

stomata on stem present; periderm formation retarded; brachyblast leaves present.
Fusiform-aggregated sclereids are also found in P. nemorosa and P. sacharosa (SE South
America, see Map 12), which are indicated as potentially basal members of this Group in
molecular phylogenies produced by R. Wallace (1995) and C. Butterworth (in litt., 10
Feb. 2000) at Iowa State University, USA.
2. Pereskia grandifolia Haw., Rev. pl. succ.: 85 (1819). Neotype (Leuenberger

1986): ‘raised from seed collected by Bowie & Cunningham in 1816 in the
neighbourhood of Rio de Janeiro’, drawing by T. Duncanson, 11 June 1824 (K).
Cactus grandifolius (Haw.) Link, Enum. pl. hort. Berol. 2: 25 (1822). Rhodocactus grandifolius (Haw.) F. M.
Knuth in Backeberg & Knuth, Kaktus-ABC: 97 (1935, publ. 1936).
Shrub or tree, 2–10 m, forming a trunk to 80 cm diam.; main shoots erect to arching. Areoles rounded,
cushion-like, initially 3–7(–8) mm diam., later to 12 mm diam., producing straight spines, and 1–4
brachyblast leaves. Spines 0–8(–11) on twigs, fasciculate to spreading, 1–4 cm, increasing in number and
size on older branches. Leaves elliptic or narrowly elliptic-oblong, narrowly ovate to obovate-
lanceolate, (6–)9–26(–30) × (3–)4–6(–9) cm, petiole to 1 cm or more; base of lamina attenuate, apex
acute to acuminate, sometimes recurved; lateral veins (7–)10–13 or more. Inflorescence terminal,
densely cymose-paniculate by proliferation from the receptacle, 10–15- to sometimes 30(–50)-flowered.
Flowers 3–6(–7) cm diam., rose-like, showy; pedicels 1–3 cm, stout; receptacle turbinate, with
conspicuous furrows and podaria, areoles on the upper half only; bracts small or leaf-like, fleshy, lower
bracts 4–7, spreading to ascending, upper bracts 2–5, adpressed in bud; outer perianth-segments 2–5,
asymmetric; inner segments 5–12, obovate to spathulate, delicate, rose, pale or purple-pink, apex often
emarginate; stamens 5–10 mm, filaments white; ovary distinctly inferior, with conical roof narrowing
into the style, which is 8–12 mm, cylindric, stigma-lobes 5–8, suberect, white to pale pink.
Infructescences large, usually pendent; fruit 5–10 × 3–7 cm, pyriform or turbinate, with conspicuous
angles, green or reddish to yellowish, bracts usually deciduous when ripe. Seeds 20–60, obovate-elliptic,
5–7 × 4.5–5.0 mm, laterally compressed.
conservation status. Least Concern [LC]; see subspecies below.
This species is divisible into two subspecies:
1. Receptacular bracts green, the lowermost ones rarely with recurved apices; outer
perianth-segments greenish to pink, inner segments 15–32 mm; anthers golden
yellow (NE & SE Brazil) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2a. subsp. grandifolia
1. Receptacular bracts and outer perianth-segments purplish-pink to dark purplish, the
lowermost bracts with recurved apices; inner perianth-segments 10–18 mm; anthers
pale yellow (cent.-S Minas Gerais to W Espírito Santo, cultivated in N Minas Gerais
& S Bahia) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2b. subsp. violacea
2a. subsp. grandifolia
vernacular names. Sabonete, Ora-pro-nobis, Quiabento, Entrada-de-baile, Rosa-mole, Sem-vergonha.
Leaves thin, green, concolorous, shortly acuminate to acute. Inflorescence and receptacular bracts green,
flat, rarely recurved; receptacular bracts shorter to longer than the inner perianth-segments, outer
segments greenish-pink, inner segments 15–32 mm, pink to deep pink; anthers golden yellow. Fruit
PERESKIA
variable, pyriform to obtriangular, angled; often nearly as broad as long; ovary locule cup-shaped to
broadly obovate-triangular. Chromosome number: 2n = 22 (Leuenberger 1986: 43).
Humid/subhumid evergreen forest element: perhaps native in agreste and Mata atlântica (including that on
the E slopes of the Chapada Diamantina, brejo and riverine forest), c. 100–1140 m, southernmost Ceará,
Pernambuco (native and cultivated, fide Andrade-Lima 1966: 1454), central-eastern and south-eastern
Bahia to southern Espírito Santo and south-western Minas Gerais, but widely cultivated; perhaps also
native in Rio de Janeiro and São Paulo; commonly introduced in the neotropics. Map 17.
maranhão: Mun. Barra do Corda, [cult.?], 20 July 1909, Arrojado Lisboa 2467 (MG).
ceará: S Ceará, Chapada do Araripe, 30 Aug. 1971, Gifford & Fonseca G 335 (E, K, photo).
pernambuco: Mun. Olinda, cult., 12 Feb. 1925, Pickel 518 (SP); Mun. São Lourenço da Mata, Engenho
São Bento, cult., Pickel 1911 (IPA, fide Andrade-Lima 1966: 1454); Mun. Recife, Dois Irmãos, 3 Oct. 1968,
I. Pontual 68/879 (PEUFR); Mun. São Benedito do Sul, 9 Feb. 1994, A.M. Miranda 1286 (PEUFR).
bahia: Mun. Rodelas, Volta do Rio, 27 Jan. 1987, L.B. Silva & G.O. Mattos e Silva 98 (HRB); Mun.
Juazeiro, cult., 11 Feb. 1990, Zappi 221A (K, photos); Mun. Palmeiras, Morro do Pai Inácio, 12°27'37"S,
41°28'40''W, 1100–1140 m, M. Alves et al. in EBNN 1016 (PEUFR); Porto Castro Alves, vale dos Rios
Paraguaçu e Jacuípe, 10 Dec. 1980, Iscardino et al. in Grupo Pedra do Cavalo 1011 (ALCB, HRB, BAH);
Mun. Andaraí, Km 0 da rodovia que liga à BR 242, 12 Jan. 1983, L.A. Mattos Silva et al. 1592 (CEPEC,
K); l.c., Rio Paraguaçu ao longo da rodovia para Mucugê, 12°50'S, 41°19'W, 19 Oct. 1997, M. Alves et
al. 1116 (HRB); Mun. Cruz das Almas, May 1973, G.C.P. Pinto s.n. (ALCB), G.C.P. Pinto 29/81
(HRB); Mun. Salvador, 27 May 1915, Rose & Russell 19656 (NY, US); Mun. Vitória da Conquista, 20
km SE of Anajé, [cult.?], 13 Apr. 1983, Leuenberger et al. 3062 (CEPEC, K); Mun. Ilhéus, 2 Dec. 1982,
T.S. dos Santos 3827 (HUEFS); Mun. Porto Seguro, Monte Pascoal, road to ‘Caraíba’ [Caraíva], 14 May
1971, T.S. dos Santos 1641 (CEPEC, K).
minas gerais: S Minas Gerais, Mun. Juiz de Fora, 18 Oct. 1969, Krieger 7528 (CESJ), l.c., near
Paraíbuna, Apr. 1970, Krieger 8244 (SPF); SW Minas Gerais, Mun. Lavras, road BR 381, 8 km S of bridge
over Rio Grande, 10 Oct. 1988, Taylor & Zappi in Harley 24802 (SPF, K).
espírito santo: Mun. Barra de São Francisco, cult., 16 Dec. 1990, Taylor & Zappi 779A (K, photo);
Mun. Cachoeiro do Itapemirim, Bello 653 (R); unlocalized [perhaps from Campos, Rio de Janeiro],
‘Campo Novo’, [1815/1816], [Wied-]Neuwied s.n. [ex Herb. Martius anno 1869] (MO).
rio de janeiro: Sertão de Cacimbas, on right embankment of Rio Itabapoana, 10 Oct. 1909, Sampaio
985 (R); between Morro do Coco and Campos, 8 Aug. 1964, Trinta 1051 (R); Monte Alegre, Feb. 1927,
Vidal s.n. (R); Mun. Cantagalo, Peckolt 121 (W).
conservation status. Least Concern [LC] (1), but perhaps Near Threatened, since, although of very
wide range, much of its original habitat has been destroyed; PD=1, EI=1, GD=1. Short-list score (1×3)
= 3. It has been taken into cultivation as a hedge plant in the region where it may be native.
The natural range of this taxon remains poorly known, probably through early
destruction of its habitat and for the uncertainty as to its native status caused by its
widespread introduction as a cultivated ornamental.
2b. subsp. violacea (Leuenb.) N. P. Taylor & Zappi in Cact. Consensus Initiatives 3: 7
(1997). Holotype: Brazil, Minas Gerais, Mun. Santana do Riacho, Serra do Cipó,
19°21'S, 43°36'W, 23 Sep. 1981, F. C. F. Silva 89 (HRB; B, iso.).
Pereskia grandifolia var. violacea Leuenb. in Mem. New York Bot. Gard. 41: 116, fig. 46 (1986).
vernacular name. Ora-pro-nobis.
Leaves thickish, dark green above, often slightly discolorous and purple below, nearly always distinctly
acuminate and with recurved apex. Inflorescence bracts all dark purplish to bright purplish pink, uppermost
PERESKIA
almost resembling open flowers from a distance; lower receptacular bracts thick, keeled, recurved, nearly as
long as the inner perianth-segments, outer segments purplish to pink, inner segments 10–15(–18) mm, pale
pink to purple-pink; anthers pale yellow. Fruit and ovary locule narrowly obtriangular, bracts purplish
tinged with green or becoming nearly green. Chromosome number: 2n = 22 (Leuenberger 1986: 43).
Southern humid/subhumid forest element: drier phases of Mata atlântica, c. 50–1400 m, drainage of the
Rio Doce, central-southern and eastern Minas Gerais to central Espírito Santo, apparently cultivated
elsewhere. Endemic to the core area within South-eastern Brazil. Map 17.
bahia: Mun. Vitória da Conquista, cult., 13 Apr. 1983, Leuenberger et al. 3060 (CEPEC).
minas gerais: Mun. Januária, 4 Nov. 1978, Krieger 16130 (CESJ) — cult.?; Mun. Pedra Azul, near
town, 16°1'S, 41°17'W, cult., 19 Oct. 1988, Taylor & Zappi in Harley 25218 (K, SPF); Mun. Itaobim, 8
km W of town, 6 km S of road Itaobim–Araçuaí, cult., 9 Apr. 1983, Leuenberger & Martinelli 3057
(CEPEC, RB); cent.-S & E Minas Gerais, Mun. Capelinha/Malacacheta, Uebelmann HU 226 (M); Mun.
Conselheiro Pena, 16.7 km SE of Galiléia, N bank of Rio Doce, 15 Dec. 1990, Taylor & Zappi 773 (K,
HRCB, ZSS, BHCB); ibid., 2 km N of Penha do Norte, regenerating from cut-over forest, 15 Dec.
1990, Taylor & Zappi (K, photo); Mun. Santana do Riacho, Serra do Cipó, 19°21'S, 43°36'W, 23 Sep.
1981, F.C.F. Silva 89 (RB), 25 Mar. 1991, Pirani et al. in CFSC 12042 (SPF, K); Mun. Itambé do Mato
Dentro, 24 Jan. 1922, G. Santos s.n. (R); Mun. Aimorés, Terreno do Varejão, 14 Sep. 1997, M.F. de
Vasconcelos s.n. (K, BHCB); Mun. Itabira, Cauê, 11 Feb. 1934, Sampaio 7068 (R); Mun. Marliéria, Parque
Est. do Rio Doce, 19°46'S, 42°36'W, 27 Sep. 1975, Heringer & Eiten 15252 (UB, US).
espírito santo: SW of Colatina, Braun 476 (B, photo); Mun. Santa Teresa, Rio Cinco de Novembro,
cult. Mus. Bot. Mello Leitão, 17 Dec. 1990, H.B.Q. Fernandez in Taylor & Zappi 789 (HRCB, K, ZSS).
conservation status. Data Deficient [DD] (1); PD=1, EI=1, GD=1. Short-list score (1×3) = 3. It has
been taken into cultivation as a hedge plant in the region where it is native. Its native habitat continues
to decline.
Even though the native distribution of subsp. grandifolia is poorly understood, there are
no records of it as other than a cultivated plant within the extensive area in which subsp.
violacea is found, thereby justifying recognition of the latter as a subspecies rather than a
variety. Indeed, recent studies of plastid DNA gene sequences, conducted by Wallace
(1995: 9) and Butterworth (unpubl.), have indicated that this taxon may be worthy of
specific status and that it is the basal element amongst the East Brazilian taxa belonging to
the P. GRANDIFOLIA Group. Whatever the true status of subsp. grandifolia, it seems
reasonably certain that subsp. violacea is native within the area drained by the Rio Doce,
where it has been observed regenerating from stumps remaining in recently cut primary
forest (see above and Plate 2.4).
3. Pereskia bahiensis Gürke in Monatsschr. Kakt.-Kunde 18: 86 (1908). Type

(Leuenberger 1986: 119): Brazil, Bahia, ‘Calderão’ [Caldeirão, Mun. Maracás], Oct. 1906,
Ule 7050 (HBG, lecto.; L, lectopara.).
vernacular names. Quiabento (Quiá-bento), Inhabento, Jumbeba, Surucurú, Flor-de-cera, Espinho-

de-Santo-Antônio, Entrada-de-baile, Ora-pro-nobis.
Shrub to small tree, 1–6 m, forming a trunk to 33 cm diam.; main shoots erect to arching, stout.
Areoles rounded, cushion-like, initially 3–6(–8) mm diam., later to 20 mm diam., producing straight
spines, and 1–4 brachyblast leaves. Spines 0–6 on twigs, fasciculate to spreading, 2–5 cm, increasing in
number and size on older branches. Leaves elliptic or obovate, often narrowly obovate on main shoots,
flat or folded upwards and recurved, (4–)5–12 × 2–7 cm, petiole short to indistinct, base of lamina
PERESKIA
attenuate, apex rounded to broadly acute, lateral veins (3–)4–7(–8). Inflorescence terminal, cymose-
paniculate by proliferation from the receptacle, 2–12-flowered, dense to lax, or flowers sometimes
solitary. Flower perigynous, 4–7 cm diam., rose-like, showy; pedicel 1 cm, stout; receptacle turbinate,
with prominent podaria and 3–8 areoles mainly on the upper half; bracts sepal-like, fleshy, lower bracts
4–6, spreading, upper bracts 4–6, erect, green with purplish to nearly black margin, keeled; outer
perianth-segments 2–5, asymmetric, greenish red to reddish purple, inner segments 7–9, 20–30 mm,
obovate to spathulate, delicate, pink to reddish purple, pale at base, apex rounded-truncate to
emarginate; stamens 5–10 mm, clustering around the style, filaments white; ovary half-inferior to
inferior, with distinct septal ridges at the roof; style 7–15 mm, cylindric, stigma-lobes 5–10, suberect.
Infructescences lax, forming a cymose-paniculate cluster of up to 12 fruits, often pendent, or fruit
solitary; 3–6 × 3–5 cm, pyriform or turbinate, ± angled, green to yellow when ripe, scented like
pineapple, bracts usually deciduous. Seeds > 30, obovate, 3.8–5.5 × 2.8–3.5 mm, laterally compressed.
Chromosome number: 2n = 22 (Leuenberger 1986: 43).
Central-southern (Bahian) caatinga element: caatinga surrounding the Chapada Diamantina, planalto de
Maracás, northern Serra do Espinhaço and Serra Geral (Bahia), 300–900 m, east and south of the Rio São
Francisco. Endemic to Bahia. Map 17.
bahia: Mun. Juazeiro, Nov. 1913, Zehntner 743 (US); Mun. Jaguarari, 64 km SE of Juazeiro towards
Senhor do Bonfim, 8 Jan. 1991, Taylor et al. 1377 (K, HRCB, ZSS, CEPEC); l.c., Barrinha, 7 June 1915,
Rose et al. 19786 (NY, US), 25 Feb. 1968, I. Pontual 68-738 (IPA, PEUFR); Mun. Ourolândia, NE of
Olho D’Água do Fagundo, 4 Aug. 2002, Taylor (obs.); Mun. São Gabriel/João Dourado, APA Gruta dos
Brejões, 4 Aug. 2002, Taylor (obs.); Mun. Morro do Chapéu, NE of Gruta dos Brejões, 4 Aug. 2002,
Taylor (obs.); l.c., W of Icó, N of Brejão (Formosa), 5 Aug. 2002, Taylor (obs.); l.c., 20 km W of Morro
do Chapéu, 3 Aug. 2002, Taylor (obs.); l.c., Bambuí limestone, 20 Jan. 1984, Fotius 3707 (IPA); Mun.
Irecê, 26 Oct. 1970, A.L. Costa s.n. (IPA), 7 Oct. 1980, M.S.G. Ferreira 118 (HRB, ALCB, BAH); Mun.
América Dourada, Nova América, 17 May 1975, Costa & Barroso s.n. (ALCB); Mun. Oliveira dos
Brejinhos, 123 km W of Seabra towards Ibotirama, 14 Jan. 1991, Taylor et al. 1420 (K, HRCB, ZSS,
CEPEC); Mun. Iraquara, Faz. Boa Vista, 12°16'S, 41°38'W, 14 Mar. 1984, O.A. Salgado 365 (HRB, RB);
Mun. Iraquara/Palmeiras, 16 km from Rio Preto to Riacho do Mel, 19 Mar. 1980, Brazão 165 (HRB);
Mun. Santa Teresinha, 20 Nov. 1986, L.P. de Queiroz 1341 (HUEFS); Mun. Itaetê, 19 Oct. 1978, Almeida
28 (HRB, RB, F); Mun. Milagres, 2 km from the town, 15 Apr. 1995, E. de Melo 1204 (HUEFS); Mun.
Marcionílio Souza, Machado Portella, 19–23 June 1915, Rose & Russell 19901 (NY, US); l.c., 18 Dec.
1912, Zehntner 677 (RB, US); Mun. Maracás, towards Tamburi, 24 Jan. 1965, E. Pereira 9719 & G. Pabst
8608 (HB); l.c., Km 26 on road from Maracás to Tamburi, 20 Apr. 1983, Leuenberger et al. 3086 (CEPEC,
K); l.c., 24 Jan. 1965, Brazão 109 (HRB, RB); Mun. Rio de Contas, 19 June 1978, Araújo 47 (RB); l.c.,
caminho para Lagoa Nova, 13°48'S, 41°47'W, 5 Feb. 1997, E. Saar et al. in PCD 5106 (ALCB, K,
CEPEC, UEFS, SPF); Mun. Livramento do Brumado, 11 km S of the town, 13°45'S, 41°49'W, 23 Nov.
1988, Taylor & Zappi in Harley 25542 (K, SPF, CEPEC); Mun. Ituaçu, 13°48'S, 41°16'W, 22 June 1987,
L.P. de Queiroz 1649 (HUEFS); l.c., Morro da Mangabeira, 22 Dec. 1983, E.P. Gouveia 46/88 (ALCB);
Mun. Jequié, 4 km N of town, 28 Mar. 1965, M. Magalhães s.n. (HB 37923); l.c., 11–17 km W of town
towards Lafaiete Coutinho, 19 Nov. 1978, Mori et al. s.n. (CEPEC, K); Mun. Brumado, Serra das Éguas,
23 Sep. 1989 & 8 Feb. 1990, A.M. Miranda 41 & 120 (PEUFR); Mun. Jequié/Poções, A.L. Costa s.n.
(ALCB, in spirit); Mun. Malhada de Pedras, Km 11 on road from Brumado to Caetité, 14 Apr. 1983,
Leuenberger et al. 3063 (CEPEC, MO); Mun. Lagoa Real, 81 km W of Brumado towards Caetité, 2 Feb.
1991, Taylor et al. 1529 (K, HRCB, ZSS, CEPEC); Mun. Caetité, 14 km towards Brumado, 19 Feb. 1992,
A.M. de Carvalho et al. 3761 (K, CEPEC); l.c., Brejinho das Ametistas, cult., 15 Apr. 1983, Leuenberger et
al. 3071 (CEPEC, K); Mun. Manuel Vitorino, Serra da Pipoca, 15 Jan. 1984, G.C.P. Pinto 65/84 (HRB,
CEPEC); Mun. Boa Nova, 20 km S of Manuel Vitorino, 19 Apr. 1983, Leuenberger et al. 3083 (CEPEC,
K); Mun. Aracatu, road BR 262, 50 km from town towards Vitória da Conquista, 12 Dec. 1984, G.P.
Lewis et al. in CFCR 6716 (SPF, SP, K); Mun. (?), Vitória da Conquista to Brumado, 31 Mar. 1966,
Castellanos 25947 (HB); Mun. Cacule, 14°31'S, 42°12'W, 2001, J. Jardim 3221 (CEPEC); Mun. Poções,
Faz. Boa Esperança, road BR 116, 9 km S of Poções, 5 Apr. 1988, L.A. Mattos Silva et al. 2317 (CEPEC,
K); Mun. Jânio Quadros, 8 km SW of Maetinga, 19 Feb. 1992, S.C. Sant’Ana et al. 227 (K, CEPEC);
Mun. Caatiba, Faz. Bom Jardim, 10 Feb. 1975, A.L. Costa s.n (ALCB).
PERESKIA
conservation status. Least Concern [LC] (1); PD=2, EI=1, GD=1. Short-list score (1×4) = 4. It has
been taken into cultivation as a hedge plant in the region where it is native.
The very close relationship between this species and the following deserves further
investigation. A plant encountered near the border of municípios Piatã and Boninal,
Bahia (Taylor & Zappi in Harley 25598, SPF, CEPEC, K), well beyond the known range
of P. stenantha, had somewhat intermediate flowers, and in the region of Caitité the two
species seem to hybridize or intergrade. As already noted by Leuenberger (1986), there
are scarcely any vegetative differences to separate them, although P. stenantha seems
capable of producing much larger leaves (especially in western Bahia where it inhabits a
region of higher rainfall).
4. Pereskia stenantha F. Ritter, Kakt. Südamer. 1: 21, figs 3–4 (1979). Holotype:
Brazil, Bahia, Caitité, Ritter 1251 (U, not found). Lectotype (Leuenberger 1986: 123):
Ritter, l.c., fig. 3 (1979).
vernacular names. As for P. bahiensis.
Shrub to small tree, 2–4(–6) m, branching mostly near the base, sometimes forming a trunk to 15 cm
diam.; main shoots erect to arching. Areoles rounded, cushion-like, initially 2–5 mm diam., later to 15
mm diam., producing straight spines, and 1–3 brachyblast leaves. Spines 0–7 on twigs, fasciculate to
spreading, 1–5 cm, increasing in number and size on older branches. Leaves obovate to elliptic, often
folded upwards along the midrib, (5–)7–11(–15) × (2–)4–6(–9) cm, petiole 2–10 mm, base of lamina
cuneate to attenuate, apex broadly acute to rounded, lateral veins 5–7. Inflorescence terminal, densely
cymose-paniculate by proliferation from the receptacle, c. 15-flowered, or flowers sometimes solitary.
Flowers perigynous to epigynous, 1–2 cm diam., urceolate-campanulate, opening very little; flower-buds
orange-red; pedicels hardly distinguishable; receptacle turbinate, with 5–8 areoles, bracteate in the upper
half, bracts sepal-like, fleshy, lower bracts 3–5(–7), spreading, upper bracts (1–)3–5(–6), erect, green to
reddish, keeled, outer perianth-segments 4–5, erect, concave, red to orange-red, inner segments 6–8,
20–25 mm, linear lanceolate, erect, spreading at the tip only, pink to purplish pink, acute to broadly acute
at apex; stamens 12–15 mm, surrounding the slightly longer style, filaments white; ovary half-inferior to
inferior, with septal ridges at the roof; style 13–20 mm, thick at base, stigma-lobes 4–7, suberect to
spreading. Infructescence a densely cymose-paniculate cluster of up to 15 fruits, erect or arching at
maturity, or fruits solitary; fruit 3–7 × 2–6 cm, pyriform or turbinate, ± angled, otherwise like P. bahiensis.
Seeds c. 30(–50) per fruit, obovoid, 4.5–5.5 × 3.2–3.6 mm, laterally compressed. Chromosome number:
2n = 22 (Leuenberger 1986: 43).
Southern Rio São Francisco caatinga element: caatinga, 450–750 m, valley of the Rio São Francisco,
western and central-southern Bahia and central-northern Minas Gerais. Endemic to the core area of
Eastern Brazil. Map 17.
bahia: W Bahia, Mun. Barreiras, estrada para São Desidério, 23 Dec. 1954, G.A. Black 54-17686 (IAN);
Mun. Vanderlei, 19 July 1989, Zappi 148A (SPF); Mun. Brejolândia, 5 km N of Tabocas do Brejo Velho,
12°39'S, 44°2'W, Harley et al. 22011 (K, U, SPF, UEC); Mun. Santana, Porto Novo do Rio Corrente,
12 Nov. 1912, Zehntner 563 (US), ibid. in IFOCS 5048 (M); l.c., 28 km S of town towards Santa Maria
da Vitória, 15 Jan. 1991, Taylor et al. 1423 (K, HRCB, ZSS, CEPEC); cent.-S Bahia, Mun. Macaúbas,
Veredinha, 12°54'S, 42°38'W, 20 Mar. 1984, D.A. Salgado & H.P. Bautista 283 (HRB); Mun. Bom Jesus
da Lapa, road to Caldeirão, Faz. Imbuzeiro da Onça, 13°9'S, 43°22'W, 19 Apr. 1980, Harley 21543 (K,
SPF, CEPEC); l.c., 2 km E of town, 16 Apr. 1983, Leuenberger et al. 3076 (CEPEC, K); l.c., near Santo
Antônio, Nov. 1912, Zehntner 631 (US); ibid. (?), ‘Lapa’, Nov. 1912, Zehntner in IFOCS 5047 (M); Mun.
Botoporã, 50 km SE of Macaúbas towards Paramirim, 26 Aug. 1988, Eggli 1297 (ZSS); Mun. Bom Jesus
PERESKIA
da Lapa/Riacho de Santana, 28 Dec. 1965, A. Duarte 9531 & E. Pereira 10442 (HB, RB); Mun. Riacho
de Santana, 10 km W of Igaporã, 18 Apr. 1983, Leuenberger et al. 3081 (CEPEC, K); Mun. Igaporã, 6 km
W of town, 16 Apr. 1983, Leuenberger et al. 3072 (CEPEC, K, MO); Mun. Caetité, type locality of Ritter
(1979); l.c., 23 km S of town towards Brejinho das Ametistas, cult., 2 Feb. 1991, Taylor et al. 1539 (K,
HRCB, ZSS, CEPEC); Mun. Guanambi, 9.5 km from town towards Caetité, 25 July 1989, Taylor (obs.);
Mun. Iuiú, 2 km S of town, 21 July 1989, Taylor & Zappi (obs.).
minas gerais: Mun. Montalvânia, S of the ferry crossing of the Rio Carinhanha, 1999, Klaassen et al.
(photo); Mun. Manga, 1999, illustrated by Braun & Esteves Pereira (2002: 155, Abb. 169); Mun. Monte
Azul, 5 km E of town, E of Vila Angical, 15°10'S, 42°49'W, 9 Nov. 1988, Taylor & Zappi in Harley 25523
(K, SPF); Mun. Mato Verde, 4 km E of town towards Rio Pardo de Minas, 30 Jan. 1991, Taylor et al.
1496 (K, HRCB, ZSS, BHCB); Mun. Varzelândia, 36 km ENE of town, 23 July 1989, Taylor & Zappi
(obs.); Mun. Porteirinha, 32 km from Mato Verde, 30 Jan. 1991, Taylor et al. 1497 (K, HRCB, ZSS,
BHCB).
conservation status. Least Concern [LC] (1); PD=2, EI=1, GD=1. Short-list score (1×4) = 4. It has
been taken into cultivation as a hedge plant in the region where it is native.
Almost indistinguishable from P. bahiensis (see above) when not in flower, but with a
distinct range.
PERESKIA PORTULACIFOLIA Group (No. 5): stone cells present; stomata on

stem 0; periderm formation early; brachyblast leaves present. Other members of this
group are from northern South America (P. guamacho, sister species of P. aureiflora; see
Map 11) and the eastern Caribbean islands (4 spp.).
5. Pereskia aureiflora F. Ritter, Kakt. Südamer. 1: 22, fig. 5 (1979). Holotype: Brazil,
Minas Gerais, Itaobim, Ritter 1413 (U, not found). Lectotype (designated here): Ritter,
l.c., fig. 5 (1979).
vernacular names. Facho, Ora-pro-nobis-da-mata.
Small tree or shrub, to 6 m, trunk to 20 cm diam.; branches erect to arching. Areoles rounded,
cushion-like, initially 3 mm diam., later to 6 mm diam., producing straight spines, and 1–3 brachyblast
leaves. Spines 0–3 on twigs, fasciculate to spreading, 1–3 cm, increasing in number and size on older
branches. Leaves obovate to elliptic-suborbicular in shade, elliptic to lanceolate and thicker in sun, flat
or somewhat boat-shaped, dimorphic, auxoblast leaves to 11 × 2–3 cm, brachyblast 4–10 ×
(1.5–)2.5–5.0 cm, petiole 2–4 mm, base of lamina broadly cuneate, apex acute, lateral veins 3–5(–7).
Flowers solitary, yellow, terminal or on short lateral twigs 3–13 cm, c. 3–4 cm diam., opening widely,
showy, epigynous; pedicels short; receptacle turbinate, with areoles, bracteate; bracts leafy, fleshy,
green, sometimes reddish at margin, lower bracts 5–8, spreading, upper bracts 5–8, erect, keeled; outer
perianth-segments c. 3, obovate, inner segments 10–12, 15–20 mm, obovate, spreading to reflexed,
bright yellow, obtuse at apex; stamens 5–10 mm, filaments white; ovary half-inferior; style 10 mm,
cylindric, stigma-lobes 4–7, erect. Fruit 10–15(–20) mm diam., globular, reddish green to brown or
‘chocolate-purple’ when ripe, bracts leafy, deciduous. Seeds c. 1–3, obovoid, (4–)5–5.5 × 4.2–5.1 mm,
laterally compressed.
Southern caatinga (inselberg) element: in caatinga/agreste, especially in association with gneiss/granite

inselbergs or derived substrates, 300–920 m, central-southern Bahia to central-northern and north-eastern
Minas Gerais. Endemic to the core area of Eastern Brazil. Map 17.
bahia: Mun. Rio de Contas, 13 km E of town on road to Marcolino Moura, 13°35'S, 41°45'W, 25 Mar.
1977, Harley 19992 (K, B, U, SPF, UEC); Mun. (?), between Jequié and Poções, Oct. 1950, G. Pinto
OPUNTIOIDEAE
628 (IAN); Mun. Guanambi, 9.5 km NE of town, granite/gneiss outcrop beside road to Caitité, 25 July
1989, Zappi 171 (SPF, HRCB); Mun. Urandi (?), Km 11, Saco da Onça – Urandi, 16 Oct. 1970, D.
Andrade-Lima 70-6067 (IPA).
minas gerais: cent.-N Minas Gerais, Mun. Monte Azul, 7 km E of town, beyond Vila Angical, 920
m, 28 Jan. 1991, Taylor et al. 1470 (K, HRCB, ZSS, BHCB); NE Minas Gerais, Mun. Itaobim, 8 km W
of town, 9 Apr. 1983, Leuenberger & Martinelli 3056 (CEPEC, RB); l.c., W end of town, 16°33'S,
41°30'W, 18 Nov. 1988, Taylor & Zappi in Harley 25525 (K, ZSS, B, SPF); l.c., 4 km E of town towards
Jequitinhonha, 9 Mar. 1977, Shepherd et al. 4422 (UEC, MG); Mun. Itinga, 4 Mar. 1984, Mattos-F. &
Rizzini 1622 (RB); l.c., 21 km from road BR 116 towards Itinga, 14 Dec. 1990, Taylor & Zappi 767 (K,
HRCB, ZSS, BHCB); l.c., 15 km SW of town towards Araçuaí, Taquaral, 8 Apr. 1983, Leuenberger &
Martinelli 3054 (CEPEC, K, NY, RB, MO); Mun. Virgem da Lapa/Araçuaí, 17 Dec. 1967, A. Duarte
10534 (RB); Mun. Araçuaí, 12 Nov. 1981, G.C.P. Pinto 398/81 (HRB, RB).
conservation status. Vulnerable [VU A2c + 3c] (2); extent of occurrence = 60627 km2; PD=2,
EI=1, GD=1. Short-list score (2×4) = 8. This species appears to be rare within its extent of occurrence,
except in north-eastern Minas Gerais, in the Rio Jequitinhonha valley, where considerable habitat
modification is taking place.
In southern Bahia and central-northern Minas Gerais P. aureiflora grows in close proximity
to, or sympatric with, P. stenantha and/or P. bahiensis, but seems to be much rarer than
either. However, it is the commonest pereskia in the middle part of the Rio
Jequitinhonha drainage system, north-eastern Minas Gerais (Itaobim/Itinga), whence it
was originally described, and where its above-mentioned congeners are absent. Its status
as sister species to P. guamacho (N South America) was earlier suspected on purely
morphological grounds, but has more recently been confirmed by gene sequence
phylogenies obtained by Wallace (1995: Fig. 10) and Butterworth (unpubl.).
5.4 OPUNTIOIDEAE K. Schum.

Areoles bearing barbed spines and/or glochids; pericarpel scarcely differentiated from
stem-segments; pollen mostly polyporate; seed enclosed in a pale, usually bony funicular
envelope (aril).
Cladistic analyses employing DNA gene sequence data (Wallace & Dickie 2002) and
seed anatomical, pollen, gross morphological data and pragmatic considerations (Stuppy
& Huber 1991, Stuppy 2002; Taylor et al. 2002) indicate that c. 16 genera deserve
recognition in this subfamily. Austrocylindropuntia Backeb. and its sister group
Cumulopuntia F. Ritter (both Andean) appear to be basal amongst Opuntioideae. The
first genus treated below is the only South American representative of a tribe of 4 genera
otherwise native to Central & North America and the Caribbean, comprising Pereskiopsis
Britton & Rose, Quiabentia, Cylindropuntia (Engelm.) F. M. Knuth and Grusonia F.
Rchb. ex Britton & Rose (incl. Corynopuntia). The broad flattened leaves of Quiabentia
and Pereskiopsis may be a synapomorphy (and both are decaploids; Pinkava 2002: 61,
Table 1), since the Austrocylindropuntia-Cumulopuntia clade and the subfamily’s
plesiomorphic potential sister group, the Maihuenioideae, possess cylindric to awl-
shaped leaves.
QUIABENTIA
Tribe CYLINDROPUNTIEAE Doweld
2. QUIABENTIA Britton & Rose
Cact. 4: 252 (1923). Type: Quiabentia zehntneri (Britton & Rose) Britton & Rose.
A genus of only 2 species, the second being Q. verticillata (Vaupel) Vaupel, a sometimes
treelike plant (2–15 m high), from the western Chaco and its periphery, in Argentina,
Paraguay and Bolivia (Map 13). Its Brazilian counterpart, treated below, is of restricted and
presumably relictual distribution, representing a marginal floristic element of the caatinga,
since it occurs only near the south-western limits of this vegetation type. Plate 4.2
1. Quiabentia zehntneri (Britton & Rose) Britton & Rose, Cact. 4: 252 (1923). Type:
Brazil, Bahia, Bom Jesus da Lapa, Rio São Francisco, 15–16 Nov. 1912, Zehntner 630
(US, lecto. designated here; NY, lectopara.).
Pereskia zehntneri Britton & Rose, Cact. 1: 14 (1919).
vernacular names. Quiabento, Flor-de-cera, Espinho-de-Santo-Antônio, Cai-cai.
Shrubby, erect, 2–3 m, leafy when growing; branches, to 30 cm (or more), 1–2.5 cm diam., cylindric,
the stoutest sometimes faintly ribbed, especially towards apex when dry, grey-green to grey, glaucous;
areoles 20–40 mm apart, with white felt. Leaves 20–55 × 5–25 mm, to 3.5 mm thick, elliptic-ovate,
orbicular or elliptic-lanceolate, pointed at both ends, concave above, gradually decurved towards apex,
sessile, fleshy, stiff, green. Spines 7 or more per areole, the longest to 50 mm, white, reddish or golden,
flexible, glochids few, spine-like and poorly differentiated from the spines apart from their position at the
upper edge of the areole. Flowers apical or subapical, 7 × 7–8 cm; pericarpel narrowly turbinate to
cylindric, to 40 mm, grey-green, with areoles and fleshy bracts; tube short, to 1 cm; perianth-segments
to 40 mm, ovate or spathulate, retuse at apex, dark-pink, spreading; stamens exposed; stigma-lobes
exserted. Fruit 4–7.5 cm long, narrowly turbinate, umbilicus not very deep, greenish to reddish or purple
outside; funicular pulp translucent, whitish, placenta yellowish. Seeds many per fruit, to 4 mm diam.
Southern Rio São Francisco caatinga element: on ± naked or thinly wooded limestone (Bambuí) outcrops
or gneiss/granite inselbergs amidst high caatinga forest, 450–750 m, both sides of the Rio São Francisco
valley, west-cent./southern Bahia and cent.-northern Minas Gerais. Endemic to the core area of Eastern
Brazil. Map 18.
bahia: Mun. Bom Jesus da Lapa, Morro da Lapa, 15 & 16 Nov. 1912, Zehntner 567 & 630 (US, RB),
28 Jan. 1958, Mendes Magalhães s.n. (IAN), 17 Apr. 1983, Leuenberger et al. 3078 (CEPEC, K); l.c.,
13°15'S, 43°26'W, 15 Apr. 1980, Harley et al. 21398 (K, CEPEC); l.c., gruta do Bom Jesus, 10 June 1992,
A.M. de Carvalho et al. 3964 (K, CEPEC); l.c., W of Rio São Francisco, reported by Andrade-Lima
(1977: 191, fig. 5); Mun. Cocos, c. 5 km W of town, 2000, L. Aona, cult. UNICAMP, Aug. 2000, Zappi
(obs.); Mun. Guanambi, 9.5 km NE of town on road to Caitité, gneiss/granite outcrop S of road, 25 July
1989, Taylor & Zappi (obs.); Mun. Palmas de Monte Alto, hill with TV tower, gneiss/granite rocks in
caatinga above base of tower, 21 July 1989, Zappi 154 (SPF); Mun. Iuiú, 2 km S of town, 21 July 1989,
Zappi 155A (SPF, fr.).
minas gerais: Mun. Januária, Brejo do Amparo, Morro Itapiraçaba, 11 Nov. 1989, P.E. Nogueira Silva
et al. 125 (K); 10 km W of Januária, 15°30'S, 44°30'W, 20 Oct. 1972, Ratter et al. 2632 (UB); Mun.
Itacarambi/Manga, 15°26'S, 43°55'W, Sep. 1999, I. Ribeiro (photos); Mun. Varzelândia, c. 10 km N of
town, ‘Serra São Felipe’ 12 Aug. 1988, Eggli 1148 (ZSS); Mun. Capitão Enéas, limestone outcrop visible
from road BR 122, June 2002, M. Machado & G. Charles (K, photos).
TACINGA
Rizzini & Mattos-Filho (1992) claim that this species is found in the ‘região de
Brumado–Condeúba, Bahia’ and is there used as hedges. However, the plant commonly
used for this purpose in this region and called ‘quiabento’ is Pereskia bahiensis and it may
be doubted that Quiabentia is native there. Braun & Esteves Pereira (2002: 170) state that
Q. zehntneri is exclusively found on limestone outcrops, which is not correct (see Bahian
localities cited above).
Tribe OPUNTIEAE
The following 4 genera form a monophyletic lineage culminating in Opuntia, sens. str.,
which is the most derived, possessing sensitive stamens and pollen with a reticulate exine.
Other genera included in this lineage, but absent from Brazil, are Miqueliopuntia F. Ritter
(N Chile), Tunilla D. R. Hunt & Iliff (southern Andes) and Consolea Lem. (E Caribbean).
The following genus has a relictual distribution almost confined to Eastern Brazil, where its
extensive range does not overlap at any point with that of Opuntia sens. str.:
3. TACINGA Britton & Rose
Cact. 1: 39 (1919). Type: Tacinga funalis Britton & Rose.

Literature: Taylor et al. (2002).
Shrubs, subshrubs and lianas, 0.2–5.0 m; branches not dimorphic, pith often ± chambered or quite
hollow when > 1–2 years old; stem-segments cylindric or compressed, then orbicular, obovate to elliptic
or rhomboid in outline; areoles borne on non-existent to very low tubercles in the axils of caducous
leaves, with abundant glochids and felt, spines present or absent. Leaves minute, subulate, sessile, fleshy,
early deciduous. Flowers solitary, from the margin or apex of the stem-segments; flowers epigynous,
pericarpel globose, turbinate or elongate, sometimes deeply depressed and forming a tube at apex, with
areoles subtended by leaf-like, fleshy, green or coloured bract-scales, often with active extra-floral
nectaries before and during anthesis (attracting ants); perianth multiseriate, tube relatively short, outer
perianth-segments short, stiff, erect, patent or strongly reflexed, coloured, fleshy, inner segments
coloured, delicate, erect, somewhat spreading or strongly reflexed; stamens numerous, erect, at least at
first, sometimes strongly exserted, not sensitive, those adjacent to the perianth represented by hairlike
staminodes in species 1–3; pollen exine not reticulate, very finely punctate/spinulate. Fruit solitary or
clustered-proliferous, globose, turbinate or elongate clavate with a very deep umbilicus, flower remnants
deciduous; funicular pulp translucent or opaque and coloured, fibrous or almost lacking; placenta white,
greenish or coloured. Seeds few, globular to reniform, to 5 mm diam., enclosed in a hard, bony
funicular envelope.
As recently amplified, a genus of 6 species named as an anagram of caatinga, the dry thorn
forest of Eastern Brazil, where it is frequent, also ascending into the included campos rupestres
(see No. 6) and extending slightly west of the core area into north-western Minas Gerais
on limestone outcrops (No. 5). Species Nos 1–4 are endemic to the core area of Eastern
Brazil as defined here. Hybrids between Nos 2 & 3, 3 & 6, 4 & 6 and 5 & 6 are known or
suspected and sometimes not uncommon, but are not keyed out below. Plates 4.3–8.1.
TACINGA
1. Perianth-segments erect and forming a tube or only spreading slightly at apex, deep
magenta-pink to orange-red . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
1. Perianth-segments ± spreading to strongly reflexed and lying against the pericarpel,
yellow, orange-reddish, green or purplish . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
2. Stamens exserted; style strongly thickened near base (cultivated) . . . . . *Nopalea cochenillifera
2. Stamens included; style not as above (eastern caatingas) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
3. Fruit greenish to reddish or purple outside, to 3 cm, funicular pulp yellowish . . . . 4. palmadora
3. Fruit greenish white, sometimes with faint pink shades, 4–5.5 cm, funicular pulp
bright pink . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3. werneri
4. Plants low-growing, rarely exceeding 1 m, mostly subshrubs; stem-segments always
flattened; spines 0 or fine and slender, to 15 mm (widespread, especially on rocks) . . . . . . . . . . . 5
4. Plants taller, to 2 m or more, scandent, with at least the lower parts of stems perfectly
cylindric; spines 0 (NE Minas Gerais to SW Pernambuco, caatinga) . . . . . . . . . . . . . . . . . . . . . . . 6
5. Stem-segments spineless at maturity; areoles well spaced, 10–20 mm apart . . . . . . . . 6. inamoena
5. Stem-segments with small, persistent spines; areoles congested, 1–14 mm apart . . . . . . 5. saxatilis
6. All stem-segments perfectly cylindric; flowers green to purplish (S Bahia to SW
Pernambuco) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1. funalis
6. Ultimate stem-segments ± flattened; flowers green (Rio Jequitinhonha valley,
NE Minas Gerais) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2. braunii
1. Tacinga funalis Britton & Rose, Cact. 1: 39–40 (1919). Type (Taylor et al. 2002):
Brazil, Bahia, Juazeiro, 1915, Rose & Russell 19723 (US, lecto.; NY, lectopara.).
? T. luetzelburgii Küpper ex Luetzelb., Estud. bot. Nordéste 3: 70, 111 (1926), nom. nud.
T. atropurpurea Werderm. in Notizbl. Bot. Gart. Berlin-Dahlem 12: 223–224 (1934). T. funalis var.
atropurpurea (Werderm.) P.J. Braun & E. Esteves Pereira in Kakt. and. Sukk. 43(5): Karteiblatt 14, [2]
(1992). T. funalis subsp. atropurpurea (Werderm.) P.J. Braun & E. Esteves Pereira in Succulenta 74:
134 (1995). Type: Brazil, Bahia, south of Boa Nova and [Vitória da] Conquista, 600 m, May 1932,
Werdermann 3999 (B†).
T. zehntneri Backeb. & Voll in Backeb., Blätt. Kakteenf. 1935(3): unpaged [p. 4] & suppl. [p. 6], with
illus. (1935). Type: not cited. Lectotype (Taylor et al. 2002): Backeberg, l.c. (illus., flower at left).
T. atropurpurea var. zehntnerioides Backeb., Descr. Cact. Nov. [1:] 10 (1956, publ. 1957). Type: Brazil,
Bahia (assumed not to have been preserved).
vernacular names. Rabo-de-rato, Rabo-de-gato, Cipó-de-espinho, Quipá-voador, Trança-perna.
Shrubby, decumbent or scandent, stems to 12 m long, hollow inside and scarcely succulent once > 1 year-
old. Stem-segments to 30–100 cm or more, 0.8–1.5 cm diam., cylindric, grey-green to purplish or strongly
reddish, often somewhat glaucous. Areoles 5–20 mm apart, prominent, felt white to brownish, with long
woolly hairs at first, spineless, glochids abundant, brownish to grey. Flowers from near the apex of terminal
segments, 57–80 × 20 mm, green to purplish; pericarpel narrowly turbinate, to 30–50 mm, grey-green, with
areoles; outer perianth-segments 10, 5–15 mm, narrowly ovate, acute, spreading to reflexed; inner segments
c. 7, to 20–40 mm, acute, revolute; stamens erect, exserted c. 17 mm or more, the outermost replaced by
abundant hairlike staminodes; anthers sometimes reddish; style to 45 mm, stigma-lobes exserted beyond
stamens. Fruit 40–50 × 20 mm, bottle-shaped, with deep umbilicus, greenish to reddish or purple outside;
funicular pulp translucent, whitish, placenta yellowish. Seeds 3–5 per fruit, to 4 mm diam.
Central-southern caatinga element: in caatinga, 380–950 m, south-western Pernambuco (also vaguely

reported from adjacent Piauí) and western and northern to southern Bahia. Endemic to the core area
within North-eastern Brazil. Map 19.
TACINGA
piauí: fide Andrade-Lima (1989: 26), but unlocalized.

pernambuco: W Pernambuco, Mun. Afrânio, 20 Apr. 1971, Heringer, Andrade-Lima et al. 212 (IPA);
Mun. Petrolina, 17 km SSE of Rajada on road BR 407, 8°55'S, 40°44'W, 6 Apr. 2000, E.A. Rocha et al.
(K, photos); l.c., between the town and the lock (‘eclusa’), 14 Aug. 1966, Hunt 6498 (see Taylor et al.
2002); Mun. Santa Maria da Boa Vista, 4 Aug. 1955, Andrade-Lima 55-2116 (IPA).
bahia: W Bahia, Mun. Barra, Luetzelburg (1925–26, 3: 70); N–S Bahia, Mun. Juazeiro, 1915, Rose &
Russell 19723 (US, NY); l.c., near Carnaíba do Sertão, 8 Aug. 1964, Castellanos 25387 (GUA); Mun.
Sento Sé, [Zehntner in] Rose 19734 (US); l.c., 10°7'S, 41°25'W, 11 July 2000, G. Charles (photos); Mun.
Jaguarari [‘Curaçá’], 9°48'S, 40°3'W, 19 Aug. 1983, S.B. Silva & G.C.P. Pinto 300 (HRB); l.c.,
Barrinha, 7–8 June 1915, Rose & Russell 19792 (K, US); l.c., 64 km S of Juazeiro on road to Senhor do
Bonfim, between Maçaroca and Barrinha, 8 Jan. 1991, Taylor et al. 1378 (K, HRCB, ZSS, CEPEC);
Mun. Campo Formoso, 32 km S of Junco, near Curral Velho, 9 Jan. 1991, Taylor et al. (obs.); Mun.
Senhor do Bomfim [‘Villa Nova’], Aug. 1912, Zehntner 306 (US); Mun. (?), 10°53'S, 41°35'W, 13 July
2000, G. Charles (photos); Mun. João Dourado & Mun. Morro do Chapéu, APA Gruta dos Brejões, 4
Aug. 2002, Taylor & Machado (K, photos); Mun. Jacobina, 27–28 km WNW towards Lajes, 12 Jan.
1991, Taylor et al. (obs.); l.c., c. 18 km S of town, ‘Buraco do Possidônio’, 2001, M. Machado (obs.);
Mun. Cafarnaum, 26 Oct. 1965, A.P. Duarte 9252 (HB, RB); Mun. Seabra, Apr. 1978, E. Esteves Pereira
71 (UFG); Mun. Iaçu, Nov. 1987, E. Esteves Pereira 273 (UFG); Mun. Marcionílio Sousa, Machado
Portella, 19 June 1915, Rose 19919 (US); Mun. Abaíra, road Piatã–Abaíra, 13°17'S, 41°42'W, 22 Sep.
1992, W. Ganev 1167 (K, HUEFS); Mun. Bom Jesus da Lapa, reported by Luetzelburg (1925–26, 3:
70); Mun. Livramento do Brumado, 12 km S of town, 23 Nov. 1988, Taylor & Zappi in Harley 25545
(SPF, K); Mun. Ituaçu (?), 13°52'S, 41°17'W, 9 Aug. 1979, J.E.M. Brazão 119 (HRB, RB); Mun.
Maracás, Porto Alegre, fide Uebelmann (1996): HU 384; Mun. Caetité, hill with cross above town, 27
Aug. 1988, Eggli 1304 (ZSS); Mun. Brumado, 19 km from town on road BR 030 towards Sussuarana,
3 Feb. 1991, Taylor et al. (obs.); Mun. Aracatu, near border with Mun. Brumado, 25 km on road BR
030, 3 Feb. 1991, Taylor et al. 1543 (K, HRCB, ZSS, CEPEC); ibid., c. 20 km N of Anajé towards
Sussuarana, 17 Aug. 1988, Eggli 1184 (ZSS); Mun. Cordeiros, 15°4.5'S, 41°54'W, Sep. 2003, M.
Machado (obs.).
conservation status. Least Concern [LC] (1); PD=2, EI=1, GD=2. Short-list score (1×5) = 5. Least
Concern at present, but habitat destruction is continuing throughout its range. It is protected inside the
A.P.A. Gruta dos Brejões, Bahia.
This species remains poorly known as fertile material, but appears to be variable. Its
flowers may be either green or purple, the latter colour variant apparently being
characteristic in the southern part of its range, but recorded northwards to at least Mun.
Cafarnaum, central Bahia. These variants do not appear to exhibit obvious vegetative
differences and the flowers of Backeberg’s var. zehntnerioides, as described, seem to be of
somewhat intermediate colour. More collections and observations during its late winter
(August/September) flowering period are needed. Marlon Machado (pers. comm.) has
observed hummingbirds visiting the flowers of this species near Morro do Chapéu, Bahia.
2. Tacinga braunii E. Esteves Pereira in Kakt. and. Sukk. 40: 134–135 (1989). Holotype:
Brazil, Minas Gerais, surroundings of the Rio Jequitinhonha, P. J. Braun 864 (ZSS).
[Opuntia rubescens sensu K. Schum. in Martius, Fl. bras. 2(4): 306 (1890), quoad Glaziou 14865, non Salm-
Dyck ex DC. (1828).]
[Tacinga funalis sensu Rizzini & Mattos-Filho, Contrib. Conhecim. Fl. NE. Minas Gerais Bahia Medit.:
38 (1992) non Britton & Rose.]
vernacular names. Rabo-de-rato, Cipó-de-espinhos, Rabo-de-espinhos.

TACINGA
Shrubby or semi-scandent with the ultimate stem-segments sometimes pendulous, to 6 m, partially

hollow inside; primary stem cylindric, erect, subsequent stem-segments rounded at base, somewhat
flattened towards apex, ultimate segments to 35 × 1.5–3.5 × 0.5–0.8 cm, rarely broader, ± strongly
flattened, grey-green to purplish, glaucous; areoles to c. 23 mm apart, prominent, felt white, with long
hairs when young; spines 0 (seen only in a plant thought to show introgression with T. werneri), glochids
abundant, greyish. Flowers apical or subapical, c. 7 × 2 cm; pericarpel narrowly turbinate, to 40 mm,
grey-green to purplish, bearing areoles; outer perianth-segments green, c. 10, 5–15 mm, narrowly ovate,
acute, erect to ± reflexed; inner segments pale green, to 40 mm, acute, revolute; stamens erect, the
outermost replaced by abundant hairlike staminodes; style to 45 mm, stigma-lobes exserted. Fruit 40 mm,
narrowly urceolate, with a deep umbilicus, greenish; funicular pulp translucent, whitish, placenta
yellowish. Seeds 4–5 per fruit, to 4 mm diam.
South-eastern caatinga (inselberg) element: on gneiss/granite outcrops/inselbergs in caatinga-agreste,

170–350 m, Rio Jequitinhonha valley, north-eastern Minas Gerais. Endemic to the Rio Jequitinhonha
valley. Map 19.
minas gerais: Mun. Jacinto, near the Rio Jequitinhonha, 16°9'S, 40°13'W, 25 May 1979, P. Vaillant
78 (HRB, RB); Mun. Itaobim, 1 km W of town, 0.5 km N of Rio Jequitinhonha, 18 Nov. 1988, Taylor
& Zappi in Harley 25530 (K, SPF); ibid., on road BR 357, 8 km W of town towards Itinga, 14 Dec.
1990, Taylor & Zappi 760 (K, HRCB, ZSS, BHCB); Mun. Itinga, on road BR 357, 2 km E of town
towards Itaobim, 14 Dec. 1990, Taylor & Zappi 769 (K, HRCB, ZSS, BHCB); Mun. Coronel Murta, 5
km SE of town towards Araçuaí, 21 Feb. 1988, Supthut 8865 (ZSS); município unknown, surroundings
of the Rio Jequitinhonha, July 1987, E. Esteves Pereira 264 (UFG), P.J. Braun 864 (ZSS); unlocalized,
[‘Serra de São José’, according to Glaziou (1909: 327), but not from there, see below], ‘1883–1884’,
[anon. in] Glaziou 14865 (K, C, F, photo).
conservation status. Vulnerable [VU B1ab(iii) + 2ab(iii)] (2); extent of occurrence = 1515 km2;
PD=2, EI=1, GD=1. Short-list score (2×4) = 8.
Although correctly named only in 1989, this distinctive species was in fact collected (see
Glaziou 14865) before its better-known sister-species treated above. The early collection,
made 120 years ago and claimed by Glaziou, was misidentified by Schumann (1890) in
the Flora brasiliensis as Opuntia rubescens Salm-Dyck ex DC., which is the type of Consolea
Lem., an opuntioid genus endemic to the Caribbean (the original description of O.
rubescens erroneously suggested it might have come from Brazil, so Schumann can,
perhaps, be forgiven). Glaziou himself should not be credited with the first collection,
however, since his locality data are clearly false (the Serra de São José, where he is known
to have collected other plants, is in perhumid southern Minas Gerais). It seems that, as in
other well-documented cases (Wurdack 1970), Glaziou appears to have appropriated
material from an anonymous collector and given it his own number. For a further
example of this deplorable practice, see T. inamoena, below.
3. Tacinga werneri (Eggli) N. P. Taylor & Stuppy in Succ. Pl. Res. 6: 111 (2002).
Holotype: Brazil, Bahia, Mun. Jequié, N. P. Taylor et al. 1555 (CEPEC; HRCB, K,
ZSS, isos.).
Opuntia werneri Eggli in Bradleya 10: 90, with illus. (1992).
Shrubby, mostly < 1 m high; stem-segments flat, elongate-ellipsoid to broadly ellipsoid, 10–20 × 4–10 ×
0.5–2.0 cm, dark green, those of seedlings and juvenile growth nearly orbicular; areoles c. 3 mm diam.,
TACINGA
on scarcely raised podaria, mostly 12–18 mm apart, often displaying indeterminate growth near base of
plant; glochids dirty whitish, few at first; spines (0–)3–8 at first (some forms remaining almost spineless),
3–63 × 1.5 mm near base, ash-grey tipped yellow to yellowish brown, diverging. Flowers c. 35–50 mm
long; perianth-segments ± erect to slightly expanded, to c. 7 mm long, bright red, outer series rather
fleshy, darker red; stamens numerous, the outermost represented by staminodes (with rudimentary
anthers or these entirely lacking), filaments orange-yellow to orange above; style narrowly obclaviform,
yellowish; stigma-lobes 4, white; pericarpel 28 × 20 mm, green, areoles with numerous glochids but
spineless. Fruit elongate ovoid, broadly beaked at apex, 40–55 × 25–35 mm, with a narrow, deeply
sunken umbilicus, pale greenish white to white, pericarp thick, pale green; funicular pulp pink when ripe.
Seed c. 3 × 3.5 mm, nearly spherical, funicular envelope pale ochre.
Eastern caatinga element: margins of gneiss/granite outcrops, on inselbergs and occasionally on

calcareous soils in caatinga of the Chapada Diamantina, 100–950 m, drainage of the Rios Paraguaçu, de
Contas and Jequitinhonha, eastern Bahia and north-eastern Minas Gerais. Endemic to the core area of
bahia: Mun. Morro do Chapéu, between ‘Vermelho’ and ‘Mulungu da Gruta’, 11°0'27"S,
41°20'36"W, Aug. 2003, M. Machado (obs.); l.c., 15–20 km W of town, Sep. 2002, M. Machado (K,
photos); l.c., nr Ventura, Sep. 2002, M. Machado (K, photos); l.c., nr ‘Buraco do Possidônio’, Sep. 2002,
M. Machado (K, photos); Mun. Rui Barbosa, near granite quarry, Aug. 1988, Eggli 1259 (CEPEC, ZSS),
7 Feb. 1991, Taylor et al. 1568 (CEPEC, HRCB, K, ZSS); Mun. Rafael Jambeiro, Km 30 of road BR
242, 11 May 1975, A.L. Costa s.n. (ALCB, in spirit); l.c., Km 33, Rio Coruja, 29 Sep. 1973, A.L. Costa
s.n. (ALCB, in spirit); Mun. Iaçu, Morro do Coité, 12°45'S, 39°53'W, 28 Sep. 1997, E. de Melo 2259
(HUEFS); Mun. Itatim, Morro da Torre, 12°43'S, 39°42'W, 9 Nov. 1996, E. de Melo 1831 (HUEFS;
K, photo); Mun. Elísio Medrado, S of Monte Cruzeiro, 25 June 1915, Rose & Russell 20051 (US, NY);
Mun. Jequié, 41.5 km E of Porto Alegre (Mun. Maracás) towards Jequié, 4 Feb. 1991, Taylor et al. 1555
(CEPEC, HRCB, K, ZSS).
minas gerais: Mun. Pedra Azul, 8 km from town towards road BR 116, 18 Oct. 1988, Taylor & Zappi
in Harley 25187 (K, SPF), 1 Feb. 1991, Taylor et al. 1520 (BHCB, HRCB, K, ZSS).
conservation status. Vulnerable [VU B2ab(ii, iii, iv, v)] (2); extent of occurrence >20000 km2, but
area of occupancy estimated to be < 2000 km2; PD=2, EI=1, GD=2. Short-list score (2×5) = 10. This
species has a local and rather disjunct distribution and is nowhere common. Its habitat at Rui Barbosa
(BA) is being destroyed by mining operations.
The flowers and fruit of this species provide a clear link between T. funalis, T. braunii and
the following species, with which it is frequently sympatric. It is likely that it has been
under-recorded due to confusion with T. palmadora. The collections from Rui Barbosa
cited above (Eggli 1259, Taylor et al. 1568) represent a variant population with rather weak
or almost absent spination. As just noted, its long term future looks doubtful while granite
quarrying activities continue.
Plants of probable hybrid origin, involving T. werneri and T. inamoena, have been
observed by the authors near Pedra Azul, Minas Gerais and west of Morro do Chapéu,
Bahia (eg. 15–20 km W of Morro do Chapéu, 5 Aug. 2002, Taylor & Machado [K,
photos]). The following names, as suggested by Ritter’s original illustration, could possibly
refer to this hybrid, although T. werneri has not yet been found near Urandi: Platyopuntia
inamoena f. spinigera F. Ritter, Kakt. Südamer. 1: 32 (1979). Opuntia inamoena f. spinigera
(Ritter) P. J. Braun & E. Esteves Pereira in Cact. Succ. J. (US) 61: 272–273 (1989).
Holotype: Bahia, Urandi, Ritter 1252A (U). If they do not belong here, then it is assumed
that they will be identifiable with T. quipa (F. A. C. Weber) N. P. Taylor & Stuppy.
TACINGA
4. Tacinga palmadora (Britton & Rose) N. P. Taylor & Stuppy in Succ. Pl. Res. 6: 112
(2002). Lectotype (Krook & Mottram 2001: 94): Brazil, Mun. Jaguarari, Barrinha, 7 June
1915, Rose & Russell 19787 (NY; NY, US, isos.).
Opuntia palmadora Britton & Rose, Cact. 1: 202 (1919).
vernacular names. Palmatória, Palmatória-de-espinho, Palma-de-espinhos, Palma, Palminha, Quipá-

de-espinho, Quipá, Palmatória-de-quipá, Rabo-de-onça.
Shrubby, to 2(–4) m tall, main stem flattened, transversely elliptic, sometimes with brownish bark, very
spiny, the areoles showing considerable indeterminate growth. Branches erect, stem-segments 8–17 × 3–7
cm, elliptic to obovate, grey-green, glaucous. Areoles 10–20 mm apart, prominent, felt white; spines
(2–)3–9, (7–)15–35 mm, pungent, golden at first, pale grey with age. Flowers from the apex of the
terminal segments, 30 × 15–20 mm; pericarpel turbinate, to 20 mm, pale to grey-green, with orange,
acute, fleshy bract-scales subtending bristles; outer perianth-segments 10 mm, narrowly lanceolate, deep
red or pinkish orange, erect; innermost 15 mm, spathulate, orange, spreading only at apex; stamens
included; stigma-lobes exserted. Fruit to 3 cm long, turbinate to obovoid, greenish to reddish or purple
outside; funicular pulp translucent, whitish, placenta yellowish. Seeds 3–5 per fruit, to 5 mm diam.
Eastern caatinga element: in caatinga, agreste and carrasco, frequent on deep sandy substrates but not
restricted to these, c. 200–1020 m, Rio Grande do Norte to southern Bahia (from the Chapada
Diamantina/northern Serra do Espinhaço eastwards). Endemic to North-eastern Brazil. Map 19.
rio grande do norte: Mun. São José do Campestre, 58 km N of Guarabira (PB) on road to Tangará,
6°21'S, 35°40'W, 1 Apr. 2000, E.A. Rocha et al. (K, photos); Mun. Tangará, between the town and
Currais Novos, 6°14'S, 35°51'W, 1 Apr. 2000, E.A. Rocha et al. (K, photos).
paraíba: Mun. Areia, ‘no agreste’, 22 Sep. 1956, J.C. Moraes s.n. (SPSF 5008); Mun. Soledade, see Eiten
(1983: figs 67 & 68); Mun. Itaporanga, Serra Água Branca, 7–10 Jan. 1994, M.F. Agra et al. 2560 (JPB);
Mun. Campina Grande, Boa Vista, 27–29 Apr. 1994, M.F. Agra et al. 3125 (JPB); Mun. São João do
Cariri, road BR 412, 27–29 Apr. 1994, M.F. Agra et al. 3122, 3123, 3124 (JPB).
pernambuco: Mun. Araripina, Dec. 1963, Ritter 1253 (SGO 122051, fide Eggli et al. 1995: 505); Mun.
Ouricuri, 26 km ESE towards Parnamirim, 7°58'S, 39°52'W, 16 July 1962, G. Eiten & L.T. Eiten 4929 (SP,
US); ibid., Faz. Tabuleiro, July 1984, G. Costa-Lima 21 (HRB, IPA); Mun. Sertânia, Moderna, 8 Apr. 2000,
E.A. Rocha et al. (obs.); l.c., 13 km W of Arcoverde on road from Custódia, 12 Feb. 1991, Taylor & Zappi
(obs.); Mun. Pesqueira, 5 Oct. 1934, Pickel 3661 (IPA); Mun. Caruaru, 24 Nov. 1991, F.A.R. Santos 11
(HUEFS, PEUFR), l.c., 9 km NE of town, 8°14'21"S, 35°54'52''W, Dec. 1999, E.A. Rocha et al. (K, photos);
Mun. Taquaritinga do Norte, Gravatá do Ibiapina, 8 Aug. 2001, Zappi (obs.); Mun. Gravatá, Russinha, 13
Oct. 1934, Pickel 3656 (IPA 4486); Mun. Buíque, Fazenda Laranjeiras, 10 July 1995, K. Andrade & L.
Figueiredo 126 (K, PEUFR); Mun. Alagoinha, Sítio Cajueiro, Sep. 1991, J. Hamburgo Alves & R. Silva s.n.
(PEUFR); l.c., Sítio Lagoa Seca, 5.5 km from road PE 217, 13 Feb. 1993, F.A.R. Santos 74, 75 (ALCB,
HUEFS, PEUFR); Mun. Belém de São Francisco, near border with Mun. Floresta, 8°40'S, 38°44'W, 8 Apr.
2000, E.A. Rocha et al. (K, photos); Mun. Floresta, 8°36'S, 38°27'W, 8 Apr. 2000, E.A. Rocha et al. (K,
photos); l.c., N of the Serra Negra, 8°33'S, 38°1'W, 8 Apr. 2000, E.A. Rocha et al. (K, photos); l.c., road to
Ibimirim, 30 June 1952, Andrade-Lima 52-1156 (R, IPA), 24 July 1994, L.P. Felix et al. 6702 (PEUFR); l.c.,
Serra de São Gonçalo, no alto da serra, 23 May 1971, Heringer & Andrade-Lima 856 (RB, IPA); Mun.
Ibimirim, Lagoa de Areia, 23 July 1994, L.P. Felix et al. 6696 (PEUFR); l.c., 12 Sep. 1954, I.I.A. Falcão et
al. 1056 (RB); Mun. Nova Petrolândia, 12 km NW of town, 12 Feb. 1991, Taylor & Zappi (obs.).
alagoas: W Alagoas, Mun. Mata Grande, Serra Verde, 30 July 1981, R. de Lyra & G.L. Esteves 692
(MAC); Mun. Delmiro Gouveia, near bridge over Rio São Francisco at border with Bahia, 12 Feb. 1991,
Taylor & Zappi 1618C (K, photo); Mun. Piranhas, Xingó, 9 Oct. 1993, R. de Lyra-Lemos et al. 2812 (MAC).
sergipe: Mun. Canindé de São Francisco, 11 Feb. 1991, Taylor & Zappi (obs.); Mun. Poço Redondo,
13 July 1983, M. Fonseca 570 (IPA 30960, ASE 2959), 1 Aug. 1986, Fonseca s.n. (ASE 4452); Mun. Porto
da Folha, 9.5 km NW of Monte Alegre de Sergipe, 11 Feb. 1991, Taylor & Zappi (obs.); Mun. Nossa
Senhora da Glória, 8–9 km NW of town, 11 Feb. 1991, Taylor & Zappi (obs.).
TACINGA
bahia: Mun. Rodelas, Salgado do Melão, 3 Aug. 1994, M.C. Ferreira 600 (HRB); Mun. [Nova] Glória,
Brejo do Burgo, 30 Nov. 1992, F.P. Bandeira 112 (ALCB, HUEFS), 27 Aug. 1995, F.P. Bandeira 265
(HUEFS); Mun. Paulo Afonso, 20 Mar. 1959, Castellanos 22451 (R); Mun. Jaguarari, 64 km S of Juazeiro
towards Senhor do Bonfim, between Maçaroca and Barrinha, 8 Jan. 1991, Taylor et al. 1379, 1380 (K,
HRCB, ZSS, CEPEC); l.c., Barrinha, 7–8 June 1915, Rose & Russell 19787 (US); ibid., Flamengo, July
1988, E. Esteves Pereira 274 (UFG); l.c., Faz. Suçuarana, 9°58'S, 39°52'W, 8 Aug. 1983, G.C.P. Pinto & S.B.
Silva 195/83 (HRB, CEPEC); Mun. Uauá, c. 51 km N of Monte Santo, 25 Aug. 1996, L.P. de Queiroz
4635 (HUEFS, K); Mun. Jeremoabo, Xuquê [Chuquê], May 1921, Luetzelburg 12673 (M); Mun. Monte
Santo, c. 5 km N of town towards Uauá, 25 Aug. 1996, L.P. de Queiroz 4608 (HUEFS, K); Mun. Itiúba,
road to Filadélfia, 10°39'S, 39°44'W, 11 May 2002, Nascimento & Nunes 101 (HUEFS); Mun. Queimadas,
9–11 June 1915, Rose & Russell 19865 (US); Mun. Tucano, Aug. 1961, Pabst s.n. (HB); Mun. São Gabriel
& Mun. João Dourado, APA Gruta dos Brejões, 4 Aug. 2002, Taylor & Machado (obs.); Mun. Morro do
Chapéu, NE of Gruta dos Brejões, 4 Aug. 2002, Taylor (K, photo); l.c., W of Icó, Brejão (Formosa),
‘Lagedo Bordado’, 5 Aug. 2002, Taylor (K, photos); l.c., 2.5 km E of América Dourada, 0.5 km from SE
bank of Rio Jacaré, 17 July 1989, Taylor & Zappi (obs.); l.c., c. 18 km S of town, ‘Buraco do Possidônio’,
2002, M. Machado & Taylor (obs.); Mun. Araci, 18 Jan. 1994, M.A.S. das Neves 84 (HUEFS); l.c., 14.5 km
N of town towards Tucano, 5 Jan. 1991, Taylor et al. 1357 (K, HRCB, ZSS, CEPEC); Mun. Jacobina,
27–28 km WNW of Jacobina towards Lages, 12 Jan. 1991, Taylor et al. 1397 (K, HRCB, ZSS, CEPEC);
Mun. Miguel Calmon, 28 km S of Jacobina, 26 Apr. 1992, Taylor & Zappi (obs.); Mun. Olindina, Mata da
Faz. Olhos D’agua, 13 July 1993, O.B. Borges 38 (HRB); Mun. Riachão do Jacuípe, 10 km SE of town on
road BR 324, 10 July 1985, Noblick & Lemos 4011 (K, HUEFS); l.c., road BR 324, 26 km NW of
Tanquinho, 25 Apr. 1992, Taylor & Zappi (K, photo); Mun. Ipirá, 12°22'S, 39°41'W, 4 Oct. 1986, L.P. de
Queiroz 959 (HUEFS); l.c., near Rio do Peixe, 12°13'S, 39°48'W, 6 Aug. 1979, J.E.M. Brazão 115 (HRB);
l.c., Faz. Caldeirão, 18 July 1984, E.L.P.G. Oliveira 756 (BAH 6115); Mun. Ibitiara, c. 52 km W of Seabra
towards Ibotirama, Serra Malhada, 14 Jan. 1991, Taylor et al. 1418A (HRCB, K, fr.); Mun. Muritiba (?),
Vale dos rios Paraguaçu e Jacuípe, Ilha do Umbuzeiro, 12°32'S, 39°5'W, Aug./Sep. 1980, Iscardino et al. in
Grupo Pedra do Cavalo 595 (CEPEC, HRB, ALCB, BAH 3383); Mun. Itatim, Morro do Agenor,
12°42/43'S, 39°42'W, 14 Oct. 1995, F. França 1405 (HUEFS), 31 Mar. 1996, E. de Melo 1566 (HUEFS);
Mun. Milagres, 18 July 1979, Hatschbach 42449 (CEPEC, MO), 16 July 1982, Hatschbach & Guimarães 45071
(MO); l.c., road BA 046 c. 4 km from junction with road BR 116 towards Amargosa, 12°51'S, 39°46'W,
2 June 1993, L.P. Queiroz & T.S.N. Sena 3196 (K, HUEFS); l.c., 13°56'S, 39°43'W, 26 Oct. 1978, A.
Araújo 122 (RB); Mun. Botuporã, c. 60 km SE of Macaúbas towards Paramirim, 26 Aug. 1988, Eggli (obs.);
Mun. Maracás, c. 20 km N of town towards Planaltino, Faz. Tanquinho, 29–30 June 1993, L.P. Queiroz &
V.L.F. Fraga 3260 (K, HUEFS); l.c., 6–7 km E of Porto Alegre on road to Jequié, 4 Feb. 1991, Taylor et al.
1552 (K, HRCB, ZSS, CEPEC); Mun. Livramento do Brumado, 11 km S of town towards Brumado, 23
Nov. 1988, Taylor & Zappi in Harley 25543 (SPF, K); Mun. Ituaçu, 13°48'S, 41°16'W, 22 June 1987, L.P.
de Queiroz 1656 (HUEFS); Mun. Jequié, 4 km from town on road BR 116, 13 Oct. 1977, T.S. dos Santos
3146 (CEPEC); Mun. Caitité, Brejinho das Ametistas, 1 km from village, 26 July 1989, Taylor & Zappi (K,
photo); Mun. Urandi, near railway line, c. 3 km S of town, 22 July 1989, Taylor & Zappi (obs.); Mun.
Vitória da Conquista, Gameleira, 14°50'25"S, 41°W, 17 April 2003, Taylor & Zappi (obs.).
conservation status. Least Concern [LC] (1); PD=2, EI=1, GD=2. Short-list score (1×5) = 5. A
widespread and frequent species.
The pollination biology of this species has been studied by Locatelli & Machado (1999a)
and their observations of hummingbird visitation are confirmed by the present authors. It
exhibits considerable regional and local variation.
5. Tacinga saxatilis (F. Ritter) N. P. Taylor & Stuppy in Succ. Pl. Res. 6: 115 (2002).
Holotype: Brazil, Minas Gerais, Montes Claros, 1959, Ritter 1035 (U).
Platyopuntia saxatilis F. Ritter, Kakt. Südamer. 1: 32–33 (1979). Opuntia saxatilis (Ritter) P.J. Braun & E.
TACINGA
Esteves Pereira in Cact. Succ. J. (US) 63: 82 (1991).

O. saxatilis var. pomosa P. J. Braun & E. Esteves Pereira in ibid. 124–129 (1991). O. saxatilis subsp. pomosa (P.
J. Braun & E. Esteves Pereira) P. J. Braun & E. Esteves Pereira in Schumannia 3: 188 (2002). Holotype:
Brazil, Minas Gerais, border region of Goiás and Minas Gerais, 1979, E. Esteves Pereira 156 (UFG 12368).
O. saxatilis var. occibahiensis P. J. Braun & E. Esteves Pereira in ibid. 311–315 (1991). O. saxatilis subsp.
occibahiensis (P. J. Braun & E. Esteves Pereira) P. J. Braun & E. Esteves Pereira in Schumannia 3: 188
(2002). Holotype: Brazil, W Bahia, 1982, E. Esteves Pereira 188 (UFG 12376).
O. saxatilis var. minutispina P. J. Braun & E. Esteves Pereira in Cact. Succ. J. (US) 67: 108 (1995). O.
saxatilis subsp. minutispina (P. J. Braun & E. Esteves Pereira) P. J. Braun & E. Esteves Pereira in
Schumannia 3: 188 (2002). Holotype: Brazil, Minas Gerais, [‘próximo Manga/Januária’], W of Rio
São Francisco, c. 550 m, [‘VII.1978’], E. Esteves Pereira 228 (UFG 12382).
Shrubby, to 50 × 400 cm, without main stem. Branches erect, stem-segments 10–16 × 5–11 cm, 7–17
mm thick, orbicular to ovate, grey-green, sometimes glaucous. Areoles 1–14 mm apart, to 0.5 mm diam.,
sunken into raised podaria, with felt, glochids immersed in the stem; spines 0–6, delicate, generally bristly,
off-white to dark brownish, erect, 1–19 mm, very thin. Flowers at the apex and margins of the terminal
stem-segments, 3–4 × 2.6–4.2 cm; pericarpel 17–25 × 17 mm, green, with yellowish or dull red, acute,
fleshy bract-scales subtending bristles; perianth-segments 12–22 × 9–14 mm, outer narrowly lanceolate,
deep red to yellow-orange, spreading; innermost spathulate, yellow to orange, spreading; stamens 2–9
mm, exposed; style 7–20 mm; stigma-lobes exserted. Fruit to 38 × 28 mm diam., globose to depressed-
globose, slightly beaked, the beak 2–6 mm, brownish green to wine-red or brownish red when ripe;
funicular pulp translucent, greenish, placenta yellowish. Seeds many per fruit, to 4.8 × 3.6 × 2.8 mm.
The spiny sister-species of T. inamoena (see below), with which it is narrowly sympatric
in western Bahia, replacing it in the Rio São Francisco valley further south on limestone
outcrops. It links T. inamoena to the preceding species in its tendency to having somewhat
beaked fruits. Its range extends westwards on limestone outcrops into north-western
Minas Gerais, slightly beyond the limits of Eastern Brazil as defined here. Two subspecies
are recognized:
1. Areoles 7–14 mm apart; perianth-segments spathulate, commonly yellowish (W &

cent.-N Minas Gerais, and W of the Rio São Francisco in SW Bahia) . . . . . . 5a. subsp. saxatilis
1. Areoles very densely disposed, 1–6 mm apart; perianth-segments
lanceolate, deep orange (Mun. Iuiú, Bahia) . . . . . . . . . . . . . . . . . . . . . . . . . . 5b. subsp. estevesii
5a. subsp. saxatilis
vernacular name. Palma.
Stem-segments 10–12 × 5–6 cm.
Southern Rio São Francisco caatinga element: on ± forest-covered limestone (Bambuí) outcrops
surrounded by caatinga, mata seca semidecídua and cerradão, c. 450–700 m, western Bahia (west of the Rio
São Francisco) to north-western, northern and central Minas Gerais (to c. 17°55'S). Map 19.
bahia: Mun. Santana, 28 km S of town towards Santa Maria da Vitória, 15 Jan. 1991, Taylor et al. 1429
(K, HRCB, ZSS, CEPEC); Mun. Santa Maria da Vitória, E of town, Sep. 1982, E. Esteves Pereira 188
(UFG); Mun. Bom Jesus da Lapa, Faz. Serra Solta, 17 July 1975, Andrade-Lima 75-8189 (IPA 45216);
Mun. Cocos, W of town, reported by Braun & Esteves Pereira (1999b).
minas gerais: Mun. Manga – Mun. Januária, June 1978, E. Esteves Pereira 228 (UFG); Mun. Januária,
‘distrito de Fabião, antes do cerrado do Judas, Buraco dos Macacos’, 15°7'23"S, 44°14'24''W, 16 Nov.
TACINGA
1998, J.A. Lombardi & L.G. Temponi 2224 (K, BHCB); [Mun. Unaí, N of the town, Apr. 1979, E. Esteves
Pereira 156 (UFG), Jan. 1991, Taylor et al. (obs.) — record outside of study area]; Mun.
Itacarambi/Manga, 15°26'S, 43°55'W, Sep. 1999, I. Ribeiro (Fund. Zoo-Bot. BH, photos); Mun. Capitão
Enéas, limestone outcrop visible from road BR 122, 9 June 2002, M. Machado & G. Charles (K, photos);
Mun. Montes Claros, c. 13 km NW of town on road to Januária, 6 Nov. 1988, Taylor & Zappi in Harley
25509 (SPF, K), 1959, Ritter 1035 (U); l.c. (?), 1980, E. Esteves Pereira 37 (UFG); Mun. Bocaiúva, dirt
road to Engenheiro Dolabela, 500 m from highway to Montes Claros, 17°28'S, 44°1'W, 4 Nov. 1988,
Taylor & Zappi in Harley 25504 (K, SPF); Mun. Buenópolis, 7 km S of town, 11 Oct. 1988, Taylor &
Zappi in Harley 24839 (K, SPF).
Concern at present, but found only on limestone and therefore potentially threatened by quarrying
activities in the long term at some of its sites.
At the first Bahian locality cited above the hybrid T. saxatilis subsp. saxatilis T. inamoena
has been observed and collected (Taylor et al. 1430, CEPEC, HRCB, K, ZSS) together with
both parental taxa. This subspecies shows significant regional variation (cf. synonymy above).
5b. subsp. estevesii (P. J. Braun) N. P. Taylor & Stuppy in Succ. Pl. Res. 6: 118 (2002).
Holotype: Brazil, Bahia, [Serra de Iuiú], 1984, E. Esteves Pereira 191 (UFG; ZSS, iso.).
Opuntia estevesii P. J. Braun in Cact. Succ. J. (US) 62: 165–169 (1990).
Stem-segments to 16 × 11 cm.
Southern Rio São Francisco caatinga element: on exposed Bambuí limestone outcrops in caatinga, c. 500–550
m, east of the Rio São Francisco, Mun. Iuiú, southern Bahia. Endemic to the above locality. Map 19.
bahia: Mun. Iuiú, Serra de Iuiú [south of the town], July 1984, E. Esteves Pereira 191 (UFG, ZSS).
conservation status. Vulnerable [VU D2] (2); area of occupancy < 20 km2; PD=1, EI=1, GD=1.
Short-list score (2×3) = 6. It is presently known only from the location cited above (number of
individuals unknown) and could be affected by quarrying in the future.
6. Tacinga inamoena (K. Schum.) N. P. Taylor & Stuppy in Succ. Pl. Res. 6: 119
(2002). Type (Taylor et al. 2002): Brazil, Rio de Janeiro [fide Schumann]; ‘Minas Gerais,
Serra dos Ilheos, Sítio’ [according to Glaziou 1909: 327, but see below; probably north-
eastern Minas Gerais], ‘1883–84’, [anon. in] Glaziou 14864 (B†; K, lecto.).
Opuntia inamoena K. Schum. in Martius, Fl. bras. 4(2): 306 (1890). Platyopuntia inamoena (K. Schum.) F.
Ritter, Kakt. Südamer. 1: 32 (1979).
Opuntia inamoena var. flaviflora Backeberg, Descr. Cact. Nov. [1:] 10 (1956, publ. 1957) and Die Cact. 1:
613, illus. (1958). Type: a cultivated plant (not known to have been preserved).
vernacular names. Quipá, Guibá, Guipá, Palmatória, Palmatória-miúda, Iviro, Gogóia, Palma-
de-ovelha.
Subshrub, 10–50 × 50–300 cm, without main stem; stem-segments erect or sprawling, very variable in
shape, size and thickness, 5–15 × 4–12 cm, orbicular to ovate etc., light green, with ducts exuding sticky,
whitish mucilage when cut; areoles 10–15 mm apart, sunken and expanded inwards. Spines absent except
sometimes on very young stem-segments (but see T. quipa below), glochids numerous but mostly
TACINGA
concealed inside the sunken areoles. Flowers from near the apex of terminal segments, c. 50 × 35–40 mm;
pericarpel globose, to 10 mm, green, with orange-red, acute, fleshy bract-scales subtending bristles; outer
perianth-segments 16 mm, narrowly lanceolate, deep red or pinkish orange, spreading, inner segments
18–20 mm, spathulate, orange, spreading to slightly reflexed at apex; stamens exposed, erect; stigma-lobes
exserted. Fruit to c. 40 mm diam., globose to depressed globose, yellow or orange when ripe, the
depressed umbilicus small in relation to fruit diameter, funicular pulp translucent, edible. Seeds many per
fruit, to 3 mm diam.
Widespread Eastern Brazil element: usually on rocks (including inselbergs) or very stony ground, open
caatinga and campo rupestre, c. 100–1550 m, from the middle drainage of the Rio Jequitinhonha (MG)
northwards to northernmost Piauí, and westwards on sandstone outcrops in the cerrado of western Bahia.
Endemic to Eastern Brazil. Amongst the commonest of cacti from the region. Only records marking its
approximate eastern, western and southern limits are given below. Map 19.
paraíba: Mun. Tacima, border with Mun. Caiçara, W margin of Rio Curimataú, 6°36'S, 35°28'W, 1
Apr. 2000, E.A. Rocha et al. (K, photo).
bahia: W Bahia, Mun. Barreiras, 34 km W of town, 2 Mar. 1972, Anderson et al. 36462 (MO); E Bahia,
Mun. Muritiba, nr São José do Itaporã, Pau-Ferro, Faz. Oito de Dezembro, 6 Aug. 2002, M. Machado
(HUEFS, K, photos).
minas gerais: NW Minas Gerais, Mun. Itacarambi/Januária, vale do Rio Peruaçu, illustrated in Costa
et al. (1998: Fig. 29); cent.-N Minas Gerais, Mun. Porteirinha, 8 km S of town towards Riacho dos
Machados, 7 Nov. 1988, Taylor & Zappi (obs.); Mun. Grão Mogol, subida da Trilha da Tropa, 27 May
1987, Zappi et al. in CFCR 11973 (SPF); Mun. Cristália, road to town from Grão Mogol, 14 Apr. 1981,
Pirani et al. in CFCR 917 (SPF, K); NE Minas Gerais, Mun. Itinga, on road BR 367, 41°48'W, 15 Feb.
1988, W.W. Thomas et al. 5971 (SPF, NY); Mun. Itaobim, 1 km W of town, 0.5 km N of Rio
Jequitinhonha, 18 Nov. 1988, Taylor & Zappi (K, photos).
conservation status. Least Concern [LC] (1); PD=2, EI=2, GD=2. Short-list score (1×6) = 6. This
species is locally a co-dominant element of the caatinga flora, where it is widespread in more open areas.
The type locality, as given by Glaziou (1909: 327), is assumed to be false and probably an
invention to disguise the fact that he was not the real collector (cf. Wurdack 1970). It
may be no coincidence that the Glaziou number for ‘Opuntia rubescens’ [sensu Schumann
(1890), non DC.], Gl. 14865, which is Tacinga braunii (q.v.), immediately follows that for
T. inamoena (Gl. 14864) and both species grow together in the valley of the Rio
Jequitinhonha, north-eastern Minas Gerais, where T. braunii is endemic. It seems
probable, therefore, that the type of T. inamoena came from north-eastern Minas Gerais,
where it reaches its south-eastern limit.
Braun & Esteves Pereira (1989: 272) remark that Opuntia inamoena occurs in the state
of Espírito Santo, but have so far apparently failed to substantiate this claim with a definite
locality.
Marlon Machado has pointed out a very distinct dwarf race of the species, found on
gneissic outcrops towards the limits of the caatinga in eastern Bahia and north-eastern Minas
Gerais. This plant forms a small erect shrub with miniature, almost shortly cylindrical stem-
segments of regular size and with reduced numbers of areoles; likewise the pericarpel. It
seems to be genetically determined, since it does not change habit when grown under
divergent conditions and appears to come true from seed. Its easternmost locality in Mun.
Muritiba is cited above, but identical populations occur in Mun. Itatim, Ipirá and Itaberaba
(Horst & Uebelmann HU 896; M. Machado, pers. comm.). It was first collected in the early
1970s at Araçuaí, Minas Gerais, near the south-eastern limit of the species (HU 336). It
TACINGA
probably deserves to be named as a subspecies, especially if this will ensure its conservation.
The true T. inamoena has spineless stem-segments as in T. funalis and T. braunii, but,
as in those species, its abundant fine glochids demand that it be treated with appropriate
respect and handled only with forceps. Forms with mature stems bearing occasional spines
are usually to be referred to the following hybrid (but see below T. werneri also):–
Tacinga quipa (F. A. C. Weber) N. P. Taylor & Stuppy in Succ. Pl. Res. 6: 120 (2002).
Type: Brazil, Pernambuco, without date or collector (P†).
Opuntia quipa F.A.C. Weber in Bois, Dict. hort.: 894 (1898). NB. Although previously treated as a
synonym of T. (Opuntia) inamoena, Weber’s diagnosis mentions the presence of occasional spines
indicating hybridity.
[T. inamoena T. palmadora]
vernacular name. Quipá.
Like T. inamoena, but stem-segments with occasional scattered spines, mainly at apex. Flowers not seen.
Fruit variable in shape and colour.
Caatinga, c. 200–1080 m, of sporadic occurrence throughout the range of T. palmadora where T. inamoena
is also present. Endemic to North-eastern Brazil.
pernambuco: without locality (locus classicus of accepted name).
alagoas: W Alagoas, Mun. Delmiro Gouveia, near bridge over Rio São Francisco at border with
Bahia, 12 Feb. 1991, Taylor & Zappi 1618D (K, photo).
bahia: Mun. (?), Raso da Catarina, Sede da Reserva, 25 Oct. 1982, L.P. de Queiroz 428 (ALCB); Mun.
Uauá, 1.7 km ESE of town towards Bendengó and Canudos, 6 Jan. 1991, Taylor et al. 1363 (K, HRCB,
ZSS, CEPEC); Mun. Jacobina, 27–28 km W of town towards Lages, 12 Jan. 1991, Taylor et al. 1398 (K,
HRCB, ZSS, CEPEC); Mun. Morro do Chapéu, W of Icó, Brejão (Formosa), ‘Lagedo Bordado’, 5 Aug.
2002, Taylor (K, photos).
The commonest and most widespread hybrid amongst the cacti of Eastern Brazil, but the
two parental species certainly do not hybridize as often as they co-occur, perhaps because
hummingbirds seem to prefer the flowers of T. palmadora to those of T. inamoena. Care is
needed in determining plants as this hybrid, since another involving T. inamoena and T.
werneri is very similar and it is likely that all 3 species can grow together (eg. nr Morro do
Chapéu, BA).
Insufficiently known taxa

The following are of uncertain identity for lack of extant type specimens. Werdermann’s
species has previously been referred to either the above hybrid or as a form of Tacinga
(Opuntia) palmadora (see below). It could also refer to T. werneri, for which it would be an
older name, although the fruit shape as described probably rules this out. The hybrid
between this species and T. inamoena is another plant known from the region of
Werdermann’s type to which his name could refer. Braun & Esteves Pereira’s assignment
of it may be correct, since there exists a rare aberrant form of T. palmadora in caatinga not
far from the type locality with few spines and yellowish white fruit. The details are:
BRASILIOPUNTIA
Opuntia catingicola Werderm. in Notizbl. Bot. Gart. Berlin-Dahlem 12: 223 (1934). O.
palmadora subsp. catingicola (Werderm.) P. J. Braun & E. Esteves Pereira in Succulenta 74:
133 (1995). Type: Brazil, Bahia, between Mundo Novo and Ventura, 600 m, Apr. 1932,
Werdermann 3998 (B†). O. catingicola var. fulviceps Haage in Backeberg, Kakteenlex., ed.
3: 495 (1976) and Cactus Lex.: 356 (1977), nom. inval. (Art. 36.1 & 37.1). Based on a
cultivated plant (not known to have been preserved).
4. BRASILIOPUNTIA (K. Schum.) A. Berger
Entwicklungslin. Kakt.: 94 (1926). Type: Cactus brasiliensis Willd.
A very distinct, highly specialized, monotypic, arborescent genus, which is allied to

Opuntia sens. str. on the basis of seed-envelope morphology/anatomy (Stuppy 2002), but
with a cylindrical, apical leader shoot of indeterminate (unjointed) growth, markedly
different pollen (cf. Leuenberger 1976) and non-sensitive stamens. The adult tree is
similar in habit to an old Araucaria and displays unique shoot-morphology, the erect,
cylindric leader, giving rise to progressively more flattened, lateral stem-segments, the
ultimate of which are very thin, hardly succulent, almost leaf-like and drought-deciduous.
However, the first shoot (plumule) of the seedling, which arises between massive
cotyledons, is thin, clearly flattened and early-determinate (cf. Opuntia sens. str.), soon
giving rise to one or more equally flattened, subapical secondary segments and sometimes
supplanted by stronger shoots of indeterminate growth arising from the axils of the
cotyledons. The cylindrical, indeterminate leader shoot(s) develop from either of these
sources and may be cylindric from the beginning or remain ± flattened for some time.
Such leader stems show very early periderm formation, and the production of glochids
from their areoles is very limited or absent. Plates 8.2–10.1.
This may be the tallest member of the Cactaceae, with specimens attaining heights of
between 20 and 25 metres in the mata de cipó, a part of the agreste, the Atlantic Forest and
caatinga ecotone.
1. Brasiliopuntia brasiliensis (Willd.) A. Berger, l.c. (1926). Type: probably a living

plant in the Berlin botanical garden; no material extant at B-W. Lectotype (Taylor et al.
2002): W. Piso, Historia naturalis Brasiliae: illustration, p. 100, below (1648); cf.
Willdenow, l.c. infra.
Cactus brasiliensis Willd., Enum. pl. suppl.: 33 (1814). Opuntia brasiliensis (Willd.) Haw., Suppl. pl. succ.:
79 (1819).
Cactus arboreus Vell., Fl. flumin.: 207 (1829). Opuntia arborea (Vell.) Steud., Nomencl. Bot. Ed. 2, 1: 220
(1841). Holotype: not extant. Lectotype (Taylor et al. 2002): Vellozo, tom. cit., Icones 5: tab. 28
(1831).
Cactus heterocladus A. St. Hil., Voy. Rio de Janeiro 2: 103 (1830). Type: Rio de Janeiro, not known to
have been preserved.
Opuntia bahiensis Britton & Rose, Cact. 1: 210–211 (1919). Brasiliopuntia bahiensis (Britton & Rose) A.
Berger, l.c. (1926). Opuntia brasiliensis subsp. bahiensis (Britton & Rose) P. J. Braun & E. Esteves
BRASILIOPUNTIA
Pereira in Succulenta 74: 132 (1995). Brasiliopuntia brasiliensis subsp. bahiensis (Britton & Rose) P. J.
Braun & E. Esteves Pereira in Schumannia 3: 188 (2002). Lectotype (Taylor et al. 2002): Brazil,
Bahia, near Toca da Onça [Jaguaquara], 27–29 June 1915, Rose & Russell 20068 (US; NY, lectopara.).
Brasiliopuntia subacarpa Rizzini & A. Mattos in Revista Brasil. Biol. 46(2): 323–328 (1986). Opuntia
brasiliensis subsp. subacarpa (Rizzini & A. Mattos) P. J. Braun & E. Esteves Pereira in Succulenta 74:
132 (1995). Brasiliopuntia brasiliensis subsp. subacarpa (Rizzini & A. Mattos) P. J. Braun & E. Esteves
Pereira in Schumannia 3: 188 (2002). Holotype: Brazil, Minas Gerais, Itinga, 17 Oct. 1984, Rizzini
& Mattos-Filho 38 (RB; UFMT, iso.).
vernacular names. Urumbeba, Rumbeba, Cumbeba, Mumbeca, Mumbebo, Facho-de-renda,

Palmatória-grande, Palmatória-do-diabo, Ambeba, Arumbeva, Gerumbeba, Jurubeba, Palmadora,
Palmatória, Xiquexique-do-sertão.
Tree 4–20(–25) m, trunk to 15–20 × 35 cm diam., cylindric, with clusters of spines to 9 cm long; pith
chambered. Branches dimorphic, patent; intermediate joints cylindric, 20–100 cm, ultimate joints
(4–)6–15 × 3–6(–7) cm, rhomboid to obovate, irregular in outline, narrow at base, thin, bright to dark
green, deciduous. Areoles on the lateral stem-segments 15–30 mm apart, with white felt, spineless or with
1 spine to 40 mm, those on the leader shoot with 1–3 or more spines, or those that have given rise to
lateral stem-segments spineless and often lacking glochids also; leaves early deciduous, ovoid or more
elongate on the leader shoot, fleshy, bright green. Flowers from near the apex of the leader shoot or
terminal stem-segments, or by proliferation from the pericarpels of old flowers, c. 2.5–3.5 × 4.5 cm;
pericarpel globose, obovoid, elongate or elongate-flattened, c. 8–28 × 9–12 mm, green, tuberculate,
bearing scale leaves c. 1 mm long, subtending areoles with white wool; outer perianth-segments to 15
mm, ovate, greenish to yellowish, erect to spreading, innermost to 20 mm long and 10 mm broad,
lanceolate to spathulate, yellow, spreading; stamens non-sensitive, to 7 mm, anthers c. 0.6 mm, whitish;
style c. 9 × 1.5 mm, whitish, stigma-lobes 3–6, exserted, to 4.5 mm. Fruit 2–4 cm diam., solitary or
clustered-proliferous, globose to obovoid or pear-shaped, purplish, red, orange-red or pale yellow, with
areoles bearing conspicuous clusters of dark brown glochids; funicular pulp fibrous, white, yellowish or
deep red (fide Britton & Rose), placenta greenish. Seeds 1–5, mostly 2, per fruit, to 8–10 mm.
Chromosome number: 2n = 22 (Pinkava 2002: 61, Table 1).
Widespread southern neotropical element: restinga, drier phases of Mata atlântica, agreste (mata de cipó),
caatinga (especially along temporary water courses), mata de brejo, mata seca (on limestone), mata de galeria
and mata do planalto, especially on deep sandy substrates and as a lithophyte, near sea level to c. 1000 m,
western Paraíba, eastern and central-southern Pernambuco, Alagoas, Sergipe, north-western, northern
and eastern Bahia, north-eastern and central-southern Minas Gerais and Espírito Santo; semi-humid
forests of extra-Amazonian Brazil; Atlantic drainage of Andes eastwards (Peru, Bolivia, northern
Argentina, Paraguay). Map 18.
paraíba: Mun. Cazajeiras, plant seen growing outside pharmacy in the town on road near eastern access
to highway BR 230, ex Fazenda Capoeiras, east of town, where said to be frequent and increasing
(property of owner of pharmacy), 3 Apr. 2000, E.A. Rocha et al. (K, photo).
pernambuco: Mun. Gravatá, Russinha, 27 Jan. 1933, Pickel 4480 (IPA), l.c., Pickel 3197 (IPA, fide
Andrade-Lima 1966: 1454); Mun. Caruaru, 18 May 1992, F.A.R. Santos 28 (HUEFS, PEUFR), l.c., 9
km NE of town, 8°14'21"S, 35°54'52''W, Dec. 1999, E.A. Rocha et al. (K, photos); Mun. Inajá, Reserva
Biol. Serra Negra, 17 Sep. 1995, A. Laurênio et al. 198 (K, PEUFR).
alagoas: W Alagoas, Mun. Mata Grande, 20 Dec. 1974, Andrade-Lima 74-7773 (IPA); S Alagoas, Mun.
Piaçabuçu, Ponta da Terra, 17 Nov. 1987, G.L. Esteves et al. 1954 (SPF).
sergipe: Mun. Nossa Senhora da Glória, Faz. Olhos D’agua, 30 Nov. 1981, G. Viana 261 (ASE 1778);
Mun. Frei Paulo, 5 km após o povoado de Mocambo, 26 Mar. 1981, M. Fonseca et al. 452 (ASE 851);
Mun. Simão Dias, 3 Feb. 1975, A.C. Barreto s.n. (ASE 288); l.c., Faz. Mercador, 30 Oct. 1981, G.N.
Silva 37 (ASE 1651); Mun. Riachão do Dantas, Faz. São José, 17 June 1982, E. Gomes 77 (ASE 2566);
l.c., Mata de Riachão, 27 Aug. 1982, G. Viana 632 (ASE 2742); l.c., Faz. Salobre, 2 Feb. 1983, E. Gomes
179 (ASE 3008), 19 Feb. 1987, G. Viana 1679 (ASE 4628).
bahia: NW Bahia, Mun. Formosa do Rio Preto, 4 km from Rio Riachão towards the town, white
BRASILIOPUNTIA
sandy substrate, 13 Oct. 1989, Walter et al. 478 (K); N Bahia, Mun. Senhor do Bomfim, c. 12 km N of
town, W of Estiva, Serra da Jacobina, below TV mast, 1 Mar. 1974, Erskine 132, cult. RBG Kew 1974-
1138 (K); Mun. Morro do Chapéu, a few km S of the summit, Apr. 2002, M. Machado (obs.); E Bahia,
Mun. Serrinha, 1 Dec. 1968, A.L. Costa s.n. (ALCB, in spirit); Mun. Lamarão, 48 km N of Feira de
Santana on road BR 116, 2 Nov. 1972, J. Ratter (K, photos); Mun. (?), ‘Pimenta’, between Feira de
Santana and Serrinha, 20 July 1959, Gomes & Laboriau 849 (RB); Mun. Ipirá, 12°22'S, 39°41'W, 1 Oct.
1985, L.R. Noblick 4400 (HUEFS); Mun. Serra Preta, Faz. Manoino, 12°9'S, 39°19'W, 7 Dec. 1992, L.P.
de Queiroz et al. 2923 (K, MBM, HUEFS); Mun. Feira de Santana, 16 Feb. 1994, F. França 949 (HUEFS);
l.c., 12°10'S, 39°11'W, 13 Nov. 1986, L.P. de Queiroz 1031 (HUEFS); l.c., Serra de São José, 12°38'S,
39°28'W, 18 June 1985, H.P. Bautista & G.C.P. Pinto 1024 (HRB, CEPEC, HUEFS); Mun. Rui
Barbosa, road BR 407, Faz. Buriti, 13 Oct. 1978, A.P. Araújo 74 (HRB, RB); Mun. Cachoeira, forest
in front of EMBASA, July 1980, Iscardino et al. in Grupo Pedra do Cavalo 462 (CEPEC, ALCB, HRB,
BAH, HUEFS); Mun. Castro Alves, Km 22 on road BR 242, 11 May 1975, A.L. Costa s.n. (ALCB, in
spirit); Mun. Itiruçu, 22 Jan. 1965, R.P. Belém & J.M. Mendes 220 (RB); Mun. Jaguaquara [Toca da
Onça], 27–29 June 1915, Rose & Russell 20068 (US, NY); Mun. Uruçuca, Serra Grande village, in
garden, probably from a wild plant in nearby restinga forest, 7 Feb. 1995, Taylor (obs.); Mun. Vitória da
Conquista, 12 & 20 km S of town, 17 & 18 April 2003, Taylor & Zappi (K, photos); Mun. Itapetinga,
15°13'S, 40°10'W, 24 Nov. 1979, A.P. Araújo 174 (HRB, RB); Mun. Itaju do Colônia, Km 2 on road
to Pau Brasil, 24 Jan. 1969, T.S. dos Santos 350 (K, CEPEC); l.c., Lagoa Bonita, 11 Feb. 1975, A.L. Costa
s.n. (ALCB, in spirit); Mun. Potiraguá, 15°23'S, 39°58'W, 2001, J. Jardim 3144 (CEPEC); Mun. Itapebi,
Faz. Ventania, Ventania–Itapebi road, 8 Nov. 1967, R.S. Pinheiro 378 & T.S dos Santos 41 (CEPEC); l.c.,
between Itapebi and Belmonte, 13 Apr. 1967, Castellanos 26354 (CEPEC); Mun. Itamaraju, c. 5 km NW
of town, Faz. Pau-brasil, 3 July 1979, A. Mattos Silva et al. 554 (K, RB, CEPEC).
minas gerais: NE Minas Gerais, Mun. Itinga, 17 Oct. 1984, Rizzini & Mattos-F. 38 (RB, UFMT);
cent. & S Minas Gerais, Mun. Santana do Riacho, margin of the Rio Cipó, growing as hedge, 19 Nov.
1989, Zappi 192 (HRCB, SPF); Mun. Matozinhos, Faz. Cauaia, 31 Oct. 1996, Lombardi 1446 (BHCB,
K); Mun. Lagoa Santa, A.P.A. Carste de Lagoa Santa, Oct. 1995 to Feb. 1996, Brina & Costa in BHCB
32769 (BHCB, K); Mun. Entre Rios de Minas, Fazenda de Pedra, Nov. 1969, L. Krieger 7764 (CESJ);
Mun. Arcos, fide Uebelmann (1996): HU 956.
espírito santo: Mun. Colatina, 15 km S of São Domingos towards Colatina, 16 Dec. 1990, Taylor & Zappi
784 (K, HRCB, ZSS, MBML); Mun. Linhares, Reserva Florestal da CVRD, Estação Flamengo, 20 Nov.
1987, D.A. Folli 668 (Univ. Fed. ES); Mun. Santa Teresa, S. J. de Petrópolis, Escola Agrotécnica Federal, 8
Oct. 1985, W. Boone 825 (MBML, HRCB); l.c., Barra de Santa Júlia, 14 May 1986, H.Q.B. Fernandes & W.
Boone 1975 (MBML, HRCB); Mun. Vitória, campus of Univ. Federal, May 1990, Taylor & Zappi (obs.);
Mun. Guarapari, Km 32 on road ES 060, 12 Sep. 1987, O.J. Pereira 1048 (Univ. Fed. ES); Mun. Piúma,
Monte Agá, south of the town, 20°52'23"S, 40°46'24''W, 26 Nov. 1999, Zappi et al. 461 (UEC, K).
rio de janeiro: Mun. São Francisco do Itabapoana, Faz. São Pedro, 21°24'S, 41°4'W, H.C. de Lima et
al. (poster presentation at 1998 Congresso Nac. de Botânica, Salvador, BA); Mun. Santa Maria Madalena,
access road to the town, 22°0'55"S, 42°7'30''W, 21 Nov. 1999, Zappi et al. 354 (K, UEC).
conservation status. Least Concern [LC] (1); PD=4, EI=1, GD=2. Short-list score (1×7) = 7. Much
of its habitat, especially the Atlantic Forest and agreste, has been destroyed in Brazil, but it does occur in
some national parks and has a wide distribution in the southern Neotropics, although its status outside of
Brazil is little known.
This widespread species, which is broadly circumscribed here, is rather variable in fruit
shape and colour and in the number, shape and colour intensity of its perianth-segments.
In the population sampled as Erskine 132 (see above) pericarpel shape varied from globose
to very elongate or flattened and across the full range of the taxon fruit colour does not
appear to present a consistent geographical pattern (there are disjunct occurrences of both
yellow and reddish to purple colorations). At least some forms from the Brazilian
Nordeste have reddish, ovoid fruit, and one such was distinguished as Opuntia bahiensis
OPUNTIA
by Britton & Rose (1919), whom, it seems, assumed that the type of Cactus brasiliensis
came from Rio de Janeiro, where the species has globose to depressed, yellow fruit.
Willdenow, however, did not state the origin of the plant grown at Berlin, but made
reference to Piso, ie. Historia naturalis Brasiliae (Piso 1648). One of Piso’s illustrations has
now been designated as lectotype and, since Piso was based at the Dutch colony at
Recife, Pernambuco (Stafleu & Cowan 1983: 276) and a coloured copy of the same
illustration in the contemporary work of Marcgraf (see Whitehead & Boeseman 1989:
tab. 3a) shows red fruits, it is clear that the name C. brasiliensis should be applied in its
strictest sense to a red-fruited form from the Nordeste (Plate 9). Therefore, Britton &
Rose’s Opuntia bahiensis would be a synonym, even if the species was interpreted in a
narrow sense (likewise Brasiliopuntia subacarpa Rizzini & A. Mattos).
*5. NOPALEA Salm-Dyck
Cact. hort. dyck. 1849: 63–64 (1850). Lectotype: Cactus cochenillifera L.
Seed-morphology/anatomy indicates that this small genus is very closely related to Opuntia
sensu stricto (Stuppy 2002), but it differs markedly in flower- and pollen-morphology
(Leuenberger 1976). Only the following introduced species is found in Eastern Brazil:
*1. Nopalea cochenillifera (L.) Salm-Dyck, Cact. hort. dyck. 1849: 64 (1850).
Lectotype (Howard & Touw 1982: 173): Dillenius, Hort. eltham.: tab. 297, fig. 383
(1732).
Cactus cochenillifera L., Sp. pl.: 468 (1753). Opuntia cochenillifera (L.) Mill., Gard. Dict. ed. 8: no. 6 (1768).
vernacular names. Palmatória, Palma, Palma-miuda, Palma-doce, Palma-de-engorda.
Shrubby, to 3 m tall, main stem with segments flattened at first, cylindric in age. Branches erect,
sometimes pendent, stem-segments (10–)12–26 × (3–)8–10 cm, narrowly elliptic or obovate, bright
green, sometimes glaucous. Areoles 25–30 mm apart, with white felt; spineless. Flowers from the apex
of the terminal segments, 5–6 × 2–2.5 cm; pericarpel obovoid, to 30 mm, green, with areoles with white
wool and few, dull pink bract-scales; outer perianth-segments to 15 mm, lanceolate, deep pink, erect,
innermost 20 mm, lanceolate to spathulate, deep pink, erect, clasping the stamen filaments, these
exserted, dark pink; stigma-lobes long-exserted, green. Fruit rarely formed in Brazil, to 4 cm, solitary,
obovoid, red; funicular pulp translucent, purplish, placenta greenish. Seeds few per fruit, to 4 mm diam.
Chromosome number: 2n = 22 (Pinkava 2002: 88).
Introduced: on cultivated land throughout North-eastern Brazil (native to Mexico and Central America).
The following are only representative records of the species and do not adequately account for its
extensive use in NE Brazil:
pernambuco: Mun. São Lourenço da Mata, São Bento, 4 Feb. 1933, Pickel 9216 (IPA); Mun. Poção,
21 Oct. 1991, F.A.R. Santos 1 (HUEFS); Mun. Pesqueira, 5 Nov. 1934, Pickel 3660 (IPA).
alagoas: Mun. Delmiro Gouveia, Oct. 1993, R. de Lyra-Lemos s.n. (MAC).
bahia: W Bahia, Mun. Vanderlei, between the town and road BR 242, 19 July 1989, Taylor & Zappi (K,
photos); N Bahia, Mun. Glória, povoado de Brejo do Burgo, 26 Aug. 1995, Bandeira 257 (K, HUEFS); E
Bahia, Mun. Conceição da Feira, Porto Castro Alves, 12°32'S, 39°5'W, Dec. 1980, Iscardino et al. in Grupo
OPUNTIA
Pedra do Cavalo 1037 (ALCB, HUEFS); Mun. Iramaia, 27.5 km NW of Pé-de-Serra, 5 Feb. 1991, Taylor
et al. (ZSS, photos); Mun. (?), between Gandu and Itaibó, 9 Oct. 1972, R.S. Pinheiro 2004 (CEPEC).
Like Opuntia ficus-indica (see below), this species is cultivated as cattle fodder for use
during drought and is also suitable as a host for the cochineal insect. Plate 10.2.
6. OPUNTIA Mill.
Gard. Dict. Abr. ed. 4: [unpaged] (1754). Type (cf. Leuenberger 1993): Cactus opuntia L.
(= O. ficus-indica (L.) Mill.).
Shrubs or occasionally treelike, 0.5–6.0 m; branches not dimorphic; stem-segments compressed, orbicular,
obovate to elliptic or rhomboid in outline, sometimes becoming cylindric through prolonged secondary
growth at base of plant; areoles in the axils of deciduous leaves, with glochids and felt, spines usually
present, or absent in some forms of No. 3. Leaves minute, subulate, sessile, fleshy, early deciduous. Flowers
solitary, mainly on the margin of the joint or born by proliferation from the flower receptacle; epigynous,
pericarpel globose or turbinate, with areoles and leaf-like, fleshy, green or coloured bract-scales; perianth
multiseriate, tube 0, outer perianth-segments coloured, fleshy, inner segments coloured, delicate,
spreading; stamens numerous, spreading, sensitive and closing around the style when touched; pollen with
reticulate exine (see Cact. Syst. Initiatives 14: 3 (2002)). Fruit solitary or clustered-proliferous; turbinate
with a pedicillate base or globose, with a broad, not very deep umbilicus and deciduous flower remnants;
funicular pulp translucent or opaque, fibrous; placenta white, greenish or coloured. Seeds few to
numerous, lenticular reniform, to 5 mm, enclosed in a hard, bony funicular envelope.
A genus of at least 150 species, even when (as here) narrowly circumscribed, ranging from
Canada to southern South America, but with only 1, marginally represented species
native to sandy places in the Mata atlântica zone within the core area of Eastern Brazil (plus
2 spp. introduced from the Northern Hemisphere). Plates 10.3–11.2.
1. Areoles with clusters of numerous golden spines; segments ± orbicular (cultivated and
naturalised) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . *2. dillenii
1. Areoles spineless or with few or dark brownish spines; segments obovate, elliptic or rhomboid . . 2
2. Segments dark green; fruits proliferating (cent.-E Minas Gerais and SE Espírito Santo;
occasionally cultivated elsewhere) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1. monacantha
2. Segments generally somewhat glaucous; fruits never proliferating (cultivated everywhere)
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . *3. ficus-indica
1. Opuntia monacantha Haw., Suppl. pl. succ.: 81 (1819). Type: ‘Lesser Antilles,
Barbados’, fide Haworth, not extant. Neotype (Leuenberger 2002): Lindley, Bot. Reg. 20:
tab. 1726 (1835). NB. Leuenberger’s neotypification was published on 26 March 2002,
c. 2 days before that proposed by Taylor et al. (2002).
Cactus urumbeba Vell., Fl. flumin.: 207 (1829). Opuntia urumbeba (Vell.) Steud., Nomencl. Bot. Ed. 2: 222
(1841) (‘O. umbrella’). Type: not extant. Lectotype (Taylor et al. 2002): Vellozo, Fl. flumin., Icones
5: tab. 32 (1831).
O. arechavaletae Speg. in Anales Mus. Nac. Hist. Nat. Buenos Aires, III, 4: 520 (1905) (‘O. arechavaletai’);
Scheinvar, Fl. Ilustr. Catarin., pt I (Cactáceas), 61, 63–68, tt. 27 & 28 (1985) (‘O. arechevaletai’).
Type: Uruguay, near Montevideo, Pan de Azúcar, Nov. 1903, ex Herb. J. Arechavaleta (LPS 14705,
holo., n.v.).
OPUNTIA
[O. vulgaris sensu Britton & Rose (1919) et auctt. p.p. (cf. Leuenberger 1993), non Mill. (1768).]
vernacular names. Urumbeba, Palmatória, Monducuru.
Shrubby, to 3 m tall, main stem flattened. Branches erect to decumbent; stem-segments 8–20 × 4–10 cm,
rhomboid to obovate, narrow at base and irregular in outline, dark green. Areoles 20–40 mm apart, with
white to pale brown felt; spines 0–1(–2), (5–)15–50(–60) mm, pungent, pale grey. Flowers arising at the
apex of the terminal segments or born by proliferation from the flower receptacle, 7–10 × 3–6 cm;
pericarpel narrowly turbinate, to 30–70 mm, bright green, with areoles with white wool and few, dull red
bract-scales; outer perianth-segments to 20 mm, lanceolate, bright yellow with deep red shades outside,
spreading; innermost 25 mm, spathulate, bright yellow, spreading; stamens exposed, sensitive; stigma-lobes
exserted; pollen exine reticulate (illustrated in Cact. Syst. Initiatives 14: 3 (2002)). Fruit to 5–10 cm,
solitary or clustered-proliferous, narrowly turbinate to obovoid, curved at base, greenish to reddish outside,
eventually falling to the ground and turning yellow; funicular pulp translucent, whitish, placenta yellowish.
Seeds many per fruit, to 4 mm diam. Chromosome number: 2n = 22, 32, 33 (Pinkava 2002: 103).
Southern humid/subhumid forest element: sand-dunes in open carrasco, c. 1000 m, central-eastern Minas
Gerais, and open restinga near sea level, southern Espírito Santo (and presumably northern Rio de
Janeiro); South-eastern and Southern Brazil; Paraguay, Uruguay and northern and eastern Argentina;
frequently naturalized or planted elsewhere (including North-eastern Brazil). Map 18.
sergipe: Mun. Arauá, Faz. Thuy, [cultivated], 17 Sep. 1982, E. Gomes 125 (ASE 2781).
bahia: Mun. Salvador (‘Bahia’), [cultivated], 24 Apr. 1918, Curran 23 (US, photo).
minas gerais: Mun. Diamantina, Mercês, Apr. 1959, Ritter 1037 (see Eggli et al. 1995: 454); Mun. Rio
Vermelho, Pedra Menina, Faz. do Sr José Batista, 7 Mar. 1988, Zappi & Prado in CFCR 11823 (SPF, K).
espírito santo: Mun. Itapemirim, Rodovia do Sol, S of Marataízes, 21°4'51"S, 40°50'24''W, 26 Nov.
1999, Zappi et al. 470 (UEC, K), 18 Jan. 1970, Krieger 7608 (CESJ, SPF).
This species has previously been known as either Opuntia monacantha Haw. or O. vulgaris
Mill., but both of these names are beset with nomenclatural difficulties. The former,
which is maintained here, was based on a collection from Barbados (Lesser Antilles),
whence only O. dillenii (Ker Gawl.) Haw. is currently recorded as native (Howard 1989).
In order to maintain its use for the plant now widely associated with Haworth’s name,
neotypification has been necessary. This assumes that the provenance data given by
Haworth were erroneous, or that O. monacantha as now understood had been introduced
to Barbados by the early years of the 19 th Century. In any case, Haworth’s brief and
unsatisfactory diagnosis does not agree with O. dillenii and so neotypification or rejection
are the only realistic options open. The above action seems marginally preferable to
taking up the next available, well-typified name, O. urumbeba (Vell.) Steud., which
unfortunately has never been used, even though its epithet repeats a distinctive,
vernacular name for the plant. The name Cactus monacanthos Willd. 1814 was cited by
Haworth with a question mark as a possible synonym of his Opuntia monacantha. This
indication of doubt rules out any consideration of Haworth’s name as a combination
based on that of Willdenow, which cannot be typified.
Opuntia vulgaris Mill. has been used in two quite different senses (O. humifusa Raf. and
O. monacantha), but is now considered to be a renaming of Cactus opuntia L., a taxonomic
synonym of O. ficus-indica (L.) Mill. (Leuenberger 1993).
O. monacantha has been recorded only rarely as a native plant within the core area
covered here, where it is at its north-eastern limit. The collections from central-eastern
OPUNTIA
Minas Gerais are rather disjunct on present knowledge, but at least that from Rio
Vermelho was from a population which appeared to be native and in an edaphically
appropriate sand-dune habitat. Similar disjunct populations are known from localities
remote from the coast in the state of São Paulo and it is probably this species that is
depicted growing near Lorena, São Paulo state, in Martius, Flora brasiliensis 1 (1, Tabulae
Physiognomicae): tab. VII (1841).
*2. Opuntia dillenii (Ker Gawl.) Haw., Suppl. pl. succ.: 79 (1819). Lectotype (Benson
1969: 126): l.c. infra, tab. 255 (1818).
Cactus dillenii Ker Gawl. in Edwards Bot. Reg. 3: tab. 255 (1818). Opuntia stricta (Haw.) Haw. var. dillenii
(Ker Gawl.) L. D. Benson in Cact. Succ. J. (US) 41: 126 (1969).
O. dillenii var. reitzii Scheinvar in Feddes Repert. 95: 277 (1984) & Fl. Ilustr. Catarin., pt I (Cactáceas),
81–88, tt. 26 (as ‘O. arechevaletai’) & 38–39 (1985). Holotype: Brazil, Bahia, Mun. Salvador, Itapoã,
6 Jan. 1982, I. & L. Scheinvar 2892 (MEXU, n.v.). Synon. nov. [An escape from cultivation.]
vernacular names. Palmatória, Palma-de-espinho.
Shrubby, to 3 m tall, main stem flattened. Branches erect; stem-segments 12–23 × 8–20 cm, orbicular or
obovate, grey-green, glaucous. Areoles 30–60 mm apart, with pale brown felt and a crown of glochids;
spines (3–)4–12, the largest flattened, to 30 mm, pungent, golden yellow or reddish. Flowers from the
apex of the terminal segments, 7–10 × 6–9 cm; pericarpel turbinate, 40–70 mm, grey-green, with areoles
with white wool and few, dull red bract-scales; outer perianth-segments to 30 mm, lanceolate, bright
yellow, spreading; innermost 35 mm, spathulate, bright yellow, spreading; stamens exposed; stigma-lobes
exserted, green. Fruit to 7–9 cm, solitary, turbinate to obovoid, purplish red; funicular pulp translucent,
purplish, placenta greenish. Seeds many per fruit, to 4 mm diam.
Introduced and sometimes escaping by the sea, becoming invasive; planted inland for hedging; native of
Caribbean coasts and southwards to Ecuador; widely introduced elsewhere in warmer regions.
paraíba: Mun. João Pessoa, commonly planted in and around the town and escaping in the vicinity of
Cabo Branco, 9 Feb. 1995, Taylor (obs.).
pernambuco: Mun. Jaboatão dos Guararapes, Oct. 1964, D. Andrade-Lima 64-4267 (IPA).
alagoas: Mun. Maceió, commonly planted near the sea shore, 13 Feb. 1995, Taylor (obs.).
sergipe: Mun. Aracaju, Feb. 1991, Taylor & Zappi (obs.).
bahia: Mun. Araci, 11.5 km N of town towards Tucano, planted as hedge, 5 Jan. 1991, Taylor et al.
1350 (K, HRCB, ZSS, CEPEC); Mun. Salvador, 1 Feb. 1989, G.C.P. Pinto 03/89 (HRB), 22 Apr. 1992,
F.A.R. Santos 31 (HUEFS).
Benson (1982: 497–501) treats O. dillenii as a variety of the scarcely spiny, more narrowly
segmented O. stricta (Haw.) Haw., and on such authority they were synonymized in the
first edition of the CITES Cactaceae Checklist (Hunt 1992b), but subsequently retained
as separate species (Hunt 1999a). Further studies are needed in the Caribbean region
where these taxa are native (cf. Howard 1989). An opuntia agreeing closely with O. stricta
is common about the region of Cabo Frio, RJ, just outside the area covered here, where
it appears to have been introduced and has successfully naturalised.
*3. Opuntia ficus-indica (L.) Mill., Gard. Dict. ed. 8: no. 2 (1768). Neotype
(Leuenberger 1991: 625): ‘Cactus articulato-prolifer, articulis ovatis-oblongis: spinis
setaceis. Lin. Spec. plant. 468. 16’ (S).
HYLOCEREUS
Cactus ficus-indica L., Sp. pl.: 468 (1753).

C. opuntia L., l.c. (1753). Opuntia vulgaris Mill., Gard. Dict. ed. 8 (1768). O. opuntia (L.) H. Karst., Deut.
Fl.: 888 (1882), nom. inval. Lectotype (Leuenberger 1993): Burser Herbarium 24: 26 (UPS).
? O. fusicaulis Griffiths fide Scheinvar, Fl. Ilustr. Catarin., pt I (Cactáceas): 68–73, tt. 31 & 32 (1985).
vernacular names. Palmatória, Palma, Palma-de-gado, Palma-gigante, Figo-da-Índia, Figo-da-

Espanha, Orelha-de-onça, Jamaracá, Jurumbeba.
Shrubby, to 6 m tall, main stem with segments flattened at first, cylindric and forming greyish bark in
age. Branches erect; stem-segments 20–40 × 12–20 cm or more, elliptic or obovate, grey-green,
sometimes glaucous. Areoles 20–50 mm apart, with white or brown felt; spineless or with occasional
spines to 10 mm. Flowers from the apex of the terminal segments, 6–9.5 × 5–6 cm; pericarpel obovoid,
to 50 mm, green, areoles with dull red bract-scales; outer perianth-segments to 20 mm, rounded,
yellow to orange, sometimes reddish, spreading; innermost 30 mm, lanceolate to spathulate, yellow to
orange, spreading; stamens exserted, filaments dark pink; stigma-lobes exserted. Fruit 5–10 × 4–8 cm,
solitary, obovoid, yellow or deep orange (sometimes reddish); funicular pulp usually orange or
yellowish, placenta greenish. Seeds many per fruit, to 5 mm diam. Chromosome number: 2n = 66, 88
(fide Kiesling 1999).
Introduced and increasingly planted about houses and on farms everywhere in North-eastern Brazil, often
at the expense of other cactus vegetation. According to Kiesling (1999) originally domesticated in Mexico
about 9000 years ago, having back-crossed with its putative wild ancestors, such as O. streptacantha and
O. megacantha; subsequently introduced throughout the warmer parts of the world and sometimes
becoming a serious pest. The following are representative records only, the plant being found in all states
in Eastern Brazil.
pernambuco: Mun. Triunfo, 21 Nov. 1992, F.A.R. Santos 57 (HUEFS); Mun. Alagoinha, 12 Feb.
1993, F.A.R. Santos 73 (HUEFS); Mun. Floresta, N of Serra Negra, 8°33'S, 38°1'W, 8 Apr. 2000, E.A.
Rocha et al. (K, photos).
bahia: Mun. Morro do Chapéu, W of Icó, Brejão (Formosa), 5 Aug. 2002, Taylor (K, photos); Mun.
Iramaia, 27.5 km NW of Pé-de-Serra, 5 Feb. 1991, Taylor et al. (ZSS, photos); Mun. Poções, road BR
116, 9 km S from the town, Fazenda Boa Esperança, 800 m, 5 Apr. 1988, L.A. Mattos Silva et al. 2334
(K, CEPEC).
minas gerais: Mun. Santana do Riacho, Serra do Cipó, Pensão Chapéu de Sol, cult., 22 Nov. 1989, Zappi
204 (HRCB, SPF); Mun. Entre Rios de Minas, Fazenda de Pedra, Nov. 1969, L. Krieger 7696 (CESJ).
An important source of cattle fodder during drought in the sertão. Also producing
delicious fruits for human consumption, but sometimes replacing stands of native cacti,
which is to be regretted.
5.5 CACTOIDEAE
Stems leafless or leaves replaced by minute fugitive scales; glochids lacking, spines never
barbed; pollen mostly tricolpate; seeds with the testa exposed.
The deletion of an approximately 700 base-pair intron in the chloroplast-encoded
gene rpoC1 supports a monophyletic origin for the subfamily Cactoideae of the Cactaceae
(Wallace & Cota 1996). The tribes and their generic composition adopted here is that
employed by Barthlott & Hunt (1993) modified on the basis of unpublished phylogenies
derived from DNA gene sequence data, presented at IOS meetings or otherwise
communicated by R. Wallace (Iowa State Univ., USA) since 1993.
HYLOCEREUS
Tribe HYLOCEREEAE Buxb.
DNA gene sequence data indicate that this tribe is one of the basal elements of
Cactoideae. Its Brazilian representatives are robust climbers or large epiphytes, with
flattened or trigonous stems and large or elongate flowers, the latter > 20 cm long.
7. HYLOCEREUS (A. Berger) Britton & Rose

in Contr. U.S. Natl. Herb. 12: 428 (1909). Type: Hylocereus triangularis (L.) Britton &
Rose (Cactus triangularis L.).
Including Selenicereus sect. Salmdyckia D. R. Hunt (1989) (Mediocactus Britt. & Rose, excl.
typ.).
Sprawling and clinging to rocks or climbing in trees; freely producing aerial roots; stems stout, trigonous
or 3-winged, irregularly constricted, the juvenile bearing numerous, fine bristle-like spines, the adult with
few, very short, stout, conical spines per areole. Flowers nocturnal, very large, to 32 × 25–34 cm, the
pericarpel and tube bearing conspicuous podaria subtending large bract-scales and/or spiny areoles, the
scales or spines sometimes becoming more pronounced on the reddish, ± globose fruit, spines ultimately
deciduous. Seeds to 3 mm, testa smooth, black.
The circumscription of Hylocereus adopted here is influenced by an unpublished phylogeny,

based on DNA gene sequence data, presented by R. Wallace (Iowa State Univ., USA) at
the IOS Congress, Bologna, September, 1996. This indicates that when Selenicereus is
circumscribed to include Sect. Salmdyckia D. R. Hunt (1989) it is paraphyletic in respect of
Hylocereus. Hitherto Hylocereus has been distinguished from the very similar trigonous-
stemmed members of Selenicereus sect. Salmdyckia on the basis of large scales versus spiny
areoles on the pericarpel, flower-tube and fruit. However, such a separation was weakened
from the start by the occurrence of occasional pericarpel spines in the otherwise typical
Hylocereus species, H. trigonus (Haw.) Saff. (syn. Cereus plumieri Gosselin, see Hunt 1984: 41
& fig. 12), native of the SE Caribbean. The gene sequence data imply that the shared
trigonous stems, whose similarity has been the cause of confusion between the two taxa
treated below, are in fact a character uniting them generically and that their floral differences
are perhaps only significant at subgeneric or sectional level. Plates 11.3–12.1.
1. Pericarpel, flower-tube and fruit bearing spines; stem edges green, never horny (native,
widespread) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1. setaceus
1 Pericarpel, flower-tube and fruit bearing large bract-scales only; stem edges often with a
horny margin (introduced, common near habitations) . . . . . . . . . . . . . . . . . . . . . . . *2. undatus
1. Hylocereus setaceus (Salm-Dyck) R. Bauer in Cact. Syst. Initiatives 17: 29 (2003).

Type: assumed not to have been preserved. Neotype (Bauer, l.c.): Pfeiffer & Otto,
Abbild. Beschr. Cact. 1: tab. 16 (1839).
Cereus setaceus Salm-Dyck in De Candolle, Prodr. 3: 469 (1828).
Selenicereus setaceus (Salm-Dyck) Werderm., Bras. Säulenkakt.: 87 (1933); Hunt in Bradleya 10: 31–32,
tab. V (1992). Mediocactus setaceus (Salm-Dyck) Borg, Cacti, ed. 2: 213 (1951).
Cactus triangularis Vell., Fl. flumin.: 206 (1829), Icones 5: tab. 24 (1831), non L. (1753).
HYLOCEREUS
Selenicereus rizzinii Scheinvar in Revista Brasil. Biol. 34: 249–256 (1974). Holotype: Brazil, Rio de
Janeiro, Araruama, Barra de Maricá, Rizzini 29L-1974 (RB).
[Mediocactus coccineus sensu (Salm-Dyck) Britton & Rose, Cact. 2: 211, fig. 290 (1920) non Cereus coccineus
Salm-Dyck ex DC.; see Hunt (1989).]
[C. triangularis sensu K. Schum. in Martius, Fl. bras. 4(2): 208–209 (1890), pro parte, tab. excl., non (L.) Haw.]
[C. undatus sensu Luetzelburg, Estud. Bot. Nordéste 3: 68, 111 (1926), non Haw.]
vernacular names. Rainha-de-noite, Mandacaru-de-três-quinas, Espada-de-jacaré.
Scrambling, climbing or epiphytic on trees or on rocks, sometimes forming large clumps, with basitonic
or mesotonic branching. Stems 3(–5)-angled, to 100 × 2–10 cm, sometimes constricted, very woody at
base, margins straight or undulate, never horny (unless very old), podaria sometimes decurved for
climbing; epidermis bright green, yellowish when exposed. Areoles 4–6 mm diam., 1.5–4.5 cm apart (in
juvenile forms as close as 4 mm or less, with very slender, bristly spines and hairs). Central spines 3–6,
conic, 1–6 mm, occasionally accompanied by bristly, more delicate radial spines. Flowers nocturnal,
20–32 × 22–25 cm; pericarpel greenish, densely covered in tubercles/podaria bearing areoles with 10–15
pungent spines; flower-tube infundibuliform, 10–13 × 1.5–2.5 cm, with scattered woolly areoles and
acute, spreading bract-scales; perianth-segments to 12 cm, outer segments reflexed, lanceolate to linear,
fleshy, dark red-green, inner segments to 43 mm wide, erect to spreading, lanceolate and fimbriate,
delicate, white; stamens exserted in relation to the perianth-segments, curved, anthers linear; style 15–17
× 0.5 cm, stigma-lobes c. 16, exserted; ovary locule ovoid to oblong in longitudinal section. Fruit ovoid,
to 9 × 7–8 cm, floral remnants deciduous; pericarp greenish then bright red when ripe, covered in spiny
areoles at first, spines to 2 cm; funicular pulp white. Seeds c. 2.5–3.0 mm, cochleariform, black, shiny;
testa-cells flat, smooth.
Widespread neotropical element: epiphyte, climber or lithophyte (on limestone or on gneiss/granite

inselbergs) in caatinga-agreste, cerradão, Mata atlântica, mata de brejo, mata de encosta, mata de planalto and
restinga, near sea level to c. 900 m, widespread in Eastern Brazil from northern Piauí southwards; Northern,
Central-western and Southern Brazil (southern Pará southwards to Mato Grosso do Sul and Paraná);
Central (?) and South America (southwards to E Bolivia, N Argentina and Paraguay). Map 20.
maranhão: (?), near Peritoró, on the road connecting Teresina (PI) with Belém (PA), Feb. 1995,
Taylor (obs.).
piauí: N Piauí, Mun. Cocal, mata de encosta, W of junction between roads PI 211 and BR 343, 19 Feb.
1995, J.B. Fernandes da Silva 365 (MG).
ceará: between Pacajús and Fortaleza, 16 Feb. 1966, Andrade-Lima 66-4446 (IPA).
pernambuco: Mun. São Lourenço da Mata, São Bento and Veneza, Apr. 1928, Pickel 1686 (IPA); Mun.
Moreno, Tapera, 29 Oct. 1932, Pickel 3120 (IPA); Mun. Bezerros, Serra Negra, 9 Nov. 1971, Andrade-
Lima 71-6746 (IPA); Mun. Escada/Vitória de Santo Antão, 6 Apr. 1973, Andrade-Lima 73-7299 (IPA).
alagoas: Mun. Marechal Deodoro, restinga S of Ponta do Cavalo Ruço, 14 Feb. 1995, Taylor & Zappi
(K, photos); Mun. Piaçabuçu, restinga, 30 Sep. 1981, R.F.A. Rocha s.n. (MAC).
sergipe: Mun. Carmópolis, Faz. Sta Bárbara, 1 Feb. 1983, E. Gomes 169 (ASE 2998).
bahia: Mun. Jaguarari, Itumirim, 8 June 1915, Rose & Russell 19813 (US); Mun. Rio Real, 18 km N
of Esplanada, 10 Feb. 1991, Taylor & Zappi (obs.); Mun. Morro do Chapéu, in forest at base of the
Cachoeira do Ferro Doido, 2002, fide M. Machado; l.c., a few km S of the summit of Morro do Chapéu,
Apr. 2002, M. Machado (obs.); Mun. Candeal, 18 Jan. 1994, M.A.S. das Neves 90 (HUEFS); Mun. Iaçu,
27 Apr. 1994, L.P. de Queiroz 3865 (HUEFS); Mun. Salvador, on trees in Praça Visconde de Cairu,
beside the Mercado Modelo, 2 Aug. 1998, Taylor (obs.); Mun. Rio de Contas, 16 km E of town towards
Jussiape, 25 Nov. 1988, Taylor & Zappi in Harley 25548 (K, SPF); Mun. Iramaia, 32 km N of Contendas
towards Maracás, 4 Feb. 1991, Taylor et al. 1548 (K, HRCB, ZSS, CEPEC); Mun. Ituaçu, c. 2 km SW
of town, limestone, 18 Aug. 1988, Eggli 1194 (ZSS); Mun. Ilhéus, Castelo Novo, Faz. Santa Rita, 12
Dec. 1968, Castellanos 26952 (HB); Mun. Vitória da Conquista, c. 7 km N of town, 3 km W of road
BR 116, 17 Aug. 2002, M. Machado (obs.); l.c., 12 km from town towards Itambé, 16 Aug. 2002, M.
Machado (obs.); Mun. Itaju do Colônia, Km 8 on road from Itaju to Pau Brasil, 13 Jan. 1969, T.S. dos
Santos 359 (CEPEC).
HYLOCEREUS
minas gerais: Mun. Januária, Pandeiros, fide Uebelmann (1996): HU 719; Mun. Pedra Azul, 10 km E
of town towards Almenara, 16°8'S, 41°12'W, 19 Oct. 1988, Taylor & Zappi in Harley 25189 (K, SPF);
Mun. Montes Claros, 13 km N of town towards Januária, 16°38'S, 43°55'W, 6 Nov. 1988, Taylor & Zappi
in Harley 25511 (K, SPF); Mun. Augusto de Lima, Santa Barbara, 6 Aug. 1988, Eggli (ZSS, photo); Mun.
Santana do Riacho, calcareous rocks at foot of Serra do Cipó, nr Cardeal Mota, 28 Oct. 1988, Taylor &
Zappi (obs.); Mun. Iapu, on road BR 458, 14 km W of crossing to the town, 13 Dec. 1990, Taylor &
Zappi 752 (K, HRCB, ZSS, BHCB); Mun. Aimorés, Fazenda do Sr Hulbert Schumacher, 21 Oct. 1997,
M.F. de Vasconcelos s.n. (K, BHCB); Mun. Matozinhos, Faz. Cauaia, 31 Oct. 1996, Lombardi 1466
(BHCB, K); Mun. Lagoa Santa, reported by Warming (1908: 146, 149, fig. 36; cf. Warming & Ferri
1973); Mun. Santa Bárbara, s.d., C. Paganini, cult. Santuário do Caraça, Catas Altas, Aug. 2001, Taylor
(K, photo); Mun. São João del Rei, Oct. 1970, Krieger 7529 (CESJ); Mun. Laranjal, 8 km N of town
towards Muriaé, 18 Dec. 1990, Taylor & Zappi 793 (K, HRCB, ZSS, BHCB); on road BR 040, between
Juiz de Fora and Três Rios (RJ), ‘Paraibuna’, 25 July 1991, Zappi (obs.); Mun. Além Paraíba, road BR
116, 16 Oct. 1985, Hatschbach & J.M. Silva 49858 (MBM).
espírito santo: Mun. São Mateus, road BR 101, Ponta do Ipiranga, 13 Oct. 1992, Hatschbach et al.
58057 (MBM); Mun. Mantenópolis, 5 km NE of town towards Mantena, 15 Dec. 1990, Taylor & Zappi
778 (K, HRCB, ZSS, MBML); Mun. Barra de São Francisco, 2 km W of town, 5.5 km E of the border
ES/MG, 16 Dec. 1990, Taylor & Zappi 779 (K, HRCB, ZSS, MBML); Mun. Santa Teresa, fide
Uebelmann (1996): HU 244; Mun. Afonso Cláudio, road to Itarana, 20°3'8"S, 41°3'3''W, 24 Nov. 1999,
Zappi et al. 411 (K, UEC), l.c., 6 km NE of town towards Serra Pelada, 17 Dec. 1990, Taylor & Zappi
790 (K, HRCB, ZSS, MBML); Mun. Vila Velha, Jacaranema, 12 Oct. 1988, O.J. Pereira 1852 (Herb.
Univ. Fed. ES); Mun. Guarapari, Setiba, near Laguna de Corais, restinga, 12 Sep. 1987, O.J. Pereira 1049
(Herb. Univ. Fed. ES); Mun. Itapemirim, Rodovia do Sol, S of Marataízes, 21°4'51"S, 40°50'24''W, 26
Nov. 1999, reported growing with Opuntia monacantha (Zappi et al. 470).
rio de janeiro: Mun. Santa Maria Madalena, access road to the town, 22°0'55"S, 42°7'30''W, 21 Nov.
1999, Zappi et al. 355 (K, UEC).
conservation status. Least Concern [LC] (1); PD=2, EI=1, GD=1. Short-list score (1×4) = 4. Much
of its presumed former habitat has been destroyed in North-eastern Brazil, but its range and frequency
can be judged from the above-cited records.
In its vegetative state this widely distributed, native species is sometimes confused with
the introduced H. undatus (see below), but the pericarpel and immature fruit is spiny and
the stem-margins seldom horny. Its fruits are edible when red and mature. It may be
related to a taxon cultivated in Colombia for the export of its yellow, egg-shaped, edible
fruit, and to another known from Roraima in Northern Brazil (A. R. Pontes s.n., K,
photo.), which is probably also the same as the plant from the Guianas illustrated and
discussed by Leuenberger (1997: 48–51) as possibly identifiable with H. extensus (Salm-
Dyck ex DC.) Britton & Rose (Cereus extensus DC., tantum quoad typ.; Selenicereus
extensus (Salm-Dyck ex DC.) Leuenb.). The Colombian plant, known in the British
supermarket trade as ‘Pitaya’, has been referred to Selenicereus (sect. Salmdyckia)
megalanthus (K. Schum. ex Ule) Moran, based on a collection from the Amazonian
drainage of Peru (Hunt 1992a). It is possible that, when better understood, Hylocereus
setaceus will prove to be the oldest name for a widespread neotropical species comprising
most of the above as regional subspecies, amongst which Selenicereus tricae D. R. Hunt
(1989) might also figure, extending the range of this complex to Central America
(including southern Mexico).
It is likely that H. setaceus, sens. str., is significantly under-recorded in the northern half
of its range within Eastern Brazil, as is suggested by the isolated collections from northern
HYLOCEREUS
Piauí and Ceará, which are more than 700 km distant from the next nearest records, in
northern Bahia and eastern Pernambuco. Plants resembling the species have been seen in
north-central Maranhão, near Peritoró, while travelling the road connecting Teresina (PI)
with Belém (PA), and it may be expected to occur in Paraíba, perhaps in brejo forest. Its
southern limit in Brazil is indicated as Paraná above, but Braun & Esteves Pereira (2002:
79) suggest it reaches Rio Grande do Sul.
*2. Hylocereus undatus (Haw.) Britton & Rose in Britton, Fl. Bermuda: 256 (1918);
Cact. 2: 187–188 (1920). Type: a plant cultivated at the London Horticultural Society,
originating from China, assumed not to have been preserved. Neotype (Taylor 1995:
119–120; superseding that designated by Scheinvar 1988): Bot. Mag. 44: tab. 1884
(1817), as ‘Cactus triangularis’.
Cereus undatus Haw. in Philos. Mag. Ann. Chem. 7: 110 (1830).
Scrambling, climbing or epiphytic, forming large woody masses of branches, with basitonic or mesotonic
branching. Stem-segments trigonous/3-winged, to 100 × 4–7(–8) cm, sometimes constricted, narrow and
very woody at base, margins crenate, horny in age; epidermis bright green, yellowish when exposed to
full sun. Areoles 4–6 mm diam., (2.5–)3–6(–8) cm apart, with 2–6 conic, c. 4 mm spines. Flower-bearing
areoles in the axils of the stem’s marginal crenations; flowers nocturnal, very showy, to 30 × 34 cm;
pericarpel greenish, densely covered in rounded, broad-based, green or reddish bract-scales, but lacking
spines; flower-tube infundibuliform, c. 12 × 1.5–2.5 cm, covered in broadly triangular, adpressed, to
acute, spreading bract-scales; perianth-segments up to 14 cm, outer segments reflexed, lanceolate to
linear, fleshy, dark red with green bases, inner segments erect to spreading, spathulate and apiculate,
fimbriate, delicate, white; stamens disposed in a broad ring around the style, sometimes curved, anthers
linear; style 14–20 × 0.8 cm, stout, stigma-lobes c. 20 or more, exserted; ovary locule oval in longitudinal
section. Fruit ovoid to globose, to 12 cm diam., flower remnants deciduous; pericarpel covered in
broadly triangular, fleshy, orange-yellow to bright red bract-scales; funicular pulp white. Seeds c. 3 mm,
cochleariform, black, shiny; testa-cells flat, smooth (Barthlott & Hunt 2000: pl. 20.5–6).
Introduced as garden plant and sometimes escaping into roadside trees and maritime scrub; perhaps native
in Mexico and Central America, commonly introduced elsewhere in the tropics and subtropics worldwide.
pernambuco: Mun. Nazaré da Mata, 21 Feb. 1955, J.C. Moraes (SPSF 4788); Mun. Caruaru, W of
town, on roof of house beside road BR 232, 11 Feb. 1995, Taylor (obs.); Mun. Ibimirim, growing outside
a house in the town, 8 Apr. 2000, E.A. Rocha et al. (obs.).
alagoas: Mun. Maceió, Avenida Fernandes Lima, covering the top of a wall, 16 Feb. 1995, Taylor (obs.).
bahia: Mun. Morro do Chapéu, 3 Aug. 2002, Taylor (K, photos); Mun. Marcionílio Sousa, Machado
Portella, 1915, Russell s.n. (US); Mun. Ilhéus, Pontal, near Praia do Sul, 20 m from beach, 20 Nov. 1981,
L.A. Mattos-Silva 1386 (CEPEC).
minas gerais: Mun. Santana do Riacho, Serra do Cipó, 19 Nov. 1989, Zappi et al. 192C (HRCB, SPF).
espírito santo: Mun. Colatina, 21 km S of Ângelo Frechiani towards Colatina, in a roadside tree, 16
Dec. 1990, Taylor & Zappi 787 (HRCB, K, ZSS).
Widely cultivated in South America, this species may be found at the sites of old houses,
where it often scrambles to the tops of trees. However, it does not seem to be able to
reproduce by means of seed in Eastern Brazil, perhaps because all or most individuals
belong to the same clone.
Some published reports of this species actually refer to the native H. setaceus (see above).
SELENICEREUS
*8. SELENICEREUS (A. Berger) Britton & Rose
in Contr. U.S. Natl. Herb. 12: 429 (1909).
Including Cryptocereus Alexander; Selenicereus sect. Cryptocereus (Alexander) D. R. Hunt

(1989).
An ill-defined genus of c. 11 species, native of Mexico, the Caribbean and northern

South America. Typically, the genus has cylindrical, scandent stems with 5 or more low
ribs (Sect. Selenicereus), but the species encountered in Brazil have flattened stems. One
such is native of Northern Brazil, Selenicereus (sect. Strophocactus) wittii (K. Schum.) G. D.
Rowley (Amazônia, igapó), but some authors now refer this back to Strophocactus Britton
& Rose. A detailed account of this species can be found in Barthlott et al. (1997).
Only the following Mexican plant is encountered in Eastern Brazil:
*Selenicereus anthonyanus (Alexander) D. R. Hunt in Bradleya 7: 93 (1989), native

of southern Mexico, is the cactus most frequently cultivated as a house plant (in pots and
hanging baskets) in Eastern Brazil. It is more rarely planted outdoors, where it has been
observed climbing trees in a semi-naturalized state in South-eastern Brazil, outside the
area treated here. It resembles the following genus in its vegetative state (especially the
Mexican Epiphyllum anguliger), but has flowers with a stouter, much shorter tube and
bristly pericarpel.
9. EPIPHYLLUM Haw.
Syn. pl. succ.: 197 (1812). Type and only species native of Eastern Brazil:
1. Epiphyllum phyllanthus (L.) Haw., l.c. (1812). Lectotype (Leuenberger 1997):

Dillenius, Hort. eltham.: tab. 64, fig. 74 (1732).
Cactus phyllanthus L., Sp. pl. 1: 469 (1753).
Epiphytic, forming woody clumps, with basitonic and mesotonic branching. Stems flat, alate, sometimes
trigonous at base, 15–100(–150) × 3–6(–8) cm, to 1 cm thick, lanceolate, narrow at base, margins crenate,
teeth 2.5–4.5 cm apart, apex obtuse, midrib rather woody; epidermis green, yellow-green to reddish or
purplish when exposed, margins sometimes horny in age; areoles spineless, but sometimes bristly on old
branches. Flower-bearing areoles in the axils of the teeth; flowers nocturnal, 15–25 × 4–6(–8) cm;
pericarpel greenish, with areoles and minute triangular bract-scales, ridged; flower-tube more than 14 cm,
cylindric, very narrow, 6–8 mm diam., with scattered linear bract-scales; perianth-segments 1.5–3.0 cm,
reflexed, lanceolate to linear, pale pink to white; stamens disposed in a ring around the style, anthers
linear; style 14–20 cm, stigma-lobes 6–8, exserted; ovary locule narrowly oblong in longitudinal section.
Fruit narrowly ovoid to pear-shaped, apiculate, to 10 × 3.5 cm, dehiscent by lateral slit, flower remnants
deciduous; pericarpel ridged, bright red or pink, with acute bract-scales; funicular pulp white,
mucilaginous. Seeds 4–4.5 mm, cochleariform, black, shiny, with a broad white hilum; testa-cells flat,
smooth (Barthlott & Hunt 2000: pl. 21.1–2).
Widespread neotropical element: epiphyte in Mata atlântica, caatinga-agreste, mata do planalto, mata ciliar and
cerrado, near sea level to at least 1300 m, ± common throughout the more humid parts of Eastern Brazil;
EPIPHYLLUM
Neotropics from Central America southwards. It is the most widespread cactus species in Eastern Brazil
and only a few representative collections/observations are cited here. There are currently no records from
Rio Grande do Norte.
ceará: Mun. Maranguape, Serra de Maranguape, 24 Nov. 1955, Andrade-Lima 55-2404 (IPA 8631).
pernambuco: Mun. Nazaré da Mata, 7 Mar. 1955, J.C. Moraes (SPSF 4951); Mun. Moreno, Tapera, 17
Jan. 1930, B. Pickel 2309 (IPA); Mun. Recife, Jardim Botânico, 21 Mar. 1966, E. Tenório 66-86 (IPA),
l.c., 30 Mar. 2000, Taylor & Griffiths (obs.).
alagoas: Mun. Colônia Leopoldina, 1 km from border with Pernambuco, 21 Nov. 1966, D. Andrade-
Lima 66-4739 (IPA); Mun. Murici, 15 Feb. 1995, Taylor & Zappi (obs.).
bahia: Mun. Rio Real, 18 km N of Esplanada, 10 Feb. 1991, Taylor & Zappi (obs.); Mun. Itatim, 1997,
F. França 1494 (HUEFS; K, photo.); Mun. Conceição de Almeida, 8 Feb. 1981, G.C.P. Pinto 45/81
(HRB); Mun. (?), Ilha de Itaparica, 22 Dec. 1976, Pereira de Souza s.n. (ALCB); Mun. Ilhéus, área do
CEPEC, Km 22 Ilhéus–Itabuna road, 7 Dec. 1988, T.S. dos Santos 4431 (CEPEC); Parque Nacional de
Monte Pascoal, 21 Mar. 1968, S.G. da Vinha & T. S. dos Santos 83 (CEPEC).
minas gerais: Mun. Monte Azul/Espinosa, c. 12 km E of Monte Azul, Serra Geral, top of serra on
path to village of ‘Gerais’, 28 Jan. 1991, Taylor et al. 1477 (BHCB, HRCB, K, ZSS); Mun. Mirabela, c.
10 km SSE of town on road BR 135, 10 Aug. 1988, Eggli 1132 (ZSS); Mun. Grão Mogol, Riacho
Ribeirão, 16°33'S, 42°54'W, 28 May 1988 & 16 June 1990, Zappi et al. in CFCR 12058, 13124 (SPF);
Mun. Rio Vermelho, Pedra Menina, Faz. Sr José Batista, 42°6'W, 17°54'S, 6 Mar. 1988, Zappi & Prado
in CFCR 11820 (SPF); Mun. Conceição do Mato Dentro, Serra do Cipó, 26 Apr. 1988, Zappi in CFSC
11178 (SPF); Mun. Laranjal, 6 Feb. 1971, Krieger 9960 (CESJ).
espírito santo: Mun. Santa Teresa, Fazenda Santa Lúcia, 19°57'24"S, 40°32’37''W, 25 Nov. 1999,
Zappi et al. 438 (K, UEC).
rio de janeiro: Mun. Santa Maria Madalena, access road to the town, 22°0'55"S, 42°7'30''W, 21 Nov.
Frequency has been reduced due to destruction of forest formations, but nevertheless very widespread
and records for Eastern Brazil too numerous to list.
Widespread in South America in various habitats, this is perhaps the only epiphytic cactus
likely to be found in, or bordering on cerrado vegetation. Plate 12.2.
*Epiphyllum oxypetalum (DC.) Haw. from southern Mexico and Central America is
occasionally cultivated and has been recorded planted or escaped outdoors in the states of
Pernambuco: Mun. Moreno, 7 Mar. 1936, B. Pickel 4121 (IPA); Mun. Arcoverde, 14
May 1966, D. Andrade-Lima 66-18 (IPA); and Bahia: Mun. Vera Cruz, Ilha de Itaparica,
31 Mar. 1997, Aline 1 (ALCB). It is easily distinguished from E. phyllanthus by its very
long, slender, cylindric to angled, basal extension shoots and flowers to 27 cm in diameter,
the outer perianth-segments conspicuously reddish to deep pink.
Tribe ECHINOCEREEAE Buxb.
The tribal placement of the following genus is rather uncertain at present and awaits the
results of analysis of DNA gene sequence data. It is placed in this tribe, including the
former Leptocereeae, only because the genus from which it was separated, ie.
Acanthocereus, is currently placed there.
PSEUDOACANTHOCEREUS
10. PSEUDOACANTHOCEREUS F. Ritter
Kakt. Südamer. 1: 47 (1979). Type: P. brasiliensis (Britton & Rose) F. Ritter.
Literature: Taylor et al. (1992).
An isolated genus of only two species, comprising P. sicariguensis (Croizat & Tamayo) N.
P. Taylor (NE Colombia & NW Venezuela, Map 11) and the following (Plates 12.3–13.1):
1. Pseudoacanthocereus brasiliensis (Britton & Rose) F. Ritter, l.c. Type: Brazil,

Bahia, Mun. Marcionílio Sousa, Machado Portella, 1915, Rose & Russell 19903 (US,
lecto. designated here; NY, lectopara.).
Acanthocereus brasiliensis Britton & Rose, Cact. 2: 125 (1920). Cereus brasiliensis (Britton & Rose) Luetzelb.,
Estud. bot. Nordéste 3: 111 (1926).
Pseudoacanthocereus boreominarum Rizzini & A. Mattos in Revista Brasil. Biol. 46(2): 327 (1986). P.
brasiliensis f. boreominarum (Rizzini & A. Mattos) P.J. Braun & E. Esteves Pereira in Schumannia 3:
188 (2002). Holotype: Brazil, Minas Gerais, Mun. Itinga, Rizzini & Mattos-Filho 40 (RB 232057).
vernacular name. Catana-de-jacaré.
Shrubby, scrambling or decumbent, sometimes forming large clumps, with basitonic or mesotonic
branching. Roots tuberous. Stems erect to decumbent, 1.5–4.5 cm diam.; 2–6(–7)-ribbed, ribs high,
acute, somewhat undulate; epidermis dark green to greyish. Areoles 3–6 mm diam., 2–7 cm apart, felt
white, soon glabrescent, areolar growth ± indeterminate. Spines 8–14, central and radial difficult to
distinguish, 5–50 mm. Flower-bearing region of stem not differentiated; flowers nocturnal, 15–17 ×
11–12 cm; pericarpel green with dull pink tinge, bearing tubercles/podaria and areoles with
bristles/spines to 5 mm; flower-tube infundibuliform, 9–10 × 1.4–2.0 cm, bearing scattered areoles with
minute spines; perianth-segments to 60 mm, outer segments reflexed, lanceolate, fleshy, dull pink to
reddish brown, inner segments spreading, lanceolate to spathulate and fimbriate, delicate, white; stamens
exserted in relation to the perianth-segments, curved, anthers linear; style c. 100 × 3 mm, stigma-lobes
8–11, exserted; ovary locule rectangular to oblong in longitudinal section. Fruit globose, to 5.5–8.0 cm
diam., floral remnants deciduous; pericarp greenish, becoming yellow when ripe and falling to the
ground, smelling of pineapple, bearing podaria and deciduous areoles with white to pale brown felt;
funicular pulp white. Seeds c. 4.5 mm, cochleariform, light brown, dull; testa-cells flat, elongate, cuticular
folds coarse, dense (Taylor et al. 1992, Barthlott & Hunt 2000: pl. 9.1–2).
Eastern caatinga-agreste element: within and at the margins of caatinga-agreste, 40–700 m, east of the
Chapada Diamantina crestline in northern and central-eastern Bahia, and in the drainage of the Rio
Jequitinhonha (associated with gneiss/granite inselbergs and lajedos) of north-eastern Minas Gerais
(apparently disjunct, but possibly under-recorded through destruction of habitat in intervening areas).
Endemic to the core area of Eastern Brazil. Map 20.
bahia: Mun. Jaguarari, Ritter 1230 (SGO 125358, fide Eggli et al. 1995: 499); l.c., Catuni, 8 June 1915,
Rose & Russell 19819 (US, NY); Mun. Senhor do Bomfim [‘Villa Nova’], Zehntner in Rose 19735 (US,
NY); Mun. Itiúba, road to Filadélfia, 10°39'S, 39°44'W, 11 May 2002, Nascimento & Nunes 102
(HUEFS); Mun. Morro do Chapéu, near Ventura, by the Rio Jacuípe, 24 Dec. 1988, Taylor & Zappi
(obs.); Mun. Riachão do Jacuípe, 17 km NW of the town towards Nova Fátima, 7 Feb. 1970, A.L. Costa
s.n. (ALCB, in spirit); Mun. Candeal, beside road BR 324, 20.5 km NW of Tanquinho, 25 Apr. 1992,
Taylor & Zappi (obs.); Mun. Tanquinho, W. Uebelmann (mss.); Mun. Feira de Santana, Serra de São José,
Nov. 2001, Zappi (obs.); l.c., 8 km SW of the town on road BR 116, margin of Rio Jacuípe, 19 Feb.
1981, A.M. de Carvalho et al. 571 (CEPEC); Mun. Rui Barbosa, 7–8 km SE of the town on road to
village of Alagoas, 23 Aug. 1988, Eggli 1265 (ZSS); Mun. Itaberaba, 21 km SE of Rui Barbosa on road
LEPISMIUM
to Alagoas, 7 Feb. 1991, Taylor et al. 1579 (K, ZSS, HRCB, CEPEC); Mun. Conceição da Feira, NE of
Barragem de Bananeiras, 12°32'S, 39°5'W, 16 Feb. 1981, Iscardino et al. in Grupo Pedra do Cavalo 1053
(ALCB); Mun. Iaçu, fide Uebelmann (1996: HU 138); Mun. Marcionílio Sousa, Machado Portella, Dec.
1912, Zehntner in IFOCS 5037 (M), 1915, Rose & Russell 19903 (NY, US); Mun. Jaguaquara [Toca da
Onça], 27 June 1915, Rose & Russell 20107 (US).
minas gerais: Mun. Pedra Azul, 8 km W of the town towards road BR 116, 18 Oct. 1988, Taylor &
Zappi in Harley 25185 (K, SPF); Mun. Itinga, 2 km E of the town, S of Rio Jequitinhonha, 14 Dec. 1990,
Taylor & Zappi 768 (K, ZSS, HRCB, BHCB); Mun. Itaobim, 1 km W of Itaobim, 0.5 km N of Rio
Jequitinhonha, 16°34'S, 41°31'W, Taylor & Zappi in Harley 25531 (K, SPF); Mun. Coronel Murta, 5 km
SE of the town on road BR 342 to Araçuaí, 21 Feb. 1988, Supthut 8863 (ZSS).
conservation status. Vulnerable [VU C1 + 2a(i)] (2); PD=3, EI=1, GD=2. Short-list score (2×6) =
12. Much of its preferred habitat has been destroyed and its original range may have been more extensive
and less markedly discontinuous than it appears today.
This inconspicuous and, when out-of-flower, rather ugly plant appears to be of erratic
occurrence, which may in part be the result of the widespread destruction of the
agreste/caatinga, under whose shade it prefers to grow. While it is a variable species, there
are no reliable differences between plants from Bahia and Minas Gerais and P.
boreominarum Rizzini & A. Mattos cannot be justified even at infraspecific rank.
The report by Uebelmann (1996), under ‘HU 1197’, of a ‘Pseudoacanthocereus sp.’ from
Penedo, Alagoas [‘Sergipe’] is considered improbable.
Tribe RHIPSALIDEAE DC.
In terms of numbers of species, this is the second largest tribe of Cactoideae in Eastern
Brazil (after Cereeae). All the species treated here are epiphytes and/or lithophytes in the
core area of Eastern Brazil, with flowers < 6 cm long and seeds with a mucilage sheath
(either restricted to the hilum area or covering the entire surface).
11. LEPISMIUM Pfeiff.
in Allg. Gartenzeitung 3: 314–315 (1835). Lectotype: L. commune Pfeiff. (= L. cruciforme

(Vell.) Miq.).
Literature: Barthlott (1987), Barthlott & Taylor (1995: 44) — now excluding subgenera
Pfeiffera, Acanthorhipsalis & Lymanbensonia; see Nyffeler (2000c).
Epiphytic, branching mesotonic or basitonic, or branching subacrotonically only when damaged, freely
producing aerial roots; stem-segments often angled or flat with the areoles subtended by crenate or serrate
marginal teeth, not deciduous as units when old. Areoles sometimes densely hairy, scale leaves often clearly
visible; composite terminal areoles lacking. Flower-buds always somewhat erumpent, the flowers either
± immersed in the sunken stem areoles or pendent, often campanulate; stamens generally much shorter
than the perianth-segments, never exserted; pericarpel ridged or angled, rarely almost terete; tube
inconspicuous. Fruit globose, ellipsoid or turbinate; pericarp ridged when immature (except in No. 3), white,
magenta, red, dark wine-coloured or blackish.
LEPISMIUM
In the narrower sense suggested by Nyffeler’s recent molecular studies, a genus of only 6
species, from the eastern Andes (Bolivia & NW Argentina: 2 endemic, Map 14) and
south-eastern South America (Brazil, Paraguay, Uruguay, NE Argentina: 3 endemic).
Species of Lepismium sens. str. can be distinguished from other Brazilian Rhipsalideae by
the combination of a basi- to mesotonic (not acrotonic) branching pattern and non-
deciduous stem-segments. The genus is restricted to the Mata atlântica in Eastern Brazil.
Plates 13.2–14.1.
1. Distal parts of stems flattened, with conspicuous marginal teeth to at least 4 mm long . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1. houlletianum
1. Stems flattened or 3–6-winged/angled/ribbed, with strongly adpressed marginal teeth/
crenations to at most 2 mm long . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
2. Areoles woolly at anthesis and subsequently; flowers immersed in the areoles, 1 or more
per areole, whitish to deep pink (Pernambuco southwards) . . . . . . . . . . . . . . . . . . 3. cruciforme
2. Areoles not woolly; flowers 1 per areole, not sunken, whitish (SE Brazil) . . . . 2. warmingianum
1. Lepismium houlletianum (Lem.) Barthlott in Bradleya 5: 99 (1987); Barthlott &

N. P. Taylor, ibid. 13: 49, pl. 7 (1995). Type: Brazil, Houllet, s.n. (assumed not to have
been preserved). Neotype (Barthlott & Taylor 1995: 46): Gürke, Blühende Kakt. 2: tab.
111 (1909).
Rhipsalis houlletiana Lem. in Ill. Hort. 5, misc.: 64 (1858). R. houlletii Hook. f. in Bot. Mag. 100: tab.
6089 (1874). Type: as above.
R. regnellii G. Lindb. in Gartenflora 39: 118 (1890). Lepismium houlletianum f. regnellii (G. Lindb.) Barthlott
& N. P. Taylor in Bradleya 13: 47 (1995). Rhipsalis houlletiana var. regnellii (G. Lindb.) Kimnach in
Cact. Succ. J. (US) 68: 156 (1996). Based on syntypes. Lectotype (Barthlott & Taylor, l.c.): Minas
Gerais, Caldas, 1861, Regnell [III] 626 (S; S, NY, P, US, F, lectoparas.).
vernacular name. Rabo-de-arara.
Plant 1–2 m, pendent; basal shoots erect at first; stem-segments 15–40 × 2–7.5 cm (towards apex),
cylindric and woody at base, flattened distally and only to c. 2.5 mm thick (away from the vascular
midrib), margins strongly serrate-dentate, teeth mostly acute, not adpressed but sometimes incurved, to
6 mm or more long, epidermis dark green to glaucous, sometimes purple tinged at the stem margin;
seedlings with stems flattened at first. Areoles in the axils of the teeth, glabrous or with 1–3 minute
bristles. Flower-buds somewhat erumpent; flowers scented, pendent, c. 20 × 15–26 mm, perianth-
segments 10–15, white (yellowish post anthesis), c. 17 × 3 mm, not expanding fully; stamens in two
weakly defined groups, the innermost clustered around the style, the remainder spreading out against the
perianth and the largest slightly > half as long as the inner perianth-segments, white, but strikingly orange
reddish at base (yellowish at base in the forma regnellii); style white, exceeding stamens, stigma-lobes 3–6,
to c. 3 mm, outspread, white becoming yellow in age; pericarpel cylindric, c. 6–8 × 2.5–4 mm,
longitudinally ridged. Fruit 5–8 mm diam., ovoid to globose, magenta to deep pink or black. Seed
(Barthlott & Hunt 2000: pl. 1.7–8). Chromosome number: 2n = 22 (Barthlott 1976).
Southern humid forest element: epiphyte in Mata atlântica, including mata de neblina, 500–1900 m, central
and south-eastern Minas Gerais; common in South-eastern and Southern Brazil; North-eastern Argentina
(Misiones). Map 21.
minas gerais: Mun. Conceição do Mato Dentro, Rio Santo Antônio, 25 Apr. 1978, Martinelli 4298
(RB); Mun. Caparaó, Parque Nacional do Caparaó, 30 Apr. 1988, Krieger et al. in FPNC 93 (CESJ, SPF),
LEPISMIUM
15 June 1991, Hatschbach et al. 55545 (MBM); Mun. Lima Duarte, Parque Estadual do Ibitipoca, 8 July
1990, R.C. Oliveira 25194 (CESJ).
Least Concern taking its total range into account.
Luetzelburg (1925–26, 3: 111) reports this species from eastern Bahia, but it has not been
encountered subsequently and his record may be an error or misidentification.
As in the case of Schlumbergera kautskyi (see below), the pointed, marginal, stem-
segment teeth (podaria) in this species are assumed to function as, and represent the
equivalent of leaf drip tips. It is rather variable in stem and floral characters.
2. Lepismium warmingianum (K. Schum.) Barthlott in Bradleya 5: 99 (1987). Type

(syntypes): Brazil, Minas Gerais, Lagoa Santa, Warming s.n. (B†, lectotype designated by
Britton & Rose 1923: 238); ibid., Caldas, Lindberg 511 (B†). Lectotype (Barthlott &
Taylor 1995: 46): Minas Gerais, Caldas, 18 Oct. 1854, Lindberg 511 (S; BR & MO
lectoparas. numbered ‘611’).
Rhipsalis warmingiana K. Schum. in Martius, Fl. bras. 4(2): 291 (1890); G.A. Lindberg in Gartenflora 41:
8–12, figs 5–7 (1892).
R. linearis K. Schum. in ibid.: 296 (1890). Lepismium lineare (K. Schum.) Barthlott in Bradleya 9: 89
(1991). Lectotype (Barthlott & Taylor 1995): Paraguay, Nov. 1874, Balansa 2500 (K; P, lectopara.).
Rhipsalis gonocarpa F. A. C. Weber in Rev. Hort. 64: 427 (1892). Type: Brazil, São Paulo (assumed not
to have been preserved).
Stems pendulous to 3 m; stem-segments 25–150 × 0.3–1.5(–2.5) cm, flat or 3–4-angled, flat segments only
1 mm thick away from the vascular midrib, margins serrate-crenate, teeth low, obtuse, to 2 mm proud of
the stem and 1–2.5 cm apart on its margins, the pale, scarious scale leaves strongly adpressed, epidermis dark
olive to yellow-green. Areoles in the axils of the scale-bearing teeth, felt 0 or very sparse prior to flowering,
bristles 0. Flowers sweetly scented, pendent, to 27 × 23 mm; perianth-segments 8–12, largest to 15 × 5 mm,
cream or white, not expanding fully; pericarpel green, 5–8 × 4 mm, distinctly angled; stamens white, yellow
or reddish at base, arranged in two groups, the outermost longest, to c. 10 mm, the inner series clustered
around the style; style c. 12 mm, stigma-lobes c. 5, 3 mm, cream. Fruit c. 12 × 10 mm, ridged and green
to red when unripe, rounded, smooth, dark wine-coloured to black and opaque when ripe.
Southern humid forest element: epiphyte or lithophyte, Mata atlântica, c. 750 m (MG), central-southern
Minas Gerais and southern Espírito Santo (Domingos Martins); common in South-eastern and southern
Brazil; eastern Paraguay and north-eastern Argentina (Misiones). Map 21.
minas gerais: Lagoa Santa, reported by Warming (1908: 146 & 151; cf. Warming & Ferri 1973); Mun.
Arcos, fide Uebelmann (1996): HU 957.
espírito santo: Mun. Domingos Martins, May 1990, R. Kautsky in Rauh 70693, cult. Univ. Bonn,
Germany, Accn. No. 05815 (K).
Concern taking its total range into account, but apparently rare in the region treated here.
Also reported from Bahia and Paraíba by Luetzelburg (1925–26, 3: 111), but these are
assumed to be misidentifications.
Lepismium warmingianum is the sister species of L. lorentzianum (Griseb.) Barthlott, from
eastern Bolivia (Santa Cruz & Tarija) and north-western Argentina.
LEPISMIUM
3. Lepismium cruciforme (Vell.) Miq. in Bull. Sci. Phys. Nat. Néerl.: 49 (1838).
Type: not extant. Typ. cons. (Taylor 1994, Greuter et al. 2000: 385): Brazil, Rio de
Janeiro, Vellozo, Fl. flumin., Icones 5: tab. 29 (1831).
Cactus cruciformis Vell., Fl. flumin., 207 (1829), nom. cons. (Brummitt 1996). Cereus cruciformis (Vell.)
Steud., Nomencl. bot., ed. 2, 1: 333 (1840). Hariota cruciformis (Vell.) Kuntze, Revis. gen. pl. 1: 262
(1891). Rhipsalis cruciformis (Vell.) A. Cast. in Anales Mus. Nac. Hist. Nat. Buenos Aires 32: 496
(1925). Cereus squamulosus Salm-Dyck in DC., Prodr. 3: 469 (1828), nom. rej. Type: not known to
have been preserved. Neotype (Taylor 1994): Vellozo, Fl. flumin., Icones 5: tab. 29 (1831).
C. tenuispinus Haw. in Philos. Mag. Ann. Chem. 1: 125 (1827), nom. rej. Type: not known to have been
preserved. Neotype (Taylor 1994): Loddiges, Bot. Cab. 19: tab. 1887 (1832).
C. tenuis DC., Prodr. 3: 469 (1828), nom. rej.; Lepismium tenue (DC.) Pfeiff. in Allg. Gartenzeitung 3:
315 (1835). Type: a living plant seen by de Candolle. Neotype (Taylor 1994): ‘Received in 1827
from Mr. Hitchin of Norwich. Cereus myosurus.’, water colour illustration by Bond, 20 Jan. 1829
(K). C. myosurus Salm-Dyck in DC., Prodr. 3: 469 (1828), nom. rej. Type: not known to have been
preserved. Neotype (Taylor 1994): as above. [Pfeiffer, l.c., treated C. myosurus as a synonym of L.
tenue.]
Lepismium radicans Vöcht. in Jahrb. Wiss. Bot. 9: 399 (1874). Rhipsalis radicans (Vöcht.) F. A. C. Weber
in Bois, Dict. hort.: 1047 (1898). Type: not known to have been preserved.
Note: for further synonymy, see Britton & Rose (1923: 215). All synonyms relating to this species are
included in the Index to botanical names and epithets (Chapter 7.2).
Stems pendent, creeping or trailing, to 3 m, 15–40(–60) × 1–3(–7.5) cm, flat or 3–6-angled/ribbed, crenate,
often irregular, epidermis green, red tinged around the edges and areoles, or entirely red when fully exposed.
Areoles in the axils of the marginal crenations, subtended by conspicuous scales prior to flowering, sunken,
becoming rather large in age, with abundant wool in tufts to 5 mm or more long and weak bristles to 6(–10)
mm, to 20 mm or more apart on the stem margins/angles. Flowers immersed in the areoles, to c. 15 × 10
mm; perianth-segments 5–6 visible from within, white, cream or pale to deep pink, opening widely. Fruit
broadly turbinate to depressed-globose, to 8.5 × 8.5 mm, magenta to deep red, shiny. Seed (Barthlott &
Hunt 2000: 35, pl. 2.1–2) 1.4 × 0.7 mm, shiny brown. Chromosome number: 2n = 22 (Barthlott 1976).
Disjunct humid forest element: epiphyte or lithophyte, Mata atlântica, including agreste (mata de cipó) and
mata de brejo (NE Brazil) and restinga, sea level to 1200 m, eastern Pernambuco to south-eastern Minas
Gerais; common in South-eastern and Southern Brazil; south-eastern Paraguay and north-eastern
Argentina. Map 21.
pernambuco: Mun. Taquaritinga do Norte, alto da serra, 12 Nov. 1983, V.C. Lima et al. in CFPE 819
(IPA), Aug. 2001, Zappi (obs.).
bahia: E & SE Bahia, Mun. Ubaíra, Três Braços (Ilha Formosa), Rio Risada, c. 5 km from the village,
13°30'S, 39°42'W, 16 Jan. 1994, França et al. 934 (K, HUEFS); Mun. Vitória da Conquista, 12 km from
town towards Itambé, 16 Aug. 2002, M. Machado (K, photos); Mun. Camacã (Camacan), road to
Canavieiras, 23 Jan. 1971, T.S. dos Santos 1408 (CEPEC), Feb. 2003, J. Jardim (obs.).
minas gerais: SE Minas Gerais, Mun. Marliéria, Parque Estadual do Rio Doce, 20 Oct. 1979, Heringer
13992 (UB); Mun. Viçosa, Rio Casca, 7 Sep. 1957, A.G. Andrade 27 (R); l.c., Distrito de São Miguel,
7 Sep. 1957, E.C. Rente 440 (R); Mun. Coronel Pacheco, Est. Experimental Cel. Pacheco, 10 Aug.
1944, Heringer 1507 (SP); Mun. Lima Duarte, Parque Estadual do Ibitipoca, 17 July 1987, R.F. Novelino
et al. 21538 (CESJ); l.c., Serra do Ibitipoca, July 1987, H. Magalhães s.n. (R).
espírito santo: Mun. Vitória, May 1990, Zappi (obs.); Mun. Venda Nova, W of town, c. 850 m, J.
Fowlie, cult. HNT Accn. No. 28676, Apr. 1974 (HNT 2001).
Concern taking its total range into account.
A common and highly variable plant in South-eastern Brazil, but apparently rare or
seldom collected in the brejo and agreste forests of the Nordeste. Braun & Esteves Pereira
RHIPSALIS
(2002: 79) indicate that Lepismium subg. Lepismium (ie. L. cruciforme) occurs in the states
of Rio Grande do Norte, Paraíba, Alagoas, Sergipe and Goiás, but this may only be
guesswork and should not be taken as fact in the absence of concrete records.
Its sister species is L. incachacanum (Cárdenas) Barthlott, from Bolivia (Cochabamba &
La Paz).
12. RHIPSALIS Gaertn.
Fruct. sem. pl. 1: 137, tab. 28 (1788). Type: R. cassutha Gaertn. (= R. baccifera (J. S.
Muell.) Stearn).
Including Erythrorhipsalis A. Berger (1920); Rhipsalis subg. Erythrorhipsalis A. Berger (1920).
Literature: Barthlott (1987), Barthlott & Taylor (1995: 48).
Habit pendent when epiphytic, or erect to decumbent when on rocks; stems segmented, always ±
acrotonically branched when apex undamaged (basal extension shoots excepted) and lacking sharp spines,
though sometimes bristly; stem-segments terete (unribbed), ribbed, angled, winged or flat (2-ribbed), the
non-basal segments deciduous when old. Areoles small to almost absent (R. pulchra) or sunken and
conspicuous and/or woolly only after bearing flowers, subtended by small to minute scale leaves; terminal
composite areoles often present. Flowers 1–13 per areole, whitish or tinged yellow or pink, rarely red, buds
erumpent from sunken areoles or developing at the stem surface; pericarpel ± naked, tube almost 0,
perianth-segments c. 5–18, reflexed or semi-patent (campanulate flowers); stamens strongly exserted or
included. Fruit subspherical to ellipsoid, perfectly rounded (never angled), naked, white, orange, pink or
purplish, pulp highly mucilaginous. Seeds 1–1.7 mm, black-brown; testa ± smooth.
As currently circumscribed a genus of 35 species (Barthlott & Taylor 1995, Taylor &
Zappi 1997) with a centre of diversity in South-eastern Brazil (especially southern
Espírito Santo, Rio de Janeiro and São Paulo). The genus is divided into 5 subgenera,
each of which is represented in Eastern Brazil, mainly distributed in the Mata atlântica and
in cloud forests associated with the campos rupestres. A minimum of 19 species is treated
here (but see Nos 14 & 16, and subg. Calamorhipsalis below), of which only 4 taxa (2 spp.,
2 subspecies) appear to be endemic, but the group remains poorly collected in North-
eastern Brazil, where most of its habitat has been destroyed. Six species (Nos 2, 13–16 &
18) only just enter the area covered, having the major part of their ranges further south
and west. Plates 14.2–22.1.
Besides the taxa treated below, Luetzelburg (1925–26, 3: 111) records R. clavata F.
A. C. Weber (SE Brazil) and R. robusta Lem. (= R. pachyptera Pfeiff., SE to S Brazil) from
eastern Bahia, which probably represent misidentifications (the former may be R.
cereuscula, the latter R. russellii or R. crispata). A record for R. trigona Pfeiff. from Espírito
Santo is likewise assumed to be an error (Barthlott & Taylor 1995: 55). Braun & Esteves
Pereira (2002: 79) indicate many state records for Rhipsalis subgenera in the Nordeste
that seem improbable and can in most cases be put down to guesswork. Luetzelburg also
uses the names R. cribrata (Lem.) N. E. Br. and R. platycarpa Pfeiff., both being of
uncertain application.
RHIPSALIS
1. Flower red; flower-buds conspicuously erumpent (splitting the stem’s epidermis); stems
perfectly terete (S Espírito Santo) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19. hoelleri
1. Flower not red and/or flower-buds not erumpent; stems various . . . . . . . . . . . . . . . . . . . . . . . . 2
2. Stems flat, angled, winged, strongly to weakly ribbed or terete with ± raised podaria
subtending ± fleshy scale leaves (or cylindric and smooth but with conspicuously erumpent
flower-buds exposing floccose areoles after the fall of the fruit); flower-buds etc. 1–many
per lateral areole . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
2. Stems perfectly terete or only the shortest, ultimate segments somewhat angled; scale leaves
minute, not fleshy or soon scarious; flower-buds etc. 1 per lateral areole or flowers terminal . . . 11
3. Flower-buds conspicuously erumpent from sunken, often very woolly areoles that were
absent, hidden by scale-leaves or minute prior to flower development, solitary . . . . . . . . . . . . . . 4
3. Flower-buds not erumpent or only so on close inspection, the areoles not more obviously
woolly after flowering, or flowers 2 or more per areole, at least on the older stem-segments . . . . 7
4. Stem-segments of indeterminate growth, branching often only subacrotonic, with numerous
discontinuous ribs (S Espírito Santo) . . . . . . . . . . . . . . 7a. pacheco-leonis subsp. catenulata
4. Stem-segments of determinate growth, branching acrotonic from terminal composite
areoles, terete, angled or with low ± continuous ribs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
5. Stem-segments terete, not ribbed, but sometimes with raised podaria subtending the
scale-leaves (widespread) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5. floccosa
5. Stem-segments ribbed or angled . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
6. Stem-segments weakly to more strongly (3–)5-ribbed, ribs continuous (S Espírito Santo) . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9. sulcata
6. Stems 3–4-angled in cross-section, the angles discontinuous and not forming ribs
(widespread) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6a. paradoxa subsp. septentrionalis
7. Fruit white or faintly tinged pink in part; stems mostly flat . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
7. Fruit entirely pink to red or purplish; stems flat or 3–4(–5)-angled/winged . . . . . . . . . . . . . . . . 9
8. Ultimate stem-segments very thin, only c. 1 mm thick (excluding midrib) in living material,
margins shallowly crenate, the areoles 2–4 mm from the outermost part of margin; flowers
mostly 1 per areole, developing during the rainy season (SE Bahia, Espírito Santo) . . 3. oblonga
8. Ultimate stem-segments stouter, to 2 mm or more thick (excluding midrib) in living
material, margins strongly crenate-lobed, the areoles to 5–6 mm from the outermost
part of margin; flowers 1–5 per areole, developing during the dry season (Pernambuco,
E Bahia) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4. crispata
9. Stems mostly pendulous or sprawling, > 25 mm wide, flat or 3–5-winged . . . . . . . . . . . . . . . . 10
9. Stems mostly erect, to 25(–30) mm wide, 3–5-angled (S Espírito Santo) . . . . . . . . . 8. cereoides
10. Flowers/fruits 1–5 per areole, flower 12–20 mm diam., yellowish, fruit globose to oblong . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2. elliptica
10. Flowers/fruits (1–)3–9 per areole, flower to 9 mm diam., whitish, fruit globose . . . . . 1. russellii
11. Flowers < 8 mm long, with 4–7, patent to reflexed perianth-segments visible from within . . . . 12
11. Flowers > 8 mm long, campanulate, with 8–15 concave perianth-segments visible from within . . 14
12. All stems of ± indeterminate growth, not forming composite areoles at apex; flowers and fruits
always lateral (widespread) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10. lindbergiana
12. Higher order stem-segments of determinate growth, forming composite areoles at apex;
flowers/fruits both lateral and terminal . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
13. Pericarpel shorter than perianth; fruit globose to shortly barrel-shaped, to c. 5 × 4 mm;
flowers/fruit mostly lateral (SE Brazil) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11. teres
13. Pericarpel as long or longer than perianth both in bud and at anthesis; fruit ovoid,
c. 7 × 6 mm; flowers/fruit mostly from the terminal composite areoles (Bahia &
Pernambuco northwards) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12. baccifera
RHIPSALIS
14. Basal extension shoots of indeterminate growth apparently lacking or to only c. 12 cm

long, all other stem-segments determinate and < 7 cm; stamens white at base . . . . . . 16. clavata
14. Basal and higher order extension shoots of indeterminate growth present and > 12 cm
long, other stem-segments usually decreasing in size towards distal parts of plant; stamens
yellow, red or purplish at base . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
15. Stem-segments > 7 cm long, extension shoots and shorter higher order segments not
markedly different, terminal composite areoles mostly lacking; flowers lateral and
subterminal (S Minas Gerais, > 1500 m) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13. pulchra
15. Stems clearly differentiated into long extension shoots and much shorter secondary
segments, some < 7 cm long and all with terminal composite areoles; flowers terminal
on the higher order stem-segments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
16. Ultimate stem-segments swollen and sometimes angled/ribbed; fruit white, rarely with
reddish scales (fruit sometimes red outside Brazil) . . . . . . . . . . . . . . . . . . . . . . . . 17. cereuscula
16. All stems terete, never angled/ribbed; fruit purple, magenta, red, orange or greenish
tinged maroon . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
17. Stems clothed in semi-adpressed, bristle-like spines; fruit bristly . . . . . . . . . . . . . . 18. pilocarpa
17. Stems and fruit naked . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
18. Flower 20–25 × 20 mm; fruit globose-ovoid, c. 10 mm long, unscented . . . . . . . 14. burchellii
18. Flower c. 15 × 12 mm; fruit truncate, c. 6 mm long, smelling of blackcurrants . . . . 15. juengeri
Subg. Phyllarthrorhipsalis Buxb. (Nos 1–4): seedlings, where known, flattened (2-ribbed)
at first; adult stem-segments of determinate size and acrotonically branched (except
secondary segments arising laterally from the somewhat indeterminate basal extension
shoots); new stem-segments and flower-buds scarcely erumpent; lateral areoles visible
before flowering; stem-segments mostly flattened, or with 3–5 continuous angles or
wings, relatively thin; flowers lateral and terminal, remaining open day and night, one to
many at a time per areole, pericarpel exposed, areoles flowering repeatedly, enlarging and
bearing more flowers each time; fruit white, pink or purplish. Central and South
America. Type: R. pachyptera Pfeiff.
1. Rhipsalis russellii Britton & Rose, Cact. 4: 242 (1923). Type (Barthlott & Taylor
1995: 60): Brazil, Bahia, Mun. Jaguaquara [Toca da Onça], 27–29 June 1915, Rose &
Russell 20106 (NY, lecto.; US, lectopara.).
vernacular name. Mandacaru-da-serra.
Sprawling lithophyte to 80 cm, or pendent epiphyte with mostly dichotomous branching, to 200 cm or
more; stem-segments usually flattened in epiphytic plants, or mostly 3–4(–5)-angled in lithophytes,
irregularly elliptic, often narrowly so, to c. 15 × 3–5.8 cm, 2–3 mm thick away from the vascular midrib;
areoles exposed from the start, indented 3–11 mm from the crenate stem margin and to c. 40 mm apart,
with conspicuous felt and 0–2 minute bristle-spines. Flower-buds whitish or pinkish, flowers (1–)3–9 per
areole, inconspicuous, to c. 7 × 5–9 mm, sometimes scarcely opening, with only the stigma-lobes visible;
perianth-segments whitish to greenish white, 4–5 visible from above, boat-shaped and rounded to
acuminate at apex, patent to ± reflexed; stamens few, c. 12–14, but sometimes only 2, 2–3 mm, anthers
minute, whitish; style c. 2.5 mm, stigma-lobes 3–6, c. 1–1.5 mm, whitish; pericarpel larger than the
perianth until just before anthesis, green, c. 2 × 3 mm, sometimes with a broad bract-scale below apex.
Fruit densely clustered, globose, to c. 7 mm diam., whitish to pale pink at first, later purplish/pinkish
magenta or red.
RHIPSALIS
Widespread humid forest/campo rupestre element: epilithic or epiphytic, campo rupestre, mata de brejo and
Mata atlântica and their ecotones with caatinga-agreste, 50–1050 m, from the Chapada Diamantina and
Serra do Espinhaço eastwards, Bahia, Minas Gerais and Espírito Santo; Central-western Brazil (Goiás and
Mato Grosso) and Southern Brazil (Paraná: Guaíra [HU188]). Map 22.
bahia: cent. Bahia (Chapada Diamantina), Mun. Lençois, Morro da Chapadinha, 12°27'24"S,
41°27'10''W, 23 Nov. 1994, Melo et al. in PCD 1283 (K, ALCB, HUEFS); Mun. Andaraí, 12°46'59"S,
41°20'58''W, 6 Aug. 2001, F.R. Nonato et al. 1015 (HUEFS 55760); l.c., 5 km S of town, near the Rio
Paraguaçu, 21 Aug. 1988, Eggli 1240 (ZSS, CEPEC); l.c., Igatu, 28 Jan. 1988, L.P. Bautista s.n. (HRB);
Mun. Mucugê, 0.5 km W of town, 22 Dec. 1988, Taylor & Zappi in Harley 25600 (CEPEC, SPF, K);
Mun. Ituaçu, c. 2 km NE of town, 18 Aug. 1988, Eggli 1200 (ZSS); E & SE Bahia, Mun. Jaguaquara [Toca
da Onça], 27–29 June 1915, Rose & Russell 20106 (US, NY); Mun. Ipiaú, just west of town on road to
Jequié, Aug. 2001, Zappi (obs.); Mun. Ilhéus, Faz. Almada, 12 Dec. 1967, S.G. Vinha & A. Castellanos 4
(CEPEC); l.c., grounds of CEPLAC, 7 Dec. 1988, T.S. dos Santos 4433 (ALCB); Mun. Floresta Azul, 500
m, [before 1966], Martins, ex coll. Brieger, ESALQ, Piracicaba, SP, cult. R.B.G. Kew (K, spirit coll.
32129, 39004); Mun. Vitória da Conquista, 12 km from town towards Itambé, 16 Aug. 2002, M. Machado
(K, photos); Mun. Camacã (Camacan), road to Pau Brasil, 19 Jan. 1971, T.S. dos Santos 1352 (CEPEC);
Mun. Itabela, 16 km S of Eunápolis, 22 Sep. 1966, P.R. Belém & R.S. Pinheiro 2644 (CEPEC, IAN); Mun.
Teixeira de Freitas, vale do Rio Alcobaça, 12 May 1971, T.S. dos Santos 1627 (RB).
minas gerais: cent. Minas Gerais, Mun. Conceição do Mato Dentro, 11 km S of town on road to
Morro do Pilar, 20 Nov. 1989, Zappi et al. 195 (HRCB); Mun. Lima Duarte, Parque Estadual de
Ibitipoca, 4 Aug. 2003, Taylor (obs.).
espírito santo: reported from the vicinity of Domingos Martins by Rauh (Rauh & Kautsky 67568, det.
W. Barthlott from living material grown at Univ. Bonn, Germany, n.v.).
conservation status. Vulnerable [VU B2ab(iii, iv)] (2); area of occupancy estimated to be < 2000
km2; PD=2, EI=1, GD=2. Short-list score (2×5) = 10. It is markedly disjunct and its habitat has
contracted severely and especially in south-eastern Bahia. Elsewhere, it is found within some reasonably
secure protected areas (eg. Parque Nacional Chapada Diamantina, BA), but its ecological niches are
invariably limited.
A distinctive species with flowers minute in relation to the stem-segments, which

somewhat resemble those of the larger-flowered R. elliptica (see below) and R. pachyptera
(Rio de Janeiro to Rio Grande do Sul). Its clustered fruits are quite variable in colour. In
cultivation, at least, the collection from Minas Gerais sometimes produces flowers almost
devoid of fertile stamens.
A living collection, at the University of Bonn’s botanic garden, said to be from Goiás
(P. J. Braun s.n.), which is yet to flower, seems referable here, and unlocalised living material
of this species from ‘Mato Grosso’ (Uebelmann s.n.) has recently flowered and fruited at
Zürich, ZSS (Accn No. 82-1444; BONN, photo.). Clearly, its range westwards from
Eastern Brazil is incompletely understood at present and could well include Mato Grosso
do Sul and eastern Paraguay also, Guaíra (PR) being on the border with these (HU 188).
2. Rhipsalis elliptica G. A. Lindb. ex K. Schum. in Martius, Fl. bras. 4(2): 293 (1890).
Type (syntypes): Brazil, Mosén 3630, Glaziou 14859 & Schenck 1218 (B†). Lectotype
(Barthlott & Taylor 1995: 60): São Paulo, Sorocaba, Mosén 3630 (S).
R. chloroptera F. A. C. Weber in Bois, Dict. hort.: 1045 (1898). Type: Brazil, São Paulo, Santos, assumed
not to have been preserved.
Pendent to 1 m or more; basal extension shoots to 35 cm; higher order stem-segments mostly flattened,
RHIPSALIS
rarely 3-winged, broadly to rather narrowly elliptic, rounded to narrowly truncate at apex, mostly
8–14(–18) × 3–6(–8.5) cm, c. 2–2.5 mm thick away from the vascular midrib, dark green, but often
turning a striking shade of magenta or purple when drought-stressed or in strong light, usually at least the
margins purplish tinged, margins undulate to broadly crenate; areoles ± inconspicuous or hidden prior to
anthesis, glabrous or bearing minute bristles, indented up to c. 3 mm from the crenate stem margin and
to 2 cm apart. Flowers 1–5 per areole, buds sometimes pinkish, at anthesis to 14 mm long (from tips of
reflexed perianth-segments to stigma-lobes), 12–20 mm diam. before full anthesis, perianth-segments pale
to golden yellow, the mostly 5 principal to 9 × 4.5 mm, strongly reflexed hiding the pericarpel, bluntly
rounded at apex; stamens conspicuous to 8 mm, white; style white, scarcely exserted from the stamens,
stigma-lobes 3–4, to 3.5 × 1 mm, white; pericarpel 3.5–4.0 × 3.5–4.0 mm, pale green. Fruit deep reddish
pink, globose to elongate.
Southern humid forest element: epiphyte or lithophyte in Mata atlântica, including mata de neblina, near
sea level to c. 1500 m, central, south-eastern and southern Minas Gerais; common elsewhere in South-
eastern and Southern Brazil. Map 22.
minas gerais: Mun. Catas Altas, Reserva Particular de Patrimônio Natural (RPPN) do Caraça,
between Bocaina and Gruta da Bocaina, 2000, J.O. Rego et al. (BHCB, Herb. Fund. Zoo-Bot. de Belo
Horizonte), l.c., 1 Aug. 2001, Taylor (K, photos); Parque Nacional do Caparaó, Córrego do Inácio, 17
Dec. 1988, Krieger et al. in FPNC 706 (CESJ); Mun. Lima Duarte/Rio Preto, Chapadão da Serra Negra,
3 Oct. 1959, Castellanos 22541 (R).
Concern taking its total range and abundance into account.
Luetzelburg (1925–26, 3: 111) reports the synonymous R. chloroptera F. A. C. Weber from

Sergipe, but in the absence of documented material this record must remain rather
doubtful and no species of Subg. Phyllarthrorhipsalis is currently known from that state.
3. Rhipsalis oblonga Loefgr. in Arch. Jard. Bot. Rio Janeiro 2: 36–37, tab. VIII
(1918). Type: Brazil, Rio de Janeiro, Ilha Grande, 1915, Löfgren & Rose (holo. not found
at SP, RB or R; material at US is doubtfully authentic and judged to be R. elliptica/R.
goebeliana; material at NY is dated 1917). Lectotype (Barthlott & Taylor 1995: 61):
Löfgren, l.c., tab. VIII (1918).
R. crispimarginata Loefgr., l.c. tab. IX (1918). Type: l.c., 1915, Löfgren, Rose & Campos-Porto (holo. not
found at SP, RB or R). Lectotype (designated here): l.c., 11 July 1915, Rose et al. 20401 (US).
Pendent to 1 m or more; stem-segments dark to pale green or yellow-green, usually flat (rarely 3-angled),
the margins sometimes tinged reddish when exposed to sun, frequently emitting aerial roots along the
vascular midrib; basal or sub-basal extension shoots to 10–20 × 1–2 cm, gradually cuneate towards the
almost cylindric base, ultimate segments to 12(–23) × 2–6(–7) cm, cuneate to truncate at base, ± truncate
to rounded at apex, only c. 1 mm thick away from the vascular midrib, often contorted, margins crisped
and tortuous, crenate beneath the areoles; areoles c. 1 mm diam., 14–35 mm apart, sunken 2–4(–10) mm
from stem margin, bearing 1 or more fine, bristly, pale, inconspicuous spines. Flowers 1–2(–3) per areole,
to 15 × 18 mm when fully expanded (maximum diameter of perianth-segments when patent, but before
full expansion when the spreading stamens reach a maximum of 15 mm diam., at which stage perianth-
segments are strongly reflexed back and hiding pericarpel); perianth-segments c. 8, the 5 largest to 8 × 4
mm, greenish-yellow or sometimes tipped reddish in bud, turning yellow post-anthesis; stamens to 6–8
mm, white, conspicuous; style to 7 mm, white, stigma-lobes (3–)4–5, to 2–3 × 1 mm, white; pericarpel
c. 4 × 4 mm, ovoid (narrowed to apex), greenish yellow. Fruit c. 6–7 × 7–8 mm, entirely white or pale
pink at apex.
RHIPSALIS
Humid forest element: epiphyte in perhumid Mata atlântica, low elevations, south-eastern Bahia (and
Espírito Santo?); to South-eastern Brazil (São Paulo, Serra do Mar). Map 22.
bahia: Mun. Una, 8.7 km E of São José [da Vitória?] on road to Una, 4 Mar. 1986, T.S. dos Santos & E.
Judziewicz 4143 (US, CEPEC, ASE 4495); ibid., 9 km on the road from São José da Buararema, 28 Oct.
1983, R. Callejas et al. 1561 (CEPEC); ibid., ± 35–40 km S of Olivença on the Ilhéus–Una road, 15°16'S,
39°4'W, 2 Dec. 1981, G.P. Lewis & A.M. de Carvalho 729 (CEPEC, RB, K, NY, M); Mun. Camacã,
road to Pau Brasil, 19 Jan. 1971, T.S. dos Santos 1352 (CEPEC, RB); Mun. Canavieiras, road BA 001, 4
km N of junction to Santa Luzia, 23 April 2003, Taylor & Zappi (obs.); Mun. Teixeira de Freitas, Vale
do Rio Alcobaça, 12 May 1971, T.S. dos Santos 1627 (CEPEC).
espírito santo: unlocalized, s.d., Luetzelburg 12 (M).
conservation status. Near Threatened [NT] (1); PD=2, EI=1, GD=2. Short-list score (1×5) = 5.
Taking its total range into account the taxon does not qualify for any of the threatened Red List
categories, but much of its former habitat in southern Bahia has suffered ongoing destruction since the
above cited collections were made in 1971.
R. oblonga is very similar to R. goebeliana Backeb. from Bolivia (Yungas) and to R.

occidentalis Barthlott & Rauh from northern Peru, southern Ecuador and Suriname. They
differ from R. oblonga in their stem-segments being consistently narrowly cuneate at base
and in their flowers with much shorter and less conspicuous stamens.
4. Rhipsalis crispata (Haw.) Pfeiff., Enum. cact.: 130 (1837); Britton & Rose, Cact.
4: 245, fig. 232, tab. XXXV fig. 3 (1923). Type: Brazil, not known to have been
preserved. Neotype (Barthlott & Taylor 1995: 61): Brazil, São Paulo, Mun. Rio Claro,
Fazenda São José, edge of lake, 1991, A. Cardoso in Zappi 249, cult. N. P. Taylor, 10 Jan.
1994 (K, in spirit).
Epiphyllum crispatum Haw. in Philos. Mag. Ann. Chem. 7: 111 (1830).
Rhipsalis rhombea sensu Loefgr. in Arch. Jard. Bot. Rio Janeiro 1: 89, tab. 16 (1915) non Cereus rhombeus
Salm-Dyck.
Like the preceding, but flat or 3-angled, aerial roots produced mainly from the junctions of the stem-
segments, extension shoots larger, to 40 cm or more, ± erect or ascending at first; ultimate segments to
7(–10) cm wide, yellow-green to dark green, much stouter, c. 2 mm or more thick, the areoles sunken
5–6 mm or more (to 20 mm exceptionally) into the marginal crenations; flowers 1–5 per areole, 12–20
mm diam., greenish white; stamens shorter than perianth-segments; stigma-lobes 2–4; fruit 6 × 5 mm,
white or greenish, ripening slowly. Chromosome number: 2n = 22 (Barthlott 1976).
Humid/subhumid forest element: epiphyte in mata de brejo and Mata atlântica, 200–800 m, eastern
Pernambuco (Mun. Caruaru) and eastern Bahia; South-eastern and Southern Brazil (coast E of Rio de
Janeiro between Niterói and Cabo Frio, inner São Paulo, Santa Catarina). Map 22.
pernambuco: Mun. Caruaru, [Brejo dos Cavalos?], Faz. Caruaru, 28 Nov. 1970, Andrade-Lima 70-6200
(IPA 23024, K, photos).
bahia: Mun. Amélia Rodrigues, road BR 324, epiphyte on mango tree in front of roadside restaurant,
‘Café da Manha’, 1999, M. Machado, cult. ibid., Cruz das Almas, 2002 (HUEFS); Mun. Santa Teresinha,
Serra da Jibóia, Nov. 2001, Zappi & Jardim (K, photos), 7 Aug. 2002, Taylor & Machado (K, photos).
conservation status. Vulnerable [VU B2ab(iii)] (2); area of occupancy estimated to be < 2000 km2;
PD=2, EI=1, GD=1. Short-list score (2×4) = 8. Apparently widely distributed, but rarely observed and
Vulnerable despite its range in view of the widespread destruction of the habitat it favours. Only the last
of the above cited populations is known to be extant. Occurring at only 7 other localities, all subject to
anthropogenic change.
RHIPSALIS
The 3 records from NE Brazil are markedly disjunct from other known sites in Rio de
Janeiro (Cabo Frio, Silva Jardim, Saquarema & Itacoatiara), São Paulo (Rio Claro &
Altinópolis) and Santa Catarina (30 km E of Blumenau), but similar disjunctions are
known in Araceae from Pernambuco (eg. Philodendron eximium Schott and P. corcovadense
Kunth, fide S. Mayo, pers. comm.). Rhipsalis crispata is a species of more markedly seasonal
or drier habitats, where it has been found as an epiphyte or lithophyte, both near the coast
and far inland, but it can also grow in the sand of the restinga (eg. at Cabo Frio, RJ). It is
closely related to R. oblonga, but has thicker and often broader stem-segments, which are
deeply crenate-sinuate at the margin and, in cultivation, appears to flower following a dry
or cool period, whereas R. oblonga flowers during the warm, humid growing period.
The use of the name R. crispata for the plant described here is clearly supported by
early herbarium records from the 19th Century in Europe, where it was becoming
widespread in cultivation under Haworth’s epithet (eg. in the garden at Leuven
(Louvain), Belgium, 1837, ex Herb. Martens [BR!], and from Munich, 3 Jan. 1850, ex
Herb. Kummer [M!]).
Subg. Epallagogonium K. Schum. (including Subg. Trigonorhipsalis A. Berger and Subg.

Goniorhipsalis K. Schum.) (Nos 5–9): stems ± angled to ribbed or terete with raised
podaria, never flattened (except in the first shoot of seedlings of No. 5 and very rarely in
shade forms of No. 8); all adult stem-segments determinate, except in R. pacheco-leonis;
flower-buds 1 per areole in the species treated below (except R. cereoides), strongly
erumpent and the pericarpel sunken into the stem (less so in Nos 8–9), areoles sometimes
flowering > once. South America. Type: R. paradoxa (Pfeiff.) Salm-Dyck.
5. Rhipsalis floccosa Salm-Dyck ex Pfeiff., Enum. cact.: 134 (1837). Type: a living
plant, presumed not to have been preserved. Neotype (Barthlott & Taylor 1995: 55):
Brazil, Bahia, Mun. Ilhéus, grounds of CEPLAC, 3 Mar. 1974, C. Erskine 164, cult.
R.B.G. Kew, 11 Nov. 1992 (K, in spirit).
Lepismium floccosum (Salm-Dyck ex Pfeiff.) Backeb. in Backeberg & Knuth, Kaktus-ABC: 155 (1935,
publ. 1936).
Pendent to 3 m or more; basal stems to 1.5 cm thick, new stem-segments arising singly or in groups of
2–5 from the apex of existing segments, at maturity to 30 cm long, pliable and rubbery in feel, but not
flaccid; primary stem of seedling flattened, subsequent segments angled, mature growth terete but
commonly with prominent podaria subtending the scale leaves, sometimes falsely ribbed in dried material,
c. 6–10 mm thick, grey-green, matt, minutely roughened, podaria to c. 25 mm apart; areoles sunken and
hidden until flowering, to 4 mm diam. and floccose after bearing fruits. Flowers to 12–25 mm diam. or
slightly larger, perianth-segments pale green to greenish white, pericarpel immersed in the sunken, woolly
areole. Unripe fruits light green, sometimes dark reddish above, ripe fruit turbinate, c. 7 × 6 mm, white,
with a faint reddish ring around the perianth scar.
conservation status. Least Concern [LC]; found in various countries in South America (see under
subsp. floccosa below) and frequently seen in Eastern Brazil.
Together with R. lindbergiana, the most widespread and commonest Rhipsalis species in
Eastern Brazil, from Pernambuco southwards, but rather variable and requiring further
RHIPSALIS
study in the field. The following, which includes the type of the species (subsp. floccosa),
is a frequent plant of the Mata atlântica and brejo forest.
5a. subsp. floccosa

Ultimate stem-segments to 10 mm thick, podaria subtending the leaf-scales usually very prominent.
Flowers small, to c. 12 mm diam. Fruit whitish, often with a reddish ring around the perianth remains,
half-immersed in stem.
Widespread humid forest element: epiphytic or epilithic in agreste and Mata atlântica, including mata de
brejo, near sea level to c. 1200 m, eastern Pernambuco, eastern foot of the Chapada Diamantina and
coastal Bahia to Espírito Santo; re-appearing south-westwards in the interior of South-eastern Brazil
(Minas Gerais and São Paulo), where frequent; replaced by subsp. oreophila in the East Brazilian
Highlands, by subsp. pulvinigera in mountains and coastal zones of South-eastern and Southern Brazil, by
subsp. hohenauensis in eastern Paraguay and north-eastern Argentina, by subsp. tucumanensis in the eastern
Andes of Argentina, Bolivia and Peru (Junín) and by subsp. pittieri in northern Venezuela. Map 23.
pernambuco: Mun. Taquaritinga do Norte, epiphyte on fruit trees, Aug. 2001, Zappi (obs.); Mun. São
Lourenço da Mata, Engenho São Bento, 26 Dec. 1963, D. Andrade-Lima 63-4190 (IPA); Mun. Moreno,
Tapera, 12 Mar. 1931, B. Pickel 2560 (IPA), 23 Aug. 1934, B. Pickel 3630 (IPA, F); Mun. Brejo da Madre
de Deus, 19/20 Aug. 1980, A. Perrucci & M.A. Maia Filho 19 (IPA), 5 Feb. 1965, D. Andrade-Lima 65-
4300 (IPA); Mun. Caruaru, Brejo dos Cavalos, 28 Oct. 1971, D. Andrade-Lima 71-6585 (IPA), 26 Nov.
1991, F.A.R. Santos 14 (HUEFS, PEUFR); Mun. Bezerros, Serra Negra, 8°10'S, 35°46'W, 17 Oct. 1999,
Krause & Liebig 250 (PEUFR); Mun. Cabo de Santo Agostinho, Gurjaú, 8 Dec. 1964, D. Andrade-Lima
64-4268 (IPA); Mun. Águas Belas, Serra do Cumanati, 29 Dec. 1969, D. Andrade-Lima 69-5627 (IPA);
Mun. Brejão, 7 km N of Teresinha, 1 km S of junction to Brejão, 13 Feb. 1991, Taylor & Zappi 1628
(UFP, HRCB, K, ZSS).
sergipe: Mun. Siriri, 1 Sep. 1983, G. Viana 750 (ASE 3409); Mun. Itabaiana, forest at edge of Estação
Ecológica da Serra da Itabaiana, 5 km W of Areia Branca on road BR 235, 10°41'15"S, 37°23'20''W, 22
Jan. 1992, W.W. Thomas et al. 8894 (CEPEC, HNT, HUEFS); Mun. Estância, 0.5 km N of the Rio
Piauí, 10 Feb. 1991, Taylor & Zappi 1615 (ASE, HRCB, K, ZSS).
bahia: Cent. Bahia, Mun. Mundo Novo, 6 km E of town on road BA 052, 26 Apr. 1992, Taylor & Zappi
(obs.); E & SE Bahia: Mun. Alagoínhas, 24 km S of Inhambupe, 15 Feb. 1991, Taylor & Zappi 1634
(CEPEC, HRCB, K, ZSS); Mun. Araçás, 2 km S of town, 10 Feb. 1991, Taylor & Zappi 1612 (CEPEC,
HRCB, K, ZSS); Mun. Feira de Santana, Serra de São José, Nov. 2001, Zappi (obs.); Mun. Conceição do
Jacuípe, 87 km NW of Salvador on road to Feira de Santana, 12 Mar. 1989, reported in Barthlott & Taylor
(1995: 53, pl. 17); Mun. Muritiba, Faz. 8 de Dezembro, epiphyte on Syagrus coronata, 6 Aug. 2002, Taylor
& Machado (K, photos); Mun. Nazaré [‘Nazareth’], 30 June 1915, Rose & Russell 20145 (US); Mun. (?), Ilha
de Itaparica, Sep. 1974, A. Leal Costa s.n. (ALCB 022913); l.c., Mun. Vera Cruz, Mar Grande, 2 Jan. 1992,
C.S.N. Guimarães 8 (ALCB); Mun. Ubaíra, 13°30'S, 39°42'W, 16 Jan. 1994, F. França 932 (HUEFS); Mun.
Valença, exit to the town from road BR 101, Aug. 2001, Zappi (obs.); Mun. Jitaúna/Jequié, c. 30 km NW
of Ipiaú, Aug. 2001, Zappi (obs.); Mun. Ipiaú, road to Ibirataia, 30 Oct. 1970, T.S. dos Santos 1244
(CEPEC); Mun. Ilhéus, grounds of CEPEC, Km 22 on the Ilhéus–Itabuna road (BR 415), 2 Sep. 1981,
J.L. Hage & H.S. Brito 1279 (CEPEC), 7 Dec. 1988, T.S. dos Santos 4444 (ALCB, HUEFS); Mun. Vitória
da Conquista, 12 km from town towards Itambé, 16 Aug. 2002, M. Machado (K, photos); Mun. Barra do
Choça, Barra Nova, Morro do Criminoso, c. 15°S, 44°33.5'W, Sep. 2003, M. Machado (obs.); Mun.
Belmonte, Estação Experimental de CEPLAC, 18 Sep. 1970, T.S. dos Santos 1109 (CEPEC).
espírito santo: Mun. São Mateus, Reserva Biológica de Sooretama, Lagoa do Macuco, 15 May 1977,
Martinelli et al. 2084 (RB, UEC); Mun. Santa Teresa, road Itarana – Santa Teresa, 19°55'50"S, 40°41'7''W,
24 Nov. 1999, Zappi et al. 424 (K, UEC); l.c., Fazenda Santa Lúcia, 19°57'24"S, 40°32'37''W, 25 Nov.
1999, Zappi et al. 439 (K, UEC); Mun. Santa Leopoldina, road to Santa Teresa, estate of Floriano
Bremekamp, 20°3'24"S, 40°32'55''W, 25 Nov. 1999, Zappi et al. 450 (K, UEC).
RHIPSALIS
Although much of its former habitat has been destroyed, it has managed to colonize cultivated trees such
as mangoes, as well as other long-lived species planted as street trees and in village squares.
Variable in the length and especially thickness of its stem-segments.
5b. subsp. oreophila N. P. Taylor & Zappi in Cact. Consensus Initiatives 6: 7 (1998).
Holotype: Brazil, Minas Gerais, mountains E of Monte Azul, 1964, Ritter 1247 (SGO
125604, lectotype of R. monteazulensis F. Ritter, l.c. infra).
Rhipsalis monteazulensis F. Ritter, Kakt. Südamer. 1: 42 (1979). Type (Taylor & Zappi, l.c.): as above.
Ultimate stem-segments to 6 mm thick, almost perfectly cylindrical, without obvious podaria subtending
the scale-leaves. Flowers small, to c. 12 mm diam., greenish white. Fruit white, scarcely immersed in stem.
Northern campo rupestre element: epiphytic or epilithic in mata de neblina (capão de mato), campo rupestre, c.
1200–1750 m, Chapada Diamantina and Serra do Espinhaço, Bahia and northern Minas Gerais. Endemic
to the core area of Eastern Brazil. Map 23.
bahia: cent. Bahia (Chapada Diamantina), Mun. Lençóis, W towards Caeté-Açu, Serra Larga, 23 May
1987, B. Stannard et al. in CFCR 10931 (SPF, K, B, ZSS); Mun. Abaíra, below Tanquinho, 13°17'S,
41°53'W, 1480 m, on rocks, 14 Sep. 1992, W. Ganev 1094 (HUEFS, K); l.c., Distrito de Catolés, Mata
do Tijuquinho, 19 Apr. 1998, L.P. de Queiroz 5001 (HUEFS); Mun. Rio de Contas, Pico das Almas,
1750 m, on rocks, 12 Nov. 1988, Harley et al. 26400 (CEPEC, SPF, K); l.c., NW of Campo do Queiroz,
1500 m, 13 Dec. 1988, Taylor in Harley 25566 (CEPEC, SPF, K).
minas gerais: N Minas Gerais (N Serra do Espinhaço), Mun. Monte Azul/Espinosa, c. 12 km E of
Monte Azul, E of Vila Angical, Serra Geral, E side near top of serra on path to village of Gerais, epiphyte,
1220 m, 28 Jan. 1991, Taylor et al. 1476 (BHCB, HRCB, K, ZSS).
conservation status. Least Concern [LC] (1); PD=1, EI=1, GD=1. Short-list score (1×3) = 3. Only
the southernmost of the 4 sites listed above is undergoing habitat modification. The northernmost site is
within a national park.
The forms of this species from above 1200 metres in the northern sector of the East
Brazilian Highlands (Chapada Diamantina, BA, and northern Serra do Espinhaço, MG)
have almost perfectly cylindrical stems devoid of podaria and seem sufficiently distinct to
be treated as subspecifically different from those of the lowland forests and South-eastern
campos rupestres. They have smaller flowers than plants from the latter area, which are
referred to the following:
5c. subsp. pulvinigera (G. A. Lindb.) Barthlott & N. P. Taylor in Bradleya 13: 55 (1995).
Lectotype (Barthlott & Taylor, l.c.): Gartenflora 38: 184, fig. 34 (1889).
R. pulvinigera G. A. Lindb. in Gartenflora 38: 182 (1889). Lepismium pulvinigerum (G. A. Lindb.) Backeb.
in Backeb. & F. Knuth, Kaktus-ABC: 155 (1935, publ. 1936).
Rhipsalis gibberula F. A. C. Weber in Rev. Hort. 64: 426 (1892). Lepismium gibberulum (F. A. C. Weber)
Backeb. in Backeb. & F. Knuth, Kaktus-ABC: 155 (1935, publ. 1936). Type: see Barthlott & Taylor,
l.c. 57.
vernacular name. Dedinho.
Like subsp. floccosa, but with generally smaller stem-segments and larger flowers, 18–20(–30) mm diam.;
fruit deep magenta-pink or pure white.
RHIPSALIS
Southern humid forest element: epiphytic and epilithic, to 1850 m, southern Serra do Espinhaço, Serra
da Mantiqueira and Serra do Caparaó, central and southern Minas Gerais to southern Espírito Santo;
abundant in South-eastern and Southern Brazil (to Rio Grande do Sul). Map 23.
minas gerais: cent.-S Minas Gerais, Mun. Santana do Riacho, Serra do Cipó, 20 Oct. 1973, A.B. Joly
et al. in CFSC 4665 (UEC, SP), 20 Feb. 1968, H.S. Irwin et al. 20574 (UB), 20 June 1987, Zappi & C.
Kameyama in CFSC 10189 (SPF); l.c., Serra das Bandeirinhas, 27 July 1991, Giulietti et al. in CFSC 12513
(SPF, K); l.c., caminho para Cachoeira das Flores, 9 Sep. 1987 J. Prado et al. in CFSC 10644 (SPF, K);
Mun. Caeté, Serra da Piedade, 29 Oct. 1971, L. Anne & Strang 1920/1461 (GUA), 15 Jan. 1971, Irwin
et al. 30425 (NY), Nov. 1915, Hoehne 6352 (SP, R), 24 Feb. 1987, Zappi et al. in CFCR 10343A
(HRCB), 20 July 1987, Zappi et al. in CFCR 11167 (SPF), 3 Nov. 1988, Taylor & Zappi in Harley 25499
(K, SPF); Mun. Catas Altas, Serra do Caraça, Gruta do Padre Caio, 23 May 1987, Zappi & Scatena in
CFCR 10931 (SPF), 11 Sep. 1990, Zappi et al. 241 (K, HRCB, BHCB, SPF); Mun. Lima Duarte, Parque
Estadual do Ibitipoca, 15 May 1970, Krieger 8589 (CESJ), 27 July 1991, Zappi 261 (K, SPF, CESJ); l.c.,
Gruta do Fugitivo, 16 Oct. 1986, H. C. Sousa et al. 9829 (BHCB, K, SPF); SE Minas Gerais, Parque
Nacional do Caparaó, 15 Oct. 1988, M. Brugger et al. in FPNC 356 (CESJ, HRCB).
espírito santo: Mun. Domingos Martins, May 1986, Rauh 67567, cult. Univ. Bonn, Germany, Accn.
No. 04504 (K), also observed in flower at Bonn, 18 March 2000.
Widespread and common outside of the area treated here.
Typical R. floccosa subsp. pulvinigera is distinguished from its northern relatives by its
generally smaller and more evenly sized, somewhat shiny, smoother, purple mottled stem-
segments (usually to only 20 cm), larger, more expanded flowers (18–25 mm diam. or
more), less woolly flower-bearing areoles and strongly exserted fruits, usually turning
bright pinkish red or pure white when ripe. Beyond Eastern Brazil it ranges from Rio de
Janeiro (Serra dos Órgãos) to Rio Grande do Sul (Mun. Guaíba) and replaces subsp.
floccosa in the coastal regions of South-eastern and Southern Brazil. They can be reliably
distinguished only on the basis of fertile material. The form of subsp. pulvinigera from the
Reserva Particular de Patrimônio Natural (RPPN) do Caraça is unusual in having ± erect
stems and very large flowers.
6. Rhipsalis paradoxa (Salm-Dyck ex Pfeiff.) Salm-Dyck, Cact. Hort. Dyck. 1849: 228
(1850). Type: Brazil, assumed not to have been preserved. Neotype (Barthlott & Taylor
1995: 57): Lem. in Hort. Universel 2: tab. 50 (1840).
Lepismium paradoxum Salm-Dyck ex Pfeiff., Enum. cact.: 140 (1837). Hariota paradoxa (Salm-Dyck ex
Pfeiff.) Kuntze, Revis. gen. pl. 1: 263 (1891). H. alternata Lem. in Hort. Universel 2: 39, tab. 50
(1840), nom. illeg. Rhipsalis alternata (Lem.) Lem., Cactées: 80 (1868). Type: as above.
conservation status. Least Concern [LC]; the widespread subsp. paradoxa is included in various
protected areas.
Only the following heterotypic subspecies is found in the area covered here:
6a. subsp. septentrionalis N. P. Taylor & Barthlott in Bradleya 13: 57 (1995).

Holotype: Brazil, SE Bahia, Mun. Juçari, 250 m, [before 1966], Martins in coll. Brieger,
Piracicaba, SP, cult. Royal Botanic Gardens, Kew, Accn. No. 1966.48946, 18 Nov. 1991
(K in spirit; iso. dried).
RHIPSALIS
Pendent to 3 m or more; branching strictly acrotonic from terminal composite areoles; stems of seedlings
of indeterminate growth, with 4–5 continuous ribs bearing closely set areoles and fine bristly spines, new
stem-segments arising singly or in groups of 2–5 from the apex of existing segments, to c. 30 × 0.7–1.1
cm, determinate, with prominent winglike angles subtending each scale-leaf/areole, these alternating
giving the stem a ± triangular cross-section, epidermis yellow-green to pale grey-green, matt, markedly
roughened to the touch, areoles sunken and hidden before flowering, subtended by minute, adpressed
scale leaves positioned 1.5–3.5 cm apart, areole scar to 5 mm diam. after bearing fruit, but almost
glabrous. Flower-buds strongly erumpent, flowers yellowish, 15 mm diam., perianth-segments to 7 × 2–3
mm; stigma-lobes 5, white. Fruit turbinate, to 7.5 × 7.5 mm, white.
Humid forest element: epiphyte in Mata atlântica, low elevations to c. 950 m, eastern Pernambuco,
eastern Bahia, central-eastern Minas Gerais and Espírito Santo. Endemic to Eastern Brazil. Map 23.
pernambuco: Mun. Primavera, Apr. 1988, P.J. Braun s.n., cult. Univ. Bonn, Germany, 1995 (K).
bahia: Mun. Conceição do Jacuípe, 87 km NW of Salvador on road to Feira de Santana, 12 Mar. 1989,
Supthut 8900 (ZSS); Mun. Barra do Choça, Barra Nova, Morro do Criminoso, c. 15°S, 40°33.5'W, Sep.
2003, M. Machado (obs.); Mun. Juçari, [before 1966], Martins, ex coll. Brieger, cult. R. B. G. Kew, 18
Nov. 1991 (K); Mun. Camacã, near the town, coll. before 1979, G. Daniels, cult. HNT Accn No. 41403,
May 1982 (HNT 5191).
minas gerais: Mun. Rio Vermelho, between the town and Pedra Menina, [Mun. Coluna, fide
Uebelmann (1996)], June 1993, Uebelmann 1507, cult. Univ. Bonn, Germany, 1996 (K).
espírito santo: Mun. Linhares, 6–8 km do Ramal do Lado leste, próximo ao Vale do Rio Doce, 2
Nov. 1971, T.S. dos Santos 2047 (CEPEC); Mun. Domingos Martins, May 1986, Rauh 67565, cult. Univ.
Bonn, Germany, 1995 (K).
conservation status. Endangered [EN B2ab(iii)] (3); PD=1, EI=1, GD=1. Short-list score (3×3) =
9. Endangered in view of continuing forest destruction throughout the region of its occurrence, in which
it is apparently a rarity. Only tiny fragments of its original habitat remain (< 500 km2) and at least 3 of
the above-cited populations are likely to have been destroyed.
This taxon differs from subsp. paradoxa (Rio de Janeiro to Santa Catarina) in its
consistently narrower vegetative parts and darker flowers. Its northernmost habitat, in
Pernambuco, is illustrated in Braun & Esteves Pereira (2002: 20, Abb. 30).
7. Rhipsalis pacheco-leonis Loefgr. in Arch. Jard. Bot. Rio Janeiro 2: 38 (1918)

(‘pacheco-leoni’). Type: region of Cabo Frio, 1915, Campos-Porto & Rose (not found at
RB or SP). Lectotype (Barthlott & Taylor 1995: 59): Rio de Janeiro, Iguaba Grande,
1915, Rose [& Campos-Porto] 20707 (US).
Lepismium pacheco-leonis (Loefgr.) Backeb. in Backeberg & F. Knuth, Kaktus-ABC: 154 (1935, publ. 12
Feb. 1936) (‘pacheco-leonii’).
conservation status. Data Deficient [DD]; requiring survey work in its 6 known locations, but
decline in these can be expected in view of their proximity to urban and touristic developments.
Only the following heterotypic subspecies is found in the area covered here:
7a. subsp. catenulata (Kimnach) Barthlott & N. P. Taylor in Bradleya 13: 59 (1995).
Holotype: Rio de Janeiro, Mun. Nova Friburgo, ‘2000 ft below Sansão’, 1976, Fowlie s.n.
(HNT, n.v.; US, iso.).
Rhipsalis paradoxa var. catenulata Kimnach in Cact. Succ. J. (US) 64: 91 (1992).
RHIPSALIS
Like R. paradoxa subsp. septentrionalis, but smaller in all its parts, stem-segments indeterminate, to only
c. 0.7 cm diam., branching sub-acrotonic, terminal composite areoles only rarely produced, the winglike
stem angles more crowded and compressed, giving the stem a ± quadrangular cross-section; areoles/
scale-leaves to only 1 cm apart, some bearing bristle-spines, or these developing post-anthesis, the scar
left after fruit production being c. 3 mm diam.; flowers < 10 mm diam., brownish salmon without,
pinkish salmon to orange within, anthesis associated with conspicuous nectar production from extra-
floral nectaries on pericarpel (ant-pollination syndrome?); fruit to 5 × 4.5 mm, white with a pink circle
marking the perianth scar.
Southern humid forest element: at c. 900 m, southern Espírito Santo (Domingos Martins); Rio de Janeiro
(Mun. Nova Friburgo). Endemic to South-eastern Brazil. Map 23.
espírito santo: Mun. Domingos Martins, May 1986, Rauh & Kautsky 67560, cult. Univ. Bonn,
Germany, Accn. No. 04502 (K, BONN).
conservation status. Data Deficient [DD]; known from only two localities that await detailed surveys.
Although not studied by the authors in habitat, living material from both localities has
been observed in cultivation.
The homotypic subspecies, R. pacheco-leonis subsp. pacheco-leonis, is known from the
regions of Macaé, Cabo Frio (Iguaba Grande & Búzios) and Pedra da Gávea, Rio de
Janeiro at low elevations. It has often rather weakly developed stem angles, bears deep
pinkish fruit and was confused with R. dissimilis by Britton & Rose (1923).
8. Rhipsalis cereoides (Backeb. & Voll) Backeb., Kakteen Pflanzen Samen 1927–1937,
10 Jahre Kakteenforschung [cat.]: 39 (1937–1938). Type: Brazil, Voll, assumed not to
have been preserved. Neotype (Barthlott & Taylor 1995: 61): Rio de Janeiro, Mun.
Maricá, Itaipu-Açu, Apr. 1936, Voll & Brade s.n. (RB 10258).
Lepismium cereoides Backeb. & Voll in Backeberg & F. Knuth, Kaktus-ABC: 411 (1935, publ. 12 Feb.
1936).
Shrubby, forming dense clumps of ± upright axes (or pendulous when rarely epiphytic); stem-segments
all of determinate size, to 14 × 3 cm, (2–)3–5-angled, new stem-segments arising singly or in groups of
2–3(–4) from the apex of existing segments. Flowers 1–4 per areole, 18–20 mm diam. or more, whitish,
with 6 or more perianth-segments visible from above. Fruit ovoid-oblong, to 10 × 7 mm, but usually
smaller, pink to whitish. Chromosome number: 2n = 22 (Barthlott 1976).
Southern humid forest (inselberg) element: lithophyte on gneissic inselbergs (rarely epiphytic on nearby
trees), southern Espírito Santo (Domingos Martins); Rio de Janeiro (both sides of the Baía de Guanabara).
Map 23.
espírito santo: Mun. Domingos Martins, May 1986, Rauh & Kautsky 67557, cult. Univ. Bonn,
Germany, Accn. No. 04462 (K).
PD=2, EI=1, GD=1. Short-list score (2×4) = 8. Altogether known from < 10 locations, some being
close to the cities of Rio de Janeiro and Niterói and affected by urban expansion and coastal tourism,
resulting in ongoing decline.
This species has not been seen at the site noted above, but the authors were able to study
an extensive population in fruit at Itacoatiara, Rio de Janeiro, near the type locality, in 1996.
RHIPSALIS
9. Rhipsalis sulcata F. A. C. Weber in Bois, Dict. hort.: 1046 (1898). Type: cultivated
material of unknown wild origin, assumed not to have been preserved (P, not found).
Neotype (Barthlott & Taylor 1995: 60): cult. New York Bot. Gard., Mar. 1912 (fls),
received from Simon, ex Muséum [Nat. d’Histoire Naturelle] Paris [where Weber
worked], in 1902 (NY).
Stems ± pendent, apparently all of determinate growth, the basal segments to c. 8–10 mm diam., new
stem-segments arising singly or in groups of 2–5 from the apex of existing segments, vigorous segments
to 250 × 6–7 mm, matt grey-green, weakly but recognizably (3–)5-angled/ribbed whether turgid or
drought-stressed, shorter segments sometimes almost terete, the weakest ultimate segments tapering to c.
3–4 mm diam. near their apices, scale-leaves minute, upon raised podaria, often subtending
inconspicuous, pale bristles to 1 mm, podaria/scales to 70 mm apart on the same angle/rib, 2–4 clustered
around the small, terminal composite areoles at the stem apices. Flowers like those of the preceding
species but c. 15 mm diam. Fruit pinkish.
Southern humid forest element: ecology and range poorly understood; only the following collection of
documented provenance has been seen:
espírito santo: Mun. Domingos Martins, 8 May 1986, Rauh & Kautsky 67562, cult. Univ. Bonn,
Germany, Accn. No. 04490, 1996 (K).
conservation status. Data Deficient [DD]; see below.
Until recently known only in cultivation, where it was misidentified as R. micrantha

(Andes to Central America). It seems to be a member of the R. pacheco-leonis / R.
pentaptera A. Dietr. complex, although it strongly resembles R. floccosa, differing most
obviously in its clearly angled/ribbed stem-segments, less conspicuously erumpent
flower-buds and scarcely woolly fertile areoles post-anthesis. It is possible that it was first
collected in the vicinity of Rio de Janeiro and may not be an endemic of the core area
of Eastern Brazil, although its range is presently documented only from the above-cited
collection from southern Espírito Santo.
Scheinvar (1985) misapplies the name R. sulcata to specimens of the very different R.
trigona Pfeiff. (São Paulo to Santa Catarina).
Subg. Rhipsalis (Nos 10–12): seedlings 3–6-ribbed/angled; adult stems usually terete,
branching ± acrotonic, but producing indeterminate, greatly elongated, basal extension
shoots, the ultimate stem-segments usually the shortest; new stem-segments and flower-
buds inconspicuously erumpent; flowers lateral and sometimes terminal, one per areole,
areoles flowering once only, pericarpel fully exposed; fruits as above. Neotropics and
palaeotropics eastwards to Sri Lanka.
10. Rhipsalis lindbergiana K. Schum. in Martius, Fl. bras. 4(2): 296 (1890). Type:
Rio de Janeiro, based on various syntypes (B†, LE, P, S, C etc.). Lectotype (here
designated, replacing that by Barthlott & Taylor (1995), which is inadmissible under ICBN
Art. 9.10): Widgren s.n. (S). Amongst the 6 collections cited in the protologue is that of the
Swede, ‘Widgren 84’. At Stockholm (S) Schumann has annotated Reidel 84 and Widgren
s.n. as his new species and it is here assumed that his later reference to ‘Widgren 84’ is an
error caused by the omission of the word ‘Reidel’ from the list in the protologue.
RHIPSALIS
R. densiareolata Loefgr. in Arch. Jard. Bot. Rio Janeiro 2: 41, tab. XII (1918). Type: Rio de Janeiro,
Tijuca, 1915, Rose & Campos-Porto (holo. not found RB, R, SP, US). Lectotype (designated here):
Löfgren, l.c., tab. XII (1918).
vernacular name. Enxerto.
Pendent to 4 metres or more, forming a dense mass of closely adjacent stems; branching sub-acrotonic,
new stem-segments arising singly or in groups of 2–5 from near the apex of existing segments, of
indeterminate growth, lacking terminal composite areoles, to 60–90 × 3–6(–12) mm, cylindric, but
sometimes longitudinally ridged when drought-stressed or when dried for the herbarium, light grey-green
to dark green; the soon scarious scale-leaves marking the position of the hidden areoles often arranged in
whorls, these to c. 10 mm apart; areoles apparent only after flowering, developing sparse wool and usually
a few bristle-spines, areole scar c. 1.5 mm diam. Flowers to 9 mm diam.; perianth-segments 5–9 visible
from above the flower, greenish white, at full anthesis strongly reflexed to show off the stamens; stigma-
lobes 3–4, white, c. 1 mm; pericarpel green to reddish. Fruit 3–5 × 2.5–4 mm, white or pink.
Disjunct/widespread humid/subhumid evergreen forest element: epiphyte or less commonly a

lithophyte, in high restinga forest, Mata atlântica, mata de brejo and rarely at edges of caatinga-agreste and in
campo rupestre, near sea level to c. 1000 m, southern Pernambuco southwards, mostly within 150 km of
the coast but reaching the northen parts of the Chapada Diamantina and Serra do Espinhaço (MG), rarely
on the latter’s western flanks in the south; to South-eastern Brazil (to W Rio de Janeiro & SE São Paulo).
Map 24.
pernambuco: Mun. Inajá, Reserva Biol. Serra Negra, topo da serra, c. 1000 m, 14 Sep. 1995, A. Laurênio
et al. 159 (K, PEUFR), 9/10 Dec. 1995, M. Tschá et al. 369, 387 (K, PEUFR); Mun. Brejão, 7 km N of
Terezinha, 1 km S of turning to Brejão, 13 Feb. 1991, Taylor & Zappi 1629 (UFP, HRCB, K, ZSS).
sergipe: Mun. Campo de Brito, Tapera da Serra, 11 Nov. 1981, G. Viana 202 (IPA, ASE 1658).
bahia: cent.-N Bahia, Mun. Saúde, 68.5 km S of Campo Formoso on road to Jacobina, 12 Jan. 1991,
Taylor et al. 1396 (CEPEC, HRCB, K, ZSS); Mun. Jacobina, Barracão de Cima, 11°1'7"S, 40°32'43''W,
6 July 1996, H.P. Bautista et al. in PCD 3465 (K, ALCB, HUEFS); Mun. Morro do Chapéu, as epiphyte
and lithophyte, both sides of road BA 052, nr Pousada Ecológica das Bromélias, 1000 m, 3 Aug. 2002,
Taylor & Machado (K, photos); Mun. Mundo Novo, 2 km E of town on road BA 052, 26 Apr. 1992,
Taylor & Zappi (obs.); NE & E Bahia, Mun. Candeal, 8 July 1994, F.A.R. Santos 92 (HUEFS); Mun.
Alagoinhas, 24 km S of Inhambupe, 15 Feb. 1991, Taylor & Zappi 1635 (CEPEC, HRCB, K, ZSS);
Mun. Entre Rios, 12 Feb. 1991, Taylor & Zappi 1614 (HRCB, K, ZSS); Mun. Mundo Novo, 2 km E
on road BA 052, 26 Apr. 1992, Taylor & Zappi (obs.); Mun. Baixa Grande, 1.5 km E of junction of road
between Rui Barbosa and Juazeiro towards Baixa Grande, 9 Feb. 1991, Taylor et al. 1606 (K, HRCB,
ZSS); Mun. Ipirá, 24 km E of town towards Feira de Santana on road BA 052, 9 Feb. 1991, Taylor et
al. 1611 (K, HRCB, ZSS); Mun. Rui Barbosa, 7 km N of town, Ritter 1245 (SGO 125541, fide Eggli
et al. 1995: 503–504); Mun. Feira de Santana, 12°15'16"S, 39°2'46''W, 7 Dec. 1992, L.P. Queiroz et al.
2926 (K, CEPEC, HUEFS); l.c., 12°18'S, 39°W, 6 Feb. 1997, F. França 2079 (HUEFS); Mun.
Conceição do Jacuípe, 87 km NW of Salvador on road to Feira de Santana, 12 Mar. 1989, reported in
Barthlott & Taylor (1995: 53, pl. 17); Mun. Cachoeira, Vale dos Rios Paraguaçu e Jacuípe, 39°5'W,
12°32'S, Nov. 1980, Grupo Pedra do Cavalo 931 (ALCB, IPA, HRB); Mun. Cachoeira/São Felix, 13
Feb. 1920, Zehntner s.n. (US); Mun. Cruz das Almas, 18 June 1915, Rose & Russell 19896 (US, NY);
Mun. Lajedo do Tabocal, 25.5 km W of road BR 116 towards Maracás, 5 Feb. 1991, Taylor et al. 1556
(CEPEC, HRCB, K, ZSS); Mun. Jaguaquara [Toca da Onça], 27 June 1915, Rose 21408 (US); Mun.
Ipiaú, Aug. 2001, Zappi (obs.); Mun. Vitória da Conquista, 12 km from town towards Itambé, 16 Aug.
2002, M. Machado (obs.); Mun. Juçari (Jussari), 27 May 1966, R.P. Belém & R.S. Pinheiro 2328 (UB,
CEPEC, F); Mun. Una, 5 km N of Comandatuba, 1977, Harley & Storr 69, cult. R.B.G. Kew Accn.
No. 1977-634, 1995 (K).
minas gerais: N Minas Gerais, Mun. Montes Claros, 13 km N of town on road to Januária, 16°38'S,
43°55'W, 6 Nov. 1988, Taylor & Zappi in Harley 25510 (SPF, K); E, cent.-S & SE Minas Gerais, Mun.
Conselheiro Pena, 26 km NE of the Rio Doce, road to Cuparaque via Penha do Norte, 15 Dec. 1990,
Taylor & Zappi 774 (BHCB, HRCB, K, ZSS); Mun. Lagoa Santa, A.P.A. Carste de Lagoa Santa, Oct. 1995
RHIPSALIS
to Feb. 1996, Brina & Costa in BHCB 32768 (BHCB, K); Mun. Marliéria, Parque Estadual do Rio Doce,
20 Sep. 1974, E.P. Heringer 13991 (UB); l.c., Revés de Belém, 1 Sep. 1975, D. Sucre et al. 10190 (RB);
Mun. São João del Rei, Oct. 1895, A. Silveira et al. 93 (R); Mun. Ubá, s.d., Saint-Hilaire 159 p.p. (P); Mun.
Muriaé/Laranjal, 2.5 km N of São João da Sapucaia and 8 km N of Laranjal on road BR 116, 18 Dec. 1990,
Taylor & Zappi 792 (BHCB, HRCB, K, ZSS); Mun. Laranjal, 8 km em direção a Palma, 24 Nov. 1982,
J.R. Pirani et al. 252 (SP, SPF); Mun. Leopoldina, 16 Oct. 1981, P.L. Krieger & M. Brugger 18710 (CESJ,
SPF); Mun. Lima Duarte, May 1897, F. Brandão s.n. in Herb. Com. Geog. Geol. de Minas No. 2298 (R).
espírito santo: Mun. Linhares, Canal da Lagoa Juparanã, 13 Aug. 1965, R.P. Belém 1588 (CEPEC),
21 July 1976, H.F. Martins s.n. (MBML); Mun. Santa Teresa, Museu de Biologia, 4 Oct. 1988, H.Q.B.
Fernandes 2587 (MBML); Mun. Afonso Cláudio, road to Itarana, 20°3'8"S, 41°3'3''W, 24 Nov. 1999,
In Eastern Brazil this species can be readily distinguished from R. baccifera in the living
state by its uniformly very long shoots lacking composite terminal areoles and giving rise
to subacrotonic secondary segments. Its fruits are generally smaller than those of R.
baccifera and sometimes pinkish. Unfortunately, these species are less easy to separate in the
herbarium and have often been confused, although this can be easily avoided on the basis
of provenance, since they have discrete ranges, being sympatric only in parts of the Hiléia
Baiana of eastern Bahia, where they differ markedly. R. lindbergiana is somewhat variable
in stem thickness, the stoutest forms including that described as R. densiareolata.
R. lindbergiana most closely resembles R. baccifera subsp. shaferi (Britton & Rose)
Barthlott & N. P. Taylor, which ranges westwards and south-westwards from inner São
Paulo (Campinas) to Paraguay, northern Argentina and southern Bolivia.
11. Rhipsalis teres (Vell.) Steud., Nomencl. bot., ed. 2, 2: 449 (1841). Type: not
extant. Lectotype (Barthlott & Taylor 1995: 64): Brazil, Rio de Janeiro, Ilha de Santa
Cruz, Vellozo, Fl. flumin. Icones 5: tab. 30 (1831).
Cactus teres Vell., Fl. flumin.: 196 (1829); ibid., Icones 5: tab. 30 (1831).
Hariota prismatica Lem. in Ill. Hort. 10, misc.: 84–85 (1863). Rhipsalis prismatica (Lem.) Rümpler in C.F.
Först., Handb. Cact. ed. 2: 884 (1886). R. teres f. prismatica (Lem.) Barthlott & N.P. Taylor in
Bradleya 13: 65 (1995). Type: Brazil, ex Hort. Verschaffelt, Oct. 1863, not known to have been
preserved. Neotype (Barthlott & Taylor, l.c.): Rio de Janeiro, Mun. Maricá, A.P.A., 1 Mar. 1991,
M. Botelho & L.O. Melande 507 (GUA).
R. virgata F.A.C. Weber in Rev. Hort. 64: 425 (1892). Type: Brazil, c. 1883–1884, [collector?], not
known to have been preserved.
R. tetragona F.A.C. Weber, l.c. 429 (1892), nom. inval. (Art. 34)? Type: cult. material, not known to have
been preserved.
R. gracilis N.E. Br. in Gard. Chron. ser. 3, 33: 18 (1903). Holotype: Hort. J. Corderoy, cult. Kew (K).
R. heteroclada Britton & Rose, Cact. 4: 224 (1923). R. teres f. heteroclada (Britton & Rose) Barthlott &
N.P. Taylor, l.c. 65 (1995). Type (Barthlott & Taylor, l.c.): Brazil, Rio de Janeiro, Rose 20371 (US,
lecto.; NY).
R. alboareolata F. Ritter, Kakt. Südamer. 1: 41 (1979). Holotype: Brazil, Rio Grande do Sul, 1964, Ritter
1286 (U).
R. clavellina F. Ritter, l.c. 43 (1979). Holotype: l.c., 1964, Ritter 1487 (U).
R. maricaensis Scheinvar in Arch. Jard. Bot. Rio Janeiro 31: 71–78 (1992 publ. 1993). Holotype: Rio de
Janeiro, Maricá, Casari 1143 (GUA).
[R. penduliflora sensu K. Schum. in Martius, Fl. bras. 4(2): 276 (1890), non N.E. Br.]
RHIPSALIS
conservation status. Least Concern [LC]; widespread and very common in SE & S Brazil, where at
least some of its habitat remains intact or is included within protected areas.
A variable and complex taxon like R. baccifera, of which it is assumed to be the southern
sister-species. Plants from the core area of Eastern Brazil are referred to the following form:
11a. f. capilliformis (F. A. C. Weber) Barthlott & N. P. Taylor in Bradleya 13: 65 (1995).
Type: Brazil, cult. Hort. Chantin, not known to have been preserved. Neotype (Barthlott
& Taylor, l.c.): Rio de Janeiro, Serra dos Órgãos, 1966, Hunt 6510 (K, neo. & isoneo.).
R. capilliformis F.A.C. Weber in Rev. Hort. 64: 425 (1892).
Like the following and especially R. baccifera subsp. hileiabaiana, but basal extension shoots to 100 cm,
ultimate stem-segments only 1.3–2.0 mm thick; flowers apparently self-incompatible, pericarpel c. 2–2.5
× 2–2.5 mm, smaller than the perianth during most of the bud’s development, flowers borne laterally,
laterally and terminally, or rarely terminally only, c. 7 × 7 mm or wider before the perianth-segments
become strongly reflexed, these c. 3.5 mm or larger, often turning yellow post-anthesis; stamens more
numerous, 3–4 mm or more; stigma-lobes 3–5; fruit globose to shortly barrel-shaped, to 4–5 × 4 mm,
white or pink to deep purplish red, leaving a more prominent scar on the stem. Chromosome number:
2n = 22 (Barthlott 1976).
Southern humid forest element: epiphyte or lithophyte in Mata atlântica; common in the Serra da
Mantiqueira and Serra do Mar of South-eastern and Southern Brazil (to Rio Grande do Sul). Map 24.
minas gerais: Mun. Entre Rios [de Minas], Jan. 1898, A. Jaguaribe 2894 (R); Mun. Juiz de Fora, Museu
Mariano Procópio, 30 Oct. 1986, Krieger & Coelho 77 (RB).
espírito santo: Mun. Piúma, Monte Agá, south of the town, 20°52'23"S, 40°46'24''W, 26 Nov. 1999,
reported growing with Coleocephalocereus fluminensis (Zappi et al. 469).
It is probable that R. teres will be found elsewhere in the southern part of the core area.
12. Rhipsalis baccifera (J. S. Muell.) Stearn in Cact. J. (Croydon) 7: 107 (1939).
Lectotype (Aymonin 1983): J. S. Mueller, l.c. infra, tab. 29.
Cassyta baccifera J. S. Muell., Ill. syst. sex. Linnaei. class IX. ord. 1, tab. 29 (1770–77).
Rhipsalis cassutha Gaertn., Fruct. sem. pl. 1: 137 (1788). Lectotype (designated here): Gaertner, l.c., tab.
28 (1788).
R. cassuthopsis Backeb., Die Cact. 2: 660 (1959). R. cassythoides Loefgr. in Arch. Jard. Bot. Rio Janeiro 2:
40 (1918) (‘cassythoydes’) non G. Don (1834). Type: Brazil, Pará, Belém, Simão da Costa, cult. RB
(holo. not found RB, R, SP etc.). Lectotype (designated here): Löfgren, l.c., tab. XI (1918).
vernacular names. Enxerto, Conambaia, Tripa-de-galinha, Irahuka’arã.
Pendent to 2 m or more, freely producing aerial roots, pale to dark green; all stems slender cylindric, naked
except for minute whitish scale-leaves occasionally subtending very fine whitish bristles, basal extension
shoots to 500 × 3–5 mm, eventually rather woody, subsequent branching acrotonic from felted terminal
composite areoles, the secondary segments arising in clusters of 3 or more, subsequent orders of segments
decreasing in size, the penultimate c. 90 × 2 mm, giving rise to clusters of 2 or more ultimate segments
measuring 20–65 × 2 mm. Flowers self-compatible, terminal or subterminal from or around the composite
areoles of the ultimate and lower order stem-segments, sometimes 2 or more clustered together, often also
RHIPSALIS
lateral, especially on the extension shoots; buds minutely erumpent, the pericarpel at least twice the size of
the developing perianth until shortly before anthesis; flowers 6–7 × 4–5 mm, principal perianth-segments
4–6, to 3 mm, patent to reflexed, greenish white; stamens 1–2.5 mm, relatively few, whitish; style 3 mm,
stigma-lobes 3, 1 mm, white, exserted; pericarpel barrel-shaped, 3 × 2 mm, pale green. Fruit translucent
white, sometimes pinkish, or violet near apex, ovoid, to 7.5 × 6 mm, juice extremely mucilaginous and
sticky. Seed (Barthlott & Hunt 2000: 35–36, pl. 5.7–8) 1.2 × 0.6 mm, glossy black-brown.
conservation status. Least Concern [LC]; at the level of species it has the widest range of any
member of the Cactaceae.
Note: the type of R. baccifera is assumed to have come from the Caribbean, whence it was
introduced to England by Philip Miller in 1758 (Stearn, l.c.). The above description accounts
only for what appear to be typical, slender-stemmed forms found in Northern and North-
eastern Brazil and the Caribbean (eg. Jamaica), since it is clear that the species represents a
complex entity requiring further detailed study. Thicker-stemmed forms, such as are known
from the Guianas and elsewhere (as well as Old World plants referred to its various
heterotypic subspecies) are not accounted for above, although some of these from the
Americas were previously included with subsp. baccifera by Barthlott & Taylor (1995: 63).
In Eastern Brazil this species is divisible into the following subspecies:
1. Higher order stem-segments short, densely clustered, 6 or more axes arising from the
apex of the longer lower order segments (coastal E Bahia and region of Catolés, Mun.
Abaíra, Chapada Diamantina) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12b. subsp. hileiabaiana
1. Stem-segments not as above (Amazônia; and Maranhão to E Pernambuco) . . 12a. subsp. baccifera
12a. subsp. baccifera

Higher order stem-segments 40–90 mm, solitary or in clusters of 2–3 from the apex of previous segments.
Flowers with perianth-segments mostly patent; stamens 1–2 mm.
Amazonian forest element: epiphyte in mata de brejo and mata de tabuleiro, near sea level to c. 600 m,
North-eastern Brazil southwards as far as coastal Pernambuco; replaced by subsp. hileiabaiana in central &
eastern Bahia; common throughout more humid parts of the neotropics, from Northern Brazil and the
central Andes of Bolivia etc. northwards to eastern Mexico, the Caribbean islands and Florida (replaced
by subspp. erythrocarpa, mauritiana & horrida in the palaeotropics). Map 24.
maranhão: Mun. Santa Luzia do Paruá, Nova Esperança, Rio Alto Turiaçu, 2°55'S, 45°45'W, 30 Nov.
1978, J. Jangoux & R.P. Bahia 88 (K, MG); Mun. Monção, Ka’apor Indian Reserve, 7 km from
‘Urutawy’, 17 Feb. 1985, Batée & Ribeiro 840 (K, NY).
ceará: N Ceará (Serra de Baturité), Mun. Pacatuba, ‘serra na mata ao sítio Pitaguari’, 27 Aug. 1979, J.E.
de Paula & R.C. de Mendonça 1232 (MG); Mun. Guaramiranga, 4 km W of town, Sítio Uruguaiana, 23
Mar. 1945, H.C. Cutler 8319 (F).
paraíba: Mun. Itaporororca, 27 Oct. 1998, E.A. Rocha 509 (IPA); Mun. Areia, Centro de Ciências
Agrárias, 6°58'12"S, 35°42'15''W, 29 Dec. 1980, V.P.B. Fevereiro et al. M417 (K), l.c., Campus III da
UFPB, 30 Apr. 1998, E.A. Rocha 404 (EAN, JPB, IPA, UFP); Mun. João Pessoa, on old tree near city
centre, 10 Feb. 1995, Taylor (obs.).
pernambuco: Mun. São Lourenço da Mata, Usina Tiúma, 1930, Pickel 2469 (IPA 6874, K, photo);
Mun. Olinda, 9 Apr. 2000, E.A. Rocha et al. (obs.); Mun. Recife, Dois Irmãos, 10 Dec. 1999, E.A. Rocha
(PEUFR); l.c., on trees in the campus of Univ. Federal de Pernambuco (UFPE), 13 Dec. 1999, Zappi
(K, photos), 30 Mar. 2000, Taylor & Griffiths (obs.); l.c., Jardim Botânico, 30 Mar. 2000, Taylor & Griffiths
(K, photo).
RHIPSALIS
Records of R. baccifera from South-eastern and Southern Brazil refer to R. lindbergiana and
R. teres; see above. However, R. baccifera subsp. shaferi (Britton & Rose) Barthlott & N.
P. Taylor is known from the state of São Paulo, where it has been collected in Mun.
Campinas (Ratter s.n.).
12b. subsp. hileiabaiana N. P. Taylor & Barthlott in Bradleya 13: 63 (1995). Holotype:
Brazil, Bahia, Mun. Ilhéus, 7 Aug. 1983, J. L. Hage & H. S. Brito 2113 (CEPEC; K,
MBM, HRB isos.).
Higher order stem-segments to c. 30 mm, in dense clusters (to 10 or more from the apex of vigorous,
lower order segments), strongly differentiated from the much longer extension shoots. Flowers with
perianth-segments reflexed to expose the 2–2.5 mm stamens.
Disjunct Bahian humid forest element: epiphyte (rarely lithophyte) in Mata atlântica (Hileia Baiana/mata
de brejo), at low elevations to c. 800 m, and in mata de neblina, 1650–1800 m, region of Catolés (Mun.
Abaíra), Chapada Diamantina, eastern and central Bahia. Endemic to Bahia. Map 24.
bahia: cent. Bahia (Chapada Diamantina), Mun. Abaíra, 13°15–16'S, 41°54–55'W, Catolés de Cima,
Tijuquinho, 10 Jan. 1992, Nic Lughadha et al. in Harley 50699 (SPF, CEPEC, HUEFS, K), 16 Nov. 1992,
W. Ganev 1468 (HUEFS, K), l.c., Mata do Cigano, 26–28 Feb. 1992, P.T. Sano & T. Laessøe in Harley
52352, 52395 (SPF, CEPEC, HUEFS, K), l.c., Riacho da Taquara, 3 Feb. 1992, Stannard et al. in Harley
51154 (SPF, HUEFS, CEPEC, K); E Bahia, Mun. Santo Amaro, Oliveira dos Campinhos, Mar. 1952,
A. Leal Costa 1096 (ALCB); Mun. Santa Teresinha, Serra da Jibóia, 23 Nov. 2001, Zappi (K, photos);
Mun. Maragogipe, Guapira, 16 June 1987, G.C.P. Pinto 24/87 (HRB); município unknown, basin of
the Rio Gongoji, Oct.–Nov. 1915, H.M. Curran 135 (US), Apr. 1917, Luetzelburg 8 (M); Mun.
Almadina/Coaraci, 9 Aug. 1977, L.A. Mattos 93 & J.L. Hage 179 (CEPEC, K); Mun. Ilhéus, CEPEC,
Km 22 on the road Ilhéus–Itabuna, 12 Apr. 1968, S.G. da Vinha 180 (ALCB), 19 Nov. 1970, J.L. Hage
29 (CEPEC), 23 Feb. 1981, J.L. Hage & E.B. dos Santos 465 (CEPEC), 20 May 1981 & 7 Aug. 1983,
J.L. Hage & H.S. Brito 695 (CEPEC, K), 2113 (CEPEC, MBM, K, HRB), 24 June 1965, R.P. Belém &
M. Magalhães 524 (UB); Mun. Floresta Azul, [before 1966], Martins in coll. Brieger, Piracicaba, SP (No.
43), cult. R.B.G. Kew (K, Spirit Coll. No. 52524); Mun. Vitória da Conquista, 12 km from town
towards Itambé, 16 Aug. 2002, M. Machado (K, photos); Mun. Una, 6 km SE of Una, 1977, Harley &
Storr 86, cult. R.B.G. Kew Accn. No. 1977-652 (K); Mun. Canavieiras, road BA 001, 0.5 km N of
junction with road to Santa Luzia, 23 April 2003, Taylor & Zappi (obs.); unlocalised, s.d., Blanchet 1518 (P).
conservation status. Vulnerable (2) [VU B2ab(i, ii, iii, iv)]; area of occupancy estimated to be <
2000 km2; PD=1, EI=1, GD=1. Short-list score (2 × 3) = 6. It is assumed that a significant proportion
of the above-cited localities (c. 5) have suffered severe modification.
This endemic subspecies, which is restricted to the region of Bahia receiving most rainfall
(within the 1750 mm annual isohyet along the coast and from very humid woodland in
brejo forest on serras further inland and in the highest part of the Chapada Diamantina),
strongly resembles some forms of R. teres in habit, but has flowers and fruits typical of R.
baccifera.
Subg. Erythrorhipsalis A. Berger (Nos 13–18). Like Subg. Rhipsalis, but flowers campanulate
(except sometimes in No. 18), pendent, scented, one or more at a time from or around
the margins of the terminal collective areole of ultimate and sometimes lower order stem-
segments (also occasionally lateral, especially in R. pulchra, but obliquely oriented on the
segments); perianth-segments 8–18 or more; stamen filaments usually highly coloured at
RHIPSALIS
base giving the flower a coloured throat; fruit white, pink, purplish, red or orange. SE
South America. Type: R. pilocarpa Loefgr.
13. Rhipsalis pulchra Loefgr. in Arch. Jard. Bot. Rio Janeiro 1: 75–76, tab. 5 (1915).
Holotype: Brazil, Serra da Mantiqueira, O. A. Derby 4394 in Comm. Geogr. Geol. São
Paulo 8834 (SP; US, iso.).
Rhipsalis macahensis Glaziou in Mém. Soc. Bot. France 1(3): 326 (1909), nom. nud. Intended type: Rio
de Janeiro, Alto Macahé, Glaziou 18262 (P, K).
Pendulous to 6 metres or more, stems erect at first but eventually always pendulous, weak and pliable,
almost as if flaccid, emitting aerial roots, to 5 mm diam., plain green, the scales sometimes with reddish
marks below; branching acrotonic or subacrotonic, the stem-segments of indeterminate size but generally
decreasing in length further away from the rootstock, their apices often not forming recognisable
composite areoles, when present these very inconspicuous, actively growing segments gradually tapered
to a fine point at apex; areoles almost 0 except on juvenile stems, represented by a few minute trichomes
visible at the distal edges of the inconspicuous scales, these < 0.25 × 0.5 mm, truncate, apiculate, pale at
maturity, triangular and reddish when young. Flower-buds inconspicuously erumpent, leaving a glabrous
scar to 1.5 mm diam. on the stem, pinkish to dark reddish; flowers with a distinctive sweetish scent,
downwardly directed, lateral and aligned with the stem and/or ± terminal and sometimes in groups of 3,
in lateral flowers the pericarpel attached sublaterally; perianth not expanding fully, ± campanulate, 15 ×
15–25 mm; perianth-segments c. 11, 8–9 visible from within the flower, outermost ovate, c. 6 × 5 mm,
whitish tinged purplish-pink at apex, inner segments 10–13 × 4.5 mm, creamy white, concave (boat-
shaped) from within; stamens in two groups, the outer series spreading towards the perianth, the inner
adpressed to style and varying from very short to equalling the style, all except the outermost with dark
orange filaments, anthers pale yellow; style 5 mm, stigma-lobes 3–6, to 3 × 0.7 mm, white, spreading;
pericarpel 4–5 × 4.5 mm, pale green to pinkish, naked or with 1–2 fleshy bract-scales. Fruit slightly
depressed-globose, to c. 7 × 7.5 mm, translucent whitish or purplish red.
Southern humid forest element: epiphyte in capão de mato, > 1500 m, Serra da Mantiqueira, southern
Minas Gerais; South-eastern Brazil (Rio de Janeiro & São Paulo). Map 25.
minas gerais: Mun. Lima Duarte, Parque Estadual do Ibitipoca, 27 Sep. 1970, Krieger 9269 (CESJ,
HRCB), 27 July 1991, Zappi 260 (K, SPF, CESJ), 3–5 Aug. 2003, Taylor & Zappi (K, photos).
Possibly Conservation Dependent, since it is found within various protected areas.
This poorly known species was originally described with and commonly bears purplish
magenta fruits and deep pink flowers, but the population cited here from southern Minas
Gerais (Zappi 260) has white fruits and rather pale flowers. However, there can be no
doubts about its identity.
14. Rhipsalis burchellii Britton & Rose, Cact. 4: 225 (1923). Lectotype (Barthlott &
Taylor 1995: 69): São Paulo, Mun. São Paulo, Jabaquara, 15 Aug. 1915, Rose & Russell
20857 (US; NY, K, lectoparas.).
Plant to 2 m or more; stems freely emitting aerial roots; primary shoots to 50.0 × 0.5 cm, crimson-red,
secondary to fifth order to 2.5–7.0 cm, terminal segments 2.5–3.5 cm × 1–2 mm, sometimes slightly
thickened at the tips; flowers terminal, 1–3 together, or appearing at the joints of the subterminal stem-
segments, campanulate, only half-expanded, to 20–25 × 20 mm, pale silvery pink; pericarpel c. 4 × 3.5
mm, stigma-lobes 4–5, to 5 mm, whitish. Fruit ovoid, c. 10 mm, magenta.
RHIPSALIS
Southern humid forest element: epiphyte in mata de neblina/galeria, c. 900 m, southern Espírito Santo;
South-eastern and Southern Brazil. Map 25.
espírito santo: Mun. Domingos Martins, May 1986, Rauh, living material observed in a private
collection in California, U.S.A., 1997, Taylor (obs.); l.c., R. Whitman s.n., cult. Univ. Bonn, Accn. No.
13488, 18 Mar. 2000, Taylor (obs.).
conservation status. Not Evaluated [NE] in view of taxonomic uncertainties (see below).
This complex of species, amongst which R. burchellii has the oldest typifiable name, is
difficult to resolve from herbarium materials alone, and it is possible that as many as 5
species of this relationship are present in the area. The following, recently described
species is one of these, as is No. 16.
The oldest name within this complex is R. cribrata (Lem.) N. E. Br. (Hariota cribrata
Lem. 1857), but this is too poorly typified to be applied with confidence and has been
variously misapplied by previous authors (Barthlott & Taylor 1995).
15. Rhipsalis juengeri Barthlott & N. P. Taylor in Bradleya 13: 69, 72, pl. 29 & 30
(1995). Holotype: origin unknown, cult., Bot. Gard. Univ. Bonn, Germany, Accn. No.
01700, Mar. 1995 (BONN).
Like R. burchellii in habit and R. clavata in its flowers, but plant pendulous to at least 3 m and sometimes
much longer; stems scarcely succulent, plain green or tinged dull red; primary long extension shoots to
200 × 0.3 cm, terminal segments 1.25–1.75 mm thick; flowers solitary or two together, sometimes
appearing at the joints of the subterminal stem-segments, campanulate, only half-expanded, c. 12–15 ×
12–15 mm, whitish, pericarpel c. 2 × 2.5 mm, stamens yellow at base; fruit globose-truncate, c. 6 mm or
more, purplish to translucent greenish tinged maroon, smelling strongly of blackcurrants (fide Barthlott).
Southern humid forest element: epiphyte in Mata atlântica, c. 1500–1600 m, south-eastern/southern

Minas Gerais; range and/or endemic status uncertain. Map 25.
minas gerais: Mun. Lima Duarte, Parque Estadual do Ibitipoca, 27 July 1991, Zappi 259 (K, SPF,
CESJ); (?) l.c., 12 May 1970, Krieger 8594 (CESJ), 1 Nov. 1973, Krieger 13174 (CESJ, HRCB), 29 & 30
Sep. 1970, D. Sucre et al. 7193, 7228 (RB); Mun. Lima Duarte/Rio Preto, Serra Negra, sterile living
plant shown to the authors at the July 1998 Congresso Nac. de Botânica, Salvador, Bahia.
conservation status. Not Evaluated [NE] in view of taxonomic uncertainties.
Of the above, only the collection by Zappi and that from the Serra Negra have been seen
as living plants and suggest the above identity, although neither has been seen in flowering
condition.
16. Rhipsalis clavata F. A. C. Weber in Rev. Hort. 64: 429 (1892); N. P. Taylor in Bot.
Mag. 19: 160–164, tab. 446 (2002). Type: Brazil, Rio de Janeiro, Petrópolis, 1886, Binot,
probably a living plant (assumed not to have been preserved). Neotype (Barthlott & Taylor
1995: 72): l.c., Teresópolis, Granja Comari, 11 Feb. 1964, Castellanos 24569 (GUA).
Hariota clavata F.A.C. Weber, l.c. (1892). Hatiora clavata (F.A.C. Weber) Moran in Gentes Herb. 8: 343
(1953).
Pendent to 3 m or more, basal extension shoots of indeterminate growth rare, but when present to c. 12
cm; primary determinate segments 3–7 cm long, secondary to fifth order segments often swollen at tips,
RHIPSALIS
up to 2.5(–6.0) cm long, terminal segments 1–4 cm long, 1.0–1.5 mm thick, thickened at the tips to
1.5–3.0 mm, up to 10 or more arising at once from the collective areole at apex of the previous segment.
Flowers terminal, 1–3 together, sometimes appearing at the joints of the subterminal segments, pendent,
campanulate to fully expanded, 13–15 × 13–23 mm; perianth-segments oblanceolate, to 10 × 2.5 mm, pure
white, (7–)8–11 visible from within; stamens differentiated into long and short, not coloured at base; style
8 mm long, stigma-lobes 3–5, 2–3 mm long, slender, spreading; pericarpel pale whitish green, c. 2–3 × 2.5
mm; fruit (excluding withered perianth remains) c. 4.5–5.0 × 4.5–5.0 mm, greenish white with a magenta
ring at apex around the perianth remains/scar, rarely red. Chromosome number: 2n = 22 (Barthlott 1976).
Southern humid forest element: epiphyte in Mata atlântica, c. 800–1140 m, southern Espírito Santo and
northern Rio de Janeiro (and perhaps south-eastern Minas Gerais); South-eastern Brazil, from sea level
to high elevations (westwards to Ilha São Sebastião, São Paulo). Map 25.
minas gerais (all of the collections cited from this state are only provisionally identified as this species
in the absence of studies based on living material): SE Minas Gerais, Parque Nacional do Caparaó, 15
Oct. 1988, M. Brugger et al. in FPNC 328, 330 (CESJ, HRCB), Oct. 1941, Brade 17115 (RB); l.c.,
Cachoeira Bonita, 18 Sep. 1988, Krieger in FPNC 233 (CESJ, HRCB, SPF).
espírito santo: Mun. Domingos Martins, 1986, Kautsky in Rauh 68384 (BONN) — red-fruited form.
rio de janeiro: Mun. Santa Maria Madalena, Parque do Desengano, Santa Clara, 21 Jan. 1976, Araujo
950 (GUA, MEXU); l.c., ‘subida da Pedra Dubois’, 21°56'36–42"S, 41°59'29–33''W, 22 Nov. 1999,
conservation status. Not Evaluated [NE] in view of taxonomic uncertainties.
The collections cited above from Rio de Janeiro and Espírito Santo are definitely R.
clavata, but those from Minas Gerais cannot be confidently identified at present from dried
specimens and may include representatives of the preceding species as well as possibly the
orange-fruited R. campos-portoana Loefgr., which appears to be wide-ranging in Southern
and South-eastern Brazil. A specimen of R. clavata said to have been collected in the
grounds of CEPLAC, between Itabuna and Ilhéus, Bahia (coll. 1975, G. Daniels, cult.
Huntington Bot. Gard.) and conserved in HNT is suspected as having incorrect
provenance data (probably due to switched labels during cultivation). This Bahian locality
is amongst the most well-collected by local botanists, who presumably would have
obtained the distinctive R. clavata by now, if it occurred there.
17. Rhipsalis cereuscula Haw. in Philos. Mag. Ann. Chem. 7: 112 (1830). Type:
Brazil, a living plant not known to have been preserved or illustrated. Neotype (Barthlott
& Taylor 1995: 69): Brazil, São Paulo, Mun. Piracicaba, campus of ESALQ, 3 Dec. 1993,
V. C. Souza 4970 (ESA; K, isoneo.).
Hariota cereuscula (Haw.) Kuntze, Revis. gen. pl. 1: 262 (1891). Erythrorhipsalis cereuscula (Haw.) Volgin in
Vestn. Moskovsk. Univ., Ser. 16, Biol. 36(3): 19 (1981).
Hariota saglionis Lem., Cact. aliq. nov.: 39 (1838). Rhipsalis saglionis (Lem.) Otto in Walp., Repert. bot.
syst. 2: 936 (1843). Type: a living plant of unknown provenance in the collection of Monville;
assumed not to have been preserved.
R. penduliflora N.E. Br. in Gard. Chron. ser. 2, 7: 716 (1877). Holotype: Hort. Pfersdorff, 1875, cult.
Kew (K).
[R. cribrata sensu K. Schum. in Martius, Fl. bras. 4(2): 277–278 (1890), non (Lem.) N.E. Br.]
Stems pale green, freely emitting aerial roots, these attaching the basal extension shoots and terminal
clusters of stem-segments to the trunk and branches of the host tree in new positions; extension shoots
to c. 60 cm × 3 mm, erect then spreading and finally pendent, densely clothed in fine adpressed bristle-
RHIPSALIS
spines at first, later mostly caducous, these subtended by minute reddish scales, the bristles forming a small
erect tuft at the stem-segment apices, second order segments arising in clusters of up to 10 or more at or
near apex of extension shoots, c. 4–10 cm, or these lacking and extension shoots bearing specialized
ultimate segments directly; highest order stem-segments densely clustered, sausage-like to ± globose, of
irregular thickness and shape, 5–15 × 2–4 mm or thicker, sometimes slightly angled or ribbed, often
partially shrunken or appearing swollen with sap, branching strictly acrotonic, with 1–5 new segments
developing at apex from the bristly composite areole. Flowers in early spring, c. 17 × 16 mm,
campanulate, whitish or very pale creamy yellow, 1–3 together from the terminal areole or its vicinity;
pericarpel obovoid, c. 5.5 × 4.0 mm, pale yellow-brown to pale green, ± naked or bearing reddish bract-
scales with trichomes and/or occasional bristles in their axils; perianth-segments c. 15, to 9 × 4 mm,
partially differentiated into an outer somewhat spreading series and an inner series remaining erect and
tending to enclose the stamens and style at the start of anthesis; stamens whitish, ± reddish at base, the
innermost clustered around the whitish, exserted style, some of the outermost almost as long as the
perianth-segments; stigma-lobes 3–5, c. 1 mm, whitish, exserted well beyond the perianth and stamens.
Fruit c. 7 × 6 mm, white (occasionally red outside Brazil), sometimes marked with reddish bract-scales.
Chromosome number: 2n = 22 (Barthlott 1976).
Disjunct humid forest element: epiphyte in Mata atlântica, including mata de brejo and agreste (NE Brazil)
and mata do planalto, c. 500–1400 m, north-eastern Pernambuco, eastern Bahia and central-southern to
southern Minas Gerais; South-eastern and Southern Brazil; Bolivia (La Paz), Argentina, Paraguay and
Uruguay. Map 25.
pernambuco: Mun. Taquaritinga do Norte, Sítio Cafundó, 28 Dec. 1972, Andrade-Lima 72-7127
(IPA); l.c., road to Vertentes, 6 Mar. 1966, Andrade-Lima 66-4483 (IPA); Mun. Brejo da Madre de Deus,
Fazenda Bituri, 25 May 1995, D.C. Silva et al. 60 (K, UFP).
bahia: Mun. Almadina, Faz. Beija Flor, Serra da Pancadinha, entrada a 1 km da cidade na estrada para
Floresta Azul, 11 Aug. 1972, R.S. Pinheiro 1910 (CEPEC); Mun. Vitória da Conquista, 12 km from town
towards Itambé, 16 Aug. 2002, M. Machado (K, photos).
minas gerais: Mun. Lagoa Santa, reported by Warming (1908: 146; cf. Warming & Ferri 1973) as R.
saglionis; Mun. Catas Altas, Reserva Particular de Patrimônio Natural (RPPN) do Caraça, Capela do
Sagrado Coração (“Capelinha”), 1997, Padre C.M. Dell’Amore, cult. Santuário do Caraça, Aug. 2001,
Taylor (K, photos); Mun. São João del Rei, Sep. 1893, A. Silveira in Herb. Com. Geogr. Geol. Minas
1593 (R), Mar. 1994, Rui Alves, cult. Jard. Bot. Rio de Janeiro, Feb. 1995, Taylor (obs.).
Concern in terms of its overall range, but frequency assumed to have been much reduced in North-
eastern Brazil due to destruction of its habitat. Apparently rare, but nevertheless protected, in the RPPN
do Caraça, Catas Altas, Minas Gerais.
This is another good example of a Rhipsalideae with a markedly disjunct distribution in

the brejos/agrestes (mata de cipó) of North-eastern Brazil. The sometimes irregularly swollen
ultimate stem-segments may function as a water store, permitting the development of
flowers at their apices during the close of the dry winter season, when they become
visibly shrunken through water loss.
18. Rhipsalis pilocarpa Loefgr. in Monatsschr. Kakt.-Kunde 13: 52 (Apr. 1903) and
in Revista Centro Sci. Campinas 1903(4): 188 (July 1903); N. P. Taylor in Bot. Mag. 14:
125–129, tab. 320 (1997). Type: Brazil, São Paulo, Mun. Itu, forests of Ipanema and Itu,
cult. Hort. Bot. São Paulo, Feb. 1903, Löfgren (not found at SP, RB or R). Lectotype
(Barthlott & Taylor 1995: 69): Löfgren, l.c.: 55, illus. (Apr. 1903).
Erythrorhipsalis pilocarpa (Loefgr.) A. Berger in Monatsschr. Kakt.-Kunde 30: 4 (1920).
RHIPSALIS
Like the preceding in overall habit, but all stem-segments cylindric, of ± uniform thickness and
conspicuously white or yellowish bristly, seldom emitting aerial roots, often strongly tinged with dark red,
otherwise dark green except the actively growing pale green apices, the lower parts of basal extension
shoots c. 5 mm or more diam., the penultimate and ultimate flower-bearing segments c. 35 × 3.5–4.0
mm, lateral areoles bearing c. 8–15 strongly adpressed bristle-spines to 6 mm, these forming an erect
cephalium-like tuft at the terminal collective areole and becoming displaced and patent at the joints of
the earlier-formed segments (forming and performing like the terminal and ring cephalia of Arrojadoa
rhodantha); flowers in early winter, mostly 1–4 together at or around the terminal composite areole,
fragrant (moth-pollination syndrome?), buds densely bristly, the perianth strongly tinged pinkish red at
first, flowers rather variable in size and form, c. 10–30 mm diam., pendent, all perianth-segments patent
to reflexed, silvery cream, tinged pink to purple at apex, c. 18 or more in total, to 9 × 2.5 mm, linear-
tapered, minutely apiculate, pericarpel ± obconic, clothed in areoles and bristles identical to the stem-
segments, green with minute reddish scales, stamens numerous, to 7 mm or more, whitish, reddish at
base, expanded as in a loose brush and very conspicuous when the perianth-segments are strongly reflexed
(cf. Subg. Rhipsalis & Subg. Phyllarthrorhipsalis etc.), the innermost shorter, tightly clustered around the
style and effectively protecting the nectar-chamber around its base, style and 6–8 spreading, whitish 2 mm
stigma-lobes exceeding the brush of stamens by 1–2 mm; fruit spherical, to 8–12 mm diam., pericarp
bright red to crimson, bearing numerous areoles and whitish bristles. Seed 1.6–1.7 mm (Barthlott & Hunt
2000: pl. 4.5–6). Chromosome number: 2n = 22 (Barthlott 1976).
Southern humid forest element: epiphyte (rarely lithophyte) in Mata atlantica, 500–900 m, southern Minas
Gerais (Rio Preto) and southern Espírito Santo (Domingos Martins); South-eastern and Southern Brazil
(to Paraná). Map 25.
minas gerais: Mun. Rio Preto, F.S. Pires & M. Brugger 21615 (CESJ, SPF, HRCB); Mun. Bocaina de
Minas, 1994, J.C.E. Correia, cult. Jard. Bot. Rio de Janeiro, Feb. 1995, Taylor (obs.).
espírito santo: Mun. Domingos Martins, May 1986, Rauh & Kautsky 67561, 67564, cult. Univ. Bonn,
Germany, Accn. Nos 04512, 04508; l.c., July 1989, Rauh 68383 (BONN, photo).
conservation status. Vulnerable (2) [VU B2ab(i, ii, iii, iv)]; area of occupancy estimated to be <
2000 km2; PD=2, EI=1, GD=2. Short-list score (2×5) = 10. Known from only 8 localities throughout
its entire range and apparently rare (Taylor, l.c.).
This, the rare and geographically more restricted and variable sister species of R. cereuscula,
appears to have evolved the stamen-brush floral syndrome convergently with Rhipsalis
subgenera Rhipsalis, Epallagogonium, Trigonorhipsalis & Phyllarthrorhipsalis, and contrasts
strongly with other members of Subg. Erythrorhipsalis in lacking truly campanulate
flowers. As in its sister species, the bristly stem-segments may assist in the collection of
moisture from mists and night-time dews.
Subg. Calamorhipsalis K. Schum. (No. 19): seedlings 3–4-ribbed/angled; adult branching

sub-acrotonic or acrotonic; flower-buds and new stem-segments conspicuously
erumpent; trichome-bearing, composite terminal and normal lateral areoles apparently
lacking or hidden at first, visible only after flowering, scale-leaves minute, not fleshy;
stem-segments perfectly terete, of indeterminate growth in the species treated below;
flower-buds strongly erumpent, leaving a prominent scar on the stem, lateral to
subterminal, solitary, areoles flowering only once; fruit red, magenta or orange. Lectotype
(Backeberg 1942): R. neves-armondii K. Schum.
Only the following species is currently accepted for Eastern Brazil. However, a recent
collection by Jomar Jardim (Feb. 2003, photo), from near Prado, Bahia, strongly suggests
that R. neves-armondii is native here.
RHIPSALIS
19. Rhipsalis hoelleri Barthlott & N. P. Taylor in Bradleya 13: 50, pl. 9 (1995).
Holotype: Brazil, Espírito Santo, (?) Mun. Domingos Martins, 1987, Orssich s.n., cult.
Univ. Bonn, Germany, Accn. No. 04841 (BONN).
Pendent, to 150 cm or more; stems perfectly terete, limp and growing vertically downwards, only 3–4 mm
diam., green with faint purplish marks at the minute and very inconspicuous scales, growth indeterminate.
Flowers c. 10 mm diam., not opening widely; perianth-segments, style and stigma-lobes intense carmine
red, stamens colourless, at least distally. Fruit subglobose, 8 mm diam., not translucent, dull olive-red when
immature, ripening to intense tomato red.
Southern humid forest element: habitat details unknown; awaiting rediscovery in the region indicated
below. Map 25.
espírito santo: Mun. (?) Domingos Martins, 1987, Orssich s.n., cult. Univ. Bonn, Germany (BONN,
holo.).
conservation status. Data Deficient [DD], but to date collected only once.
This recently described species is closely related to R. puniceodiscus (Rio de Janeiro to

Santa Catarina), which it strongly resembles in vegetative characters. Its unusual red
flowers are presumed to be an adaptation for pollination by hummingbirds.
13. HATIORA Britton & Rose
in Bailey, Stand. cycl. hort. 3: 1433 (1915). Type: Hatiora salicornioides (Haw.) Britton &
Rose.
Literature: Barthlott & Taylor (1995, 1996).
Like Rhipsalis, but habit ± erect, all stem-segments of strictly determinate growth and
flowers always terminal from composite areoles, highly coloured.
A genus of 6 species endemic to the Mata atlântica of Brazil, between Bahia and Rio
Grande do Sul. Two species from Subg. Hatiora are represented here, although H.
epiphylloides (Porto & Werderm.) Buxb. (Subg. Rhipsalidopsis) may occur in Minas Gerais
(Bocaina de Minas), just outside the southern limits of the core area of Eastern Brazil. The
genus is weakly distinguished from Rhipsalis, especially in view of the floral similarities
presented by H. cylindrica. Plate 23.
1. Stem-segments globose, ovoid, cylindric or inverted bottle-shaped; inner perianth-

segments erect; fruit whitish, segments erect; often with a reddish ring at apex . . 1. salicornioides
1. Stem-segments uniformly cylindric, or somewhat thicker at base; all perianth-segments
spreading; fruit entirely red or purplish . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2. cylindrica
1. Hatiora salicornioides (Haw.) Britton & Rose in Bailey, Stand. cycl. hort. 3: 1433
(1915). Type: ‘Ind. Occident.’ (assumed not to have been preserved). Neotype (Barthlott
& Taylor 1995: 73, pl. 27): ‘Brazil [Rio de Janeiro], Messrs Bowie & Cunningham,
Duncanson del.’, water colour illustration (K).
PLATE 14
1 2
3 4
Plate 14. 14.1 Lepismium cruciforme, in fruit. Cult. Univ. Bonn (WB). 14.2 Rhipsalis russellii. Bahia, SE of
Vitória da Conquista, 2002 (MM). 14.3 Ibid., in fruit. Cult. Univ. Bonn (WB). 14.4 Rhipsalis elliptica. Minas
Gerais, RPPN do Caraça, 2001 (NT).
PLATE 15
1 2
3 4
Plate 15. 15.1 Rhipsalis elliptica, in fruit. São Paulo, Zappi 246, cult. Kew (K). 15.2 R. oblonga. São Paulo,
Ilhabela, Taylor & Zappi 1645A, cult. Kew (K). 15.3 R. crispata, in young fruit. Bahia, Serra da Jibóia, 2001 (DZ).
15.4 R. floccosa subsp. floccosa. Bahia, Muritiba, 2002 (NT).
PLATE 16
1 2
3 4
Plate 16. 16.1 Rhipsalis floccosa subsp. oreophila. Bahia, Harley et al. 25566 (NT). 16.2 R. floccosa subsp.
pulvinigera. Minas Gerais, Zappi et al. 241, cult. Kew (K). 16.3 R. paradoxa subsp. septentrionalis (Pernambuco,
Braun s.n.), with subsp. paradoxa (right). Cult. Univ. Bonn (WB). 16.4 Ibid (subsp. paradoxa, left). Ibid. (WB).
PLATE 17
1 2
3 4
Plate 17. 17.1 Rhipsalis pacheco-leonis subsp. catenulata. Type collection, cult. Kew (K). 17.2 Ibid. Ibid. (K). 17.3
Ibid., in fruit. Ibid. (K). 17.4 R. cereoides. Cult. Univ. Bonn (WB).
PLATE 18
1 2
3 4
Plate 18. 18.1 Rhipsalis cereoides, in fruit. Rio de Janeiro, Itacoatiara, 1996 (AJ). 18.2 R. sulcata. Cult. Univ.
Bonn (WB). 18.3 Ibid. Ibid. (WB). 18.4 R. lindbergiana. Minas Gerais, Taylor & Zappi 792 (NT).
PLATE 19
1 2
4 5
Plate 19. 19.1 Rhipsalis lindbergiana. Bahia, Harley & Storr 69, cult. Kew (K). 19.2 R. teres f. capilliformis. Cult.
Univ. Bonn (WB). 19.3 R. baccifera subsp. baccifera. Pernambuco, Recife, 1999 (DZ). 19.4 R. baccifera subsp.
hileiabaiana. Bahia, Serra da Jibóia, 2001 (DZ). 19.5 R. pulchra. S Minas Gerais, Zappi 260, cult. T. Hewitt (K).
PLATE 20
1 2
3 5
Plate 20. 20.1 Rhipsalis pulchra. S Minas Gerais, Zappi 260, cult. T. Hewitt (K). 20.2 R. burchellii. Cult. Univ.
Bonn (WB). 20.3 R. juengeri. Type collection, cult. Univ. Bonn (WB). 20.4 Ibid., in fruit. Ibid. (WB). 20.5 R.
cereuscula. SE Bahia, S of Vitória da Conquista, 2002 (MM).
PLATE 21
1 2
3 4
Plate 21. 21.1 Rhipsalis clavata. Cult. Univ. Bonn (WB). 21.2 Ibid., in fruit. Ibid. (WB). 21.3 R. pilocarpa.
Cult. Univ. Bonn (WB). 21.4 Ibid. Espírito Santo, Domingos Martins, Rauh 68383 (WR).
PLATE 22
Plate 22. Rhipsalis hoelleri. Type collection, cult. Univ. Bonn (WB).
PLATE 23
1 2
3 4
Plate 23. 23.1 Hatiora salicornioides. Cult. Univ. Bonn (WB). 23.2 H. cylindrica. Bahia, Itapoã, cult. R. Augusto,
2002 (MM). 23.3 Schlumbergera kautskyi. Type collection, cult. Univ. Bonn (WB). 23.4 S. opuntioides. Cult.
Univ. Bonn (WB).
PLATE 24
1 2
3 4
Plate 24. 24.1 Schlumbergera microsphaerica, in fruit. Minas Gerais, Pico da Bandeira, Orssich (WB). 24.2 Ibid.,
flower. Cult. Ralf Bauer (RB). 24.3 Brasilicereus phaeacanthus. Bahia, Harley et al. 25541 (NT). 24.4 Ibid., in bud.
Minas Gerais, Taylor & Zappi 762 (NT).
PLATE 25
1 2
3 4
Plate 25. 25.1 Brasilicereus phaeacanthus, in fruit. Bahia, Jequié, 2002 (MM). 25.2 B. markgrafii. Minas Gerais,
Serra da Bocaina, 2002 (GC). 25.3 Ibid. Minas Gerais, Grão Mogol, Harley et al. 25069 (NT). 25.4 Ibid. Loc.
cit., Preston-Mafham 111, cult. A. Hofacker (AH).
PLATE 26
1 2
Plate 26. 26.1 Cereus mirabella. Minas Gerais, Mirabela, Horst & Uebelmann HU 540, cult. A. Hofacker (AH).
26.2 C. albicaulis, in fruit. Bahia, Taylor et al. 1392 (NT). 26.3 Ibid. Pernambuco, Taylor & Zappi 1622 (NT).
PLATE 27
1 2
3 4
Plate 27. 27.1 Cereus fernambucensis subsp. fernambucensis, in fruit. Bahia, Harley et al. 27874 (NT). 27.2 Ibid.
Neotype collection (NT). 27.3 C. fernambucensis subsp. sericifer. Espírito Santo, 1990 (NT). 27.4 C. insularis.
Cult. A. Hofacker (AH).
PLATE 28
1 2
3 4
Plate 28. 28.1 Cereus jamacaru subsp. jamacaru, with Marlon Machado. Bahia, APA Gruta dos Brejões, 2002
(NT). 28.2 C. jamacaru subsp. calcirupicola, juvenile spination. Minas Gerais, Engenheiro Dolabela, 2002 (GC).
28.3 C. sp. nov. (5c), seedling. SE Bahia, S of Vitória da Conquista, cult. M. Moreira, 2002 (MM). 28.4 Ibid.,
immature plant, c. 7 metres tall. Loc. cit., 2003 (NT).
PLATE 29
1 2
3 4
Plate 29. 29.1 Cereus sp. nov. (5c). SE Bahia, S of Vitória da Conquista, 2003 (NT). 29.2 C. hildmannianus,
with ripe fruit. Cult. Kew (K). 29.3 Cipocereus laniflorus. Type collection, cult. Kew (K). 29.4 Ibid., in fruit.
Minas Gerais, RPPN do Caraça, 2001 (NT).
PLATE 30
1 2
3 4
Plate 30. 30.1 Cipocereus crassisepalus. Minas Gerais, Serra do Ambrósio, Zappi & Prado in CFCR 11822 (DZ).
30.2 C. pusilliflorus, in fruit. Minas Gerais, Horst & Uebelmann HU 400, cult. Werner van Heek (WvH). 30.3 C.
bradei. Minas Gerais, Eggli 1117 (UE). 30.4 Ibid., in fruit. Minas Gerais, Harley et al. 24508 (NT).
PLATE 31
1 2
3 4
Plate 31. 31.1 Cipocereus minensis subsp. leiocarpus, in bud. Minas Gerais, SW of Datas (GC). 31.2 Ibid. Minas
Gerais, Harley et al. 25136 (NT). 31.3 Ibid., with young fruit. Ibid. (NT). 31.4 C. minensis subsp. minensis. Minas
Gerais, Serra do Cipó (WR).
PLATE 32
Plate 32. Stephanocereus leucostele. Bahia, APA Gruta dos Brejões, 2002 (NT).
PLATE 33
2 3
Plate 33. 33.1 Stephanocereus leucostele. Bahia, Taylor et al. 1521 (NT). 33.2 S. luetzelburgii. Bahia, Harley et al.
25551 (NT). 33.3 Ibid., cephalium apex. Bahia, Harley et al. 25559 (NT).
PLATE 34
1 2
3 3
Plate 34. 34.1 Arrojadoa bahiensis. Bahia, Taylor et al. 1557A (NT). 34.2 Ibid. Ibid., cult. Kew (K). 34.3 Ibid.
Ibid. (K).
PLATE 35
1 2
3 4
Plate 35. 35.1 Arrojadoa dinae subsp. dinae. Minas Gerais, N. Roque et al. in CFCR 15407 (RH). 35.2 A. dinae
subsp. eriocaulis. Minas Gerais, Harley et al. 25518 (NT). 35.3 Ibid. Bahia, cult. G. Charles (GC). 35.4 A.
albiflora (A. dinae A. rhodantha). Ibid. (GC).
PLATE 36
1 2
3 4
Plate 36. 36.1 Arrojadoa penicillata. Bahia, Zappi 124 (K). 36.2 Ibid., in fruit. Bahia, Taylor et al. 1542 (NT).
36.3 A. rhodantha, plant 3.5 metres tall! Bahia, Taylor et al. 1370 (NT). 36.4 Ibid., in bud. Minas Gerais, Taylor
et al. 1459 (NT).
PLATE 37
1 2
3 4
Plate 37. 37.1 Arrojadoa rhodantha. Pernambuco, Zappi 226 (NT). 37.2 Ibid., in fruit. Bahia, Morro do
Chapéu, June 2002 (MM). 37.3 A. sp. nov. (4b). Bahia, nr Sussuarana, 2003 (NT). 37.4 Ibid., in flower. Loc.
cit., cult. A. Oliveira Soares Filho, 2003 (NT).
PLATE 38
2 3
4 5
Plate 38. 38.1 Pilosocereus tuberculatus. Pernambuco, Mun. Floresta, 2000 (NT). 38.2 P. gounellei subsp. gounellei.
Bahia, Araci, 1991 (NT). 38.3 P. tuberculatus. Pernambuco, Taylor & Zappi 1623 (NT). 38.4 P. gounellei subsp.
gounellei. Bahia, Harley et al. 25556 (NT). 38.5 Ibid., in fruit. N Paraíba, Tacima, 2000 (NT).
PLATE 39
1 2
3 4
Plate 39. 39.1 Pilosocereus gounellei subsp. zehntneri. N Bahia, APA Gruta dos Brejões, 2002 (NT). 39.2 Ibid.
SW Bahia, Taylor et al. 1425 (NT). 39.3 P. catingicola subsp. catingicola. Bahia, Zappi 117 (NT). 39.4 P. catingicola
subsp. salvadorensis, with DZ. NW Sergipe, Taylor & Zappi 1618A (NT).
PLATE 40
1 2
3 4
Plate 40. 40.1 Pilosocereus azulensis, young plant. Type locality, 2002 (MM). 40.2 P. floccosus subsp. quadricostatus
P. multicostatus. Minas Gerais, Pedra Azul, 2002 (MM). 40.3 P. arrabidae. Espírito Santo, O.J. Pereira et al. 2119
(NT). 40.4 Ibid., in fruit. Rio de Janeiro, Arraial do Cabo, Zappi 476 (DZ).
PLATE 41
1 2
Plate 41. 41.1 Pilosocereus brasiliensis subsp. brasiliensis. Espírito Santo, O.J. Pereira et al. 2120, cult. UNESP Rio
Claro (NT). 41.2 P. brasiliensis subsp. ruschianus. E Minas Gerais, Taylor & Zappi 770 (NT). 41.3 P.
flavipulvinatus. Maranhão, Zappi 213 (NT).
PLATE 42
1 2 3
4 5
Plate 42. 42.1 Pilosocereus pentaedrophorus subsp. pentaedrophorus. Bahia, Zappi 131A (NT). 42.2 Ibid., in bud
and fruit. Bahia, Serra de São José, 2001 (DZ). 42.3 P. pentaedrophorus subsp. robustus. Bahia, Taylor et al. 1515C
(NT). 42.4 P. glaucochrous. Bahia, Morro do Chapéu, 2002 (NT). 42.5 Ibid., in bud. Ibid. (NT).
PLATE 43
1 2
3 4
Plate 43. 43.1 Pilosocereus floccosus subsp. floccosus. Minas Gerais, near Rodeador, 2002 (GC). 43.2 P. floccosus cf.
subsp. quadricostatus. NE Minas Gerais, Harley et al. 25146 (NT). 43.3 P. floccosus subsp. quadricostatus. NE Minas
Gerais, Taylor & Zappi 765 (NT). 43.4 P. subsimilis. NE Minas Gerais, Harley et al. 25535 (NT).
PLATE 44
1 2
3 4
Plate 44. 44.1 Pilosocereus fulvilanatus subsp. fulvilanatus. Minas Gerais, Grão Mogol, 2002 (GC). 44.2 P.
fulvilanatus subsp. rosae. Type locality (KH). 44.3 Ibid. Loc. cit., cult. A. Hofacker (AH). 44.4 P. pachycladus
subsp. pachycladus. N Minas Gerais, Harley et al. 25520 (NT).
PLATE 45
1 2
3 4
Plate 45. 45.1 Pilosocereus pachycladus subsp. pachycladus. Bahia, W of Morro do Chapéu, 2002 (NT). 45.2 P.
pachycladus subsp. pernambucoensis. Piauí, Mun. Padre Marcos, 2000 (NT). 45.3 Ibid. (NE form). Pernambuco,
Belém de São Francisco, 2000 (NT). 45.4 P. magnificus. NE Minas Gerais, Taylor & Zappi 755 (NT).
HATIORA
Rhipsalis salicornioides Haw., Suppl. pl. succ.: 83 (1819) (‘salicornoides’). Hariota salicornioides (Haw.) DC.,
Mém. Cact.: 23 (1834).
H. villigera K. Schum. in Martius, Fl. bras. 4(2): 265 (1890). Rhipsalis salicornioides var. villigera (K. Schum.)
Loefgr. in Arch. Jard. Bot. Rio de Janeiro 1: 85 (1915). Hatiora salicornioides f. villigera (K. Schum.)
Supplie, Rhipsalidinae: [66] (1990). Type: Brazil, Sello 5239 (B†; MO, No. ‘523’).
Epiphyte or lithophyte, rarely terrestrial on sand, to 2 m, ± erect. Stems articulate, segments globose,
ovoid, cylindric or inverse bottle-shaped and then very narrow at base, often with an irregular outline and
presenting bristly spines at apex, 10–25 × 0.8–3.0(–5.0) mm, basal segments thick, with brownish bark,
apical segments 1–3.5 cm, epidermis bright green, with red shades or spots and minute areoles at the sides,
sometimes bearing up to 6 bristly spines, terminated by a large composite areole with felt and minute
bristles. Flowers terminal from the composite areole, solitary or 2–3 together, orange or bright yellow, c.
15 × 10–15 mm; pericarpel pale greenish, obconic, smooth, c. 3.5 × 4 mm; tube inconspicuous, almost 0;
inner perianth-segments erect, to 11 × 3 mm, others sometimes more expanded, ovate, broader, to 4 mm
wide; stamens all at the same level, anthers rounded; style 5 mm, stigma-lobes 4–7, ± exserted; ovary locule
obtriangular in longitudinal section. Fruit globose to turbinate, c. 10 × 5–8 mm, translucent, whitish,
funicular pulp transparent, very mucilaginous. Seeds ovoid, 1.2 × 0.9 mm, brown-black, shiny; testa-cells
flat, smooth (Barthlott & Hunt 2000: pl. 3.5–6). Chromosome number: 2n = 22 (Barthlott 1976).
Disjunct humid forest element: usually epiphytic or lithophytic, Mata atlântica, mata de grotão and mata de
neblina, c. 800–1750 m, Bahia (E Chapada Diamantina), central-southern Minas Gerais (S Serra do
Espinhaço and Serra da Mantiqueira), Espírito Santo and northern Rio de Janeiro, but northern records
markedly disjunct; South-eastern and Southern Brazil (to Paraná). Map 26.
bahia: cent. Bahia, Mun. Lencois, Serra da Chapadinha, 8 July 1996, H.P. Bautista et al. in PCD 3486
(ALCB, K), 27 Feb. 1997, Gasson et al. in PCD 5882 (ALCB, K); Mun. (?) Barra da Estiva, Serra do
Sincorá [above Sincorá Velho?], reported by Ule (1908).
minas gerais: Mun. Santana do Riacho, Serra do Cipó, 20 Oct. 1973, Joly et al. in CFSC 4606 (SP);
l.c., Alto do Palácio, 28 Dec. 1948, Balegno & Cuezzo 3992 (R); Mun. Jaboticatubas, Serra das
Bandeirinhas, 9 Sep. 1987, Zappi et al. in CFCR 10849 (SPF); Mun. Caeté, Serra da Piedade, Sep. 1915,
F.C. Hoehne 6349 (SP), 6348 (R), 16 Jan. 1971, Irwin et al. 30460 (MO), 24 Feb. 1987, Zappi in CFCR
10343 (SPF), s.d., V.C. Souza & C.M. Sakuragui 2047 (ESA, K); Mun. Santa Bárbara/Catas Altas, Serra
do Caraça, 1936, Brade s.n. (RB); l.c., 11 Sep. 1990, Zappi 239 (HRCB, BHCB); l.c., Mun. Catas Altas,
between Bocaina and Gruta da Bocaina, 1 Aug. 2001, Taylor (K, photo); Mun. Itabirito, c. 50 km SE of
Belo Horizonte, Pico do Itabirito, 17 Feb. 1968, Irwin et al. 19845 (MO); Mun. Ouro Preto, Mar. 1892,
Ule s.n. (R 91021); Mun. Caparaó, Parque Nacional da Serra do Caparaó, 30 Apr. 1988, Krieger et al. in
FPNC 92 (CESJ, SPF), 17 Dec. 1988, Krieger et al. in FPNC 625 (CESJ), 19 Nov. 1988, Krieger et al. in
FPNC 848 (CESJ), 24 Oct. 1989, Mello-Silva et al. 103 (SPF); Mun. Viçosa, Fazenda de Aguada, 27 Nov.
1930, Y. Mexia 5373A (US, F, MO); Mun. São João del Rei/Tiradentes, June 1896, A. Silveira 1279 (R);
l.c., Serra de São José, 1300–1400 m, 3 Oct. 1987, M. Perón 291 & 315 (RB); Mun. Lima Duarte, Serra
da Capoeira Grande, 10 July 1987, F.R.S. Pires et al. 21541 (SPF); l.c., Parque Estadual do Ibitipoca, 16
Dec. 1986, H.C. Sousa et al. 9829 (BHCB); Mun. São Tomé das Letras, 27 Sep. 1968, M.C. Vianna 357
(GUA); Mun. Lima Duarte/Rio Preto, Chapadão da Serra Negra, 3 Oct. 1959, Castellanos 22537 (R).
espírito santo: Mun. Santa Teresa, Reserva Biológica de Nova Lombardia, 14 July 1976, A.G. Pedrine
s.n. (MBML); l.c., morro da estação de TV, 14 Nov. 1985, W. Pizziolo 200 (MBML); l.c., Fazenda Santa
Lúcia, 19°57'24"S, 40°32'37''W, 25 Nov. 1999, Zappi et al. 440 (K, UEC).
rio de janeiro: Mun. Santa Maria Madalena, ‘subida da Pedra Dubois’, 21°56'36–42"S,
41°59'29–33''W, 22 Nov. 1999, Zappi et al. 368 (K, UEC).
This species presents rather diverse morphology, partly related to the conditions under
which it grows (Zappi 1991). Its flowers are also unusually variable, especially in the degree
to which the outer perianth-segments expand, some forms scarcely opening except to reveal
the anthers and stigma-lobes. Most of these variations appear to be determined genetically.
HATIORA
The above synonymy is almost certainly incomplete, since many of the names of
potential variants are inadequately typified and lack details as to the key floral characters
that distinguish the two species treated here.
2. Hatiora cylindrica Britton & Rose, Cact. 4: 219 (1923); Backeberg, Die Cact. 2:
709, Abb. 652 (1959). Holotype: Brazil, Rio de Janeiro, Ilha Grande, 22–24 July 1915,
Rose & Löfgren (US).
Rhipsalis cylindrica (Britton & Rose) Vaupel, Die Kakteen 1: 39 (1925), nom. illeg., non (Vell.) Steud.
(1841) nec Kuntze (1891). Hatiora salicornioides f. cylindrica (Britton & Rose) Supplie, Rhipsalidinae:
unpaged [p. 65] (1990). Rhipsalis salicornioides var. cylindrica (Britton & Rose) Kimnach in Cact. Succ.
J. (US) 68: 156 (1996).
Small shrub to c. 50 cm with erect, strictly acrotonic branching, up to 4 new stem-segments arising from
the apex of existing segments, cylindrical, 10–40 × 2–3 mm, dark green to reddish-purple when exposed
to intense sunlight, with a conspicuous terminal composite areole, bearing a mass of very short trichomes
and few acicular spines to c. 1 mm, and sometimes with a few lateral areoles, but these inconspicuous and
inactive. Flower-buds reddish yellow, flowers one per terminal areole, orange-yellow, fully expanded,
funnelform, c. 10 × 10 mm, perianth-segments lanceolate, acute, 6 × 2 mm, the outer series c. 7, reddish
orange, inner segments c. 8, orange-yellow; stamens c. 20, to 5 mm, yellowish, anthers small, yellow;
pericarpel 4 mm, red; style creamy white, 5 mm, stigma-lobes 4, slender, curled. Fruit dark red-magenta,
turbinate to globular, 6 mm in diam.
Humid forest element: epiphytic and terrestrial (in sand-dunes and probably as a lithophyte), restinga, mata
de brejo and Mata atlântica, from near sea level to c. 1200 m, eastern Bahia and (?) southern Espírito Santo;
South-eastern Brazil (but rarely collected). Map 26.
bahia: E Bahia, Mun. Salvador, Itapuã, in sand-dunes away from the beach, Mar. 1989, R. Augusto s.n.,
cult. ibid., Cruz das Almas, 7 Aug. 2002 (HUEFS); Mun. Venceslau Guimarães, c. 40 km W of
Teolândia, near Nova Esperança, 1 Apr. 1993, S.C. Sant’Ana & L.A.M. Silva 319 (CEPEC, K), l.c., 8
km above Rio Vermelho, 2 km above junction of road to Taquara, 13°36'S, 39°47'W, 15 May 1992,
W.W. Thomas et al. 9359 (CEPEC); Mun. Barra do Choça, Sep. 2003, M. Machado (photos).
espírito santo: (?) Mun. Castelo, Forno Grande, June 1949, Brade 19978 (RB) — identification
provisional (could be referable to the preceding species).
PD=2, EI=1, GD=1. Short-list score (2×4) = 8. Bahian and doubtless other localities are declining due
to growth in tourism and deforestation, but it may be included in protected areas.
The above description, kindly prepared by Marlon Machado, is based on living material
seen by NT from Itapuã, Bahia, where the plant was discovered by Sra Rita Augusto
(Cruz das Almas) growing directly in the sand of the coastal dunes. The fully expanded
flowers of this species readily distinguish it from other, more or less cylindrical forms
referable to H. salicornioides, which have erect inner perianth-segments adpressed to the
stamens. This may indicate divergence in their pollination syndromes.
The erect stem-segments of ± constant size and absence of indeterminate extension
shoots are the only significant characters that separate this taxon from superficially similar
Rhipsalis species.
SCHLUMBERGERA
Excluded taxon
A single collection of the magenta-flowered sister-species, Hatiora herminiae (Porto & A.
Cast.) Backeb. ex Barthlott, made in October 1942, from the Estação Experimental
Coronel Pacheco, Minas Gerais (Heringer 911, SP), is assumed to represent a plant that
was in cultivation.
14. SCHLUMBERGERA Lem.
in Rev. Hort., ser. 4, 7: 253 (1858). Type: S. epiphylloides Lem. (= S. russelliana (Hook.)
Britton & Rose).
Including Epiphyllanthus A. Berger (1905).
Literature: Hunt (1969), Barthlott & Rauh (1975), Barthlott & Taylor (1995: 74), Taylor
& Zappi (1995).
Epiphytic or epilithic; stems terete, few-ribbed, 2–3-winged or flattened, segmented, all stem-segments
of determinate growth, oblong to obovate, sometimes truncate. Areoles scattered over the surface of the
joints or confined to the ribs or to the crenate or serrate margins, crowded into composite areoles at the
apex of segments and there originating new joints and flowers (except in S. opuntioides). Spines bristly,
short or absent. Flowers scarcely to strongly zygomorphic, red, pink or purplish (rarely white); pericarpel
terete or ribbed, greenish to brownish, naked, flower-tube elongate, with bract-scales only at its very
base, straight to oblique at apex; perianth-segments spreading or recurved, in various series, giving the
impression of a flower inserted within a flower; stamens numerous, comprising an outer and a distinct
inner series united at their bases to form a short tube around the style and partially enclosing the nectar-
chamber; style and stamens exserted; stigma-lobes erect, connivent. Fruit globose to obconic, ribbed or
terete, the perianth deciduous. Seeds subreniform to ovate in outline, 1–1.7 mm, testa dark brown-black,
shiny, with intercellular depressions.
An endemic Brazilian genus of 6 species, ranging from southern Espírito Santo

(Domingos Martins) and adjacent Minas Gerais (Serra do Caparaó) to Rio de Janeiro,
southernmost Minas Gerais and south-eastern São Paulo, in the mountains of the Serra
do Mar and Serra da Mantiqueira (Mata atlântica, to 2700 metres altitude). Three species
are native to the area covered here, the first being endemic. Plates 23.3–24.2.
1. Terminal stem-segments globose, short-cylindric or linear, rounded in cross-section . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2. microsphaerica
1. Terminal stem-segments orbicular to obovate or truncate in outline, laterally compressed . . . . . . 2
2. Stem-segments almost unarmed, but margins toothed; flowers scarcely zygomorphic . . 1. kautskyi
2. Stem-segments covered in areoles with pungent, very thin spines; flowers strongly
zygomorphic . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3. opuntioides
1. Schlumbergera kautskyi (Horobin & McMillan) N. P. Taylor in Bradleya 9: 90

(1991); McMillan & Horobin in Succ. Pl. Res. 4: 23–26, pl. 3.1–3.3 (1995). Holotype:
Brazil, Espírito Santo, Mun. Domingos Martins, 900 m, May 1986, R. Kautsky in Rauh
67558, cult. Univ. Bonn, Germany (BONN).
SCHLUMBERGERA
S. truncata subsp. kautskyi Horobin & McMillan in Epiphytes 15(57): 8 (1991). S. truncata var. kautskyi
Horobin & McMillan in ibid. 14: 111–115 (1990), nom. inval. (Art. 37.1).
Epilithic or less often epiphytic, pendent; stem-segments 2.2–3.5(–4.0) × 1.4–1.8(–2.5) cm; compressed,
truncate, with 1–2 sharp teeth on each side, or at least at apex on either side of the composite areole;
epidermis dark green, often with reddish marks around the margins; primary stem of seedling flattened.
Areoles in the axils of the marginal teeth and at apex; spines bristly or absent. Flowers self-compatible,
scarcely zygomorphic, to c. 5 × 2.7 cm; pericarpel 12 × 5–6 mm, narrow, cylindric, 4-angled, green to
dark reddish; flower-tube pink, with magenta outer perianth-segments, inner segments spreading or
recurved at anthesis, lanceolate, magenta-pink. Fruits 25 × 19 mm, turbinate, 4-angled, pericarp yellow-
green when ripe, with a reddish tinge. Seed 1 mm.
Southern humid forest (inselberg) element: lithophytic on inselbergs, rarely epiphytic, Mata atlântica,
900–1300 m, central-southern Espírito Santo. Endemic. Map 26.
espírito santo: Mun. Domingos Martins, Pico da Pedra Azul, Kautsky s.n. (K, in spirit), May 1986,
R. Kautsky in Rauh 67558, cult. Univ. Bonn, Germany (BONN, holo.); Mun. Alfredo Chaves, São
Bento de Urânia, R. Kautsky, reported by Horobin & McMillan, l.c. (1990).
conservation status. Endangered [EN B2ab(iii)] (3); area of occupancy estimated to be < 500 km2;
PD=2, EI=1, GD=1. Short-list score (3×4) = 12. Area being developed as a residential centre.
Disjunct from its nearest relatives in the Serra dos Órgãos (RJ) by over 250 kilometres.
2. Schlumbergera microsphaerica (K. Schum.) Hövel in Kakt. and. Sukk. 21: 186
(1970); Heath in Calyx 2: 64 (1992); McMillan & Horobin in Succ. Pl. Res. 4: 27, pl.
6.1–6.4 (1995). Type: Brazil, Rio de Janeiro, Glaziou s.n. (B†). Neotype (Taylor 1991b;
Heath 1992b): Arch. Jard. Bot. Rio de Janeiro 2: tab. 5, fig. B (1918).
Cereus microsphaericus K. Schum. in Martius, Fl. bras. 4(2): 197 (1890). Epiphyllanthus microsphaericus (K.
Schum.) Britton & Rose, Cact. 4: 181 (1923). Arthrocereus microsphaericus (K. Schum.) A. Berger,
Kakteen: 337 (1929), tantum quoad typ. Cereus damazioi Weing. in Monatsschr. Kakt.-Kunde 21:
91–94 (1911), nom. illeg. (Art. 52.1). Trichocereus damazioi (Weing.) Werderm., Bras. Säulenkakt.: 94
(1933), nom. illeg. Type: as above.
Cereus obtusangulus K. Schum., l.c. 198 (1890). Epiphyllanthus obtusangulus (K. Schum.) A. Berger in Rep.
(Annual) Missouri Bot. Gard. 16: 84 (1905). Zygocactus obtusangulus (K. Schum.) Loefgr. in Arch. Jard.
Bot. Rio de Janeiro 2: 28 (1918). Epiphyllum obtusangulum (K. Schum.) Vaupel, Die Kakteen, Lfg 2:
88 (1926). Schlumbergera obtusangula (K. Schum.) D.R. Hunt in Kew Bull. 23: 260 (1969). S.
microsphaerica f. obtusangula (K. Schum.) P. V. Heath in Calyx 2: 64 (1992). Type: Brazil, Rio de
Janeiro, Glaziou s.n. (C, missing). Neotype (Taylor 1991b): Schumann, Gesamtb. Kakt.: fig. 30 (1897).
? Cereus parvulus K. Schum., l.c. 197 (1890). Schlumbergera microsphaerica f. parvula (K. Schum.) P. V. Heath
in Calyx 2: 64 (1992). Type: Brazil, Rio de Janeiro, Glaziou s.n. (B†).
Zygocactus candidus Loefgr. in Arch. Jard. Bot. Rio de Janeiro 2: 30 (1918). Epiphyllanthus candidus (Loefgr.)
Britton & Rose, Cact. 4: 182 (1923). Schlumbergera microsphaerica subsp. candida (Loefgr.) D.R. Hunt in
Succ. Pl. Res. 4: 79 (1995). Type: Rio de Janeiro, Itatiaia, Várzea de Ayuruóca (not extant).
Epiphytic or epilithic, ± erect, to 40 cm, densely branched; stem-segments 5–40 × 2–5 mm, globose,
short-cylindric, angled or not, or linear, sometimes the terminal segments compressed, epidermis bright
green, often with reddish markings or entirely reddened. Areoles scattered over the surface, or mainly at
apex of the terminal segments; spines bristly, to 5 mm, brown or yellowish. Flowers self-compatible,
zygomorphic, to 4–4.5 cm; pericarpel rounded, smooth or ridged; flower-tube pale pink, with pink
bract-scales; perianth-segments spreading at anthesis, lanceolate, pale or deep pink; pericarpel reddish
brown. Fruits 5–10 mm diam., turbinate to rounded, indistinctly 5-ribbed, green or white. Seed c. 1 mm.
SCHLUMBERGERA
Southern humid forest element: lithophytic or epiphytic, mata de neblina, at > 2000 m, Serra do Caparaó
(Pico de Bandeira), Minas Gerais/Espírito Santo; Rio de Janeiro (Serra de Itatiaia). Map 26.
minas gerais: Serra do Caparaó, 2200 m, Sep. 1941, Brade s.n. & 17024 (RB); Parque Nacional do
Caparaó, 2300–2700 m, 29 Sep. 1977, L. Krieger 15161 (CESJ, SPF), 4 Oct. 1966, Strang 715 (HB), 29 June
1955, N. Santos & Z. Campos s.n. (R 52203); ibid., Terreirão, Cachoeira, 30 Apr. 1989, L. Krieger et al. in
FPNC 956 (CESJ).
conservation status. Data Deficient [DD], but much of its range is within protected areas.
The name S. microsphaerica must be used in preference to the more familiar S. obtusangula,
since the former, as Epiphyllanthus microsphaericus, was the name accepted by Britton & Rose
(1923: 181), the first authors to treat these two equally priorable names as synonymous.
Scheinvar (1985) reports S. obtusangula from northern Santa Catarina in Southern
Brazil, but this is assumed to be an error and it has not been possible to locate the
collection she cited. It is probably no coincidence that the locality she cites is a site for
Hatiora rosea (Lagerh.) Barthlott, which sometimes develops cylindric, ribbed stem-segments
like those of S. microsphaerica. This may also explain why Braun & Esteves Pereira (2002:
79) indicate that Schlumbergera occurs in Santa Catarina.
The distribution of S. microsphaerica is markedly disjunct between the localities cited
above and its locus classicus on Itatiaia, by some 350 kilometres. Living plants from these
two areas should be compared in view of the distances involved, but photographs of
flowering plants in habitat at the eastern locality, shown to the authors by Inês Ribeiro
de Andrade have confirmed the identity accepted here (see also Plate 24.1).
3. Schlumbergera opuntioides (Loefgr. & Dusén) D. R. Hunt in Kew Bull. 23: 260
(1969); McMillan & Horobin in Succ. Pl. Res. 4: 26–27, pl. 3.4, 3.5, 5.1 & 13.6 (1995).
Holotype: Brazil, Rio de Janeiro, Itatiaia, 2400 m, 11 June 1902, Dusén 1530 (R).
Epiphyllum opuntioides Loefgr. & Dusén in Arq. Mus. Nac. Rio de Janeiro 13: 49 (1905). Zygocactus
opuntioides (Loefgr. & Dusén) Loefgr. in Arch. Jard. Bot. Rio de Janeiro 2: 26, tab. 4 (1918).
Epiphyllanthus opuntioides (Loefgr. & Dusén) Moran in Gentes Herb. 8: 338 (1953). E. obovatus
Britton & Rose, Cact. 4: 180 (1923), nom. illeg. (Art. 52.1).
Epiphyllum obovatum Engelm. ex K. Schum., Gesamtbeschr. Kakt.: 224 (1897), nom. inval. (provisional
name based on Sellow 884, B†).
Epiphyte or lithophyte, pendent, to 50 cm; stem-segments 15–70 × 5–30 mm, to 9 mm thick, orbicular
to obovate, compressed, epidermis dark green to purplish. Areoles woolly, scattered over the surface of
the segments, composite terminal areoles lacking; spines to 5 mm, pungent but delicate, golden, more
densely developed on the older segments where up to 80 per areole. Flowers self-incompatible, 1–3 per
stem-segment, strongly zygomorphic, to 6 × 4.5 cm; pericarpel obconic or 4-angled, smooth, red or
white; flower-tube c. 3–4 cm, pale pink, with magenta outer perianth-segments, inner segments
spreading at anthesis, those on the lower side of the flower reflexed, lanceolate, magenta-pink; stamens
red, anthers bearing red-brown pollen; stigma-lobes 4–7, white; pericarpel slightly 5–7-angled, red. Fruits
c. 10 mm diam., spherical to 4–5-angled, green. Seed 1.6–1.7 × 1.0 mm, glossy brown (Barthlott & Hunt
2000: pl. 2.5–6). Chromosome number: 2n = 22 (Barthlott 1976).
Southern humid forest element: lithophytic/epiphytic, capão de mato, c. 1700 m, Serra da Mantiqueira,
southern Minas Gerais; to north-western Rio de Janeiro (Itatiaia) and eastern São Paulo (Campos do
Jordão). Map 26.
minas gerais: Mun. Lima Duarte, Parque Estadual do Ibitipoca, 27 July 1991, Zappi 258 (SPF, CESJ).
SCHLUMBERGERA
conservation status. Near Threatened [NT] (1); area of occupancy estimated to be < 500 km2;
PD=2, EI=1, GD=1. Short-list score (1×4) = 4. Included within some protected areas and dependent on
their maintenance.
Tribe CEREEAE Salm-Dyck
Columnar, treelike to shrubby or semi-erect, to low-growing and ± globose; stems ribbed;

pericarpel (and tube) of flower and fruit with minute bract-scales or naked, glabrous and
lacking bristles/spines (except in 2 Cipocereus spp. and Cereus subg. Mirabella), but sometimes
immersed in a woolly/bristly, lateral or terminal cephalium. Type: Cereus Mill.
The most important tribe in Eastern Brazil. Its circumscription follows that
employed by Taylor & Zappi (1989) and Barthlott & Hunt (1993), but this is not
supported by recent DNA gene sequence studies (Soffiatti unpubl.) and will very likely
have to be revised. Analysis of the chloroplast segment rpl16 suggests that Brasilicereus,
Praecereus and Cereus are not closely related to the remainder of Brazilian taxa here
included in the tribe, of which the Trichocereeae could be the sister group. However,
an earlier survey of the family, utilizing surface waxes (n-alkanes), lends support to the
distinctiveness of tribe Cereeae as employed here (Maffei et al. 1997). The following
genus, displaying a conspicuously scaly pericarpel, may in any case be closer to
Browningia and its relatives (Browningieae):
15. BRASILICEREUS Backeb.
in Blätt. Kakteenf. 1938(6): [20] (1938). Cereus subg. Brasilicereus (Backeb.) P. J. Braun
(1988). Type: Brasilicereus phaeacanthus (Gürke) Backeb. (Cereus phaeacanthus Gürke).
Erect plants, usually branched above the ground, 0.5–5.0 m, roots fibrous. Stems erect or inclined and
partly supported by surrounding vegetation, with seasonal growth increments separated by slight
constrictions, giving a jointed appearance; ribs 7–14, low, sinuses straight; central cylinder very woody,
cortex ± lacking mucilage. Areoles felted, longer hairs absent, developing golden to brownish, adpressed
to spreading spines, central spine(s) porrect, much longer than radials. Flower-bearing region of stems not
differentiated, apparently quite random; flowers nocturnal, shortly tubular, to 7 cm; pericarpel and tube
ridged, usually covered in acute or broad, rounded bract-scales; pericarpel and ovary locule depressed;
stamens inserted in two series, the uppermost forming a ‘throat ring’ (as in Echinopsis, Trichocereeae),
filaments slender, mobile; pollen 16-colpate (in B. phaeacanthus, Leuenberger 1976). Fruit ovoid,
indehiscent, floral remnants persistent, erect, pale brown when not decayed (not strongly blackening as in
some other Cereeae), deeply sunken into fruit apex, pericarp somewhat ridged and bearing large,
rounded, greenish, red, brownish or purplish bract-scales; funicular pulp white. Seeds black,
cochleariform, applanate, shiny; testa-cells flat.
An endemic genus of 2 species, possibly related to Praecereus Buxb. (with 2 South

American species, P. euchlorus being the most widespread, ranging into Central-western
& South-eastern Brazil) and replacing it in Eastern Brazil; Map 7 (Taylor 1992a: 25). The
genus is restricted to the South-eastern campos rupestres and Southern caatingas-agrestes.
Plates 24.3–25.4.
BRASILICEREUS
1. Shrubby to tree-like, branched above ground; flowers with ovate, truncate bract-scales;
pericarpel c. 2–3× wider than long; stems 2–6 cm diam.; hilum-micropylar region forming an
angle of 20–30° with the long-axis of seed (caatinga/agreste etc., E & S Bahia & N Minas Gerais)
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1. phaeacanthus
1. Stem solitary or poorly branched at base; flowers with acute to acuminate bract-scales;
pericarpel < 2× wider than long; stems 1.5–2.0 cm diam.; hilum-micropylar region forming
an angle of 60° with the long-axis of seed (campo rupestre/carrasco, Mun. Grão Mogol etc.,
Minas Gerais) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2. markgrafii
1. Brasilicereus phaeacanthus (Gürke) Backeb. in Jahrb. Deutsch. Kakteen-Ges.

1941(2): 50 (1942). Type: Bahia, nr Maracás, Sep. 1906, Ule 7022 (HBG, lecto.
designated here [K, photo.]).
Cereus phaeacanthus Gürke in Monatsschr. Kakt.-Kunde 18: 57 (1908). Cephalocereus phaeacanthus (Gürke)
Britton & Rose, Cact. 2: 57 (1920). Pilocereus phaeacanthus (Gürke) Backeb. in Backeberg & F. M.
Knuth, Kaktus-ABC: 333 (1935, publ. 1936).
Brasilicereus breviflorus F. Ritter, Kakt. Südamer. 1: 228 (1979). Cereus phaeacanthus var. breviflorus (F. Ritter)
P. J. Braun in Bradleya 6: 87 (1988). Brasilicereus phaeacanthus subsp. breviflorus (F. Ritter) P. J. Braun
& E. Esteves Pereira in Succulenta 74: 83 (1995). Holotype: Brazil, Minas Gerais, Itaobim, 1965,
Ritter 1337 (U).
Shrubby to tree-like, mostly somewhat branched above ground level, to 5 m. Stems erect, inclined on
surrounding vegetation or sometimes semi-decumbent on rocks, arching over and rooting at apex, 2–6
cm diam.; ribs 7–14, to 5 × 6 mm; epidermis dull or dark to grey-green. Areoles 2–3 mm diam., 4–8 cm
apart on the ribs. Central spines 1–3, to 3 cm; radials 10–12, to 1.5 cm. Flowers 4–7 × 4–5.2 cm;
pericarpel very strongly depressed, 5–7 × 15–18 mm, pale green; flower-tube infundibuliform, c. 3.2 ×
1.7–2.0 cm, pericarpel and tube bearing broad, ovate, truncate, reddish to purplish bract-scales, these
rarely absent from the pericarpel; perianth-segments to 2.7 cm, outer segments spreading, lanceolate,
fleshy, dark red at tips, inner segments spreading to erect, spathulate, delicate, white to very pale green;
nectar-chamber 11 × 13 mm; style 40–48 × 2 mm, stigma-lobes c. 10, exserted in relation to the anthers;
ovary locule strongly depressed in longitudinal section. Fruit globose to ovoid, to 3.5 cm diam.; pericarp
greenish to dull red or slightly purplish. Seeds c. 2–3 × 1.4 mm, cochleariform, hilum-micropylar region
forming an angle of 20–30° with long-axis.
Southern caatinga element: in caatinga-agreste, often associated with granite/gneiss inselbergs, 40–920 m,
central-eastern to central-southern Bahia and central-northern and north-eastern Minas Gerais. Endemic
to the core area of Eastern Brazil. Map 27.
bahia: Mun. Macajuba, 26 km N of Rui Barbosa on road to Baixa Grande, 9 Feb. 1991, Taylor et al.
(obs.); Mun. Rui Barbosa, 1.5 km SW of town towards main road to Itaberaba, 7 Feb. 1991, Taylor et al.
1570 (HRCB, CEPEC, K, ZSS); Mun. Ibipitanga, 5–10 km N of Ibiajara towards Novo Horizonte,
12°56'S, 42°10'W, Aug. 2003, M. Machado (obs.); Mun. Cachoeira (?), Vale dos Rios Paraguaçu e Jacuípe,
Morro Belo, 12°32'S, 39°5'W, Dec. 1980, Iscardino et al. in Grupo Pedra do Cavalo 976 (CEPEC); Mun.
Elísio Medrado/Castro Alves, S of Monte Cruzeiro, 25 June 1915, Rose & Russell 20053 (US, NY); Mun.
Marcionílio Sousa, Machado Portella, 19–23 June 1915, Rose & Russell 19912 (US, NY); Mun. Iramaia,
40 km NW of ‘Pé-de-Serra’ on road from Maracás, 5 Feb. 1991, Taylor et al. (obs.); Mun. Santa Inês, nr
Caldeirão, 26 June 1915, Rose & Russell 20058 (US); Mun. Jaguaquara [Toca da Onça], 27–29 June 1915,
Rose & Russell 20095 (US, NY); Mun. Maracás, Sep. 1906, Ule 7022 (K, photo ex HBG); Mun.
Livramento do Brumado, 7 km from Rio de Contas, 13°38'S, 41°50'W, 23 Nov. 1988, Taylor & Zappi
in Harley 25541 (SPF, K); Mun. Jequié, reported as Horst & Uebelmann 746 by Hofacker & Braun (1998);
l.c., 29.5 km E of Porto Alegre towards Jequié, 4 Feb. 1991, Taylor et al. (obs.); Mun. Tanhaçu, 21 km
N of Sussuarana towards Contendas do Sincorá, 3 Feb. 1991, Taylor et al. 1547 (HRCB, CEPEC, K,
ZSS); município unknown, between Vitória da Conquista and Jequié, 22 Sep. 1946, Duarte 9310, Pereira
10023 & Barroso (RB, HB); Mun. Caitité, c. 9 km ESE from town towards Brumado, then c. 2 km S,
BRASILICEREUS
27 Aug. 1988, Eggli (ZSS, photo); Mun. Aracatu, c. 20 km N of Anajé towards Sussuarana, 17 Aug. 1988,
Eggli (ZSS, photo); Mun. Urandi, Jan. 1964, Ritter 1239 (SGO 124989, fide Eggli et al. 1995: 501–502);
Mun. Vitória da Conquista, c. 32 km SE of town towards Itambé, 16 Aug. 1988, Eggli 1176 (ZSS).
minas gerais: cent.-N Minas Gerais, Mun. Espinosa, 4 km S of BA/MG border, 25 July 1989, Taylor
& Zappi (obs.); Mun. Monte Azul, 7 km E of town, beyond Vila Angical, Serra Geral, 920 m, 28 Jan.
1991, Taylor et al. 1468 (HRCB, BHCB, K, ZSS); NE Minas Gerais, Mun. Taiobeiras / Águas
Vermelhas, c. 5 km NW of Curral de Dentro towards junction with road BR 116, 13 Aug. 1988, Eggli
1162 (ZSS); Mun. Pedra Azul, 21 Sep. 1965, Duarte 9308 (HB, RB), Duarte 9289 & Pereira 10002 (RB);
l.c., 8 km W of town, 15°57'S, 41°22'W, 18 Oct. 1988, Taylor & Zappi in Harley 25184 (SPF, K); Mun.
Itinga, 5 km E of town, road BR 367, 16°36'S, 41°43'W, 19 Nov. 1988, Taylor & Zappi in Harley 25534
(SPF, K); Mun. Itaobim, 8 km W of town on road BR 367, 14 Dec. 1990, Taylor & Zappi 762 (HRCB,
BHCB, K, ZSS); Mun. Padre Paraíso, c. 10 km N of town, 13 Dec. 1990, Taylor & Zappi (obs.).
This species is rather variable in stem thickness and rib-number, but this variation seems
to lack any kind of geographical pattern. A form with unusually short flowers has been
described from near the Rio Jequitinhonha, north-eastern Minas Gerais (B. breviflorus F.
Ritter), but it is otherwise unremarkable and does not merit recognition at any rank when
the overall variation of the species is taken into consideration. A rather different impression
is given in a recent article by Hofacker & Braun (1998), in which they distinguish two
subspecies in a key and illustrate each by a single collection. Their key implies that these
two entities are geographically separated, the heterotypic subsp. breviflorus (F. Ritter) P. J.
Braun & E. Esteves Pereira representing the species in Minas Gerais, while the homotypic
subspecies is restricted to southern Bahia. However, this is not so, since the short-flowered
population named by Ritter is only one amongst a number of variants found in Minas
Gerais, none of which differs significantly from contiguous Bahian populations. The
Bahian form they illustrate from Jequié under the number ‘HU 746’ is very far from typical
of the species, having a rather peculiar naked pericarpel.
2. Brasilicereus markgrafii Backeb. & Voll in Arch. Jard. Bot. Rio de Janeiro 9: 155
(1949, publ. 1950); Hofacker & P. J. Braun in Kakt. and. Sukk. 49: 267–268, illus. (1998).
Holotype: Brazil, Minas Gerais, Grão Mogol, Nov. 1938, Markgraf et al. s.n. (RB 65043).
Cereus markgrafii (Backeb. & Voll) P. J. Braun in Bradleya 6: 87 (1988).
Shrubby, solitary or poorly branched, to 2.5 m. Stems ± erect, 1.5–2.0 cm diam.; ribs 8–14, 2 × 3 mm;
epidermis bright to grey-green. Areoles 2 mm diam., 4–8 cm apart on the ribs. Central spines 1–2(–3),
to 4(–5) cm, radials 10–12, to 1 cm. Flowers 6 × 4 cm; pericarpel greenish, 11–12 × 8–9 mm; flower-
tube infundibuliform, 4 × 1.5–1.8 cm, pericarpel and tube bearing broad, triangular, acute to acuminate
bract-scales, green to strongly reddish at tips; perianth-segments to 2.8 cm, outer segments spreading,
lanceolate, fleshy, dark red at tips, inner segments spreading to erect, spathulate, delicate, white to very
pale green; stamens with greenish filaments; nectar-chamber 7–8 × 10–12 mm; style 60–65 × 3 mm,
stigma-lobes c. 8, at the same level as the anthers; ovary locule hemiglobose to depressed in longitudinal
section. Fruit globose to ovoid, to 3 cm diam.; pericarp greenish to dull red or slightly purplish. Seeds c.
2.4 × 1.8 mm, black, shiny; hilum-micropylar region forming an angle of 60° with long-axis (Barthlott
& Hunt 2000: pl. 47.3).
South-eastern campo rupestre (Grão Mogol) element: carrasco, 850–1000 m, region of Grão Mogol,
CEREUS
northern Minas Gerais. Endemic to the core area within Minas Gerais. Map 27.
minas gerais: Mun. Grão Mogol, Serra da Bocaina, 48 km W of bridge over Rio Vacaria, 28 km E of
Caveira, 31 Jan. 1991, Taylor et al. 1505 (HRCB, BHCB, K, ZSS); Grão Mogol, [near town], Nov. 1938,
Markgraf, Mello-Barreto & Brade s.n. (RB), 23 July, Zappi et al. in CFCR 9829 (SPF), 4 Sep. 1986, Mello-
Silva & Cordeiro in CFCR 10118 (SPF); l.c., c. 18 km W of town, 21 Feb. 1969, Irwin et al. 23684 (K,
NY, MO); 7 km S of town, 16°37'S, 42°56'W, 15 Oct. 1988, Taylor & Zappi in Harley 25069 (SPF, K);
Mun. Cristália, May 1988, Zappi (obs.).
conservation status. Endangered [EN B1ab(iii) + 2ab(iii)] (3); extent of occurrence = 434 km2;
PD=3, EI=1, GD=1. Short-list score (3×5) = 15. Ongoing deterioration of habitats in this region of Minas
Gerais is occurring. Part of its range is included in the Parque Estadual Serra do Barão, Grão Mogol.
This species is clearly differentiated from the widespread and variable B. phaeacanthus by its
acute pericarpel bract-scales and curved seeds. Its pollen-morphology merits investigation.
16. CEREUS Mill.
Gard. Dict. Abr. ed. 4 [unpaged] (1754).
Including Mirabella F. Ritter (1979); Cereus subg. Mirabella (F. Ritter) N. P. Taylor (1992a).
Literature: Taylor (1992a: 17, 25).
Treelike, shrubby or semi-scandent, with (2–)3–10-ribbed, triangular to cylindric or constricted-

cylindric stems; roots tuberous or fibrous; vascular tissues becoming very woody; epidermis and cuticle
thick and tough, often covered in whitish or bluish wax; areoles well-spaced along the ribs, usually
rather spiny at least in young plants. Flowers large, elongate funnelform to salverform, nocturnal, ±
sweetly scented, pericarpel and tube terete or somewhat angled, naked except for few minute scales,
glabrous or with small tufts of trichomes in their axils; perianth white inside; floral remnants strongly
blackening post-anthesis. Fruit ovoid to barrel-shaped, splitting laterally or opening at apex, pericarp
deep pink to reddish or yellow, sometimes strongly glaucous until ripe, floral remnant persistent or
early-deciduous leaving a depressed scar. Seeds c. 2–3 mm, black, testa smooth to ruminate.
A genus of some 20 poorly understood, South American species, divided between 4

subgenera, of which two are represented in Eastern Brazil (Subg. Cereus & Subg. Mirabella)
by a total of 6 or 7 species (one probably undescribed, see 5c). A third subgenus (Subg.
Ebneria (Backeb.) D. R. Hunt, with Praecereus excluded) is represented in Central-western
Brazil by 4 described species, C. spegazzinii F. A. C. Weber, C. kroenleinii N. P. Taylor
(1995b), C. adelmarii (Rizzini & A. Mattos) P. J. Braun and C. saddianus (Rizzini & A.
Mattos) P. J. Braun, although Braun & Esteves Pereira (2002: 79) indicate that is also found
in Bahia, which seems unlikely. Subgenera Ebneria and Mirabella are considered to be
vicariant groups, the latter replacing the former in Eastern Brazil (Taylor 1992a): Map 8.
In Eastern Brazil the genus has one or more representatives in all of the major
vegetation types (see subgenera, below). Plates 26–29.2.
1. Floral remnant early-deciduous, leaving a well-defined scar at the apex of the ± terete (not
strongly angled) developing fruit; rootstock fibrous, not rhizomatous-tuberous (in Brazilian spp.);
semi-decumbent shrubs with branches to > 5 cm diam. or erect and tree-like to > 4 m (Subg.
Cereus) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
CEREUS
1. Floral remnant not deciduous from fruit or breaking off above its base to leave a blackened
appendage on the strongly angular unripe fruit; rootstock rhizomatous-tuberous; semi-
scandent shrubs to < 3.5 m, with non-erect branches to 5 cm diam. (Subg. Mirabella) . . . . . . . . . 5
2. Fruit (when undamaged) opening from apex into c. 3 segments (S Minas Gerais, N Rio
de Janeiro) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6. hildmannianus
2. Fruit opening by a single lateral split, often from near base . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
3. Tree-like to > 4 m, with a ± well-defined, much thickened trunk; fruit pinkish- to
purplish-red (NE Brazil to cent. Minas Gerais) . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5. jamacaru
3. Semi-decumbent or low shrub, rarely treelike, 0.5–4.0 m; fruit pinkish red (coastal sand
& rocks) or yellow (W Espírito Santo, S edge of Minas Gerais & N Rio de Janeiro) . . . . . . . . . . 4
4. Ribs 3–5; flowers 14–25 cm (mainland Brazil) . . . . . . . . . . . . . . . . . . . . . . . 3. fernambucensis
4. Ribs 5–9; flowers c. 13 cm (Fernando de Noronha) . . . . . . . . . . . . . . . . . . . . . . . . . 4. insularis
5. Ribs (2–)3–4(–6), acute; wood yellow beneath the bark (caatinga, Piauí, Ceará, Pernambuco,
N & S Bahia to (?) cent.-N Minas Gerais) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2. albicaulis
5. Ribs (3–)4–6, rounded; wood whitish beneath the bark (cerrado & cerrado-caatinga ecotone,
SW Maranhão, W Bahia, N & cent.-E Minas Gerais) . . . . . . . . . . . . . . . . . . . . . . . 1. mirabella
Subg. Mirabella (F. Ritter) N. P. Taylor (Nos 1 & 2): roots tuberous; stems semi-scandent,
slender; bract-scales of pericarpel and tube with conspicuous trichomes and sometimes
fine spines in their axils; fruit bearing persistent floral remnant at apex. More or less
restricted to sandy substrates in cerrado and caatinga.
Kiesling (1994) has transferred the two species treated here to Monvillea Britton & Rose, which is typified
by Cereus cavendishii Monv. ex Lem. Hunt (1988) drew attention to the view already expressed by others
that this name and its better-known synonym, C. paxtonianus Monv. ex Salm-Dyck, had been misapplied
by J. D. Hooker, and later by Britton & Rose, to plants now correctly known as either Praecereus euchlorus
(F. A. C. Weber) N. P. Taylor or P. saxicola (Morong) N. P. Taylor (both described from central-eastern
South America). As to type, Hunt suggested that C. cavendishii and, therefore also Monvillea, were
referable to Acanthocereus Britton & Rose. Heath (1992a) characteristically disagreed with Hunt’s view and
neotypified C. cavendishii with an illustration published by J. D. Hooker (1899), to maintain the usage
established by Britton & Rose. Heath argued that the type locality given for C. cavendishii, namely
Cartagena, Colombia, was an error. However, photographs of Acanthocereus tetragonus (L.) Hummelinck
from northern Colombia preserved at NY, showing juvenile growth stages, strongly suggest that Cereus
cavendishii could have been based on a juvenile Acanthocereus, or possibly Pseudoacanthocereus, from the
region of Cartagena, rather than a Praecereus from central South America, and that Heath’s neotypification
should be superseded. It also should be noted that the genus Praecereus, represented by P. euchlorus subsp.
smithianus (Britton & Rose) N. P. Taylor, does occur in northern South America (Colombia &
Venezuela), but this subspecies has stems with 8–15 ribs (not 4–6 as required by the original description
of C. cavendishii): see Map 7.
1. Cereus mirabella N. P. Taylor in Bradleya 9: 85 (1991) et ibid. 10: 21, fig. F (1992).
Holotype: Brazil, Minas Gerais, c. 15 km W of Água Boa, Ritter 1238 (U).
Mirabella minensis F. Ritter, Kakt. Südamer. 1: 111 (1979), non Cereus minensis Werderm. (= Cipocereus
minensis (Werderm.) F. Ritter). Monvillea minensis (F. Ritter) R. Kiesling in Cact. Succ. J. (US) 66:
165 (1994).
Rootstock tuberous, with a large yam-like taproot; plant shrubby, usually ± decumbent upon or
scrambling over other vegetation, stems ascending to 1 m above the ground, to 2 m long and 3 cm
diam., strongly glaucous when new, wood whitish beneath the bark; ribs (3–)4–6, rounded (5–6 in
CEREUS
seedlings); areoles 3–5 mm diam., with blackish glandular trichomes at first, later whitish, 2–4 cm apart
on the ribs; spines 3–6 (more in seedlings and young basal shoots), needle-like, brownish, paler tipped,
the longest ± central, to 25 mm. Flower (Taylor 1992a: 21, Fig. F) narrow funnelform, c. 18.5 × 10
cm; tube angular, c. 12 × 1.3–1.5 cm, bearing occasional bract-scales with trichomes in their axils,
stamens inserted in uppermost 2–2.5 cm, to 5 cm; style c. 14 cm, stigma-lobes c. 10, to 9 mm, whitish;
perianth-segments c. 4–5 cm, narrow, white; pericarpel oblong, c. 22 × 15 mm, angular, bearing
minute bract-scales with trichomes and occasional minute spines in their axils. Fruit 3.5–4.0 × 2–3 cm,
tapered at apex, with a few minute scales bearing trichomes and sometimes minute spines near base,
pinkish, bearing the blackened perianth remains. Seed c. 2.4 × 1.6 mm, testa ± smooth, shiny black.
Cerrado element: mostly in sandy phases of the cerrado and more open places of the cerrado-caatinga ecotone,
c. 150–750 m, south-western Maranhão and western Bahia to cent.-N and W Minas Gerais (Rio São
Francisco drainage), and disjunctly in E-cent. Minas Gerais (Rio Doce/Rio Jequitinhonha watershed);
Tocantins (?) and Goiás. Map 28.
maranhão: Mun. Carolina, 16 km E of town, road BR 230, Serra da Madeira, 7°21'S, 47°25'W, 5 July
1993, T.M. Sanaiotti (K, photo).
bahia: W Bahia, Mun. Formosa do Rio Preto, Faz. Estrondo, 11°7'7"S, 45°29'36''W, 12 Nov. 1997, D.
Alvarenga et al. 1072 (K); Mun. Barreiras, 31.5 km W of town on road BR 020, 19 Jan. 1991, Taylor et al.
1442 (HRCB, CEPEC, K, ZSS); (?) Mun. Angical, 20 km W of Cristópolis on road BR 242, Horst &
Uebelmann 206 [Uebelmann, mss.]; Mun. Carinhanha, 14°14'S, 43°52'W, 2001, J. Jardim 3574 (CEPEC).
minas gerais: W Minas Gerais, Mun. Januária, ‘distrito de Fabião, 2 km na estrada partindo do Abrigo
do Malhador’, 15°7'16"–15°8'57"S, 44°15'20"–44°14'13''W, 26 Oct. 1998, J.A. Lombardi 2123 (K,
BHCB); Mun. São Francisco, near Urucuia, 1987, illustrated by Braun & Esteves Pereira (2002: 143,
Abb. 156); Mun. João Pinheiro, fide Uebelmann (1996): HU 584 [record outside E Brazil area]; cent.-N
Minas Gerais, Mun. Varzelândia, 36 km ENE of town on dirt road to Cachoerinha (near where the road
MG 401 crosses the Rio Verde Grande), 23 July 1989, Zappi 167 (K, fr & seed only); Mun. São João da
Ponte, 20 km from Lontra on dirt road to Varzelândia, 11 Aug. 1988, Eggli 1145 (ZSS); l.c., 17 km SSW
of Varzelândia, 23 July 1989, Zappi 166 (SPF, HRCB); l.c., Km 265–263 on road BR 135 from Montes
Claros, 10 Aug. 1988, Eggli 1131 (ZSS); Mun. Brasília de Minas, 27 km N of Mirabela on road BR 135,
16°2'S, 44°18'W, 6 Nov. 1988, Taylor & Zappi in Harley 25513 (SPF, K); Mun. Mirabela, c. Km 310 on
road BR 135 from Montes Claros, 10 Aug. 1988, Eggli 1129 (ZSS); Mun. Francisco Dumont, fide
Uebelmann (1996): HU 710; cent.-E Minas Gerais, Mun. São Sebastião do Maranhão (?), c. 15 km W
of Água Boa, Ritter 1238 (U).
PD=2, EI=1, GD=1. Short-list score (2×4) = 8. Cerrado habitats in Eastern Brazil are suffering
considerable modification by agricultural development and the populations of this species, although
numbering > 10, are severely fragmented.
Together with Arrojadoa dinae, Cipocereus crassisepalus and 3 Discocactus spp., this is one
amongst few terrestrial cacti that inhabits cerrado, although it cannot be said to be a
common component of this vegetation, being of erratic occurrence in sandy places. The
record of C. mirabella from south-western Maranhão is separated from those from western
Bahia by c. 500 km, but this probably reflects the lack of collections from this relatively
inaccessible region and it is possible that its distribution is ± continuous in the cerrados in
between. Of greater interest is the apparent disjunction in its distribution across the Serra
do Espinhaço, between central-eastern (W of Água Boa, the type locality) and central-
northern Minas Gerais — c. 250 km. Nevertheless, it is possible that geographically
intermediate populations await discovery. Braun & Esteves Pereira (2002: 79) indicate
that ‘Mirabella’ occurs in Tocantins, which is well possible given the distribution of this
species recorded above.
CEREUS
2. Cereus albicaulis (Britton & Rose) Luetzelb., Estud. bot. Nordéste 3: 111 (1923,
publ. 1926); N. P. Taylor in Bradleya 10: 24, pl. III (1992). Type: Brazil, Bahia, Barrinha,
1915, Rose & Russell 19808 (US, lecto. designated here; NY, lectopara.).
Acanthocereus albicaulis Britton & Rose, Cact. 2: 125–126 (1920). Mirabella albicaulis (Britton & Rose) F.
Ritter, Kakt. Südamer. 1: 110 (1979). Monvillea albicaulis (Britton & Rose) R. Kiesling in Cact. Succ.
J. (US) 66: 165 (1994).
Very similar to the preceding, but scrambling to at least 3 m high; tuberous taproot to 40 cm; stem to 5
cm diam., cortex yellow beneath the bark; ribs (2–)3–4(–6), usually sharply triangular-acute; central spine
better defined. Flower (Taylor, l.c.) nocturnal, 13.5–18.3 cm (more slender in all its parts than C. mirabella
in the few examples seen), stamens c. 20 mm. Fruit to 7.5 × 4 cm, funicular pulp white. Seed c. 2 × 1.4
mm (Barthlott & Hunt 2000: pl. 46.5–6).
Widespread caatinga element: in caatinga de altitude, caatinga and carrasco, often on sandy substrates of the
Cipó soil series, c. 470–1020 m, northern and south-eastern Piauí, north-western Ceará, western and
central-southern Pernambuco to western, northern and north-eastern Bahia and southwards through the
Chapada Diamantina to the Serra do Espinhaço, southern Bahia and (?) northernmost Minas Gerais.
Endemic to Eastern Brazil. Map 28.
piauí: N Piauí, Mun. Piracuruca, Alto Alegre, fide Uebelmann (1996): HU 941; SE Piauí, Mun. Dom
Expedito Lopes, Gaturiano, Dec. 1963, Ritter 1236 (SGO 121755, fide Eggli et al. 1995: 501); Mun. Padre
Marcos, Serra Velha, 27 Dec. 1993, M.E. Alencar 37 (PEUFR), without date, M.E. Alencar 375
(PEUFR); Mun. Simões, 7°36'S, 40°45'W, 5 Apr. 2000, E.A. Rocha et al. (K, photos); Mun. Canto do
Buriti, fide Uebelmann (1996): HU 837; Mun. São Raimundo Nonato, near foot of Serra da Capivara,
6 Feb. 1990, Zappi 218 (SPF, HRCB); l.c., Zabelê, 8°44'1"S, 42°29'21''W, 20 Dec. 1998, J.R. Lemos 74
(PEUFR).
ceará: NW Ceará, Mun. Ubajara, Planalto da Ibiapaba, 21 July 1994, F.S. Araújo s.n. (UEC 96574).
pernambuco: Mun. Araripina, fide Uebelmann (1996): HU 932; Mun. (?), between Jutaí and Cruz de
Malta (Santa Cruz), 26 Nov. 1959, Andrade-Lima 59-3411 (IPA); Mun. Petrolina, Rajada, fide
Uebelmann (1996): HU 1755; Mun. Floresta, N of Serra Negra, 8°33'S, 38°1'W, 8 Apr. 2000, E.A.
Rocha et al. (K, photos), l.c., 38 km from town on road PE 360 to Floresta, 18 Jan. 1992, F.A.R. Santos
45 (HUEFS, PEUFR); Mun. Nova Petrolândia, 18 km W of town towards Floresta, 12 Feb. 1991, Taylor
& Zappi 1622 (UFP, HRCB, K, ZSS).
alagoas: reported by Luetzelburg (1925–26, 3: 111), without locality.
bahia: W Bahia, Mun. Barra, Ibiraba, dunes by the Rio São Francisco, s.d., P. Rocha 47 (SPF); Mun.
Santa Maria da Vitória, near airstrip, 4 Oct. 1985, Horst & Uebelmann 833, cult. 1986 (K, photo); N & cent.
Bahia, Mun. Remanso, 9°29'S, 42°16'W, 27 Feb. 1978, Miranda 315 (IPA); l.c., 11 km E of Remanso,
9°33'S, 42°W, 7 Apr. 2000, E.A. Rocha et al. (K, photos); l.c., 47 km E, 9°22'S, 41°48'W, 7 Apr. 2000,
E.A. Rocha et al. (obs.); Mun. Juazeiro, 16 July 1959, Castellanos 22444 (R); Mun. Sento Sé, c. 0.5 km N
of Cabeluda, 11 Jan. 1991, Taylor et al. 1392 (CEPEC, HRCB, K, ZSS); Mun. Jaguarari, 64 km S of
Juazeiro on road to Senhor do Bomfim, 8 Jan. 1991, Taylor et al. 1376 (CEPEC, HRCB, K, ZSS); l.c.,
Barrinha, 1915, Rose & Russell 19808 (US, NY); l.c., Flamengo, Ritter 1237 (SGO 121756, fide Eggli et al.
1995: 501); Mun. Uauá [‘Curaçá’], Lagoa da Fulô, 9°51'S, 39°41'W, 16 Aug. 1983, G.C.P. Pinto & S.B.
Silva 187/83 (HRB); l.c., c. 10.5 km NW of town towards Poço de Fora, 7 Jan. 1991, Taylor et al. 1367
(CEPEC, HRCB, K, ZSS); Mun. Jacobina, 27–28 km WNW of town towards Lajes, 12 Jan. 1991, Taylor
et al. 1399 (CEPEC, HRCB, K, ZSS); Mun. Euclides da Cunha, fide Uebelmann (1996): HU 731; Mun.
Morro do Chapéu, 19 km W of town on road BA 052, 11°28'S, 41°19'W, 25 Dec. 1988, Taylor & Zappi
in Harley 27390 (CEPEC, SPF, K); Mun. Cafarnaum, 9 km N of town towards road BA 052, 24 Aug.
1988, Eggli 1278 (ZSS); Mun. Seabra/Iraquara, 29 km N of town on road to Água de Rega, 27 Feb. 1971,
Irwin et al. 31216 (K, UB); Mun. Seabra, 8 km E of Seabra, 20 Mar. 1989, Supthut 8951 (ZSS); Mun. Rio
do Pires, Ibiajara, fide Uebelmann (1996): HU 129; NE Bahia, Mun. Biritinga, 18 Jan. 1994, M.A.S. das
Neves 81 (HUEFS); S Bahia, Mun. Iaçu, Morro da Garrafa, 12°45'S, 39°51'W, 22 Feb. 1997, E. de Melo
2076 (HUEFS); Mun. Guanambi, 19 km from the town on road to Caitité, 25 July 1989, Zappi 170 (SPF,
HRCB); Mun. Caitité, Brejinho das Ametistas, 26 July 1989, Zappi 173 (SPF); Mun. Jacaraci, 2 km NW
CEREUS
of town, 1982, Horst & Uebelmann 577 (K, ZSS); Mun. Condeúba, ‘Serra de Condeúba’, reported by
Werdermann (1933: 99); Mun. Cordeiros, 15°4'S, 41°53.5'W, Sep. 2003, M. Machado (obs.).
minas gerais: (?) cent.-N Minas Gerais, reported by Braun & Esteves Pereira (1991b: 190; 2003:
32A, illus.).
Concern in view of its extensive range, but the plant is not common and its habitat is declining.
This species has a patchy geographical distribution, being widely spread in the caatinga
towards its northern limits, but more or less restricted to the East Brazilian Highlands in
the southern parts of its range. This distribution pattern may in part correspond with that
of the ‘Cipó’ soil series, upon which a distinct type of caatinga vegetation is found (type
No. 5 of Andrade-Lima 1981: 159), but it has also been seen growing upon inselbergs and
occasionally on other substrates. It is variable in the robustness of its stems.
It was originally described by Britton & Rose as an Acanthocereus, whose species it
vaguely resembles in vegetative characters, although not in flower and fruit-morphology,
which Britton & Rose unfortunately did not know.
Subg. Cereus (Nos 3–6): rootstock fibrous (in Brazilian taxa); treelike, semi-decumbent
or creeping, stems stout; bract-scales of pericarpel and tube glabrous in their axils (rarely
with inconspicuous trichomes in C. fernambucensis); floral remnant early-deciduous from
fruit apex.
This is the least understood subgenus, species delimitation being hampered by lack of
data on fruit and seedling morphology. Found in various phases of the Mata atlântica, caatinga-
agreste, cerrado (but only on limestone outcrops) and South-eastern campos rupestres (rare).
3. Cereus fernambucensis Lem., Cact. gen. nov. sp.: 58 (1839). Type: ‘patria est
Fernambuco’ [Pernambuco = Recife]; not known to have been preserved. Neotype
(designated here): Brazil, Pernambuco, Mun. Jaboatão dos Guararapes, Candeias, by the
sea, 20 Feb. 1990, Zappi 228 (HRCB) — nearest extant population (in 1990), since
already extinct in Mun. Recife.
? C. obtusus Haw., Rev. pl. succ.: 70 (1821). Piptanthocereus obtusus (Haw.) F. Ritter, Kakt. Südamer. 1:
231 (1979). Type: not known to have been preserved.
? ‘C. neotetragonus’ Backeb., Die Cact. 4: 2363 (1960), nom. inval. (Art. 37). Supposedly based on a plant
from Brazil (Rio de Janeiro?), but not typified. The plant in the plate 77 that Backeberg cites from
Werdermann, Blühende Kakt. und andere Sukk. (1934), does not match his description and probably
represents the Andean C. hankeanus F. A. C. Weber.
[C. variabilis auctt. non Pfeiff., Enum. cact.: 105 (1837), nom. illeg. (Art. 52.1).]
[C. pitajaya sensu Hook. in Bot. Mag. 70: tab. 4084 (1844), non (Jacq.) DC. (= Acanthocereus tetragonus
(L.) Hummelinck).]
Shrubby or caespitose, erect or sprawling, always branched at or very near base, 0.5–4.0 m; stems with
numerous constrictions, pale green to glaucous, non-mucilaginous; ribs 3–5, to 65 × 10–30 mm; areoles
to 10 mm diam., 15–35 mm apart on mature stems, rather woolly, felt often blackish at first, hairs whitish
and conspicuous on young growth. Spination variable, pale yellow-brown to brown when young, later
grey, sometimes almost lacking; central spines 1–4, to c. 50 mm; radials to 8 or more, to c. 25 mm.
CEREUS
Flowers slender funnel-form, 14–25 × 9–15 cm, white, sometimes pinkish without; pericarpel bract-
scales sometimes bearing trichomes in their axils, tube pale green. Fruit 5–7.2 × 3–6 cm, pinkish red or
yellow, laterally dehiscent by a single split in the pericarp, the base of the floral nectar-chamber forming
a deep depression (to 8 mm) at its apex. Seed c. 2–3 × 1.5 mm, black, shiny, cuticular folds almost absent.
conservation status. Least Concern [LC].
The specific epithet is correctly spelled with ‘f’ and should not be corrected to
‘pernambucensis’ as advocated by various authors, since Pernambuco was written
Fernambuco by some europeans during the 19th Century (see Brummitt & Taylor 1990:
302–303; Werdermann 1933: 89–90). The illegitimate name C. variabilis Pfeiff. as to
type belongs in the synonymy of Acanthocereus tetragonus (L.) Hummelinck, a Caribbean
taxon, and cannot be used even though Pfeiffer’s concept included elements referable
to the Cereus treated here (cf. ICBN Arts 7.5 & 52.1/2). Haworth’s C. obtusus has also
been used for this species, but its original description and typification are unsatisfactory
and its provenance is uncertain. Prior to conservation of the name C. jamacaru DC.
with a new type (see below), De Candolle’s name was in fact based on an illustration
of C. fernambucensis.
Two subspecies are recognized and are quite separate geographically, except in the
lowlands of the northern half of Rio de Janeiro:
1. Fruit pinkish red; flower to c. 17 cm (sand & rocks by the sea) . . . . . 3a. subsp. fernambucensis
1. Fruit yellow; flower to c. 25 cm (rocks inland, W Espírito Santo, S Minas Gerais &
N Rio de Janeiro) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3b. subsp. sericifer
3a. subsp. fernambucensis
vernacular names. Cardo-ananá, Cardo-de-praia, Cardo-vinagre, Figueira-do-inferno.
Low-growing, caespitose, mostly < 80 cm tall, occasionally semi-pendent or scrambling through

vegetation and reaching 2.5 m high; stems very variable in size, 4–11 cm diam., pale to blue-green but
not strongly glaucous; areoles spiny. Flowers c. 14–17 × 9–11 cm; pericarpel 13–18 × 7.5–12.0 mm, tube
narrowed to 7–8.5 mm, then flared to 13–30 mm diam. at apex; perianth-segments to 4.5–6 × 1.4–2.0
cm; style 11–16.5 cm; stigma-lobes 8–13, 8–15 mm, green. Fruit pinkish red.
Humid/subhumid forest (restinga) element: on sand-dunes, rocks and growing through shrubs of the
restinga, sometimes within reach of the sea spray, to c. 100 m, throughout the coast of Eastern Brazil from
Rio Grande do Norte southwards (to São Paulo: Ilha do Cardoso). Endemic to North-eastern and South-
eastern Brazil. Map 29.
rio grande do norte: Mun. Natal, 5 km S of Ponta Negra, 2 Sep. 1946, Wurdack B-136 (US); Mun.
Búzios, E.A. Rocha (K, photo).
paraíba: Mun. Mataraca, reported by Oliveira-Filho (1993: 224); Mun. Cabedelo, 6°58'40"S,
34°50'18''W, reported and illustrated by Locatelli & Machado (1999b); Mun. João Pessoa, Cabo Branco,
edge of restinga forest, 9 Feb. 1995, Taylor (K, photos).
pernambuco: Mun. Jaboatão dos Guararapes, Prazeres [Estação], 21 Feb. 1933, B. Pickel 3233 (IPA);
l.c., Candeias, by the sea, 20 Feb. 1990, Zappi 228 (HRCB); Mun. Ipojuca, Praia de Porto de Galinhas,
4 Sep. 1992, H.B. Martins 2 (PEUFR).
alagoas: Mun. Maragogi, Peroba, 7 June 1993, I.A. Bayma 68 (MAC); Mun. Marechal Deodoro,
Pontal da Barra, 5 Dec. 1977, G.L. Esteves 162 (MAC); l.c., A.P.A. de Santa Rita, Saco da Pedra, 20 Feb.
1991, C.S.S. Barros et al. 108 (MAC), 23 Feb. 1989, Esteves & Lyra-Lemos 2190 (MAC); Mun. Penedo,
CEREUS
7 km S of turnoff to town on road to Piaçabuçu, 14 Feb. 1991, Taylor & Zappi (obs.); Mun. Piaçabucu
30 Sep. 1981, R.F.A. Rocha s.n. (MAC).
sergipe: Mun. Barra dos Coqueiros, 4 Aug. 1998, Taylor (obs.); Mun. Aracaju, Praia de Atalaia, 11 July
1980, M. Fonseca s.n. (ASE 766).
bahia: Mun. Conde, Praia de Seribinha, 28 Apr. 1996, E.C. Medeiros Neto 36 (HUEFS); Mun. Mata de
São João, Sauípe, 27 Aug. 1996, R. Soeiro 39/95 (HRB); Mun. Camaçari, estrada do Coco, próx. a
Guarajuba, 19 Nov. 1981, G.C.P. Pinto 419/81 (HRB); Mun. Salvador, Itapuã, 12°56'S, 38°21'W, 2
Mar. 1980, L.R. Noblick et al. 1695 (ALCB, HUEFS); ibid., Jardim de Alá, 1959, A.L. Costa s.n. (ALCB);
ibid., Lagoa de Abaeté, 4 Jan. 1991, Taylor et al. 1345 (CEPEC, HRCB, K, ZSS); Mun. Itacaré, between
the praias do Farol and da Ribeira, 14 Dec. 1992, A.M. Amorin et al. 963 (CEPEC, K); Mun. Ilhéus,
Sambaituba, 17 Mar. 1993, W.W. Thomas et al. 9554 (CEPEC); l.c., 6 km S of town, W of coast road, 3
Jan. 1989, Taylor & Zappi in Harley 27874 (CEPEC, K, SPF); Mun. Una, 3 km N of Comandatuba,
15°20'S, 39°2'W, 3 Jan. 1989, Taylor & Zappi in Harley 27875 (CEPEC, SPF, K); Mun. Canavieiras, 32
km from town on road to Camacã, 8 Sep. 1965, R.P. Belém 1729 (CEPEC); Mun. Porto Seguro, 7 km
N of town towards Santa Cruz Cabrália, Taperapuã, 20 Apr. 1982, A.M. de Carvalho et al. 1225
(CEPEC); Mun. Alcobaça, road BA 001, 5 km S of Alcobaça, 17 Mar. 1978, S.A. Mori et al. 9619
(CEPEC, NY); Mun. Nova Viçosa, 18 June 1985, Hatschbach 49455 (CEPEC, K).
espírito santo: Mun. Itaúnas, 29 Mar. 1995, J. Araújo (K, photo); Mun. Linhares, Vale do Rio Doce,
Km 30–40 from Linhares, 5 Oct. 1971, T.S. dos Santos 2065 (CEPEC); ibid., Res. Biológica dos
Comboios, IBDF, Regência, próx. à reserva da FUNAI, 24 Jan. 1990, D.A. Folli 1075 (Herb. Univ. Fed.
ES); Mun. Guarapari, Setiba, 14 Oct. 1988, O.J. Pereira 1844 (Herb. Univ. Fed. ES); Mun. Anchieta,
Rodovia do Sol, between Meaípe (Mun. Guarapari) and Anchieta, 20°45'13"S, 40°32'37''W, 25 Nov.
1999, reported growing with Pilosocereus brasiliensis (Zappi et al. 451); Mun. Itapemirim, Rodovia do Sol,
S of Marataízes, 21°4'51"S, 40°50'24''W, 26 Nov. 1999, reported growing with Opuntia monacantha
(Zappi et al. 470), Jan. 1970, L. Krieger 7607 (CESJ).
rio de janeiro: Mun. Quiçamã (Quissamã), road to Barra do Furado, 22°6'5"S, 41°26'W, 23 Nov.
Concern because of its frequency, although the majority of littoral habitats in Eastern Brazil are undergoing
significant modification and substantial reductions in numbers of adult individuals can be projected.
The above key and description of this variable subspecies does not account for southern
forms from western Rio de Janeiro and São Paulo, which are considerably larger in their
stems and flowers. Its pollination biology has been studied by Locatelli & Machado (1999b).
3b. subsp. sericifer (F. Ritter) N. P. Taylor & Zappi in Cact. Consensus Initiatives 3: 7
(1997). Holotype: Brazil, Rio de Janeiro, Três Rios, 1965, Ritter 1410 (U).
Piptanthocereus sericifer F. Ritter, Kakt. Südamer. 1: 232–233, Abb. 185–187 (1979). Cereus sericifer (F.
Ritter) P. J. Braun in Bradleya 6: 86 (1988).
Erect or semi-decumbent shrub to 4 m tall, branching mainly near base or from decumbent parts of the
branches; stem 10 cm diam. or more, green to strongly glaucous; areoles almost naked to very spiny.
Flowers to c. 25 × 15 cm; pericarpel c. 25 × 20 mm, tube narrowed to not < 10 mm, then flared to 35
mm or more at apex; perianth-segments to 11 × 4 cm; style c. 20 cm; stigma-lobes c. 15, 15–20 mm,
pale yellow. Fruit yellow.
Southern humid/subhumid forest (inselberg) element: on ± naked rock outcrops (especially

gneiss/granite inselbergs) inland in the Mata atlântica zone, 50–400 m, western and central Espírito Santo
and southernmost Minas Gerais, to adjacent Rio de Janeiro (Rio Paraíba drainage). Endemic to South-
eastern Brazil. Map 29.
CEREUS
minas gerais: on road BR 040, between Juiz de Fora and Três Rios (RJ), ‘Paraibuna’, 25 July 1991,
Zappi (obs.).
espírito santo: Mun. Águia Branca, 29.5 km from Barra de São Francisco on road ES 080, 16 Dec.
1990, Taylor & Zappi 781 (HRCB, MBML, K, ZSS); l.c., c. 5 km S of Águia Branca, 16 Dec. 1990,
Taylor & Zappi 782 (HRCB, MBML, K, ZSS); Mun. São Gabriel da Palha, photo of plant in habitat with
ripe yellow fruit, supplied by A. Hofacker, 28 Sep. 2003 (by e-mail); Mun. Colatina, 15 km S of São
Domingos on road to Colatina, 16 Dec. 1990, Taylor & Zappi (obs.); município unknown, Rio Doce,
Feb. 1917, Luetzelburg 7227 in part (M); Mun. Itarana, road to Praça Oito, 19°54'35"S, 40°49'46''W, 24
Nov. 1999, Zappi et al. 416 (K, UEC); Mun. Santa Teresa, 8 km S of Santo Antônio coming from
Colatina, 17 Dec. 1990, Taylor & Zappi 788 (HRCB, K); l.c., Rio Cinco de Novembro, H.Q.B.
Fernandes 2259, cult. HRCB, May 1990 (HRCB); Mun. Jerônimo Monteiro, illustrated by Uebelmann
(1996) together with Coleocephalocereus pluricostatus; Mun. Bom Jesus do Norte, 7 km N of town coming
from São José do Calçado, 18 Dec. 1990, Taylor & Zappi 791 (HRCB, MBML, K, ZSS).
rio de janeiro: Mun. Itaocara, 21 Aug. 1963, Castellanos 24010 (GUA); Mun. Santa Maria Madalena,
access road to the town, 22°0'55"S, 42°7'30''W, 21 Nov. 1999, Zappi et al. 353 (K, UEC); Mun. Tres
Rios road BR 040, at crossing of Rio Preto, rocks W of road, 3 Aug. 2003, Taylor (obs.).
Habitats not significantly declining at present.
This subspecies represents a distinct inland race of the otherwise littoral C. fernambucensis.
Apart from its yellow fruit it differs from subsp. fernambucensis, as seen in Eastern Brazil,
by being larger in all its parts, although forms of the latter from further south approach it
in their stems and flowers.
4. Cereus insularis Hemsl., Rep. Challenger, Bot. 1(2): 16 (1884). Type: Brazil,
Fernando de Noronha (St Michael’s Mount), Moseley s.n. (K, BM).
? C. ridleii Andrade-Lima ex Backeb., Die Cact. 4: 2352, Abb. 2247 (1960). Holotype: Brazil, Fernando
de Noronha, Oct. 1955, Andrade-Lima 55-2221 (IPA).
Like C. fernambucensis subsp. fernambucensis, but stems more evenly cylindrical (less obviously constricted/
jointed), to 2(–5) m tall; ribs 5–9, 1–2 cm high, spines usually more numerous, to 15 or more, partially
hiding the stem; flowers very similar, but to only c. 13 cm. Seed: Barthlott & Hunt (2000: pl. 45.3–4).
Humid/subhumid forest (restinga) element: rocky habitats and cliffs, Fernando de Noronha, North-
eastern Brazil. Endemic. Map 1.
pernambuco: Archipelago of Fernando de Noronha, 1887, Ridley et al. 23 (K, BM; IPA, photos), Oct.
1955, Andrade-Lima 55-2221 (IPA), 1987, Duranton s.n. (K, SPF 48093, 48094, 48095); l.c., próximo a
salina, 20 Oct. 1955, Andrade-Lima 55-2222 (IPA); Praia do Bode, 3°50'S, 32°25'W, 3 June 1993, A.M.
Miranda et al. 999 (PEUFR); l.c., Praia Ataleia, 4 June 1993, A.M. Miranda 1042, 1043 (PEUFR, HUEFS).
conservation status. Least Concern [LC] (1); PD=2, EI=1, GD=2. Short-list score (1×5) = 5. The
Archipelago of Fernando de Noronha is an effectively managed protected area.
This taxon is clearly very similar to C. fernambucensis, from which it is separated by c. 350
km of Atlantic Ocean, but further studies are required to determine how close they are and
also the status of the much larger C. ridleii. Cereus insularis is apparently very variable in habit
and spination, some forms being completely covered in fierce spines, the stems 5–9-ribbed
(cf. Duranton s.n.). Its flowers, which are known from carefully preserved material, scarcely
differ from those borne by smaller forms of C. fernambucensis, such as are found along the
CEREUS
coast of Pernambuco. The poorly known C. ridleii, which was discovered in the archipelago
by the respected Dárdano Andrade-Lima, remains something of an enigma (Braun 1990).
It is known for certain only from his original habitat photograph, depicting a distinctive
tree-like plant (c. 3–5 m), and from its holotype preserved at Recife (IPA). Contrary to
Backeberg’s statements (l.c.), the type specimen, collected in October 1955 (Andrade-Lima
55-2221), is 6-ribbed and bears flowers. (It is possible that wild material was cultivated at
Recife until it flowered, but the label does not give any clue about this.) Were it not for
the photograph and the 6-ribbed stem, the holotype could easily be identified as C.
fernambucensis. The photograph published by Backeberg (1960: Abb. 2247) shows the tree-
like C. ridleii surrounded by plants of the low-growing C. insularis. It is therefore tempting
to suppose that C. insularis could be some kind of stabilized juvenile, yet reproductive
(neotenic) form, which occasionally develops into the erect adult stage represented by C.
ridleii, such behaviour being known, for example, in the Madagascan cactus relative, Didierea
trollii (Rowley 1992). An alternative explanation would be that an arborescent Cereus,
presumably C. jamacaru, was introduced to the islands in historic times and what Andrade-
Lima observed was a rare product of hybridization of this with C. insularis. This is not such
an unlikely scenario, because ‘mandacaru’ is regarded as a useful plant in the Nordeste and
is frequently transported and planted around habitations as a living hedge.
Recently, Brazilian botanist, Emerson Rocha (pers. comm.), has visited the
archipelago and has observed distinctively tall, somewhat branched forms of C. insularis,
although these are much more spiny than Andrade-Lima’s C. ridleii. In various respects,
apart from habit and size, C. fernambucensis and C. jamacaru are morphologically
comparable, and C. ridleii, if it is not the product of hybridization, would appear to
potentially link them even further, combining the small flowers of the former with the
treelike habit of the latter.
5. Cereus jamacaru DC., Prodr. 3: 467 (1828), nom. cons. Holotype (Taylor & Zappi
1992c): Brazil, Bahia, Mun. Curaçá, 7 Jan. 1991, Taylor et al. 1369 (CEPEC; HRCB, K,
ZSS isos.), typ. cons. (Greuter et al. 2000: 387).
Piptanthocereus goiasensis Ritter, Kakt. Südamer. 1: 234 (1979). Cereus goiasensis (Ritter) P. J. Braun in
Bradleya 6: 86 (1988). C. jamacaru subsp. goiasensis (Ritter) P. J. Braun & E. Esteves Pereira in
Succulenta 74: 84 (1995). Holotype: Brazil, Maranhão, Carolina, 1963, Ritter 1282 (U).
vernacular names. Mandacaru, Mandacaru-de-boi, Mandacaru-facheiro, Mandacaru-de-faixo,

Cardeiro, Jamacaru, Jamaracurú, Jumucurú, Jumarucú, Cumbeba, Urumbeba.
Tree-like, to at least 12 m (rarely to 18 m), with a spread of up to 10 m or more, often with a short but
well-defined, bark-covered, greyish trunk 50–200 × 30–80(–100) cm, regrowing from base to form a
shrub when cut down, densely branched; stems 7–20 cm diam., often strongly glaucous when young,
later dark blue-green, epidermis very tough, vascular cylinder extremely woody when old; ribs 4–10,
sometimes only 3 in juvenile plants < 1.5 m tall, to 60 × 18–45 mm, high and narrow to broader and
more rounded, with or without transverse fold lines alongside the areoles, margin weakly to more
strongly crenate; areoles ± circular, 4–8 mm diam. at first, 15–40 mm apart, dirty white to grey woolly
felted and hairy at first, often showing marked indeterminate growth in age, to 15 mm diam. Spination
very variable, often ± absent on mature branches high above the ground, but sometimes spiny throughout
and/or very spiny near base and then with many more spines per areole than described below, spines pale
CEREUS
yellow-brown to reddish brown at first, later grey to blackish; central spines 1–4 or more, very variable
in size, occasionally to 150 × 3 mm; radials 7–12, to c. 35 mm. Flowers developing from areoles almost
anywhere on the branches, funnel-form, 21–30 × c. 10–20 cm; pericarpel 20–28 × 12–18 mm, like the
tube bearing green or reddish bract-scales glabrous in their axils, apex of tube to 50 mm diam.; nectar-
chamber to 65 mm; perianth-segments 50–100 × 20–27 mm, outer greenish sometimes tinged crimson
near apex, inner pure white; style white to pale green, stigma-lobes 12–16, 11–19 × 1 mm, greenish.
Fruit 6–10 × 4–8 cm, crimson to pinkish red, dehiscent by a longitudinal split to reveal white funicular
pulp. Seed black, testa with cuticular folds.
This well-known taxon, whose name has already been conserved for purely
nomenclatural reasons so that it can continue to be used in its familiar sense (Taylor &
Zappi 1992c), may in future be liable to taxonomic union with C. hexagonus (Linnaeus)
Miller. The Linnaean species is traditionally known from Venezuela southwards to the
Guianas and northernmost Brazil (Roraima), but there exist populations further to the
south-east, in the states of Pará and Maranhão, whose identity has not been determined.
Since, on present knowledge, there do not appear to be any clear characters to distinguish
between this pair of taxa, the absence of any significant geographical disjunction makes it
tempting to suppose that there may be only a single widespread and variable species
involved. Nevertheless, the name C. jamacaru is maintained for the time being, at least
until the situation in the field has been thoroughly researched.
Even in the restricted sense adopted here, this is a very wide-ranging and rather
variable species divisible into the following subspecies:
1. Juvenile plants (between 10 cm and 1 m high) passing through a stage with only 3–7 ribs
and yellow to orange-brown spines of variable length; mature stem-segments variously
shaped; flower 15–20 cm diam. or more; pericarpel and tube to c. 16 cm, bract-scales red,
conspicuous; largest perianth-segments 8–10 cm (NE Minas Gerais northwards, from
c. 18°25'S in the drainage of the Rio Jequitinhonha) . . . . . . . . . . . . . . . . . . 5a. subsp. jamacaru
1. Juvenile plants (between 10 cm and 1 m high) passing through a stage with 5–8 ribs
and uniformly short, dark red-brown spines; mature stem-segments broadest near base;
flower 10–15 cm diam.; pericarpel and tube to 21 cm, bract-scales green or brownish,
inconspicuous; largest perianth-segments 5–7 cm (W Bahia to central Minas Gerais, to
20°S, Rio São Francisco drainage, often on limestone) . . . . . . . . . . . .5b. subsp. calcirupicola
5a. subsp. jamacaru

Widespread caatinga element: in stony to sandy soil and on rocks of various kinds, caatinga-agreste, entering
into Mata atlântica (and restinga) in NE Brazil, c. 50–1000(–1200) m, widespread in Eastern Brazil, but less
frequent west of the Rio São Francisco in south-western Bahia (where replaced by subsp. calcirupicola on
limestone outcrops) and uncommon within the southern part of the Chapada Diamantina, southwards to
central-northern and north-eastern Minas Gerais (extending as a frequent plant in the Rio Jequitinhonha
valley to at least 17°S and beyond in the region of Diamantina, at c. 18°25'S), sometimes cultivated
further south, as well as within its natural range, ranging north-westwards to Maranhão, northernmost
Piauí and Ceará; Northern Brazil (Tocantins & Pará)? Only a few records marking the approximate
western, eastern and southern limits of natural range of the subspecies are cited (Map 29):
maranhão: Mun. Carolina, 1963, Ritter 1282 (U); Mun. Loreto, Rio Balsas, 26 May 1962, G. & L.T.
Eiten 4722 (SP).
pernambuco: Mun. Moreno, Tapera, 29 Feb. 1930, Pickel 2260 (IPA 4469).
CEREUS
alagoas: S Alagoas, road BR 101, between Miaí de Cima and Piaçabuçu, restinga forest, Aug. 2001,
Zappi (obs.).
sergipe: Mun. Japoatã, Aug. 2001, Zappi (obs.).
sergipe/bahia: Mun. Cristinápolis to Mun. Jandaíra, surviving in fields derived from cut-over Atlantic
Forest along road BR 101, 2 & 4 Aug. 1998, Taylor (obs.).
bahia: NW Bahia, Mun. Santa Rita de Cássia, Monte Alegre, reported by Andrade-Lima (1975: 229);
E Bahia, Mun. Conceição da Feira, Porto Castro Alves, 12°32'S, 39°5'W, Dec. 1980, Iscardino et al. in
Grupo Pedra do Cavalo 1038 (ALCB, CEPEC); SE Bahia, Mun. Itambé, road BR 415 15 km E of town,
16 April 2003, Taylor & Zappi (obs.).
minas gerais: cent.-N Minas Gerais, Mun. Januária, [Pedras de] Maria da Cruz, Faz. São Francisco, 11
Aug. 1997, I. Pimenta (K, photos); Mun. Brasília de Minas, 11 km N of Lontra on road to Januária,
15°50'S, 44°20'W, 6 Nov. 1988, Taylor & Zappi in Harley 25512 (K, SPF); Mun. Porteirinha, 5 km N of
the town, 30 Jan. 1991, Taylor et al. 1500 (K, ZSS, HRCB, BHCB); ibid., 8 km S of town, 8 Nov. 1988,
Taylor & Zappi (obs.); NE Minas Gerais, Mun. Coronel Murta, 5 km SE of town on road BR 342, 21
Feb. 1988, Supthut 8862 (ZSS); Mun. Araçuaí, reported by Rizzini & Mattos-Filho in Rodriguésia
31(54): 64 (1980); Mun. Itinga, near road BR 116, c. 36 km N of Padre Paraíso, 13 Dec. 1990, Taylor
& Zappi (obs.); Mun. Diamantina, 25 km NE of town on road to Mendanha, by the Rio Jequitinhonha,
18°6'S, 43°28'W, 31 Oct. 1988, Taylor & Zappi in Harley 25485 (SPF, K), ibid., 25 km E, 17 Mar. 1970,
Irwin et al. 27786 (NY, K).
conservation status. Least Concern [LC] (1); PD=1, EI=1, GD=2. Short-list score (1×4) = 4. One
of the commonest cacti in Eastern Brazil and with various uses (see below). Sometimes regenerating from
stumps and recovering, or intentionally spared, when the forest it inhabits is cut down.
Very important as a source of cattle fodder during times of drought and sometimes
planted for hedging purposes. Formerly used as a source of timber, but most larger
specimens have gone. A widespread regional variant from western Rio Grande do Norte
and adjacent Paraíba differs from ‘typical’ plants in having more ribs, both as a seedling
and subsequently, as well as the stem being less obviously constricted, with ± parallel-
sided false seasonal growth segments. Local variations are common and include specimens
with up to 10 ribs and marked differences in number and length of spines. One of the
finest specimens seen stands in a garden behind the filling station (‘posto’) south of road
BA 052 on the west side of Morro do Chapéu, Bahia. Its height has been estimated at c.
18 metres, while its trunk is 2 × 0.8 m and bears a massive crown composed of 100s of
branches. It is obviously of considerable age and cannot have been planted, almost
certainly being a remnant of the caatinga de altitude vegetation surrounding the town.
Apparently very similar to subsp. jamacaru is a plant from Depto Florida (W Santa
Cruz), Bolivia, which Ritter (1980: 553) named Piptanthocereus colosseus. Braun & Esteves
Pereira (1995) refer this taxon to C. lamprospermus K. Schum., but its relationship deserves
further investigation.
5b. subsp. calcirupicola (F. Ritter) N. P. Taylor & Zappi in Cact. Consensus Initiatives
3: 7 (1997). Holotype: Brazil, Montes Claros, 1959, Ritter 1011 (U).
Piptanthocereus calcirupicola F. Ritter, Kakt. Südamer. 1: 234–235, Abb. 188 & 189 (1979). Cereus
calcirupicola (F. Ritter) Rizzini in Revista Brasil. Biol. 46: 782, figs 1–4 (1986); Zappi in Bol. Bot.
Univ. São Paulo 12: 46–48, figs 12–27 (1990, publ. 1991).
Piptanthocereus cipoensis F. Ritter, op. cit. 236, Abb. 191 (1979). Cereus calcirupicola var. cipoensis (F. Ritter) P. J.
Braun in Bradleya 6: 86 (1988). C. calcirupicola subsp. cipoensis (F. Ritter) P. J. Braun & E. Esteves Pereira
CEREUS
in Succulenta 74: 83 (1995). Holotype: Brazil, Minas Gerais, Serra do Cipó, 1964, Ritter 1411 (U).
Piptanthocereus cabralensis F. Ritter, op. cit. 235, Abb. 190 (1979). Cereus calcirupicola var. cabralensis (F.
Ritter) P. J. Braun in Bradleya 6: 86 (1988). Cereus calcirupicola subsp. cabralensis (F. Ritter) P. J. Braun
& E. Esteves Pereira in Succulenta 74: 83 (1995). Holotype: Brazil, Minas Gerais, Serra do Cabral,
1964, Ritter 1412 (U).
Piptanthocereus calcirupicola var. pluricostatus F. Ritter, l.c. (1979). Cereus calcirupicola var. pluricostatus (F. Ritter)
Rizzini in ibid. 784 (1986). Holotype: Brazil, Minas Gerais, Granjas Reunidas, 1964, Ritter 1011a (U).
C. calcirupicola var. albicans Rizzini, l.c. (1986). Type: Brazil, Minas Gerais, Cordisburgo, 15 July 1983,
Rizzini & Mattos s.n. (RB, not found Feb. 1989). Lectotype (designated here): Rizzini in ibid. 783,
fig. 2 (1986).
[C. coerulescens sensu Warm., Lagoa Santa: 150–151, figs 36 & 37 (1908), non Salm-Dyck (1834).]
Rio São Francisco (Rio das Velhas, MG) caatinga/mata seca element: on ± forest-covered limestone (Bambuí)
outcrops, where locally co-dominant with other arborescent cacti, more rarely on arenitic rock or sand (at
higher elevations only), amongst caatinga, cerradão, cerrado and rarely campo rupestre, c. 450–1200 m, western
Bahia to central Minas Gerais (to c. 19°40'S). Endemic to the core area of Eastern Brazil. Map 29.
bahia: W Bahia, Mun. Santana, 28 km S of town towards Santa Maria da Vitória, Faz. São Geraldo,
path to Gruta do Padre, 15 Jan. 1991, Taylor et al. (ZSS, K, photos); 7 km from Porto Novo coming from
Santana on dirt road, 16 Jan. 1991, Eggli (ZSS, photos); Mun. Cocos, W of town, reported (as ‘Cereus
sp.’) and illustrated by Braun & Esteves Pereira (1999b).
(photo); Mun. Januária, ‘distrito de Fabião, 2 km na estrada partindo do Abrigo do Malhador’,
15°7'16"–15°8'57"S, 44°15'20"–44°14'13''W, 26 Oct. 1998, J.A. Lombardi 2122 (K, BHCB); Mun.
Itacarambi/Manga, 15°26'S, 43°55'W, Sep. 1999, I. Ribeiro (photos); Mun. Varzelândia, edge of town, 23
July 1989, Taylor & Zappi (obs.), 12 Aug. 1988, Eggli s.n. (ZSS, photos); Mun. Capitão Enéas, limestone
outcrop visible from road BR 122, June 2002, M. Machado (obs.); Mun. Montes Claros, 12 km N of
town on road to Januária, 16°38'S, 43°55'W, Taylor & Zappi in Harley 25507 (SPF, K); Mun. Bocaiúva,
Engenheiro Dolabela, 0.5 km E of road BR 135, 17°28'S, 43°58'W, 4 Nov. 1988, Taylor & Zappi in
Harley 25502 (SPF, K); Mun. Buenópolis, Serra do Cabral, 7 km from town on dirt road to Barreira da
Lapa, 17°53'S, 44°15'W, 12 & 13 Oct. 1988, Taylor & Zappi in Harley 24910, 24992 (SPF, K); Mun.
Cordisburgo, near Gruta de Maquiné, see Eiten (1983: fig. 24); Mun. Paraopeba & Mun. Sete Lagoas,
reported by Rizzini (1986); Mun. Santana do Riacho, Serra do Cipó, Cardeal Mota, 8 May 1987, Zappi
et al. in CFSC 10124 (SPF), 27 Oct. 1988, Taylor & Zappi in Harley 25424 (SPF, K); Mun. Matozinhos,
Faz. Cauaia, 31 Oct. 1996, Lombardi 1463 (BHCB, K); Mun. Lagoa Santa, Lapa Vermelha, reported by
Warming (1908: fig. 37; cf. Warming & Ferri 1973).
conservation status. Least Concern [LC] (1); PD=1, EI=2, GD=2. Short-list score (1×5) = 5. Its
habitat is declining in some places due to quarrying of limestone.
This subspecies is distinguished from typical subsp. jamacaru by relatively minor yet
recognizable differences in juvenile stem-morphology and flower shape etc. It ranges to
the south-west of the region occupied by subsp. jamacaru and is at its eastern limit in the
upper drainage of the Rio São Francisco (Rio das Velhas). Some populations on isolated
limestone outcrops have developed into distinctive variants, such as that described as var.
pluricostatus by Ritter (see synonymy above), but these are matched by variations seen in
subsp. jamacaru and do not seem worthy of being named.
5c. Cereus sp. nov. (?)
Possibly keying out here to C. jamacaru (but mode of dehiscence of ripe fruits at present
unknown) is a recently discovered Bahian taxon requiring further study (Plates 28.3–29.1).
CEREUS
It is tall and treelike, with stems constantly 3-ribbed for the first 3–8 metres, the massive
trunk sometimes attaining 10 metres! The ribs are extremely thin, very broad and strongly
crenate and it has very stout flowers. It inhabits a high agreste forest (mata de cipó), including
characteristic Mata atlântica cacti (Pereskia aculeata, Lepismium cruciforme, 4 Rhipsalis spp.), as
follows: Mun. Vitória da Conquista, 20 km from town towards Itambé, 18 April 2003,
Taylor et al. (K, photos), l.c., campus of UESB, 900m, 18 April 2003, Taylor et al. (obs.);
Mun. Cândido Sales, 2002, M. Moreira (obs.) This remarkable plant, which is probably
related to C. trigonodendron K. Schum. ex Vaupel (syn. C. vargasianus Cárdenas; eastern
Peru), will be reported on further as soon as ripe fruits have been studied. These are
reported to be yellow by Avaldo de Oliviera Soares Filho.
6. Cereus hildmannianus K. Schum. in Martius, Fl. bras. 4(2): 202 (1890). Type:
Brazil, Rio de Janeiro (Schumann, l.c.), or Minas Gerais, Queluz [= Conselheiro Lafaiete]
(fide Glaziou 1909: 325), Glaziou s.n. (B†). Lectotype (designated here): Schumann, l.c.,
tab. 41, fig. I (1890) [depicting a 6-ribbed, spineless stem apex bearing a flower].
? C. alacriportanus Pfeiff., Enum. cact.: 87 (1837). Piptanthocereus alacriportanus (Pfeiff.) F. Ritter, Kakt.
Südamer. 1: 236–237 (1979). Type: Brazil, Rio Grande do Sul, Porto Alegre, assumed not to have
been preserved or lost.
Cereus milesimus E. C. Rost in Desert Pl. Life 4: 43 (1932). Type: Brazil, São Paulo, Ribeirão Preto, living
plant assumed not to have been preserved. Lectotype (designated here): Rost, l.c. 42, photo.
Piptanthocereus neonesioticus F. Ritter var. interior F. Ritter, Kakt. Südamer. 1: 237–238 (1979). Cereus
neonesioticus (F. Ritter) P. J. Braun var. interior (F. Ritter) P. J. Braun in Bradleya 6: 86 (1988).
Holotype: Brazil, São Paulo, Piedade, Ritter 1421 (U).
[C. jamacaru sensu Scheinvar, Fl. Ilus. Catarinense, I, Cact.: 96–103, tab. 45–47 (1985), non DC. (1828).]
[C. peruvianus auctt. pro parte, non (L.) Mill.]
conservation status. Not Evaluated [NE]; not studied in sufficient detail.
Only the following subspecies is found in the area covered here:
6a. subsp. hildmannianus

Very similar to C. jamacaru, but tree-like or shrubby, to 15 m or more tall, sometimes with a very well
developed trunk to > 2 m; stem tissues highly mucilaginous; ribs 5–12, very variable in number, to 50–70
× 10–30 mm, sometimes higher and thinner, juvenile plants 3–5-ribbed; flowers 10–14 cm diam.; fruit
5–12 × 7–12 cm, yellow, orange or reddish depending on exposure, usually splitting open from the apex
along c. 3 lines to fully expose the white funicular pulp (Plate 29.2); seeds c. 3 × 2.8 mm, testa with
cuticular folds (Barthlott & Hunt 2000: pl. 45.1–2).
Southern humid/subhumid forest element: in rocky places and on dry shallow soils in mata de planalto, c.
800–1000 m, southern and western Minas Gerais, from (?) Conselheiro Lafaiete southwards and
westwards, but frequently cultivated for ornament outside its natural range in Eastern Brazil; South-
eastern and Southern Brazil; Mato Grosso do Sul (?); central and south-eastern South America (E
Paraguay, Uruguay, NE & E Argentina etc., where replaced by the shorter-flowered subsp. uruguayanus
(R. Kiesling) N. P. Taylor). Map 29.
minas gerais: Mun. Conselheiro Lafaiete, fide Glaziou (1909: 325); S of Juiz de Fora near road BR 040,
25 July 1991, Zappi (obs.).
rio de janeiro: Mun. Tres Rios, W of road BR 040, c. 10 km S of border with Minais Gerais, 3 Aug.
2003, Taylor & Zappi (obs.); l.c., road to Petrópolis, 21 July 1996, Taylor (obs.); Mun. Niterói/Maricá,
CEREUS
Itacoatiara, on rocks above the sea, 28 July 1996, Taylor (obs.) — at both these sites sympatric with C.
fernambucensis subspp. [records outside of E Brazil area].
Concern, given its very extensive range in eastern South America.
This species has not been studied in the field within the core area treated here. However,
plants from Rio de Janeiro, south-western Minas Gerais, São Paulo and Paraná have been
examined in habitat, though seldom with ripe fruit. The fruits that have been observed
in the field and in cultivation generally display the characteristic mode of dehiscence
described above and, together with the highly mucilaginous stem tissues, provide an
effective means for distinguishing C. hildmannianus from C. fernambucensis subspp.
fernambucensis and sericifer, with whose ranges it slightly overlaps, and from the related but
allopatric C. jamacaru. All three species exist as spineless forms and when these are
cultivated outside their natural range their identification can be very difficult unless ripe
fruits are present. Unnecessary confusion between C. hildmannianus (C. peruvianus auctt.)
and C. jamacaru has been created by Leia Scheinvar (see synonymy above) and propagated
by her elsewhere, eg. Gutman et al. (2001: 709), but there can be little doubt that these
are two quite distinct and geographically separate, naturally occurring species.
C. hildmannianus may be widespread in drier phases of the semi-humid and humid,
subtropical and tropical planalto forests east of the Chaco in south-eastern South America,
and has a potentially much more extensive synonymy than that tentatively given above
(Hunt 1992b). However, this will remain uncertain until the Cereus species of Paraguay
and Bolivia are better understood.
Schumann (l.c.) stated that C. hildmannianus came from Rio de Janeiro, whereas
Glaziou (1909: 325), its collector, later gave a precise locality in southern Minas Gerais as
its provenance (see type citation above). There are only the above records of the plant
growing wild in the state of Rio de Janeiro, where it is much more commonly cultivated
(as elsewhere), but it is certainly native in Minas Gerais and is not uncommon in the
adjacent states of São Paulo and Paraná. The commonly cultivated form has ± spineless
stems, as does that depicted in Schumann’s plate, which is here selected as lectotype (in
the absence of any extant original herbarium material). This is in contrast to some wild
forms, which can have very spiny stems, and it is possible, therefore, that what Schumann
received from Glaziou and described was actually a cultivar from Rio de Janeiro and not
a wild plant. Braun & Esteves Pereira (2003) report it from Espírito Santo, but if it occurs
there at all, it is most likely only to be as a cultivated plant.
17. CIPOCEREUS F. Ritter
Kakt. Südamer. 1: 54 (1979); N. P. Taylor & Zappi in Bradleya 7: 16–17 (1989). Type:
C. pleurocarpus F. Ritter (= C. minensis (Werderm.) F. Ritter).
Including Floribunda F. Ritter (1979); Pilosocereus subg. Floribunda (F. Ritter) P. J. Braun
(1988); tantum quoad typ.
CIPOCEREUS
Shrubby or tree-like, 0.4–3.5 m, branched or not above the ground; vascular cylinder strongly to weakly
woody; tissues sometimes stiff, scarcely to extremely mucilaginous; branches showing slight constrictions
or ± constant growth, epidermis green, grey or pale waxy blue, smooth; ribs 4–18, acute to rounded,
sinuses straight. Areoles with felt and variable spination; spines straight, opaque to translucent. Fertile part
of stem not differentiated, random (lateral) and/or subapical. Flowers 1.5–9.0 × 1–7 cm, anthesis
commencing at evening or diurnal; flower exterior bluish, covered in wax, sometimes reddish, especially
on the distal half, but still waxy, smooth to ridged, glabrous or with areoles, and then hairs and minute
spines as well as areolar felt also present; pericarpel oblong, nearly cylindric; flower-tube straight, terete,
with some scattered acute bract-scales, more frequent towards its apex; perianth-segments reflexed, outer
segments fleshy, inner segments thin, white or cream-coloured; nectar-chamber straight, not protected
by the stamens (except in C. pusilliflorus); stamens many, anthers 1–3 mm, verrucose, forming a compact
mass; style 9–65 mm, stigma-lobes 8–10, exserted or included in relation to the anthers. Fruit ovoid to
globose, indehiscent; floral remnant persistent, blackening, erect, with rounded, deep insertion; pericarp
covered in blue wax, dark purplish, brownish, pinkish or green beneath; funicular pulp aquose,
translucent, greenish or yellowish. Seeds 1.2–1.8 mm, brown or dark brown-black, cochleariform,
hilum-micropylar region differentiated, intercellular depressions present, mainly forming craterae,
sometimes only slightly pronounced, testa-cells convex, conic or flat, cuticular folds coarse, very dense.
A genus of 5 very distinct species endemic to the Serra do Espinhaço and Serra do Cabral
of Minas Gerais (campo rupestre and sandy phases of the cerrado); see Maps 5 & 30. The
waxy and mostly light bluish, indehiscent, ovoid to globose fruits with translucent pulp
are characteristic. Species Nos 1 & 2 may be related on the basis of seed- and stem-
morphology (including seedlings); likewise Nos 3–5. However, natural hybridization
between Nos 2 and 4 has been observed (Taylor & Zappi 1989). Recent preliminary gene
sequence studies employing the chloroplast gene segment rpl16 (Soffiatti unpubl.) suggest
that Cipocereus is monophyletic and basal to the remainder of the Brazilian ‘Cereeae’
treated below (as well as to Uebelmannia). Plates 29.3–31.4.
1. Ribs 4–7, triangular in section; areoles with dense felt and long hairs; central spines larger
than radials . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
1. Ribs > 8, rounded in section; areoles with short, brown or white felt, long hairs absent;
central and radial spines alike or upper areoles unarmed . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
2. Roots not tuberous; flowers and fruits bearing areoles with spines and long hairs (Serra
do Caraça, MG) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1. laniflorus
2. Roots tuberous (Dahlia-like); flowers and fruits naked (E & N of Diamantina, MG)
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2. crassisepalus
3. Stems 5–8 cm diam., unarmed or bearing very few black spines, transverse folds between
adjacent areoles on the same rib well marked, epidermis sky blue, waxy (Serra do Cabral
& adjacent W slope of Serra do Espinhaço, MG) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3. bradei
3. Stems 2.5–5.0 cm diam., densely spiny; transverse folds absent; epidermis grey-green or
bright green . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
4. Plants to c. 50 cm tall; flowers small, < 2 cm, diurnal; fruits to 1.3 cm diam., pinkish, with
a translucent waxy bloom (W slope of Serra Geral, northern Minas Gerais) . . . . . 5. pusilliflorus
4. Plants > 50 cm tall; flowers > 4 cm, nocturnal; fruits > 2 cm diam., dark blue-black
covered in a pale blue waxy bloom, or whitish (MG: Serra da Bocaina, Grão Mogol
southwards) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4. minensis
1. Cipocereus laniflorus N. P. Taylor & Zappi in Cact. Consensus Initiatives 3: 7

(1997). Holotype: Brazil, Minas Gerais, Serra do Caraça, 11 Sep. 1990, Zappi et al. 240
(SPF; HRCB, BHCB, isos.).
CIPOCEREUS
Pilosocereus laniflorus (N.P. Taylor & Zappi) P. J. Braun & E. Esteves Pereira in Schumannia 3: 188 (2002).
Erect, 0.5–1.5(–2) m, tapered to base, unbranched or shrublike, rarely with 25 or more actively growing
stem apices; vascular cylinder strongly woody; inner tissues mucilaginous; branches 6–9 cm diam.,
epidermis olive-green, with strikingly pruinose, bluish wax on the young growth; ribs 5–7, 2 × 2 cm,
triangular in section. Areoles to 6 mm diam., ± contiguous, felt at first pale yellow to light-brown,
becoming grey to blackish, long hairs whitish to yellowish, very dense at apex. Spines opaque, dark
brown to reddish; central spines 7–9, 1.5–2.0(–3.0) cm, porrect; radials 10–14, to 1 cm, adpressed. Fertile
part of stem lateral to subapical, not differentiated, but fertile areoles producing dense hairs to 1 cm in
some individuals. Flower-buds woolly, acute 1–2 days before anthesis; flower (May–August) c. 7 × 3.5
cm, exterior dark blue, ridged, with areoles, these bearing 3–5 minute reddish spines and long white hairs;
pericarpel 18–20 × 12–15 mm; flower-tube 45 × 24 mm at its widest point, ± parallel-sided, straight, with
acute bract-scales at apex; outer perianth-segments 12–14 × 7 mm, dark blue to purplish cream, inner
segments 18 × 8 mm, white or pale creamy yellow, spathulate, apiculate; nectar-chamber c. 20 × 6–7
mm; anthers 2–2.3 mm; style 40 × 1.6 mm, tapering; stigma-lobes c. 8, 6 mm, acute at apex; ovary locule
c. 7 × 6 mm, quadrangular in longitudinal section. Fruit (July–Sep.) to 3.5–4.0 × 3.5–4.0 cm, ovoid, up
to 20 per plant; pericarp purplish-blue, strongly 7–8-ribbed, with woolly areoles and spines, the latter
growing somewhat as the fruit develops. Seeds (1.5–)1.6–1.7 × 1–1.1 mm, blackish, hilum-micropylar
region 0.6(–0.7) mm, forming an angle of 30° with long-axis; testa-cells domed, with intercellular
depressions and coarse, dense cuticular folds.
South-eastern campo rupestre element: deeply fissured quartzitic outcrops on steep hill slopes and cliffs in
campo rupestre, in constant association with Philodendron cipoense, 1000–1500 m, Serra do Caraça, south-
central Minas Gerais. Endemic to the above-cited area within Minas Gerais. Map 30.
minas gerais: Mun. Santa Bárbara, Reserva Particular de Patrimônio Natural (RPPN) do Caraça, steep
hill slopes below the Oratório and above the track to Buraco da Boiada, side valley west of the Ribeirão
Caraça and the Cascatona, 1000–1200 m, 3 Aug. 2001, Taylor (K, photos); l.c., Campo de Fora, rocks
and cliffs above the Córrego do Eugênio/Capivari, 2 Aug. 2001, Taylor (K, photos); Mun. Catas Altas,
RPPN do Caraça, Gruta do Padre Caio, 24 May 1987, Zappi & Scatena in CFCR 10987 (SPF, ZSS), 11
Sep. 1990, Zappi et al. 240 (SPF, HRCB, HBCB), ibid., cult. R.B.G. Kew, Accn. No. 1991-1422, 17 &
25 Feb. 1997 (K, fls in spirit, photos), l.c. Aug. 1996, Soffiatti et al. (K, photos), 1 Aug. 2001, Taylor (K,
photos); l.c., 17–20 minutes’ walk on the path from the Caraça Santuário towards the Grande Funil and
Cascatona, steep hill slope east of path, 3 Aug. 2001, Taylor (K, photos); l.c., between Bocaina and Gruta
da Bocaina, 1 Aug. 2001, Taylor (K, photos).
conservation status. Endangered [EN D] (3); extent of occurrence < 10 km2 and area of occupancy
probably not more than 1% of this (0.1 km2); PD=3, EI=1, GD=1. Short-list score (3×5) = 15. Presently
known from 8 small areas situated sufficiently close together to be considered a single population, which
is entirely within the Reserva Particular de Patrimônio Natural do Caraça. Field surveys during 2001 (by
Taylor and staff from the Fundação Zoo-Botânica de Belo Horizonte) and throughout 2002 (by Juliana
Rego, FZ-BBH, supported by the RBG Kew Threatened Plants Appeal), have so far documented a total
of 321 individuals, less than 250 of these being adult plants. Knowledge of the whole area suggests it is
unlikely that further populations exist and the Caraça massif itself is isolated from other parts of the Serra
do Espinhaço range by areas of lower elevation, formerly covered in high forest. Two of the small areas
have 80 or more individuals, while the remainder have smaller numbers (2–38). Only a handful of plants
is actually visible from the footpaths that traverse the Caraça RPPN and entry/exit to and from the
reserve is controlled through a single manned gate on the only paved access road. Collecting on the
reserve is strictly by permission from its director, which is not freely given, and there are signs advising
the public that such activity is prohibited. Thus, the existing population, though small, appears to be safe
from human interference at present.
In its flowers and fruits bearing well-developed, woolly and spiny areoles this species seems
to be a relict, which has the most plesiomorphic floral characters within tribe Cereeae, only
CIPOCEREUS
Cipocereus minensis subsp. minensis and Cereus subg. Mirabella having comparably primitive
floral features. The strongly glaucous young growth and approximate, darkly spined areoles
of this species are strongly convergent with those of Pilosocereus fulvilanatus, q.v.
The plant is most abundant at lower elevations, below 1400 metres. Above this
elevation the plants are few and appear to be limited by winter frosts. Plants at these
elevations occur only under the protection of rocks or bushes and exposed stems are
invariably damaged and re-sprout from below.
2. Cipocereus crassisepalus (Buining & Brederoo) Zappi & N. P. Taylor in Bradleya 9:

86 (1991). Holotype: Brazil, Minas Gerais, Diamantina, 500–1000 m, Horst & Uebelmann
169 (U).
Cereus crassisepalus Buining & Brederoo in Krainz, Die Kakteen, Lfg 53 (1973). Piptanthocereus crassisepalus
(Buining & Brederoo) F. Ritter, Kakt. Südamer. 1: 238 (1979).
Erect shrub or treelet, 1–2 m, branched above the ground; root system well developed, roots tuberous as
in a Dahlia; cauline vascular cylinder rather woody; succulent tissues mucilaginous. Young plants slender,
stems with 5–8 low ribs, the areoles with dense, white or greyish, long hairs and reddish spines. Mature
stems 5–6 cm diam., epidermis olive-green, rather glaucous when young; ribs 4–6, 1.5–2.0 × 2.5–3.0 cm,
triangular in section. Areoles to 15 mm diam., 10–30 mm apart, sometimes almost contiguous, with dense
felt and long hairs, at first golden red, later blackish. Spines opaque, yellowish, becoming dark brown or
black; central spines 1–2(–3), to 4 cm, porrect; radials 3–8, to 2 mm, adpressed. Fertile part of stem not
differentiated, subapical. Flower-buds obtuse before anthesis; flowers c. 7.5–9 × 4–5 cm, dull, dark blue,
ridged, glabrous except for scattered apiculate bract-scales; pericarpel 15–20 × 16 mm; tube 50 × 30–40
mm at its widest point, ± parallel-sided, straight, with ovate, acute bract-scales at apex; outer perianth-
segments 14 × 9–11 mm, dark blue to purplish-cream, inner segments 18 × 10 mm, white or cream,
spathulate, apiculate; nectar-chamber c. 20 × 6–7 mm; anthers 2–3 mm; style 45–50 mm, tapering;
stigma-lobes c. 9, c. 10 mm; ovary locule c. 13 × 8–9 mm, rectangular in longitudinal section. Fruit to
4.5 × 3.0 cm, narrowly oblong to ovoid; pericarp dark blue, smooth, naked. Seeds 1.6–1.7 × 1.0–1.2 mm,
blackish, hilum-micropylar region c. 0.5 mm, forming an angle of 35–40° with long-axis; testa-cells
domed, with intercellular depressions and coarse, dense cuticular folds.
South-eastern campo rupestre (cerrado) element: in sandy cerrado/carrasco associated with crystalline rock
outcrops, 500–1200 m, north of Diamantina, Serra Negra and east side of Serra do Espinhaço, Minas
Gerais. Endemic to the core area within Minas Gerais. Map 30.
minas gerais: Mun. Itamarandiba, c. 20 km SW of airport, towards Penha de França, 3 Aug. 1988, Eggli
1061 (ZSS); Mun. Diamantina, Horst & Uebelmann 169 (U); l.c., 25 km along road to Mendanha, E of road
towards Rio Jequitinhonha, 31 Oct. 1988, Taylor & Zappi in Harley 25487 (SPF, K) — population including
hybrids with sympatric Cipocereus minensis; l.c., between Mendanha and Inhaí, 17°59'44"S, 43°36'4''W, M.
Machado (obs.); Mun. Rio Vermelho, Serra do Ambrósio, 1959, Ritter s.n. [1012] (U), 14 July 1984, Varanda
et al. in CFCR 4885 (SP, SPF), 6 Mar. 1988, Zappi & Prado in CFCR 11822 (SPF). Some of the location
data used in Map 30 were supplied by M. Machado (pers. comm.), for localities where this species is
sympatric with Uebelmannia, but the coordinates are not reproduced here for the protection of the latter.
conservation status. Vulnerable [VU B1ab(iii) + 2ab(iii)] (2); extent of occurrence = 1407 km2;
PD=3, EI=1, GD=1. Short-list score (2×5) = 10. Habitat destruction by charcoal producers continues to
be a problem for this species comprising < 10 populations.
As noted above, this species hybridizes with C. minensis when they come into contact.
CIPOCEREUS
3. Cipocereus bradei (Backeb. & Voll) Zappi & N. P. Taylor in Bradleya 9: 86 (1991).
Type: Brazil, Minas Gerais, Diamantina, Brade s.n. (apparently not preserved). Lectotype
(designated here): Backeberg & Voll, l.c. infra, [p.3], photograph (1935).
Pilocereus bradei Backeb. & Voll in Jahrb. Deutsch. Kakteen-Ges. 1941: 78 (1942); Backeb. & Voll in Blätt.
Kakteenf. 1935 (1:) [p. 3] (1935), nom. inval. (Art. 36). Pilosocereus bradei (Backeb. & Voll) Byles & G.
D. Rowley in Cact. Succ. J. Gr. Brit. 19(3): 66 (1957). Cephalocereus bradei (Backeb. & Voll) Borg,
Cacti, ed. 2: 147 (1951); E. Y. Dawson in Los Angeles County Mus. Contr. Sci. 10: 6 (1957).
Pseudopilocereus bradei (Backeb. & Voll) Buxb. in Beitr. Biol. Pflanzen 44: 252 (1968).
vernacular name. Quiabo-da-lapa.
Erect shrub or treelike, 1–3.5 m, branched above the ground; vascular cylinder weakly woody; succulent
tissues mucilaginous. Seedlings and young plants with 10 low ribs and areoles with dense reddish
spination. Mature stems 5–8 cm diam., sometimes slightly constricted, epidermis covered in thick,
striking, light blue wax; ribs 8–11, 0.5 × 1.2–1.7 cm, rounded in section, produced below the areoles.
Areoles to 3 mm diam., 20–23 mm apart, felt short, at first white, the areoles frequently unarmed,
becoming blackish in age. Spines on young areoles 0 or few, if present, black or dark reddish brown, to
15 mm; old areoles with 5–7 spines, to 2.5 cm, porrect. Fertile part of stem not differentiated, flowers
numerous, often from one side of stem and/or subapical region. Flower-buds obtuse before anthesis;
flowers c. 6.5–8.0 × 3.5–4.0 cm, exterior dark blue, ridged, glabrous except for scattered bract-scales;
pericarpel 15 × 15 mm; tube 50 × 20–25 mm at its widest point, ± parallel-sided, straight, with ovoid to
spathulate, apiculate bract-scales at apex; outer perianth-segments 9–12 × 6–8 mm, dark blue to purplish
cream, inner segments 18 × 8 mm, white or cream, spathulate, fimbriate and apiculate; nectar-chamber
c. 20–23 × 8–10 mm; anthers 2.2–3.0 mm; style 45–48 mm, tapering; stigma-lobes c. 10, c. 9 mm; ovary
locule c. 9 × 7 mm, rectangular in longitudinal section. Fruit to 4.5 × 3.5 cm, ovoid; pericarp dark-blue,
smooth, naked, but covered in dense, sky blue wax. Seeds 1.6–1.8 × 1.1–1.2 mm, brownish, hilum-
micropylar region 0.7 mm, forming an angle of 35–40° with long-axis; testa-cells conic, with intercellular
South-eastern campo rupestre element: crystalline rocks in campo rupestre, carrasco or cerrado, 500–1200 m,
Serra do Cabral and west slope of Serra do Espinhaço, Minas Gerais. Endemic. Map 30.
minas gerais: Mun. Francisco Dumont, 2 km from the town, 7 Aug. 1988, Eggli 1117 (ZSS, SPF);
Mun. Joaquim Felício, road to Várzea da Palma, 7 May 1990, Zappi et al. in CFCR 13235 (SPF, HRCB);
Mun. Buenópolis, above and west of the town, 12 Oct. 1988, Taylor & Zappi in Harley 24508, 24990
(SPF, K); Mun. Augusto de Lima, Santa Barbara, 6 Aug. 1988, Eggli s.n. (ZSS, photo); Mun. Diamantina,
N of Sopa, c. 18°10'S, 43°44'W, M. Machado (obs.); Mun. Conselheiro Mata, 18°17'S, 43°58'W, 8 June
2002, G. Charles & M. Machado (photos).
conservation status. Endangered [EN B1ab(iii) + 2ab(iii)] (3); extent of occurrence = 1887 km2;
area of occupancy < 500 km2; PD=3, EI=1, GD=1. Short-list score (3×5) = 15. Of very limited range,
comprising 6 somewhat fragmented localities, none included within any protected area. Subject to
commercial seed harvest for the international trade and likely to suffer a decline in habitat quality in the
future due to local extractive activities.
At the second locality cited above C. bradei was found growing with C. minensis and there
was evidence of introgression.
4. Cipocereus minensis (Werderm.) F. Ritter, Kakt. Südamer. 1: 57 (1979). Type:

Minas Gerais, June 1932, Werdermann 3992 (B†). Lectotype (designated here):
Werdermann, l.c. infra, 112, photograph captioned ‘Cereus minensis Werderm.’ (1933).
CIPOCEREUS
Cereus minensis Werderm., Bras. Säulenkakt.: 113 (1933). Pilocereus minensis (Werderm.) Backeb. in
Backeberg & Knuth, Kaktus-ABC: 328 (1935, publ. 1936). Pilosocereus minensis (Werderm.) Byles &
G. D. Rowley in Cact. Succ. J. Gr. Brit. 19(3): 67 (1957). Cephalocereus minensis (Werderm.) E. Y.
Dawson in Los Angeles County Mus. Contr. Sci. 10: 6 (1957). Coleocephalocereus minensis (Werderm.)
F. H. Brandt in Succulenta 60: 117–118 (1981).
Cipocereus pleurocarpus F. Ritter, Kakt. Südamer. 1: 54–55 (1979). Coleocephalocereus pleurocarpus (F. Ritter)
F. H. Brandt in Succulenta 60: 117 (1981). Pilosocereus pleurocarpus (Ritter) P. J. Braun in Bradleya 6:
88 (1988). Cipocereus minensis subsp. pleurocarpus (F. Ritter) N.P. Taylor & Zappi in Bol. Bot. Univ.
São Paulo 12: 48 (1990, publ. 1991). Holotype: Brazil, Minas Gerais, Serra do Cipó, Nov. 1964,
Ritter 1327 (U).
vernacular names. Rabo-de-raposa, Quiabo-da-lapa, Quiabo-do-inferno.
Erect shrub to 2 m, branched above the ground, branches sometimes decumbent; vascular cylinder
woody; inner tissues stiff, not mucilaginous. Stems 3–5 cm diam., sometimes constricted, epidermis olive-
green; ribs 12–16, 4–5 × 5 mm, rounded in section. Areoles to 2–3 mm diam., 7–8 mm apart, felt short,
at first brown, becoming greyish and glabrescent. Spines golden brown to dark reddish brown; central
spines 4–6, 5–40 mm, porrect; radials 7–10, 5–10 mm. Fertile part of stem not differentiated, flowers
numerous, randomly from lateral and subapical regions. Flower-buds obtuse before anthesis; flowers c.
4.5–7.5 × 2.5–6.5 cm, ridged or quite terete, glabrous or with spiny areoles; pericarpel 10–18 × 15–18
mm; flower-tube to 50 × 15–30 mm at its widest point, ± parallel-sided, straight, with ovoid, apiculate
bract-scales at apex; outer perianth-segments 8–10 × 4–5 mm, inner segments 9 × 8 mm, white or cream,
spathulate, fimbriate and apiculate; nectar-chamber c. 15–20 × 8 mm; anthers c. 2.5 mm; style 45–50 mm,
tapering; stigma-lobes c. 8, c. 7 mm; ovary locule c. 6–8 × 7–8 mm, rectangular or quadrangular in
longitudinal section. Fruit ovoid. Seeds 1.5–2.0 × 1.1–1.2 mm, brownish, hilum-micropylar region
0.6–0.8 mm, forming an angle of 30–40° with long-axis; testa-cells domed to conic, with intercellular
The precise application of the name Cereus minensis Werderm. (1933, 1942) is here
changed from that established by Ritter (1979) when describing Cipocereus pleurocarpus.
Unfortunately, Werdermann did not see the plant in fruit, nor did he indicate a type
locality, and the holotype itself is no longer extant. In these circumstances it seems Ritter
may have assumed that the plants Werdermann described and illustrated (sterile) were
referable to the more widespread smooth-fruited race of the species, since Werdermann
indicates that the plant was ‘scattered throughout Minas’ (Werdermann 1942: 62).
However, the only localities actually mentioned were the Serra do Cipó and Serra do
Caraça, the former subsequently being the origin of Ritter’s new species, the latter
apparently the source of the ‘partly dried floral remnants’ described in Werdermann’s
protologue (compare l.c.: 61, 93). Curiously, the reference to Caraça seems to be
inaccurate, because this well preserved mountain habitat has been extensively botanized in
recent years and does not appear to hold a population of C. minensis (rather it is the home
of the endemic and very different C. laniflorus — certainly not that illustrated by
Werdermann, l.c.: 62). However, in 2001, a population of C. minensis came to light only
some 40 km north of the Serra do Caraça, near Cocais, which is close to or on the
historical routes connecting Diamantina, Belo Horizonte and Caraça, key places visited
and mentioned by Werdermann. Furthermore, this plant, which was recently encountered
by Taylor and staff from the Fundação Zoo-Botânica de Belo Horizonte, occurs on north-
CIPOCEREUS
facing cliffs, as described by Werdermann (l.c.: 61), and closely resembles that illustrated
by him (l.c.: 62). Besides this, his description of the floral remnants refers to the flower-
tube as ‘bearing together with the ovary sparse minute scales, these with a tufted lanose
pubescence in the axils’ (l.c.: 93). This feature is only rarely present in the more widespread
race that Ritter identified with C. minensis, but is characteristic of and constant for the
plant found at the Serra do Cipó and Cocais. Therefore, reluctantly, we have concluded
that the names C. minensis and C. pleurocarpus refer to one and the same race of the species,
leaving the more wide-ranging northern subspecies without a name. This is rectified here.
The two subspecies recognized are distinguished as follows:
1. Flower 7.5 × 6.5 cm or larger; ripe fruit not ribbed, smooth, blue (Grão Mogol to E edge
of Serra do Cabral and region of Diamantina) . . . . . . . . . . . . . . . . . . . . . . 4a. subsp. leiocarpus
1. Flower 4.5–6.0 × 2.5–4.5 cm; fruit ribbed, with a few spine-bearing areoles, brownish,
pale green, whitish or bluish (Serra do Cipó southwards) . . . . . . . . . . . . . . . 4b. subsp. minensis
4a. subsp. leiocarpus N. P. Taylor & Zappi subsp. nov. differt a subsp. minensis floribus
maioribus, pericarpio fructus laevi. Holotype: Minas Gerais, Mun. Grão Mogol, Serra da
Bocaina, 48 km W of bridge over Rio Vacaria, 28.5 km E of Caveira on road BR 251,
31 Jan. 1991, Taylor et al. 1506 (HRCB; K, ZSS, BHCB, isotypes).
[C. minensis sensu F. Ritter, Kakt. Südamer. 1: 56 (1979). C. minensis subsp. minensis sensu Taylor & Zappi
in D. R. Hunt, CITES Cactaceae Checklist, ed. 2: 170 (1999).]
Spines generally golden brown (at least in northern part of range), central spines not exceeding 15 mm.
Flowers to 65 mm diam., blue without; tube smooth, glabrous; outer perianth-segments dark blue or
purplish, patent at anthesis. Fruit not ridged, smooth, bright to deep blue-waxy, only rarely bearing areoles.
Widespread South-eastern campo rupestre element: mostly amongst crystalline rocks in campo rupestre,
500–2000 m, Serra do Espinhaço and eastern part of Serra do Cabral, Minas Gerais. Endemic. Map 30.
minas gerais: Mun. Grão Mogol, Serra da Bocaina, 48 km W of bridge over Rio Vacaria, 28.5 km
E of Caveira on road BR 251, 31 Jan. 1991, Taylor et al. 1506 (K, HRCB, ZSS, BHCB); ibid., near the
town, 4 Sep. 1985, Zappi et al. in CFCR 8356 (SPF), 5 Nov. 1986, Cordeiro & Mello-Silva in CFCR
10140 (SPF), 27 May 1988, Zappi et al. in CFCR 11971 (SPF), 15 Oct. 1988, Taylor & Zappi in Harley
25070, 25136 (SPF, K); Mun. Bocaiúva, E of Engenheiro Dolabela, 7 Aug. 1988, Eggli 1115 (ZSS, SPF);
Mun. Joaquim Felício, road to Várzea da Palma, Serra do Cabral, 7 May 1990, Zappi et al. in CFCR
13236, fl. (SPF, HRCB); Mun. Couto de Magalhães de Minas, 3 Aug. 1988, Eggli 1055 (ZSS); Mun.
São Gonçalo do Rio Preto, Parque Estadual do Rio Preto, Ribeirão das Éguas, 18°9'S, 43°22'W, 18
Nov. 1999, Lombardi 3566 (BHCB, K); Mun. Diamantina, 12 Jan. 1963, Duarte 7902 (RB), 29 May
1979, V.F. Ferreira 875 (RB); l.c., 5 km from Biri-Biri, 5 June 1985, H.F. Leitão Filho et al. 17603
(UEC); road to Curvelo 4.5 km from Diamantina, 4 Aug. 1990, Sakuragui & Souza 183 (ESA, K); road
to Conselheiro Mata, June 1934, Brade s.n. (RB 35757), 3 July 1985, Pirani et al. in CFCR 7958 (SPF),
29 Jan. 1986, Zappi et al. in CFCR 9425 (SPF), 10 Sep. 1986, T.B. Cavalcanti et al. in CFCR 10305
(SPF), 25 Feb. 1987, Zappi et al. in CFCR 10394 (SPF), 18 July 1987, Zappi et al. in CFCR 11129
(SPF), 19 July 1987, Zappi et al. in CFCR 11293 (SPF); road to Mendanha, 31 Oct. 1988, Taylor &
Zappi in Harley 25483, 25489 (SPF, K); Mun. Rio Vermelho, c. 9 km SSE from Pedra Menina, (?) Rio
Mundo Velho, 5 Aug. 1988, Eggli 1087 (ZSS); Mun. Gouvea, Ritter 1013 (SGO 125024, fide Eggli et
al. 1995: 445); Mun. Datas, Córrego Água Limpa, 23 May 1989, Hatschbach et al. 53108 (MBM); Mun.
Serro, Milho Verde, 2 May 1995, T. Grandi (K, photo); Mun. Santo Antônio do Itambé, Pico de
Itambé, c. 2000 m, 16 Aug. 1987, Zappi et al. in CFCR 11258 (SPF); Mun. Presidente Kubitschek,
Trinta Réis, 19 Apr. 1987, Zappi et al. in CFCR 10667 (SPF).
CIPOCEREUS
Modification of its habitat is relatively slow at present, but needs monitoring.
Hybridizes with C. crassisepalus and C. bradei, q.v., where they come into contact.
4b. subsp. minensis

Spines generally reddish brown to blackish, central spines to 40 mm. Flowers with distal half brownish,
reddish or yellowish without; tube with ridges and areoles with minute spines; perianth-segments
brownish or bright yellow, erect at anthesis. Fruit ribbed, with a few spiny areoles, brownish, pale green,
whitish or bluish. Seed: Barthlott & Hunt (2000: pl. 47.5–6).
South-eastern campo rupestre element: on cliffs and between rocks at 900–1300 m, Serra do Cipó and
Lapinha, southwards to Cocais and Itabirito, central-southern Minas Gerais. Endemic to the above area.
Map 30.
minas gerais: Mun. Santana do Riacho, Serra do Cipó, 7 Dec. 1949, Duarte 2398, and 16 Sep. 1950,
Duarte 3261 (RB), Nov. 1964, Ritter 1327 (U), 5 July 1972, Hatschbach 29886 (MBM), 23 July 1980,
Menezes et al. in CFSC 6390 (SPF), 30 June 1981, Giulietti et al. in CFSC 7358 (SPF), 17 Feb. 1982,
Mantovani et al. in CFSC 7786 (SPF), 25 Jan. 1988, Martinelli et al. 11348 (RB, SPF), 2 Nov. 1986,
Zappi & Kameyama in CFSC 9867 (SPF), 30 Aug. 1988, Kameyama et al. in CFSC 11206 (SPF), 27
Oct. 1988, Taylor & Zappi in Harley 25409, 25413 (K, SPF), 18 Nov. 1989, Zappi et al. 190, 191
(HRCB, SPF); l.c., Juquinha, May 1997, Soffiatti et al. (K, photos), l.c., near Cachoeira Capivara,
Soffiatti et al. (K, photo.); Mun. Jaboticatubas, Lapinha [on road to Santana do Riacho], 9 Sep. 1990,
M.H.L. Arndt & A.M. Silva s.n. (K, BHCB); Mun. Cocais, W of town, on the road near to and at
Pedra Pintada, 4 Aug. 2001, J.O. Rego et al. 850 (BHCB, Herb. Fund. Zoo-Bot. de Belo Horizonte 647
& fls/frs in alcohol [Carpoteca 761], K, photos), l.c. (?), ‘N side on the way up Mount Caraça’, 1932,
cf. Werdermann (1942: 61, photo p. 62); Mun. Itabirito, Pico de Itabirito, 9 Aug. 1993, W. Antunes
Teixeira s.n. (K, BHCB).
conservation status. Endangered [EN B2ab(iii)] (3); area of occupancy < 500 km2; PD=3, EI=1,
GD=2. Short-list score (3×6) = 18. Endangered from excessive burning of its habitat (2 populations
including one within the Parque Nacional da Serra do Cipó) and by mining operations (1 population);
an extensive part of the habitat at Cocais is private land and inaccessible sheer cliff.
Flower and fruit colour is quite variable in this taxon, the more colourful flower forms
suggesting that the plants are evolving towards hummingbird pollination.
5. Cipocereus pusilliflorus (F. Ritter) Zappi & N. P. Taylor in Bradleya 9: 86 (1991).

Holotype: Brazil, Minas Gerais, Monte Azul, Jan. 1964, Ritter 1232 (U).
Floribunda pusilliflora F. Ritter, Kakt. Südamer. 1: 58–60 (1979). Pilosocereus pusilliflorus (F. Ritter) P. J.
Braun in Bradleya 6: 90 (1988) and in Kakt. and. Sukk. 42(10): Karteiblatt 29 (1991).
Erect shrub to 50 cm, branched at ground level, branches semi-erect to decumbent; vascular cylinder
weakly woody; inner tissues mucilaginous, turning orange when cut. Stems 4–5 cm diam., not
constricted, epidermis bright green; ribs 13–18, 2–3 × 3–9.5 mm, rounded in section, sinuses straight.
Areoles to 2.5 mm diam., to 7 mm apart, felt short, brown, becoming glabrescent. Spines golden brown
to dark reddish brown; central spines 2–4, 1–2 cm, porrect; radials 10–12, 3–6 mm. Fertile part of stem
not differentiated, flowers near or below stem apex at one side, buds sometimes developing on the
underside of horizontal stems. Flowers c. 1.6 × 1.0 cm, glabrous, ridged; pericarpel 4 × 10 mm, purplish
outside; tube parallel-sided, straight, reddish pink, with small bract-scales at apex; perianth-segments
erect, whitish, inner segments acute; nectar-chamber c. 2 × 4 mm, closed at apex by the innermost layer
CIPOCEREUS
of filaments; style 9 mm, tapering. Fruit globose, 6–13 mm diam., reddish, with a blue-waxy bloom.
Seeds (1.3–)1.4–1.5 × 1.1 mm, brownish, hilum-micropylar region 0.6 mm, forming an angle of 20–30°
with long-axis of seed; testa-cells flat, with slight intercellular depressions and coarse, dense, marginal
cuticular folds (Barthlott & Hunt 2000: pl. 47.7–8).
Northern campo rupestre element: on cliffs and ledges of crystalline rocks, 800–1000 m, west slope of
Serra Geral (northern Serra do Espinhaço), east of Monte Azul, central-northern Minas Gerais. Endemic.
Map 30.
minas gerais: Mun. Monte Azul, Jan. 1964, Ritter 1232 (U), Horst & Uebelmann HU 400 (U), 29 Jan.
1991, Taylor et al. 1486 (K, ZSS, HRCB, BHCB). A further report by Uebelmann from Santa Barbara
(Mun. Augusto de Lima), under the number HU 840, cannot be confirmed with certainty and is probably
to be discounted (U. Eggli, pers. comm.).
conservation status. Critically Endangered [CR D] (4); PD=3, EI=1, GD=1. Short-list score (4×5)
= 20. Critically Endangered on present knowledge (< 5 individuals seen at the only certain locality) and
urgently in need of further field studies to determine if its range is more extensive. Maintained in a few
ex situ collections in Europe, including that at the Royal Botanic Gardens, Kew.
In its stem areoles lacking long trichomes and in its seed-micromorphology this species is
clearly allied with C. minensis and C. bradei and certainly not directly related to Arrojadoa
bahiensis (syn. Floribunda bahiensis), q.v., with which it is convergent in its floral
hummingbird syndrome and habitat.
18. STEPHANOCEREUS A. Berger
Entwicklungslin. Kakt.: 59, 97 (1926). Type: Stephanocereus leucostele (Gürke) A. Berger

(Cereus leucostele Gürke).
Including (?) Coleocephalocereus subg. Lagenopsis Buxb. (1972); Stephanocereus subg. Lagenopsis
(Buxb.) N. P. Taylor & Eggli (1991).
Columnar-segmented and ± branched, or bottle-shaped and solitary, 0.5–3.5(–6.0) m high. Juvenile

plants globose and without long-hairy stem areoles at first. Stems erect, 3–15(–20) cm diam., constricted
or not, vascular cylinder very woody; tissues highly mucilaginous; epidermis dull to grey-green; ribs
12–20, sinuses straight. Areoles with felt and abundant, white, long hairs; spines stiff, golden to greyish,
central spines porrect to deflexed, longer than the radials; indeterminate growth of basal areoles present.
Flower-bearing region of stem differentiated into a slender, elongate chlorophyllous cephalium or a
terminal cephalium tuft, this becoming transformed (through continued vegetative growth) into a lateral,
ring-like flowering region (cf. Arrojadoa spp.) with long bristles, spines and hairs. Flowers nocturnal, white
within, 3.5–10.0 × 2.5–5.5 cm; pericarpel and tube ± naked, pericarpel smooth, subglobose to
obpyramidal; tube almost terete, with broad, fleshy scales near apex; perianth-segments triangular to
lanceolate or spathulate, spreading to reflexed, innermost white to slightly pinkish; innermost filaments
curved towards the style. Fruit subglobose to ovoid-ellipsoid, impressed or not at apex, (at least in subg.
Stephanocereus) opening by a basal pore upon its detachment from the plant, naked, flower remnant
blackened, pendent to erect, deeply or shallowly inserted in fruit apex; pericarp smooth, greenish or dark
purple to blackish, with a blue waxy coating (as in Cipocereus), funicular pulp solid. Seeds 1.2–2.2 mm,
cochleariform; testa-cells convex; cuticular folds present.
A genus endemic to Bahia, comprising two monotypic subgenera, Stephanocereus and

Lagenopsis (Buxb.) N. P. Taylor & Eggli, the first characteristic of the Bahian caatinga, the
STEPHANOCEREUS
second of the campo rupestre of the Chapada Diamantina. Recent preliminary chloroplast
gene sequence studies (Soffiatti unpubl.) have cast doubt on the circumscription adopted
here, each of the species having more in common with different elements within
Arrojadoa. As currently circumscribed Stephanocereus differs from the latter in having a
globose juvenile phase and larger, strongly smelling flowers adapted for nocturnal
pollination by bats, rather than diurnal/crepuscular for hummingbirds (Taylor & Zappi
1996). This said, Aloísio Cardoso has photographed an opportunistic hummingbird
visiting the expanding flowers of S. luetzelburgii at dusk. Plates 32 & 33.
1. Columnar, segmented, at maturity at least 1.75 m, with flowering region terminal and in
rings at the articulations of the stem (caatinga surrounding the East Brazilian Highlands of
N to S Bahia) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1. leucostele
1. Bottle-shaped, to 1.5 m, rarely more, with subapical flowers on a continuous, elongate,
terminal, chlorophyllous cephalium (campo rupestre, Chapada Diamantina, central Bahia)
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2. luetzelburgii
1. S. leucostele (Gürke) A. Berger, Entwicklungslin. Kakt.: 59, 97 (1926). Holotype:

Brazil, Bahia, Mun. Maracás, ‘Calderão’, Ule 2 (B).
Cereus leucostele Gürke in Monatsschr. Kakt.-Kunde 18: 53 (1908). Cephalocereus leucostele (Gürke) Britton
& Rose, Cact. 2: 36 (1920). Pilocereus leucostele (Gürke) Werderm., Bras. Säulenkakt.: 111–112
(1933).
Single-stemmed or tree-like, 1.5–6.0 m, ± branched, branches arising at the stem-segment joints or more
rarely between these. Stems erect, 5–8 cm diam., slightly thickened below the cephalia; vascular cylinder
very woody; ribs 13–18, 7–8 × 5 mm; epidermis grey-green; areoles 3–4 mm, to 10 mm apart, with
abundant, long white trichomes partly obscuring the stem. Spines golden or whitish, stout; central spines
3–5, 12–20(–30) mm, spreading to deflexed, radials 15–20, to 10 mm. Fertile part of stem apical at first,
later forming up to 5 or more rings on the same branch as the vegetative growth overtops successive
cephalia, flower-bearing areoles with abundant, loose white wool and long, golden or whitish bristles to
40–90 mm. Flowers nocturnal, 9–10 × 5–5.3 cm, smelling strongly of cabbage; pericarpel and tube almost
naked, with only a few small, triangular scales, these sometimes slightly woolly in their axils, pericarpel
13–16 × 22 mm, green; tube 50–60 × 22–35 mm, cylindric, somewhat curved, greenish cream, with
fleshy, brownish scales at apex; perianth-segments 15–18 × 6–8 mm, spathulate to lanceolate, outer
segments reflexed, inner segments spreading to reflexed, pure white; stamens included in relation to the
perianth-segments; style to 60 mm, stigma-lobes 7–8, exserted; ovary locule 10 × 10 mm, obtriangular
in longitudinal section. Fruit globose to pyramidal, 4.4–4.7 × 3.5–4.5 cm; pericarp impressed at apex,
greenish to dark blue, wall to 8 mm thick, funicular pulp white or red. Seeds 2–2.2 mm long,
cochleariform, black, dull; testa-cells convex, with cuticular folds (Barthlott & Hunt 2000: pl. 49.1–4).
Central-southern (Bahian) caatinga element: in caatinga surrounding the Chapada Diamantina, northern
Serra do Espinhaço and Serra Geral, Bahia, 300–1100 m, east of the Rio São Francisco. Endemic.
Map 31.
bahia: Mun. Sento Sé, 10°7'S, 41°25'W, 11 July 2000, G. Charles (photos); Mun. Campo Formoso, 20
km from Maçaroca towards Fazenda Lagoa do Angico, 8 Jan. 1991, Taylor et al. 1385 (K, HRCB, ZSS,
CEPEC); Mun. Itaguaçu da Bahia, 31 km SE of town towards Irecê, 13 Jan. 1991, Taylor et al. 1411 (K,
HRCB, ZSS, CEPEC); Mun. (?), 10°53'S, 41°35'W, 13 July 2000, G. Charles (photos); Mun.
Ourolândia, E of the town and also NE of Olho D’Água do Facundo, 4 Aug. 2002, Taylor (obs.); Mun.
João Dourado & Mun. Morro do Chapéu, APA Gruta dos Brejões, 4 Aug. 2002, Taylor & Machado (obs.
& K, photos); Mun. Jacobina, Catinga do Moura, 10 km from Lages, 16 July 1989, Zappi 133 (SPF);
Mun. América Dourada, c. 17 km S of road BA 052, 2 km E of Belo Campo, 26 Dec. 1988, Taylor &
STEPHANOCEREUS
Zappi (obs.); Mun. Morro do Chapéu, Icó, Sep. 2001, M. Machado (obs.); l.c., between Cafarnaum and
road BA 052, 26 Dec. 1988, Taylor & Zappi (obs.); Mun. Canarana, 6 km E towards Cafarnaum, 24 Aug.
1988, Eggli (obs.); Mun. Ipupiara, Vanique, 11°47'S, 42°32'W, 23 June 1978, C.A. Miranda 292 (RB);
Mun. Ibitiara, 49 km W of Seabra on road to Ibotirama, 1100 m, 14 Jan. 1991, Taylor et al. (obs.); Mun.
Oliveira dos Brejinhos, 131 km W of Seabra on road BR 242, 14 Jan. 1991, Taylor et al. (obs.); Mun.
Marcionílio Sousa, Machado Portela, 19 June 1915, Rose & Russell 19902 (US); Mun. Iaçu, Jan. 1964,
Ritter 1227 (SGO 121290, fide Eggli et al. 1995: 498); Mun. Botuporã, c. 50 km S of Macaúbas on road
to Paramirim, 26 Aug. 1988, Eggli (obs.); Mun. Rio de Contas, 15 May 1983, Hatschbach 46404 (MBM);
Mun. Paramirim, near Caraíbas, 30 Nov. 1988, Taylor (K, photos); Mun. Dom Basílio, 13°59'S, 41°45'W,
26 June 1978, A.P. Araújo 43 (RB); Mun. Barra de Estiva, 17 km NE of Contendas on road to Maracás,
4 Feb. 1991, Taylor et al. (obs.); Mun. Ituaçu, c. 2 km NE of town, 18 Aug. 1988, Eggli 1202, 1203 (ZSS,
photos, seeds); Mun. Contendas do Sincorá, 27 km N of Sussuarana, 3 Feb. 1991, Taylor et al. (obs.);
Mun. Maracás, 6–7 km E of Porto Alegre, 300 m, 4 Feb. 1991, Taylor et al. (obs.); Mun. Rio do Antônio,
43 km on road from Brumado to Caitité, 14 Mar. 1983, Leuenberger et al. 3067 (CEPEC, B, K); Mun.
Guanambi, 19 km NE of town, N of road to Caitité, 25 July 1989, Taylor & Zappi (obs.); Mun. Brumado,
7 km N of town on road to Livramento, 14°8'S, 41°40'W, 22 Nov. 1988, Taylor & Zappi in Harley 25539
(K, SPF, CEPEC); Mun. Manuel Vitorino, road BR 116, 27 July 1989, Taylor & Zappi (obs.); Mun.
Palmas de Monte Alto, near Telebahia tower above the town, 21 July 1989, Taylor & Zappi (obs.); Mun.
Aracatu, 95 km W of Vitória da Conquista towards Brumado, 41 km SE of Brumado, on road BA 262,
2 Feb. 1991, Taylor et al. 1521 (K, HRCB, ZSS, CEPEC); Mun. Anajê, c. 10 km N towards Sussuarana,
17 Aug. 1988, Eggli (ZSS, photos); Mun. Poções, c. 15–20 km NE of Planalto on road BR 116, 27 July
1989, Taylor & Zappi (obs.).
Concern at present, but its habitat is continuing to decline and its status needs to be monitored.
One of the most characteristic cacti of the Bahian caatinga, but absent from north-eastern
Bahia, being closely associated with the East Brazilian Highlands, where the following
species replaces it in the campos rupestres. This suggests that their ancestor was a plant of
montane origin (cf. Cipocereus).
2. S. luetzelburgii (Vaupel) N. P. Taylor & Eggli in Bradleya 8: 91 (1991). Type: Brazil,

Bahia, Serra das Almas [Pico das Almas], July 1913, Luetzelburg 22 (B, holo., n.v.).
Cereus luetzelburgii Vaupel in Z. Sukkulentenk. 1: 57 (1923). Pilocereus luetzelburgii (Vaupel) Werderm.,

Bras. Säulenkakt.: 111 (1933). Cephalocereus luetzelburgii (Vaupel) Borg, Cacti: 111 (1937). Pilosocereus
luetzelburgii (Vaupel) Byles & G. D. Rowley in Cact. Succ. J. Gr. Brit. 19: 67 (1957). Pseudopilocereus
luetzelburgii (Vaupel) Buxb. in Beitr. Biol. Pflanzen 44: 252 (1968). Coleocephalocereus luetzelburgii
(Vaupel) Buxb. in Krainz, Kakteen, Lfg. 48–49 (1972).
Solitary, 0.5–1.0(–1.5) m, often not erect but inclined ± westwards, bottle-shaped, sometimes branched
at base, normally unbranched above unless damaged, except for populations found around Morro do
Chapéu, which are frequently branched; central cylinder very woody at base. Stems erect to decumbent
at apex, 12–15(–20) cm diam. at base, 2.5–4 cm diam. at apex, in the flower-bearing region; ribs 12–20,
20 × 18 mm on juvenile stem, gibbose beneath the areoles; epidermis grey-green to dark green. Areoles
5–6 mm, to 1 cm apart, long hairs abundant on the fertile part, white to pale cream. Spines golden or
whitish, blackish when old; central spines 4–6, 1.2–2(–3) cm, ascendent to spreading, radials 10–16, to
1.2 cm. Fertile part of stem elongate, with lateral to subapical flower-bearing areoles with abundant, loose
white to pale brown hairs and long, golden, whitish or greyish bristles, these 2–4 cm. Flowers 2–5 or
more borne together, 4–6 × 2.5–3.0 cm, smelling strongly of cabbage; pericarpel 10 × 5–6 mm, greenish;
tube 2.5 × 1.2–2.0 cm, narrowed above nectar-chamber, slightly curved, greenish cream, with fleshy,
ARROJADOA
brownish bract-scales only at apex; perianth-segments 8–10 × 3–5 mm, spathulate to lanceolate, outer
segments reflexed, inner segments spreading to reflexed, white; stamens included in relation to the
perianth-segments; style to 28 mm, stigma-lobes 6–8, exserted; ovary locule 4 × 4 mm, obtriangular in
longitudinal section. Fruit obovoid, 2.2–2.5 × 1.5–2 cm; pericarpel smooth, dark blue to blackish, but
covered in lighter blue wax, scarcely impressed at apex (not observed in fully ripe state, but claimed by
some observers to be dehiscent and reddish). Seeds 1.2–1.5 mm, cochleariform, black, dull; testa-cells
convex, with cuticular folds.
Northern campo rupestre (Chapada Diamantina) element: on and between crystalline and sandstone rocks
and gravels, campo rupestre, 380–1550 m, Bahia. Endemic. Map 31.
bahia: Mun. Jacobina, 2.5 km E of bridge over the river E of town, 16 July 1989, Zappi 129 (SPF);
Mun. Morro do Chapéu, W of Icó, N of Brejão (Formosa), nr ‘Lagedo Bordado’, 5 Aug. 2002, Taylor
& Machado (obs.); l.c., c. 3.5 km SE of Morro do Chapéu, 25 July 1993, L.P. de Queiroz & N.S.
Nascimento 3441 (HUEFS, K); l.c., c. 18 km E of town, margins of Rio Ferro Doido, 17 Feb. 1971, Irwin
et al. 32438 (K, NY, WAG); l.c., 20 km W of town, 2002, M. Machado (obs.); Mun. Seabra, c. 28 km
W of town S of road BR 242, 18 July 1989, Taylor & Zappi (K, photos); Mun. Palmeiras, Morro do Pai
Inácio, 12°27'S, 41°28'W, 13 Feb. 1994, V.C. Souza et al. 5252 (K, SPF); Mun. Lençóis, 24 Sep. 1965,
A.P. Duarte 9363 & E. Pereira 10076 p.p. (RB); l.c., Serra da Chapadinha, 12°27'S, 41°25'W, 24 Nov.
1994, Melo et al. in PCD 1315 (K, ALCB); Mun. Mucugê, c. 15 km NW of town, 12°58'S, 41°28'W,
26 Mar. 1980, Harley et al. 20976a (K, CEPEC); l.c., c. 7 km SW of town towards Barra da Estiva, 21
Aug. 1988, Eggli 1251 (ZSS); Mun. Piatã, 13°8'S, 41°50'W, 27 Aug. 1992, W. Ganev 976 (HUEFS);
Mun. Érico Cardoso [Água Quente], 26 km E of town, between Santa Rosa and the Pico das Almas,
13°32'S, 42°W, 1 Dec. 1988, Taylor in Harley 25559 (K, SPF, CEPEC); Mun. Rio de Contas, 16 km
NW of town, near the Pico das Almas, 13°32'S, 41°54'W, 27 Nov. 1988, Taylor in Harley 25551 (K, SPF,
CEPEC); l.c., ‘near Bom Jesus [do Rio de Contas]’, 1915, Luetzelburg s.n. (US); Mun. Jussiape, Cachoeira
da Fraga, próx. da cidade, 17 Feb. 1987, Harley et al. 24372 (SPF); Mun. Barra da Estiva, 3–13 km W of
town towards Jussiape, 23 Mar. 1980, Harley et al. 20837 (K, CEPEC); Mun. Ituaçu, c. 2 km NE of town,
19 Aug. 1988, Eggli 1204 (ZSS); unlocalized, Nov. 1913, Luetzelburg 742 (M).
Habitat modification affecting this species is currently limited.
Unmistakable for its bottle-shaped stem, this species and Micranthocereus purpureus are the
most characteristic elements of the Chapada Diamantina’s cactus flora. Variable in stem
shape and rib number between populations.
19. ARROJADOA Britton & Rose
Cact. 2: 170 (1920). Type: Arrojadoa rhodantha (Gürke) Britton & Rose (Cereus rhodanthus
Gürke).
Including (?) Pierrebraunia E. Esteves Pereira (1997), quoad typ.
Literature: Taylor & Zappi (1996).
Columnar, shrubby, 0.2–1.5(–3.5) m, branched below, at or above ground, but not forming a definite
trunk (except rarely in A. rhodantha); vascular cylinder rather woody; tissues very mucilaginous;
subterranean stem-tuber system sometimes present. Stems erect or rarely decumbent, 1–7(–12) cm diam.,
constricted, epidermis dull to bright green; ribs 6–14(–17), low, crenate, sinuses straight. Areoles with felt
and long hairs at first; spines stiff, very pungent, central spines porrect to deflexed, longer than the radials;
indeterminate growth of areoles near stem base present. Flower-bearing region of stem apical, more or
ARROJADOA
less dilated and forming a cephalium with bristles and wool, or cephalium lacking (A. bahiensis) or flowers
also arising partly in a lateral position (A. dinae subsp. eriocaulis). Flower-buds coloured bright pink from
the earliest stages of their development; flowers diurnal, opening at morning, or late afternoon to
evening, and sometimes remaining open through part or all of the night, brightly coloured, 15–42 × 5–20
mm; pericarpel smooth, naked, subglobose and clearly delimited from the tube; tube broadened below,
cylindric above, reddish, pink or magenta, with broad, fleshy bract-scales only near the apex; innermost
filaments curved towards the style, protecting the nectar-chamber; perianth-segments triangular to
lanceolate or spathulate, erect to spreading, opening very little, innermost white, yellow, pinkish or
purplish. Fruit obovate to turbinate, indehiscent, naked, flower remnant drying black, normally erect,
broad at base, forming a deep and often broad point of insertion at apex of fruit; pericarp smooth,
variously coloured, funicular pulp white, scarce, ± solid or aqueous. Seeds to 1.5 mm, cochleariform;
testa-cells convex; cuticular folds present or nearly absent.
A genus of 4 or 5 very distinct, but rather variable, hummingbird-pollinated species,

characteristic of the caatinga-agreste and campos rupestres (and included cerrados). The
potential fifth species is a plant discovered in August 2000, in north-eastern Goiás, by
botanist, Rafaela Forzza (RB). It appears to be related to A. dinae. Prior to this the genus
was considered to be endemic to Eastern Brazil. Another novelty of uncertain status is
mentioned under A. rhodantha (see 4b).
P. J. Braun & E. Esteves Pereira (1995b: 81) have established the Subgenus
Albertbuiningia based on the second species treated here, A. dinae Buining & Brederoo.
The following key does not attempt to distinguish the hybrids that occasionally occur
between species Nos 2–4 (2×3, 2×4 & 3×4, see below). Plates 34–37.
1. Stem 4–10 cm diam., neither strongly constricted nor terminated or interrupted by

cephalia, flowers from ± undifferentiated areoles at stem apex (cent. Bahia, campo rupestre,
1000–2000 m) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1. bahiensis
1. Stems constricted, thickened at apex or < 4 cm diam., the flowers developed in cephalia
composed of wool and long bristle-spines . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
2. Stems 10–50 × 2 cm, sometimes arising from a tuberous, rhizomatous rootstock; flowers
bicoloured (anthesis p.m.), inner perianth-segments contrasting with flower-tube (Serra do
Espinhaço: Caitité BA to Bocaiúva MG, campo rupestre, cerrado & ecotones with caatinga) . . . 2. dinae
2. Stems usually > 50 cm or > 2 cm diam., rootstock fibrous; flowers concolorous . . . . . . . . . . . . . 3
3. Stems to 1.8 cm diam., but expanded and much broader below the cephalia; anthesis p.m.,
outer perianth-segments expanding (N, E & S Bahia & NE Minas Gerais, caatinga-agreste)
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3. penicillata
3. Stems 2–6 cm or more in diameter, not as above; anthesis a.m., perianth-segments hardly
expanding, erect (widespread, caatinga and ecotones with campo rupestre) . . . . . . . . . 4. rhodantha
1. Arrojadoa bahiensis (P. J. Braun & E. Esteves Pereira) N. P. Taylor & Eggli in Kew
Bull. 49: 98 (1994); N. P. Taylor & Zappi in Bot. Mag. 13: 72–75, tab. 291 (1996). Type:
Bahia, ‘Chapada Diamantina’, E. Esteves Pereira 337 (UFG 13007, holo. n.v.).
Floribunda bahiensis P. J. Braun & E. Esteves Pereira in Pabstia 4: 11–16, illus. (1993); A. Hofacker in Kakt.
and. Sukk. 45: 120–123, illus. (1994). Pierrebraunia bahiensis (P. J. Braun & E. Esteves Pereira) E.
Esteves Pereira in Cact. Succ. J. (US) 69: 296 (1997).
Arrojadoa cremnophila Taylor et al. in sched., nom. nud.
Subglobose to elongate plants (depending on exposure), normally to 30 cm, but occasionally to 60 cm

or more, unbranched or branched below ground or at base to form clumps of stems; central cylinder not
ARROJADOA
woody. Stems erect or inclined, 4–10 cm diam.; ribs (9–)10–14(–17), 5 × 15(–25) mm, with or without
lateral folds; epidermis bright yellow-green to dark green. Areoles 2–3 mm, approximate, with felt and
long, white to grey trichomes. Spines brittle, pale pinkish brown or golden, black when old; central
spines 1–4, to 30 mm, ascendent, radials 6–8(–10), to 5 mm, spreading; weak indeterminate growth of
spines occurring at basal stem areoles. Fertile part of stem apical, flower-bearing areoles scarcely
differentiated. Flowers (reported in February and August, but produced in all months in cultivation) few
or up to 25 developing together, opening first at night and closing around noon the following day (timing
and degree of perianth expansion apparently linked to light and low temperatures), c. 29–39(–42) × to
11 mm at apex of tube, the inner perianth-segments expanding to give an opening 8–13 mm in diam.,
pericarpel and tube together to c. 30 mm, naked or the latter with a few minute bract-scales in distal half,
magenta-pink outside, pericarpel c. 5 × 7 mm, nectar-chamber externally 6 × 8.5 mm, tube constricted
above and below; outer perianth-segments short and fleshy, magenta-pink, inner thin and delicate, white,
3.9–4.7 × 2.3–2.8 mm; stamens 3–14 mm, the lowermost with thickened filament bases protecting the
nectar-chamber, anthers c. 0.8–1.1 mm, pale yellow, with abundant pollen; style to 28 × 0.8–0.4 mm,
white; stigma-lobes 5–7, white, pointed, partly exserted in relation to the anthers or hidden amongst
them, c. 4 × <1 mm. Fruit c. 17 × 15 × 12 mm, broadly ovoid, somewhat compressed, ridged and
contracted or beaked at apex, deep shocking pink, with a paler waxy bloom, funicular pulp liquid,
translucent, floral remnants ± blackened. Seeds to 1.3 mm, cochleariform, black, dull; testa-cells convex,
cuticular folds present.
Northern campo rupestre (Chapada Diamantina) element: on cliffs and rock ledges in sun or deep shade,
campo rupestre, c. 1000–2000 m, central Bahia. Endemic. Map 32.
bahia: Mun. Mucugê, 27 km NW of the town on road to Guiné, then 1 km E, 12°51'S, 41°31'W, 23
Dec. 1988, Taylor & Zappi in Harley 27382 (K, SPF, CEPEC), 7 Sep. 1981, A. Furlan et al. in CFCR 2020
(SPF); W of town, above cemitério, 15 Mar. 1989, Supthut 8937 (ZSS), 6 Feb. 1991, Taylor et al. 1557A
& cult. Zappi HRCB, then cult. RBG Kew, 1995 (K, ZSS, photos), coll. unknown, cult. A. Hofacker,
1993 (ZSS – fr., K, photos); Mun. Abaíra, ‘9 km N de Catolés, caminho de Ribeirão de Baixo a Piatã,
subida da Serra do Atalho’, 13°14'S, 41°55'W, 10 July 1995, L.P. de Queiroz 4384 (HUEFS); Mun. Rio
de Contas, Pico do Itobira, 2 Jan. 2003, Harley (K, photos); l.c., top of Pico das Almas, 1958 m, 20 Feb.
1987, Harley 24522 (SPF, K, CEPEC), 15 Dec 1988, Taylor in Harley 25565 (K, SPF, CEPEC).
conservation status. Vulnerable [VU D2] (2); extent of occurrence = 1614 km2; area of occupancy
estimated to be < 20 km2; PD=3, EI=1, GD=2. Short-list score (2×6) = 12. Known from only c. 5
populations (one within the Parque Nacional da Chapada Diamantina) and at risk from collection of
plants and seeds, even though many individuals are protected by the plant’s preference for steep slopes
and cliffs inaccessible to the collector.
This unique and remarkable Bahian endemic was first collected by scientists only in 1981,
but could prove to be quite widespread in the Chapada Diamantina when the inaccessible
cliff habitats it occupies have been further investigated. Its flowers, fruit and seed would at
first sight seem to ally it with the following allopatric-vicariant species. However, small,
sterile individuals strongly resemble juvenile plants of the sympatric Stephanocereus
luetzelburgii and in morphology (but not size and colour) their flowers are also very similar.
Furthermore, recent preliminary chloroplast gene sequence studies (Soffiatti unpubl.)
suggest that these two may indeed be sister species and that the absence of a proper
cephalium distinguishes them from the remainder of the Arrojadoa-Stephanocereus alliance.
A. bahiensis is superficially similar and convergent with Cipocereus (Floribunda)
pusilliflorus, but has rather different, woolly adult areoles (long hairs lacking in C.
pusilliflorus), mature fruits and seeds, and its perianth is clearly and abruptly differentiated
into coloured fleshy outer segments and white thinner inner segments, as in A. dinae.
ARROJADOA
2. Arrojadoa dinae Buining & Brederoo in Kakt. and. Sukk. 24: 99–101 (1973).
Holotype: Bahia, Brazil, Mun. Urandi, 900 m, Horst & Uebelmann 399 (U).
A. beateae P. J. Braun & E. Esteves Pereira in Kakt. and. Sukk. 40: 250–256 (1989). Holotype: Brazil,
Minas Gerais, Mun. Itacambira, July 1987, Esteves Pereira 261 (UFG); Braun 830 (ZSS, B, paratypi).
A. dinae var. nana P. J. Braun & E. Esteves Pereira in ibid. 42: 190–195 (1991). A. dinae subsp. nana (P.
J. Braun & E. Esteves Pereira) P. J. Braun & E. Esteves Pereira in Succulenta 74: 82 (1995).
Holotype: Minas Gerais, Braun 415 (ZSS).
A. heimenii van Heek & W. Strecker in Kakt. and. Sukk. 50: 137–138 (1999); W. van Heek in Piante
Grasse, Speciale 1997: 29, 31, ‘Foto’ 70 & 71 (1998), nom. inval. Holotype: Minas Gerais, Mun.
Engenheiro Dolabela, P. Braun 20528 (UFG).
vernacular name. Rabo-de-raposa.
Subshrub, sometimes creeping/suckering or semidecumbent, 10–60 cm, poorly branched above ground;
vascular cylinder moderately woody; subterranean stems not or slightly thickened, or strongly tuberous,
sprouting when above-ground parts are damaged. Aerial stems erect, 1–2 cm diam., slightly to strongly
thickened below the flower-bearing apical cephalium and sometimes to 3.5 cm diam. near rootstock; ribs
6–11, 3–4 × 2–4 mm; epidermis olive-green. Areoles 2 mm, to 5 mm apart, with few to many, long,
white to grey trichomes, conspicuously hairy in some forms. Spines reddish at first, paler to off-white in
age, delicate, needle-like; central spines 1–2, to 10 mm, radials 8–10, 3–5 mm. Cephalium with much
loose wool and long reddish bristle-spines to 20 mm, up to 6 cephalia per stem. Flowers arising from the
cephalia or sometimes partly from the stem beneath, 14–29 × 4–9 mm, opening late in the afternoon;
pericarpel 4 × 3.5 mm; tube to 10 × 4–5 mm, cylindric, coral-red, orange-red or magenta, with fleshy
bract-scales only at apex; perianth-segments 2–3 × 2 mm, triangular to lanceolate, outer segments erect,
coral-red to reddish pink, inner segments opening very slightly, rather variable in colour, pale creamy
yellow, white or lilac-pink to purplish; stamens included in relation to the perianth; style 8–9 mm,
stigma-lobes 7–8, included; ovary locule 1.8–2.0 × 2 mm, hemicircular in longitudinal section. Fruit
obovoid, broadly beaked to truncate at apex, to 7–21 × 5–18 mm, pericarp smooth, greenish, pink,
reddish or purplish brown at apex. Seeds 0.9–1.8 × 0.8–1.4 mm, cochleariform, black, dull; testa-cells
convex, with cuticular folds (Barthlott & Hunt 2000: pl. 49.5–7).
conservation status. Near Threatened [NT]; extent of occurrence > 20000 km2 and at least 17
locations known, but most of these are projected to decline in quality and numbers of individuals held
(the subspecies, below, are threatened).
The following subspecies are recognized — both are rather variable in flower and fruit
colour:
1. Well-developed subterranean stem-tubers present; stem areoles very woolly, giving the
stem a felted appearance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . b. subsp. eriocaulis
1. Subterranean part of stem sometimes thickened, but well-developed tubers lacking;
above-ground vegetative part of stem not as above . . . . . . . . . . . . . . . . . . . . . . . . a. subsp. dinae
2a. subsp. dinae

? A. multiflora F. Ritter, Kakt. Südamer. 1: 89–90 (1979). Holotype: Brazil, Bahia, Mun. Caitité, Brejinho
das Ametistas, Jan. 1964, Ritter 1243 (U).
Northern campo rupestre element: sandy cerrado, campo rupestre (sometimes on rocks), gerais and in the
caatinga/campo rupestre ecotone, 550–1400 m, central-southern Bahia (from c. 13°55'S southwards) and
northern Minas Gerais (south to Bocaiúva) in the Serra do Espinhaço (and Serra Geral). Endemic to the
core area within Eastern Brazil. Map 32.
ARROJADOA
bahia: Mun. Caitité, 14°10'S, 42°32'W, 27 Oct. 1993, L.P. de Queiroz 3605 (HUEFS); (?) l.c., 15 km
NW towards Riacho de Santana, Ritter 1243 (SGO 121896, fide Eggli et al. 1995: 503); l.c., hill above
the town with Telebahia tower, 27 Aug. 1988, Eggli 1305 (ZSS); ibid., 17 Mar. 1989, Supthut 8944 (ZSS)
— showing some influence (?) of introgression with A. rhodantha; (?) l.c., Brejinho das Ametistas, 1964,
Ritter 1243 (U); Mun. Urandi, Horst & Uebelmann 399 (U); l.c., SE of town, 1983, Braun 411 (ZSS); Mun.
Jacaraci, 5.5 km from town on road to Caetité, H. Britsch in I.S.I. 1691, cult. R. Mottram, 1988 (K);
Mun. Cordeíros, 15°4'S, 40°53.5'W, Sep. 2003, M. Machado (photos).
minas gerais: Mun. Rio Pardo de Minas/São João do Paraíso (?), ‘SW of Condeúba, Bahia’, 1983,
Braun 415 (ZSS); Mun. Monte Azul, Serra Geral, c. 12–15 km E of town, 24 July 1989, Zappi 169 (SPF,
K); Mun. Rio Pardo de Minas, Montezuma, Horst & Uebelmann 578 (ZSS); l.c., divisa entre Mun.
Espinosa e Montezuma, Serra do Pau d’Arco, 15°4'55"S, 42°38'27''W, 15 Mar. 1994, N. Roque et al. in
CFCR 15407 (K, SPF); Mun. Itacambira, Alto da Serra, c. 28 km de Pau D’oleo (42 km de Juramento),
16 June 1991, R. Mello-Silva et al. 524 (SPF, K); l.c. (?), Braun 830 (ZSS), July 1987, E. Esteves Pereira
261 (UFG); Mun. Bocaiúva, NE of Engenheiro Dolabela, c. 15 & 40 km towards Sítio, 7–8 Aug. 1988,
Eggli 1116, 1123 (ZSS).
conservation status. Vulnerable [VU B1ab(iii)] (2); extent of occurrence = c. 20,000 km2; PD=3,
EI=1, GD=2. Short-list score (2×6) = 12. Continuing decline in habitat quality caused by charcoal
producers and forestry activities.
Very variable in stem morphology and flower colour. On present knowledge, the typical
northern populations and the southern forms (described as A. beateae and A. heimenii)
appear to be disjunct. Ritter’s A. multiflora is clearly a synonym of the species treated here,
but its referal to subsp. dinae is currently uncertain. The plant recently discovered in
north-eastern Goiás by Rafaela Forzza (RB) is, from photographs at least, vegetatively
hard to distinguish from subsp. dinae.
2b. subsp. eriocaulis (Buining & Brederoo) N. P. Taylor & Zappi in Cactaceae Consensus
Initiatives 3: 7 (1997). Holotype: Brazil, Minas Gerais, Mun. Mato Verde, 900 m, Horst
& Uebelmann 349 (U).
A. eriocaulis Buining & Brederoo in Kakt. and. Sukk. 24: 241–244 (1973).
A. eriocaulis var. albicoronata van Heek et al. in Kakt. and. Sukk. 33: 224–227 (1982). A. eriocaulis subsp.
albicoronata (van Heek et al.) P. J. Braun & E. Esteves Pereira in Succulenta 74: 82 (1995). Type:
Brazil, Minas Gerais, Grão Mogol, 15 Aug. 1981, van Heek et al. 81/113 (KOELN, holo., n.v.).
A. eriocaulis var. rosenbergeriana van Heek & W. Strecker in Kakt. and. Sukk. 44: 258–262 (1993). Type:
Brazil, Minas Gerais, SE of Mato Verde, W. van Heek & W. Strecker 85/216 (KOELN, holo., n.v.).
A. multiflora subsp. hofackeriana P. J. Braun & E. Esteves Pereira in Kakt. and. Sukk. 53: 75 (2002).
Holotype: Bahia, nr Piatã, I. Horst HU1394 (UFG 22429, n.v.; ZSS, iso., n.v.). Synon. nov.
Northern campo rupestre element: in sandy cerrado, 700–950 m, SE edge of the Chapada Diamantina and
eastern drainage of Serra do Espinhaço, central-southern Bahia and northern Minas Gerais. Endemic to
the core area within Eastern Brazil. Map 32.
bahia: Mun. (?) Piatã, c. 95 km S of road BR 242, Horst & Uebelmann HU 1394, see Bohle (2000: 48),
l.c., 13°40'30"S, 47°16'12''W, 15 July 2000, cult. G. Charles, June 2001 (photos).
minas gerais: Mun. Mato Verde/Rio Pardo de Minas, 12 km E of Mato Verde towards Santo Antônio
do Retiro, 15°23'S, 42°45'W, 9 Nov. 1988, Taylor & Zappi in Harley 25518 (K, SPF), 13 Aug. 1988, Eggli
1154 (ZSS); Mun. Grão Mogol, on road to Francisco Sá, 5 Jan. 1986, Kameyama et al. in CFCR 8839
(SPF, MBM); l.c., Córrego Escurinha, 23 July 1986, Zappi et al. in CFCR 9828 (SPF), 28 May 1988,
Zappi et al. in CFCR 12019 (SPF).
ARROJADOA
conservation status. Endangered [EN B2ab(iii)] (3); area of occupancy assumed to be < 500 km2;
PD=3, EI=1, GD=2. Short-list score (3×6) = 18. Endangered from habitat destruction by charcoal
producers; needs regular monitoring in view of this and its apparent rarity.
Variable in flower colour. The plants described and illustrated as A. multiflora by Ritter,
l.c. and A. eriocaulis var. rosenbergeriana by van Heek & Strecker, l.c., appear to be
somewhat intermediate between the two subspecies recognized. The form from Bahia
requires further study, but does not seem to be specifically distinct as some commentators
have suggested.
Arrojadoa dinae is unique amongst the cacti of Eastern Brazil in its possession, at least
in subsp. eriocaulis, of well-developed true stem-tubers (despite claims that these also occur
in the root-tuberous Cereus subg. Mirabella by Braun & Esteves Pereira 2002: 42).
3. Arrojadoa penicillata (Gürke) Britton & Rose, Cact. 2: 171 (1920); N. P. Taylor &
Zappi in Bot. Mag. 13(2): 75–78, tab. 292 (1996). Type: Brazil, Bahia, Mun. Maracás,
‘Calderão’, Oct. 1906, Ule 7052 (B, holo. in spirit, HBG, iso. [K, photo.]).
Cereus penicillatus Gürke in Monatsschr. Kakt.-Kunde 18: 70–71 (1908). Cephalocereus penicillatus (Gürke)
Werderm., Bras. Säulenkakt.: 116 (1933).
Arrojadoa penicillata var. decumbens Backeb. & Voll in Arch. Jard. Bot. Rio de Janeiro 9: 164–166, illus. p.
159 (1949, publ. 1950). Lectotype (Taylor & Zappi, l.c.): illus., l.c. 159.
A. penicillata var. spinosior Brederoo & S. Theun. in Succulenta 59: 20–27 (1980). Holotype: Brazil,
[Minas Gerais], Rio Jequitinhonha, Horst 113 (U).
Shrubby or semi-scandent to decumbent (especially when growing in sand-dunes), to 4 m long, ±

branched; rootstock fibrous. Stems erect or more rarely decumbent, 1–1.8 cm diam., not very fleshy,
expanded to 2.5(–3.0) cm diam. below the cephalia, vascular cylinder very woody; ribs 7–12, 2–4 × 3
mm; epidermis olive-green. Areoles 2 mm, to 7 mm apart, long hairs present but soon falling. Spines
reddish brown, delicate; central spines 2–4, 4–30 mm, ± deflexed, radials 6–10, 3–5 mm. Cephalia
sometimes forming 3 or 4 rings on the same axis, with abundant, loose wool and long red to brownish
bristle-spines to 20–30 mm. Flowers to 30 × 20 mm, opening shortly before dusk, pinkish magenta or
reddish pink; pericarpel 7–8 × 5 mm; tube to 12 × 4–7 mm, cylindric, with fleshy bract-scales only at
apex; perianth-segments bright pink with a very fine darker line around the margin, outer segments to 7
× 2.8 mm, triangular to lanceolate, patent (well-expanded), innermost segments remaining erect and
forming a tube which encloses the stamens etc., its opening 3–4 mm in diam.; stamens included in
relation to the perianth-segments; style 10 mm, stigma-lobes 6–8, included; ovary locule 2–2.5 × 2 mm,
obtriangular in longitudinal section. Fruit obovate to globose, sometimes laterally compressed, 12–25 ×
12 mm; pericarp smooth, greenish to dull red, olive-brownish or whitish. Seeds c. 1.2 mm,
cochleariform, black, dull; testa-cells convex, with or without cuticular folds.
Central-southern caatinga element: on granite/gneiss inselbergs or lajedos, sand-dunes and stony ground
(rarely on limestone), growing under or through shrubs in the caatinga-agreste and caatinga/campo rupestre
ecotone, c. 200–850 m, north-western Bahia (Pilão Arcado & Barra), northern to southern Bahia east of
the Chapada Diamantina, and north-eastern Minas Gerais (Rio Jequitinhonha valley). Endemic to the
core area of Eastern Brazil. Map 33.
bahia: NW Bahia, Mun. Pilão Arcado, 42°54'W, 10°2'S, 21 Nov. 1978, Miranda 303 (IPA, HRB, F);
Mun. Barra, dunes by the Rio São Francisco, 10°47'16"S, 42°46'22''W, 23 June 1996, Hind s.n. (K,
photos); N, NE, E & S Bahia, Mun. Jaguarari, vicinity of Pilar, 9°54'S, 39°57'W, 18 Aug. 1983, G.C.P.
Pinto & S.B. Silva 191/83 (HRB); l.c., 9.5 km N of Jaguarari, 15 July 1989, Zappi 124, cult. Taylor &
ARROJADOA
Zappi, 1995 (K, K, photos); l.c., 20 km N of Senhor do Bomfim towards Juazeiro, 25 Dec. 1984, Harley
in CFCR 7577 (SPF, K); Mun. Monte Santo, c. 5 km N, na estrada para Uauá, 25 Aug. 1996, L.P. de
Queiroz & N.S. Nascimento 4609 (HUEFS, K); Mun. Jacobina, 10 km from Jacobina towards Morro do
Chapéu, 14 Mar. 1990, A.M. de Carvalho & J. Saunders 2787 (CEPEC); ibid., ‘margem esquerda do Rio
Jacuípe’, 21 Jan. 1985, Passos 26 (MO, BAH 6021); Mun. Santa Luz, 10 km from town on road BA 120
to Queimadas, 16 Nov. 1986, L.P. de Queiroz 1135 (NY, HUEFS); Mun. Valente, c. 5 km SW of town
towards São Domingos, 11°25'S, 39°30'W, 29 Dec. 1992, L.P. de Queiroz et al. 3041 (MBM, HUEFS);
Mun. Morro do Chapéu, Ventura, Sep. 2001, M. Machado (obs.); Mun. Riachão do Jacuípe, 9 km NW
of town, road BR 324, 25 Apr. 1992, Taylor & Zappi (obs.); l.c., 13 km SE of town on road BR 324, 10
July 1985, L.R. Noblick & Lemos 4106 (CEPEC, HRB, ALCB, HUEFS); l.c., 24.5 km NW of Tanquinho,
road BR 324, 25 Apr. 1992, Taylor & Zappi (obs.); Mun. Santa Barbara/Feira de Santana, 21 km ‘SW’
[SE] of Tanquinho along BR 116, 31 Mar. 1976, Davidse et al. 11672 (MO); Mun. Serra Preta, 42.5 km
E of Baixa Grande on road BA 052, 9 Feb. 1991, Taylor et al. (obs.); Mun. Feira de Santana, 10 km N of
town on road BR 101, 26 Jan. 1980, Andrade-Lima 80-8812 (IPA); Mun. Iraquara, Faz. Torrinha, entrando
13 km após a entrada para Iraquara, 18 Apr. 1991, G.L. Esteves & R. de Lyra-Lemos 2534 (SPF, K, MAC);
Mun. Rui Barbosa, 1.5 km SE of town near road to Itaberaba, 7 Feb. 1991, Taylor et al. 1569 (K, HRCB,
ZSS, CEPEC); Mun. Seabra, 6 km E of town, then 1 km behind posto Esso, 25 Aug. 1988, Eggli 1282
(ZSS); Mun. Itaberaba, 12°30'S, 39°59'W, 9 Oct. 1987, L.P. de Queiroz 1773 (HUEFS); l.c., lowlands
along river below town, 22 Aug. 1988, Eggli 1252 (ZSS); l.c., 63 km from Argoim on road BR 242, 16
Apr. 1991, G.L. Esteves & R. de Lyra-Lemos 2529 (MAC); Mun. Iaçu, Faz. Lapa, 12°42'S, 39°56'W, 26
Feb. 1983, G.C.P. Pinto 171/83 (HRB); Mun. Itatim, Morro do Agenor, 12°42'S, 39°46'W, 26 Nov.
1995, 21 Apr. 1996 & 1 Sep. 1996, F. França 1490, 1617, 1817 (HUEFS; K, photo); l.c., Morro das Tocas,
25 Nov. 1995 & 28 Sep. 1996, F. França 1453, 1853 (HUEFS); Mun. Milagres, road BA 046 c. 4 km from
junction with road BR 116 towards Amargosa, 12°51'S, 39°46'W, 2 June 1993, L.P. de Queiroz & T.S.N.
Sena 3197 (K, HUEFS); Mun. Castro Alves, Oct. 1972, G.C.P. Pinto s.n. (ALCB); l.c., 11 km from road
BR 116 (Argoim), 12°32'S, 39°55'W, 13 Mar. 1984, L.C.O. Filho 86 (HRB); Mun. Milagres, Km 5
towards Iaçu from road BR 116, 13 Oct. 1981, A.M. de Carvalho & G.P. Lewis 962 (CEPEC, K); Mun.
Milagres/Amargosa, 19 Oct. 1967, A.P. Duarte 10584 (HB, RB); Mun. Marcionílio Sousa, Machado
Portela, 19 June 1915, Rose & Russell 19913 p.p. (US); Mun. Santa Inês, Km 7–9 from town towards road
BR 116, 17 Oct. 1972, R.S. Pinheiro (CEPEC); Mun. Maracás, Faz. Tanquinho, 2 Jan. 1989, L. Paganucci
158 & M.L. Guedes 124 (ALCB); l.c., 6–7 km E of Porto Alegre, 4 Feb. 1991, Taylor et al. 1554 (K, ZSS,
CEPEC, HRCB) — showing influence from A. rhodantha; Mun. Ituaçu, c. 2 km NE, 18 Aug. 1988, Eggli
1199 (ZSS); l.c., c. 10 km SE of town towards Tanhaçu, 18 Aug. 1988, Eggli 1196 (ZSS); Mun. Jequié
(?), between the town and Porto Alegre, 1974, Gottsberger s.n., cult. (K, ZSS); between Jequié and Manuel
Vitorino, 18 Jan. 1975, Read et al. 3443 (US); Mun. Jequié, E of the town, Aug. 2001, Zappi (obs.); Mun.
Aracatu, close to the border with Mun. Brumado, 25 km on road BR 030 towards Sussuarana, 3 Feb.
1991, Taylor et al. 1542 (K, HRCB, ZSS, CEPEC); l.c., c. 20 km N of Anagê (Anajé) towards Sussuarana,
17 Aug. 1988, Eggli 1188 (ZSS); Mun. Poções, 5 km towards Boa Nova, 25 Jan. 1973, I. & G. Gottsberger
16-25173a (K); cent.-S Bahia (W of Serra do Espinhaço), (?) see hybrid, Nos 2 × 3, below.
minas gerais: NE Minas Gerais, Mun. Itaobim, 1 km W of town, 0.5 km N of the Rio Jequitinhonha,
16°34'S, 41°31'W, 18 Nov. 1988, Taylor & Zappi in Harley 25529 (K, SPF); l.c., 8 km W of town on
road BR 357 towards Itinga, 14 Dec. 1990, Taylor & Zappi 761 (K, HRCB, ZSS, BHCB), 9 Apr. 1983,
Martinelli & Leuenberger 9224 (K); Mun. Itinga, 2 km E of town on road BR 357, 14 Dec. 1990, Taylor
& Zappi (obs.).
Concern, but its extensive habitat continues to decline.
This species has a more restricted range than the following, but is not nearly as common,
nor as variable. The record by Uebelmann (1996) from western Pernambuco seems to be
a misidentification of the following species, which is abundant in the region. Their
different habitat preferences deserve further analysis — they are rarely sympatric despite
the considerable overlap in distribution.
ARROJADOA
4. Arrojadoa rhodantha (Gürke) Britton & Rose, Cact. 2: 171–172 (1920). Type: Brazil,
Piauí, Caatinga de São Raimundo [Nonato], 1907, Ule 11 (B, holo.; K [photo ex B]).
Cereus rhodanthus Gürke in Monatsschr. Kakt.-Kunde 18: 69 (1908). Cephalocereus rhodanthus (Gürke)
Arrojadoa canudosensis Buining & Brederoo in Cact. Succ. J. (US) 44: 111–113 (1972). A. rhodantha subsp.
canudosensis (Buining & Brederoo) P. J. Braun in Bradleya 6: 90 (1988). Holotype: Brazil, Bahia,
Canudos, Horst & Uebelmann 251 (U).
A. aureispina Buining & Brederoo in Kakt. and. Sukk. 23: 95–98 (1972). A. rhodantha subsp. aureispina
(Buining & Brederoo) P. J. Braun & E. Esteves Pereira in Succulenta 74: 81 (1995). Holotype: Brazil,
Bahia, Mun. Caitité, 1966, Horst & Uebelmann 154 (U).
A. aureispina var. anguinea P. J. Braun & E. Esteves Pereira in Cact. Succ. J. (US) 60: 174–180 (1988). A.
rhodantha subsp. anguinea (P. J. Braun & E. Esteves Pereira) P. J. Braun & E. Esteves Pereira in
Succulenta 74: 81 (1995). Holotype: Brazil, Bahia, Braun 80a (ZSS).
A. aureispina var. guanambensis P. J. Braun & Heimen in Kakt. and. Sukk. 31: 334–337 (1980). A. rhodantha
subsp. guanambensis (P. J. Braun & Heimen) P. J. Braun & E. Esteves Pereira in Succulenta 74: 81 (1995).
Holotype: Brazil, Bahia, Mun. Guanambi, June 1979, Braun & Heimen 80 (KOELN, n.v.; ZSS, iso.).
A. theunisseniana Buining & Brederoo in Krainz, Die Kakt., Lfg 52 (1973). A. rhodantha var. theunisseniana
(Buining & Brederoo) P. J. Braun in Bradleya 6: 90 (1988). Holotype: Brazil, Bahia, Botuporã,
Buining 1002 (U).
A. horstiana P. J. Braun & Heimen in Kakt. and. Sukk. 32: 186–190 (1981). Holotype: Brazil, Minas
Gerais, Braun & Heimen 83 (KOELN, n.v.; ZSS, iso.).
A. rhodantha subsp. reflexa P. J. Braun in Kakt. and. Sukk. 35: 34–38 (1984). Holotype: Brazil, Bahia, 1000
m, Braun 68 (KOELN, n.v.).
A. rhodantha var. occibahiensis P. J. Braun in Kakt. and. Sukk. 36: 114–115 (1985). Holotype: Brazil, Bahia,
W of the Rio São Francisco, Horst & Uebelmann HU 616 (ZSS).
vernacular names. Rabo-de-raposa, Rabo-de-onça.
Shrubby, to 2 m, rarely reaching 3.5 m, ± branched. Stems erect or semi-decumbent, 1.7–6.0 cm diam.,
not or scarcely expanded below the cephalia, vascular cylinder very woody; ribs 8–14, 4–5 × 5 mm,
epidermis olive-green. Areoles 2–3 mm, to 10 mm apart, long hairs soon lost from young growth. Spines
reddish, golden, greyish or brown, stiff, very sharp; central spines 4–5, 1.5–4.0(–5.0) cm, ascendent to
spreading, radials 6–10, to 1 cm. Fertile part of stem apical, later forming up to 4 and sometimes many
more rings on the same axis, cephalium areoles with abundant, loose white hairs and long red, golden or
brownish bristles, these 2–3 cm. Flowers 20–35 × 8–10 mm (at apex), opening at morning only, pinkish
magenta or reddish pink; pericarpel 6–7 × 4–5 mm; tube to 20 × 10–14 mm, cylindric or constricted
above the nectar-chamber, with fleshy bract-scales at apex; perianth-segments 5–6 × 5 mm, obovate,
outer segments erect, inner segments expanding partially to give an aperture of only a few mm; stamens
included in relation to the perianth-segments; style 20 mm, stigma-lobes 6–8, included; ovary locule 2–3
× 2 mm, obtriangular in longitudinal section. Fruit globose, 20–25 × 10–18 mm, often somewhat laterally
compressed; pericarp smooth, dark red. Seeds c. 1.2 mm, cochleariform, black, dull; testa-cells convex,
with or without cuticular folds.
Central-southern caatinga element: found on various substrates (including in the shade of dense caatinga
forest and in the open on inselbergs) and entering the caatinga/campo rupestre ecotone, 220–1200 m, south-
western Piauí and western Pernambuco to central-northern Minas Gerais. Endemic to the core area of
piauí: Mun. Picos, road to Jaicós, 7°13'S, 41°17'W, 6 Apr. 2000, E.A. Rocha et al. (K, photos); Mun.
Padre Marcos, Serra Velha, 14 May 1993, M.E. Alencar 33 (PEUFR, IPA), s.d., M.E. Alencar 138
(PEUFR, IPA); Mun. Simões, 7°36'S, 40°45'W, 5 Apr. 2000, E.A. Rocha et al. (K, photos); Mun.
Paulistana, 11 km N of town, 6 Apr. 2000, E.A. Rocha et al. (K, photos), l.c., 7 km N of border with
Pernambuco, 6 Apr. 2000, E.A. Rocha et al. (obs.); Mun. São Raimundo Nonato, 1907, Ule 11 (B, K,
photo), Horst & Uebelmann 422 (ZSS); Parque Nac. Serra das Confusões, cult. Vitória da Conquista,
ARROJADOA
Bahia, April 2003, Avaldo de Oliveira Soares Filho s.n., Taylor (obs.) — form with very long central spines.
pernambuco: Mun. Araripina, Sítio Jatobá, 20 Nov. 1992, F.A.R. Santos 54 (HUEFS, PEUFR); l.c.,
behind Posada do Araripe, 15 Feb. 1990, Zappi 226 (SPF, K, photos); Mun. Serrita, 7°48'S, 39°14'W, 4 Apr.
2000, E.A. Rocha et al. (K, photos); Mun. Ouricuri, 28 Feb. 1984, G. Costa-Lima 17 (IPA, HRB); Mun. São
José do Belmonte, 2 Mar. 1981, A. Peruci 1 (IPA); Mun. Parnamirim, 23 km along road to Petrolina, 25 May
1984, F. Araújo 100 (IPA, PEUFR); l.c., 18 km from the town, Sítio Favela, 11 Feb. 1993, F.A.R. Santos 72
(ALCB, HUEFS); Mun. (?) Ibimirim, [Açude] Poço da Cruz [Açude Engenheiro Francisco Saboia], 12 Apr.
1968, Andrade-Lima 68-5353 (IPA); Mun. Dormentes, road to Lagoa, 33 km from Dormentes, 19 Nov. 1992,
F.A.R. Santos 52 (PEUFR); Mun. Afrânio, 5 km N of Rajada, 8°47'S, 40°52'W, 6 Apr. 2000, E.A. Rocha et
al. (K, photos); Mun. Santa Maria da Boa Vista, 2 May 1971, E.P. Heringer 451 (IPA); Mun. Petrolina, 17
km S of Rajada, W of road BR 407, 8°55'S, 40°44'W, 6 Apr. 2000, E.A. Rocha et al. (K, photos); l.c., Distrito
de Nilo Coelho, 9°15–30'S, 40°30–45'W, 4 Apr. 1991, P.E. Nogueira et al. 285 (SPF).
bahia: W Bahia, Mun. Vanderlei, 12 km N of road BR 242 towards the town, 19 July 1989, Taylor &
Zappi (obs.); Mun. Santana, 30 km from the town towards Porto Novo, 16 Jan. 1991, Taylor et al. 1435
(K, HRCB, ZSS, CEPEC); Mun. Sítio do Mato, Horst & Uebelmann 208, cult. Braun (ZSS); Mun. Serra
do Ramalho, Serra do Ramalho, 13°31'S, 43°45'W, 2001, J. Jardim (CEPEC); Mun. Carinhanha, 14°1'S,
43°43'W, 2001, J. Jardim 3500 (CEPEC); N & NE Bahia, Mun. Casa Nova, c. 50 km from Remanso
towards Petrolina, 10 Feb. 1990, Zappi 221 (SPF); Mun. Curaçá, 43 km N of Barro Vermelho towards
Curaçá, 7 Jan. 1991, Taylor et al. 1370 (K, HRCB, ZSS, CEPEC) — giant form to 3.5 m high (!); Mun.
Juazeiro, Km 12 from town towards Sento Sé, 8 Dec. 1971, Andrade-Lima et al. 1209 (IPA, MAC); Mun.
Sento Sé, 158 km W of Juazeiro, then 15.5 km S on road to Cabeluda, 11 Jan. 1991, Taylor et al. 1393
(K, HRCB, ZSS, CEPEC); Mun. Uauá, 12.5 km N of town towards Patamuté, 6 Jan. 1991, Taylor et al.
1366 (K, HRCB, ZSS, CEPEC); Mun. Canudos, 1983, Braun 464 (ZSS); Mun. Jaguarari, S of the border
of Mun. Juazeiro, 64 km S of Juazeiro, 8 Jan. 1991, Taylor et al. 1374 (K, HRCB, ZSS, CEPEC); l.c.,
between Barrinha and Flamengo, Serra dos Icós, 27 Nov. 1968, I. Pontual 68-781 (IPA); Mun. Monte
Santo, c. 5 km N of town towards Uauá, 25 Aug. 1996, L.P. de Queiroz 4609 (HUEFS); Mun. (?), 10°53'S,
41°35'W, 13 July 2000, G. Charles (photos); Mun. Queimadas, 10°55'S, 39°36'W, 17 Nov. 1986, L.P. de
Queiroz 1159 (HUEFS); Mun. Araci, 18 Jan. 1994, M.A.S. das Neves 87 (HUEFS); l.c., 14.5 km N of
town towards Tucano, 5 Jan. 1991, Taylor et al. 1355 (K, HRCB, ZSS, CEPEC); Mun. Morro do Chapéu,
near Buracão, 19 km from the town, 11°38'57"S, 41°16'16''W, 16 Mar. 1996, Giulietti et al. 2440 (ALCB,
UEC); l.c., 15 km W of town beside road BA 052, 5 Aug. 2002, Taylor (obs.); Mun. Cafarnaum/Mun.
Morro do Chapéu, 26 Sep. 1965, A.P. Duarte 9551 (HB, RB); Mun. Canarana, 6 km E of town towards
Cafarnaum, 24 Aug. 1988, Eggli 1277 (ZSS); cent. & S Bahia, Mun. Ibitiara, Serra Malhada, 38 km W of
Seabra towards Ibotirama, 1050 m, 25 Aug. 1988, Eggli 1292 (ZSS); l.c., 52 km W, 1020 m, 14 Jan. 1991,
Taylor et al. 1419 (K, HRCB, ZSS, CEPEC); Mun. Botuporã, c. 50 km S of Macaúbas on road to
Paramirim, 26 Aug. 1988, Eggli 1296 (ZSS); Mun. Bom Jesus da Lapa, Km 8 on road to Ibotirama, then
3 km E, 17 Apr. 1983, Leuenberger et al. 3080 (CEPEC, B); 28 km SE of Bom Jesus da Lapa on road to
Caitité, 43°13'W, 13°23'S, 16 Apr. 1980, Harley et al. 21424 (SPF, K); Mun. Abaíra, Morro do Zabumba,
Engenho de Baixo, 13°18'S, 41°48'W, 13 Mar. 1992, Stannard et al. in Harley 51947 (SPF, CEPEC,
HUEFS, K); Mun. Marcionílio Sousa, Machado Portela, see collection cited for hybrid with A. penicillata,
below; Mun. Riacho de Santana, 22 km from town towards Bom Jesus da Lapa, 16 Apr. 1983, Leuenberger
et al. 3073 (CEPEC, B); Mun. Livramento do Brumado, 10 km S, 14 Mar. 1991, G.P. Lewis s.n. (K,
photos); Mun. Caitité, estrada para Igaporã, 14°1'S, 42°33'W, 31 Mar. 1984, G.B.A. Bohrer 29 (HRB);
l.c., hill above town with Telebahia tower, 27 Aug. 1988, Eggli 1303 (ZSS); Mun. Dom Basílio, 13°59'S,
41°42'W, 1999, M. Machado et al. (obs.); Mun. Tanhaçu, 44 km N of Anagé (Anajé) towards Sussuarana,
27 Jan. 1965, G. Pabst 8658 & E. Pereira 9769 (HB, RB); Mun. Guanambi, 5–10 km E on road BR 030,
15 Jan. 1995, Hatschbach et al. 61900 (MBM, K); l.c., 19 km NE on road to Caitité, 25 July 1989, Zappi
169A (SPF); Mun. Palmas de Monte Alto, hill with Telebahia tower, 28 Aug. 1988, Eggli 1313 (ZSS);
Mun. Aracatu, 33 km along road BR 030 from Brumado towards Sussuarana, 3 Feb. 1991, Taylor et al.
1544 (CEPEC, K, HRCB, ZSS); Mun. Anagé, 10 km N towards Sussuarana, 17 Aug. 1988, Eggli 1181
(ZSS); Mun. Malhada, Iuiú, c. 2 km S of town, 28 Aug. 1988, Eggli 1317 (ZSS); Mun. Urandi, c. 3 km
S of town centre, near railway line, 22 July 1989, Taylor & Zappi (obs.).
minas gerais: Cent.-N Minas Gerais, Mun. Januária/Mun. Manga, reported by Braun & Esteves
Pereira (1995a); Mun. Monte Azul/Espinosa, 5.5 km E of Monte Azul, beyond Vila Angical, 28 Jan.
ARROJADOA
1991, Taylor et al. 1459 (K, HRCB, ZSS, BHCB); l.c., ‘Serra do Espinhaço’, 14 Mar. 1995, Hatschbach
et al. 61889 (MBM, K); Mun. Rio Pardo de Minas (?), ‘W of Rio Pardo’, 1979, Braun & Heimen 83
(ZSS); Mun. Varzelândia, c. 42 km E of town towards the Rio Verde Grande, 23 July 1989, Taylor &
Zappi (obs.); Mun. Porteirinha, 8 km S of town on road to Riacho dos Machados, 15°49'S, 43°4'W, 7
Nov. 1988, Taylor & Zappi in Harley 25516 (K, SPF).
Concern, but its extensive habitat continues to decline.
This extremely variable species comprises many locally distinct forms but cannot be
conveniently divided into a manageable and meaningful number of infraspecific taxa. A
part of this variation can be attributed to gene exchange with species Nos 2 and 3 (see
below), but much of it is probably inherent and independent of such influence, occurring
in areas remote from the known ranges of its congeners. A. rhodantha is one of the most
characteristic cactus species of the central and southern caatingas, but is absent from north-
eastern Minas Gerais (Rio Jequitinhonha drainage), where A. penicillata occurs.
4b. Arrojadoa (?) sp. nov. (‘A. marylanae’, nom. prov.)
Possibly keying out here is a remarkable plant recently discovered by Marylan Coelho
from the Univ. Estadual do Sudoeste da Bahia and communicated by Marlon Machado. It
has a massive erect, 27–28-ribbed stem (cf. Coleocephalocereus goebelianus), that rarely
branches, and dense, porrect, needle-like spination. The ring cephalia are numerous and
closely spaced, but with wool and bristles similar to C. goebelianus and the flowers are quite
different to those of other Arrojadoa species, the perianth strongly resembling that of
Coleocephalocereus or Melocactus. At present it is known only from southern Bahia, where it
grows in large numbers on quartzitic ridges together with Espostoopsis dybowskii, as follows:
Mun. Tanhaçu, nr Sussuarana, Serra Escura, Sep. 2001, Avaldo de Oliveira Soares Filho s.n.
(photos), 19 April 2003, Taylor et al. (K, photos). Its generic disposition is at present
unclear, but there can be little doubt that it is a new taxon, which could be some kind of
stabilised intergeneric hybrid (between Arrojadoa and Coleocephalocereus ?). Plate 37.3–4.
2 3: Arrojadoa dinae subsp. dinae (?) Arrojadoa penicillata

Vegetatively intermediate between the presumed parents, the ends of some stems strongly expanded
beneath the cephalium. Flowers described as pink, green within.
bahia: Mun. Palmas de Monte Alto / Mun. Guanambi, road BR 030, 31 Mar. 1984, 14°14'S, 43°1'W,
J.E.M. Brazão 303 (HRB, K photo).
Although A. penicillata is not otherwise recorded west of the Serra do Espinhaço in southern
Bahia, the above-cited collection may confirm its presence. The characters displayed by this
presumed hybrid seem best explained as having come from A. penicillata and A. dinae.
2 4: Arrojadoa dinae subsp. dinae Arrojadoa rhodantha

A. albiflora Buining & Brederoo in Succulenta 54: 21–27 (1975). Holotype: Brazil, Bahia, Mun. Urandi,
1070 m, Horst 401 (U).
PILOSOCEREUS
Intermediate between the parents, but variable in flower colour. See Ritter (1979: Abb. 61) for a good
illustration.
bahia: (?) Mun. Caetité, São Francisco, caminho para Lagoa Real, 14°S, 42°12'W, 8 Feb. 1997, B.
Stannard et al. in PCD 5224 (ALCB, K, CEPEC, UEFS, SPF); Mun. Urandi, mountains above town,
between sandstone rocks, 1964, Ritter 1244 (U); l.c., 1070 m, Horst HU 401 (U), 6 Oct. 1985, cult. ZSS,
10 May 1989, Horst & Uebelmann 401 (ZSS).
Probably to be found at other sites where the species are almost or quite sympatric, eg.
east of Monte Azul (MG).
3 4: Arrojadoa penicillata Arrojadoa rhodantha

Sometimes forming a complete range of intermediates between the parents in places within the areas of
sympatry cited below.
Caatinga in the drainage of the Rio Paraguaçu and Rio Brumado/Rio de Contas, 300–450 m, southern
Bahia.
bahia: Mun. Marcionílio Sousa, Machado Portela, 19 June 1915, Rose & Russell 19913 p.p. (NY); Mun.
Maracás, c. 18 km S of the Contendas–Maracás road towards Porto Alegre, 4 Feb. 1991, Taylor et al. (obs.);
Mun. Aracatu, 33 km on road BR 030 from Sussuarana, 3 Feb. 1991, Taylor et al. 1544 (K, HRCB, ZSS,
CEPEC); (?) Mun. Brumado, 7 km N of town on road to Livramento, 14°8'S, 41°40'W, 22 Nov. 1988,
Taylor & Zappi in Harley 25538 (K, SPF, CEPEC) — could also be merely a very slender form of A. rhodantha.
20. PILOSOCEREUS Byles & G. D. Rowley
in Cact. Succ. J. Gr. Brit. 19: 66–69 (1957). Type: Pilosocereus leucocephalus (Poselg.) Byles
& G. D. Rowley [NE Mexico].
Including Pseudopilocereus Buxb. (1968).
Literature: Zappi (1994).

Shrubby or tree-like, rarely epiphytic (on carnaúba palm trunks), 0.3–10.0 m; branched or not above
ground; vascular cylinder strongly to weakly woody; tissues mostly extremely mucilaginous; branches
showing more or less constant growth, rarely constricted; ribs triangular, (3–)4–28. Areoles with felt and
generally with long hairs, spination diverse; spines straight, rarely curved at base, opaque to translucent.
Fertile part of stem not or various differentiated, when strongly so flower-bearing areoles with long hairs
and/or bristly spines, lateral cephalium superficial, rarely sunken into the branch. Flowers 2.5–9.0 × 2–7
cm, anthesis starting at evening; flowers dull or sometimes bright pink-red without, smooth to slightly
striate, glabrous; pericarpel with few and minute bract-scales, symmetry of flowers dependent on shape
of flower-tube, which can be straight or curved, terete or infundibuliform, distal half or third with broad
bract-scales; outer perianth-segments thick, inner segments thin, white or rarely pinkish; nectar-chamber
broad, ± protected by the innermost filaments; stamens many, anthers 1.2–2.5 mm, forming a compact
mass; style 25–65 mm, stigma-lobes 8–12, exserted or included in relation to the anthers. Fruit depressed-
globose, very rarely globose, always dehiscent by lateral, abaxial, adaxial or apical slit, floral remnants
persistent, blackening, pendent or rarely erect; pericarp green, purplish, brownish, magenta, deep red or
wine-coloured, smooth, striate or rugose; funicular pulp solid, white, red or magenta. Seeds 1.2–2.5 mm,
dark-brown or black, cochleariform, rarely hat-shaped, hilum-micropylar region differentiated,
intercellular depressions present to absent, testa-cells flat, convex or conic, cuticular folds fine, coarse or
absent (for seed SEM images of all species treated below, except P. azulensis, see Zappi 1994).
PILOSOCEREUS
A genus of some 37 species in Mexico, the Caribbean (incl. Florida Keys), Venezuela,
Suriname, Guyana, Peru, Ecuador, Brazil and Paraguay. Its range in Brazil includes
Roraima, Pará, Maranhão, Piauí, Ceará, Rio Grande do Norte, Paraíba, Pernambuco,
Alagoas, Sergipe, Bahia, Minas Gerais, Espírito Santo, coastal Rio de Janeiro, northern
São Paulo, Tocantins, Goiás, Mato Grosso and Mato Grosso do Sul. This is the largest
and most important genus of Cactaceae in Eastern Brazil, represented by at least 20
species, plus various heterotypic subspecies, and occurring in a wide variety of vegetation
types, though sometimes restricted to rock outcrops within habitats such as cerrado and
Mata atlântica. On occasions certain species may even dominate the vegetation in which
they occur, e.g. P. gounellei, P. catingicola and P. pachycladus. The Brazilian species are
classified into two subgenera and various species-groups. Plates 38–49.3.
1. Branching candelabriform, the new branches arising near the apex of the subtending stems;
floral remnants erect to pendent, not deeply immersed in apex of fruit, forming a circular
insertion point (Subg. Gounellea) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
1. Branching erect and/or plants branched only at base, the new branches first developing well
below the apex of the stems subtending them; floral remnants pendent, deeply immersed in
apex of fruit, forming a linear insertion point (Subg. Pilosocereus) . . . . . . . . . . . . . . . . . . . . . . . . 3
2. Treelike, with a well-defined trunk, lacking branches near base; ribs 4–7; flower-bearing
areoles without long hairs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1. tuberculatus
2. Shrublike, usually without a well-defined trunk, branching near base; ribs (8–)10–15; flower-
bearing areoles with abundant, silky, long hairs enveloping the flowers . . . . . . . . . . 2. gounellei
3. Treelike or shrubby, branching above the base; vascular cylinder strongly to moderately woody . . . 4
3. Shrubby, branching only at base when undamaged; vascular cylinder weakly woody . . . . . . . . .18
4. Apical, subapical and flower-bearing areoles without conspicuous long hairs (except when
young) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
4. Apical, subapical and flower-bearing areoles with abundant long hairs . . . . . . . . . . . . . . . . . . . . 7
5. Epidermis covered in blue wax . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8. pentaedrophorus
5. Epidermis greyish, pale or olive-green, not strongly waxy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
6. Spines brown to greyish, opaque; flower-buds acute; flower opening wide, 6–7 × 4–5 cm,
tube straight (restinga, S Bahia & SE Brazil) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5. arrabidae
6. Spines golden to translucent yellow; flower-buds obtuse; flower narrow, 5.8–6.0 × 3.5–3.7 cm,
tube curved (caatinga and littoral zone, Ceará, Piauí and S Maranhão) . . . . . . 7. flavipulvinatus
7. Epidermis rough, green to grey-green, not covered in wax . . . . . . . . . . . . . . . . . . 10. floccosus
7. Epidermis smooth, frequently bluish or obviously covered in wax . . . . . . . . . . . . . . . . . . . . . . . 8
8. Areoles contiguous, difficult to isolate from one another . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
8. Areoles > 2 mm apart, distinct . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
9. Flower-bearing areoles with scarce, white long hairs; young spines golden, flexible
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13. magnificus
9. Flower-bearing areoles with abundant dark yellow to reddish brown long hairs; spines
dark brown to blackish . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11. fulvilanatus
10. Spines golden, translucent . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
10. Spines brownish, reddish or blackish, opaque . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
11. Long hairs equally abundant on both flower-bearing areoles and vegetative areoles;
flower-tube narrow, curved, pinkish or reddish without . . . . . . . . . . . . . . . . . 9. glaucochrous
11. Long hairs more abundant on the flower-bearing areoles; tube ± straight, broad at apex,
greenish, brownish or dark purplish brown . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
PILOSOCEREUS
12. Ribs 5–12; central spines thicker and longer than the radials; stem blue
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12a. pachycladus subsp. pachycladus
12. Ribs > 12; central spines delicate, nearly indistinguishable from the radials, stem green
or bluish . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .13
13. Fertile part of stem with golden bristles to 3–4 cm and few long hairs; ribs (15–)18–26
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17. multicostatus
13. Fertile part of stem without or with few bristles, long hairs white, silky; ribs 13–19 . . . . . . . . . 14
14. Spines golden; ribs wider than high; fertile part of the stem subapical, not strongly modified
(caatinga, N Bahia northwards) . . . . . . . . . . . . . . 12b. pachycladus subsp. pernambucoensis
14. Spines pinkish yellow to brownish; ribs higher than wide; fertile part of stem lateral,
a true, sunken cephalium sometimes present (Bambuí limestone outcrops, cent.-N Minas
Gerais and SW Bahia) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20. densiareolatus
15. All stems with 7 or more ribs (seedlings excepted), and/or flowers > 4 cm in diameter at
full anthesis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
15. Ribs 4–6, or some stems with up to 7 ribs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
16. Areoles 8–16 mm apart on the ribs; seeds 2–2.3 mm (NE Brazil) . . . . . . . . . . . . . 3. catingicola
16. Areoles 5–7 mm apart on the ribs, seeds 1.5–1.6 mm (NE Minas Gerais) . . . . . . . . . 4. azulensis
17. Treelike or shrubby, primary branches 4–6-verticillate; flower-buds acute, with triangular
bract-scales, flowers ± solitary, 4.7–7.0 cm in diameter, tube straight, wide (NE Brazil)
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3. catingicola
17. Shrubby, sparsely branched, or if treelike branching not as above; flower-buds obtuse, with
obovate to truncate bract-scales, flowers aggregated, 2.5–3.0 cm in diameter, tube curved,
narrow (SE Bahia & SE Brazil) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6. brasiliensis
18. Ribs (5–)6–8, transverse folds visible; flower-bearing areoles not differentiated; spines
opaque, brown to grey, central spines well differentiated from the radials (restinga) . . . 5. arrabidae
18. Ribs 8–23, transverse folds not visible; flower-bearing areoles ± differentiated; spines
red, dark brown or golden, translucent, central spines not very different from the radials
(cerrado, caatinga and campo rupestre) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
19. Stem blue-green or strongly glaucous; flower-tube infundibuliform; ribs 8–17 . . . . . . . . . . . . . 20
19. Stem pale to dark green; flower-tube cylindric, straight to curved; ribs 12–28 . . . . . . . . . . . . . 21
20. Fruit dehiscent by an apical transverse slit; pericarp rugose, red to wine-coloured when ripe;
seeds with flat testa-cells (Serra do Espinhaço, Minas Gerais) . . . . . . . . . . . . . . . . . 15. aurisetus
20. Fruit dehiscent by lateral slit, pericarp smooth, green or bluish; seeds mostly with domed
to highly conical testa-cells (Bahia, S Piauí) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14. machrisii
21. Flower-bearing areoles strongly differentiated, forming a lateral cephalium sunken into the
branch, with abundant wool and bristles to 3–6 cm; stem epidermis pale green . . . 19. chrysostele
21. Flower-bearing areoles not strongly differentiated, appearing randomly or at apex of branches,
usually with bristles and some hairs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
22. Vegetative areoles with long hairs; fruit pulp white; seeds dull, testa-cells domed to conic
(SEM: cuticular folds coarse, dense); flowers c. 2.5 cm wide, pinkish without (Bahia, slopes
E of the Rio São Francisco) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16. aureispinus
22. Vegetative areoles without long hairs; fruit pulp magenta; seeds shiny, testa-cells slightly
domed to flat (SEM: cuticular folds scarce to absent); flowers c. 3 cm wide or more,
greenish-white without . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
23. Flowers curved, > 5 cm, flower-buds obtuse before anthesis (SE Piauí northwards)
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18. piauhyensis
23. Flowers straight, < 5 cm, flower-buds acute before anthesis (NE Minas Gerais)
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17. multicostatus
PILOSOCEREUS
Subg. Gounellea Zappi (Nos 1 & 2). Endemic to the caatinga and mata calcária of NE Brazil
and northern Minas Gerais.
1. Pilosocereus tuberculatus (Werderm.) Byles & G. D. Rowley in Cact. Succ. J. Gr.

Brit. 19(3): 69 (1957). Type: Brazil, Pernambuco, 1932, Werdermann (B†). Lectotype
(Zappi 1994): l.c., Serra Negra, ‘900 m’, Mar. 1932, Werdermann, l.c. infra, photograph,
p. 21 (iconotype).
Pilocereus tuberculatus Werderm., Bras. Säulenkakt.: 101 (1933). Pseudopilocereus tuberculatus (Werderm.)
Buxb. in Beitr. Biol. Pflanzen 44: 253 (1968).
vernacular names. Caxacubri, Mandacaru-de-laço.
Tree, 2–5(–6) m, with a spread of up to 6 m diam., trunk to c. 20 cm diam.; vascular cylinder strongly
woody up to the apex of mature axes; branches 3–6 cm diam., apex erect to inclined, older parts ±
horizontal, new branches arising subapically, epidermis olive-green, with pruinose wax on the tips when
young; ribs 4–7(–8–12, in seedlings), 1–2 × 0.5–2.0 cm, with sinuate sinuses and pronounced, oblique,
transverse folds above the areoles. Areoles 8–10 × 6–7 mm, 2–3.3 cm apart, situated upon prismatic stem-
projections (podaria), felt at first light brown, becoming black. Spines opaque, at first light brown, sharp,
later greyish, brittle, porrect, centrals 3–5, 30–65 mm, lowermost generally longest, often decurved near
base, radials 10–12, 4–7(–10) mm, adpressed, decurved. Fertile part of stem not differentiated, subapical.
Flower-buds acute 1–2 days before anthesis, with active extra-floral nectaries; flowers 6–6.7 × c. 3 cm;
pericarpel obconic, tube 45 × 20–23 mm at its widest point, ± parallel-sided, straight, nearly terete, olive-
green, striate, distal third with a few acute bract-scales; style 40–42 × 2–3 mm, tapering; stigma-lobes c. 8,
6 × 0.7 mm; ovary locule obtriangular in longitudinal section. Fruit to 4 cm diam., globose to depressed,
dehiscent by a lateral slit, floral remnant erect, with rounded, shallow insertion; pericarp olive-green to
purplish, rugose, with active extra-floral nectaries; funicular pulp magenta. Seeds (1.2–)1.3–1.4(–1.5) ×
0.9–1.0 mm, testa-cells domed, with intercellular depressions and fine cuticular folds.
Central-southern caatinga element: in dense or sparse caatinga vegetation, on fine, white or reddish, sandy
soil (especially of the Cipó series), c. 200–790 m, northern Bahia, western Sergipe and Pernambuco.
Endemic to the core area within North-eastern Brazil. Map 34.
pernambuco: Mun. Sertânia, just N of border with Mun. Ibimirim, road PE 360, 8°27'S, 37°28'W, 8
Apr. 2000, E.A. Rocha et al. (obs.); Mun. Floresta, 20 km E of Airi on road PE 360 towards Ibimirim, N
side of Serra Negra, 12 Feb. 1991, Taylor & Zappi 1623 (HRCB, K, ZSS, UFP); l.c., 8°33'S, 38°1'W, 8
Apr. 2000, E.A. Rocha et al. (K, photos); l.c., fazenda ± 35 km após Ibimirim, 4 Apr. 1989, Rodal 74
(IPA); Mun. Ibimirim, triângulo irrigado de Moxotó, Nov. 1992, F.A.R. Santos 39, 44 (PEUFR, K,
photo); Mun. Buíque, Brejo de São José, 8°37'S, 37°10'W, 15 Nov. 1994, Rodal 365 (PEUFR, K); Mun.
Nova Petrolândia, 12 km NE of turnoff to the old town, 12 Feb. 1991, Taylor & Zappi 1621 (HRCB,
K, ZSS, UFP); Mun. Petrolina, 5 km from the city towards Recife, 1966, Hunt (obs.).
sergipe: Mun. Canindé de São Francisco, povoado Curituba, 23 Apr. 1981 & 1992, M. Fonseca et al.
480 (ASE 895, K, photo).
bahia: Mun. Remanso, 20 km E of town towards Casa Nova, 10 Feb. 1990, Zappi 219 (HRCB); l.c.,
47 km E, 9°22'S, 41°48'W, 7 Apr. 2000, E.A. Rocha et al. (obs.); Mun. Glória, povoado de Brejo do
Burgo, 28 Nov. 1992, & 26 Aug. 1995, Bandeira 74 (ALCB), Bandeira 263 (ALCB, K, HUEFS); Mun.
Santa Brígida, c. 10 km W of junction for the town with road BR 110, Raso da Catarina, c. 9°29'S,
38°12'W, 19 Dec. 1993, L.P. de Queiroz & Nascimento 3736 (K, HUEFS); Mun. Canudos, Reserva
Biológica de Canudos, 10 Feb. 2001, T. Andrade (K, photos); Mun. Barra, dunes by the Rio São
Francisco, 10°47'16"S, 42°46'22''W, 23 June 1996, Giulietti et al. in PCD 2965 (K, ALCB, UEC); l.c.,
Ibiraba, road to Brejos, 10°48'S, 42°50'W, 23 Feb. 1997, L.P. de Queiroz 4813 (K, HUEFS); l.c., dunes
by the Rio São Francisco, 5 Oct. 1987, M.T. Rodrigues s.n. (SPF 48476), 1988, P.L. Bernardes 189
(SPF); Mun. Morro do Chapéu, between Gruta dos Brejões and Olho D’Água do Facundo, 4 Aug.
2002, M. Machado (photos); Mun. Ourolândia, NE of Olho D’Água do Facundo, 4 Aug. 2002, Taylor
PILOSOCEREUS
(obs.); Mun. Araci, 14.5 km N of town towards Tucano, 5 Jan. 1991, Taylor et al. 1353 (K, HRCB,
ZSS, CEPEC).
Concern, but some habitat is declining due to agricultural development, eg. for plantations of Opuntia
ficus-indica.
This species shares unique apomorphic characters with P. gounellei, such as the sub-apical
branching-pattern and the morphology of the fruits, lacking a deeply sunken floral
remnant. This pair of species are the only representatives of Subg. Gounellea and are
endemic to Eastern Brazil.
P. tuberculatus is characteristic of the sand-dunes and sandy soils of the Rio São
Francisco, but is also known from NE Bahia, between Araci and Glória, including the
Raso da Catarina. A collection by Luetzelburg (No. 19, M!), bearing a label indicating
southern Paraíba (Monteiro), is clearly a labelling error, since the vernacular name given
on the label, ‘Rabo de Raposa’, was applied to Harrisia adscendens by Luetzelburg, which
is common about Monteiro, where P. tuberculatus is absent. Andrade-Lima (1981: 159)
lists this species as one of those characteristic of his caatinga type No. 5, which is found
on the ‘Cipó’ soil series.
The biology of P. tuberculatus is interesting in that it is able to secrete nectar from the
outer bract-scales of the flower-buds, tube and fruit, which attracts ants. In large
specimens, these insects inhabit the hollow pith of old, dead branches, suggesting a
symbiotic relationship of attraction/defence. The slight damage or sudden movement of
a branch of this plant is immediately followed by a quick defence reaction by the ants,
which run out from inside the dead branches to attack the supposed agressor.
2. Pilosocereus gounellei (F. A. C. Weber) Byles & G. D. Rowley in Cact. Succ. J.

Gr. Brit. 19: 67 (1957). Type: Brazil, Pernambuco, Gounelle s.n. (P†). Neotype (Zappi
1994): Brazil, Paraíba, São Gonçalo [Várzea de Souza], Jan. 1936, P. Luetzelburg 26921
(M; K, photo.; IPA).
Pilocereus gounellei F.A.C. Weber in K. Schum., Gesamtbeschr. Kakt.: 188 (1897). Cephalocereus gounellei
(F.A.C. Weber) Britton & Rose, Cact. 2: 34 (1920). Pseudopilocereus gounellei (F.A.C. Weber) Buxb.
in Beitr. Biol. Pflanzen 44: 252 (1968).
Cereus setosus Gürke in Monatsschr. Kakt.-Kunde 18: 19 (1908), nom. illeg. (Art. 53), non Loddiges
(1832). Pilocereus setosus Gürke in ibid. 18: 52–53 (1908). Type: Brazil, Bahia, Mun. Maracás,
‘Calderão’, [1906], Ule 1 (B [photo K, SPF ex B]).
Shrub, more rarely treelike, 0.5–4.0 m, with many branches arising from the main stem; vascular cylinder
moderately woody; branches with apex erect to inclined, older parts horizontal, new branches arising
sub-apically, epidermis olive-green, or strongly glaucous, especially when young; ribs 8–15, with sinuate
sinuses and transverse folds above the areoles. Areoles ovate, felt cream, grey or brownish, with white
long hairs to 35 mm, growth indeterminate near base of plant. [Spination: see subspecies.] Flower-buds
completely encircled by silky long hairs, obtuse 2 days before anthesis; flower 4–9 × 2.5–6.0 cm;
pericarpel obconic; tube 32–60 × 16.5–17.0 mm at base, widening towards apex to 20–25 mm, curved,
greenish brown to pinkish, slightly striated, with decurrent triangular bract-scales, these more frequent
and larger towards tube apex; style 2.6–3.5 × 2.5–1.2 mm, tapering, stigma-lobes c. 10, 4 mm long; ovary
locule obtriangular in longitudinal section. Fruit 3.5–4.5 × 4.5–6.0 cm, globose to depressed, dehiscent
PILOSOCEREUS
by adaxial or abaxial slit, floral remnant erect or pendent, with rounded, shallow insertion; pericarp
pinkish red, smooth or striate. Seeds (1.6–)1.7–1.8 × 1.4–1.6 mm, hat-shaped, symmetric, hilum-
micropylar region expanded, forming an angle of 70–80° with long axis of seed, testa-cells flat, without
intercellular depressions or cuticular folds.
The neotype chosen (Zappi 1994) was collected by Luetzelburg in the state of Paraíba
(Luetzelburg 26921), and is a form which agrees with Weber’s description, where the
spines are given as relatively short (1 cm long), this being somewhat atypical for the
species as a whole.
P. gounellei is the type of Subg. Gounellea, including its sister species, P. tuberculatus,
from which it differs in its mostly shrubby, not treelike habit, mature branches with a
higher number of ribs (8–)9–15, only moderately woody vascular cylinder, flowering
areoles with white silky hairs and very distinct seeds, which may be adapted for dispersal
by water (Zappi 1994). In the Brazilian Nordeste it is commonly known by the
vernacular name of xique-xique, and represents one of the most characteristic plants of
the caatinga, where, together with Cereus jamacaru DC. (vernacular: mandacaru) and
Tacinga inamoena (K. Schum.) N. P. Taylor & Stuppy (quipá), it is one of the most
common and widespread cacti of Eastern Brazil.
This species is divided in two subspecies:
1. Spines stout, strong, brownish to greyish, opaque, (0.9–)1.0–1.9 mm diam. near base, centrals
distinctly longer than the radials (NE Brazil) . . . . . . . . . . . . . . . . . . . . . . . . 2a. subsp. gounellei
1. Spines slender, fragile, golden to reddish, translucent, 0.25–0.6(–0.8) mm diam. near base, centrals
and radials ± equal (N Minas Gerais to W & N Bahia) . . . . . . . . . . . . . . . . 2b. subsp. zehntneri
2a. subsp. gounellei
vernacular names. Xique-xique, Alastrado.
Shrub; branches 5–9 cm diam.; ribs (8–)9–13, 7–12 × 12–20 mm. Spines (0.9–)1.0–1.9 mm diam. at base,
brown to greyish, opaque; centrals 1–5(–6), 1–13(–15) cm, ascending to porrect, the lowermost porrect
to reflexed, radials 12–15, 6–30 mm, adpressed. Flower-bearing areoles occupying the subapical region
of 1–2 ribs, with silky long hairs covering the flower-buds only. Fruit with magenta (rarely white)
funicular pulp.
Caatinga element: widely distributed, common and locally dominant in low, sparse caatinga and along road
sides, on shallow, rocky or sandy soils and gneissic/granitic outcrops/inselbergs, including those
surrounded by more humid vegetation, rarely as epiphyte in seasonally flooded carnaúba (Copernicia
prunifera) forest, near sea level to c. 1200 m, E Maranhão, Piauí, Ceará, Rio Grande do Norte, Paraíba,
Pernambuco, Alagoas, Sergipe and Bahia (N of 15°S). Endemic to North-eastern Brazil. Map 34.
maranhão: Mun. Barão de Grajaú, Morro do Cruziero, 3 Feb. 1990, Zappi 213A (HRCB).
piauí: N Piauí, Mun. Luís Correia, near Camurupim, gneissic inselberg, 19 Feb. 1995, Taylor (K,
photos); Mun. Cocal, W of junction between road PI 211 and BR 343, 19 Feb. 1995, Taylor (obs.); Mun.
Campo Maior, S of Rio Jenipapo, beside road BR 343, epiphytic in flooded carnaúba forest, 18 Feb.
1995, Taylor (K, photos); Mun. Dom Expedito Lopez, 1 Feb. 1990, Zappi 211A (SPF); Mun. Jaicós,
7°22'S, 41°9'W, 6 Apr. 2000, E.A. Rocha et al. (obs.); l.c., 74 km S of Picos, 6 Apr. 2000, E.A. Rocha et
al. (K, photos); Mun. Canto do Buriti and São Raimundo Nonato, Feb. 1990, Zappi & Taylor (obs.);
PILOSOCEREUS
Mun. Paulistana, 54 km N of town beside road BR 407, 6 Apr. 2000, E.A. Rocha et al. (K, photos).
ceará: Mun. Camocim, between Chaval and Camocim, gneissic inselbergs, 19 Feb. 1995, Taylor (K,
photos); Serra de Meruóca, between Coreaú and Alcântaras, 20 Feb. 1995, Taylor (K, photos); Mun.
Quixadá, 27 Jan. 1990, Zappi 206 (HRCB), l.c., 12 km SW of town, 5°2'S, 39°2'W, 16 Feb. 1985, Gentry
et al. 50209 (MO); Mun. Tauá, 30 Jan. 1990, Zappi 209A (HRCB); Mun. Milagres, 2 km E of town on
road BR 116, 4 Apr. 2000, E.A. Rocha et al. (obs.).
rio grande do norte: Chapada do Apodi, Pedra das Abelhas, 11 July 1960, Castellanos 22895 (RB); Mun.
Natal/Parnamirim, 5 km S of Ponta Negra, 2 Sep. 1946, Wurdack B-136 (US); Mun. Pau dos Ferros, 2 Apr.
2000, E.A. Rocha et al. (obs.); Mun. Tangará, W of town on road BR 226 to Currais Novos, 6°14'S,
35°51'W, 1 Apr. 2000, E.A. Rocha et al. (photo); Mun. Jardim do Seridó, 1 Apr. 2000, E.A. Rocha et al.
(obs.).
paraíba: Mun. Brejo do Cruz, serra W of town, 6°22'S, 37°28'W, 2 Apr. 2000, E.A. Rocha et al. (obs.);
Mun. Tacima, 6°36'S, 35°28'W, 1 Apr. 2000, E.A. Rocha et al. (K, photos); Mun. Guarabira, Serra da
Jurema, 6°50'S, 35°29'W, 1 Apr. 2000, E.A. Rocha et al. (K, photos); Mun. Patos, Vasconcellos Sobrinho
s.n. (IPA); Mun. Pocinhos, 8 July 1994, L.P. Felix & Miranda 6581 (PEUFR); Mun. Piancó, 5 km from
Sta Terezinha, 29 Nov. 1971, Andrade-Lima et al. 1085 (PA, RB); Mun. Itaporanga, Serra Água Branca,
7–10 Apr. 1994, M.F. Agra et al. 2823 (JPB); Mun. Maturéia, Serra de Teixeira, Pico do Jabre, 7°11'10"S,
37°8'22"–25'53''W, 1994, M.F. Agra et al. (JPB); Mun. Sumé, 29–30 June 1994, M.F. Agra et al. 2833,
2834, 2835 (JPB).
pernambuco: Mun. Bodocó, 22 km NE of Ouricuri, 7 Mar. 1970, Eiten & Eiten 10859 (SP); Mun.
Ouricuri, July 1984, Costa-Lima 023 (IPA); Parnamirim–Ouricuri, 4 Jan. 1961, Andrade-Lima 61-3587
(IPA); Mun. Serra Talhada, rock outcrop N of the town, 13 Feb. 1990, Zappi 222A (HRCB); Mun.
Salgueiro, 17 Feb. 1990, Zappi 227B (SPF, K); Mun. Taquaritinga do Norte, Gravatá do Ibiapina, 8 Aug.
2001, Zappi (obs.); Mun. Moreno, Tapera, Dec. 1928, Pickel 1814 (IPA); Mun. Caruaru, Sítio Gaibeira,
22 Oct. 1991, F.A.R. Santos 7, 8 (HUEFS, PEUFR); Mun. Afrânio, towards Piauí, 23 Apr. 1971,
Heringer et al. 309 (IPA, RB); Mun. Belém de São Francisco, 8°40'S, 38°44'W, 8 Apr. 2000, E.A. Rocha
et al. (obs.); Mun. Ibimirim, Poço do Ferro, 6 km from Ibimirim towards Floresta, 4 May 1989, M.J.
Rodal 38 (PEUFR); Mun. Petrolina, Distrito de Nilo Coelho, 9°15–30'S, 40°30–45'W, 4 Apr. 1991, P.E.
Nogueira et al. 284 (SPF).
alagoas: Mun. Santana do Ipanema, reported by Rizzini (1982: 67); Mun. Traipu, June 1990, R. de
Lyra-Lemos (MAC, photo).
sergipe: Mun. Canindé de São Francisco, Mun. Poço Redondo, Mun. Monte Alegre and Mun. Nossa
Senhora da Glória, Feb. 1991, Taylor & Zappi (obs.).
bahia: W Bahia, Mun. Barra, Ibiraba, road to Brejos, 10°48'S, 42°50'W, 26 Feb. 1997, L.P. de Queiroz
4873 (K, HUEFS); N, E & S Bahia, Mun. [Nova] Glória, Brejo do Burgo, 27 Aug. 1995, Bandeira 264
(HUEFS, K); l.c., Raso da Catarina, 31 Jan. 1982, E. de S.F. da Rocha et al. 799 (GUA); Mun. Remanso,
47 km E of town on road to Casa Nova, 9°22'S, 41°48'W, 7 Apr. 2000, E.A. Rocha et al. (obs.); Mun.
Juazeiro, 2–6 June 1915, Rose & Russell 19725 (US, NY); l.c., 14.5 km E of the town, 7 Jan. 1991, Taylor
et al. 1372A (K, ZSS, photos); Mun. Jaguarari, Barrinha, 7–8 June 1915, Rose & Russell 19789 (US, NY);
l.c, 71 km S of Juazeiro on road BR 407 towards Senhor do Bonfim, 15 July 1989, Zappi 126 (SPF,
HRCB); Mun. Araci, 18 Jan. 1994, M.A.S. das Neves 83 (HUEFS); Mun. Queimadas, 9–11 June 1915,
Rose & Russell 19846 (US, NY); Mun. Boa Vista do Tupim, 27 Apr. 1994, L.P. de Queiroz 3879
(HUEFS); Mun. Itatim, Morro do Agenor, 12°42'S, 39°46'W, 26 Nov. 1995, F. França 1489 (HUEFS,
K, photo.); Mun. Milagres, July 1989, Taylor & Zappi (obs.); Mun. Marcionílio Souza, Machado Portela,
19–23 June 1915, Rose & Russell 19945 (US, NY); Mun. Jaguaquara [Toca da Onça], 27–29 June 1915,
Rose & Russell 20081 p.p. (US, NY); Mun. Paramirim, 5 km N of town on road to Macaúbas, 17 Jan.
1997, Hatschbach et al. 65897 (MBM, K); l.c., 43 km E of Livramento do Brumado, road to Paramirim,
13°38'S, 42°10'W, 28 Nov. 1988, Taylor in Harley 25556 (K, SPF, CEPEC); Mun. Maracás, 29 June
1993, L.P. de Queiroz 3261 (HUEFS); l.c., ‘Calderão’, [1906], Ule 01 (K, SPF photos ex B); Mun. Dom
Basílio, 12 Mar. 1991, Lewis s.n. (K, photos); Mun. Tanhaçu, 47 km N of Anajé, road to Sussuarana, 17
Aug. 1988, Eggli 1191 (ZSS); Mun. Palmas de Monte Alto, July 1989, Zappi s.n. (K, photos).
Found throughout most of North-eastern Brazil and common.
PILOSOCEREUS
Subspecies gounellei is characterized by stout and sometimes very long spines, to nearly 2
mm diam. and to 15 cm long. It received the nickname of ‘tyre-killer’ from Werdermann
(1933, 1942) and is typical of low, very dry caatinga on sandy or stony soil and common
on gneiss/granite outcrops. It can also be seen as an epiphyte on Copernicia palms in the
northern caatinga vegetation type defined in Andrade-Lima (1981: 160).
2b. subsp. zehntneri (Britton & Rose) Zappi in Succ. Pl. Res. 3: 43 (1994). Type (Zappi,
l.c.): Brazil, Bahia, district of Chique-Chique [Xique-Xique], Serra de Tiririca, Nov.
1917, Zehntner s.n. (US, lecto.; NY, lectopara.).
Cephalocereus zehntneri Britton & Rose, Cact. 2: 35 (1920). Pilocereus gounellei var. zehntneri (Britton &
Rose) Backeb. in Backeberg & Knuth, Kaktus-ABC: 331 (1935, publ. 1936). Pilosocereus gounellei var.
zehntneri (Britton & Rose) Byles & G. D. Rowley in Cact. Succ. J. Gr. Brit. 19: 67 (1957);
Backeberg, Die Cactaceae 4: Abb. 2283 (1960). P. zehntneri (Britton & Rose) F. Ritter, Kakt.
Südamer. 1: 74–75 (1979). Pseudopilocereus zehntneri (Britton & Rose) P. V. Heath in Calyx 4: 141
(1994).
P. superfloccosus Buining & Brederoo in Cact. Succ. J. (US) 46: 60–63, excl. figs 1 & 2 (1974). Pilosocereus
superfloccosus (Buining & Brederoo) F. Ritter, Kakt. Südamer. 1: 84 (1979). Holotype: Brazil, Bahia,
Mun. Santana, 8 km NW of Porto Novo, 460 m, 1975, Horst 394 (U).
P. braunii E. Esteves Pereira in Kakt. and. Sukk. 38: 132 (1987) and ibid. 40: 6–13 (1989). Holotype:
Brazil, Bahia, 1979, P. J. Braun 70 (ZSS, K, iso.).
vernacular names. Xique-xique-das-pedras, Chique-chique, Cheque-cheque.
Shrub or treelike, with a definite trunk to 1 m or more, branches 3.7–7.0 cm diam.; ribs 9–15, 3–8 × 6–13
mm. Spines 0.25–0.6(–0.8) mm diam. at base, whitish or golden to reddish, translucent, centrals 4–10,
10–20 mm, porrect to reflexed, radials 10–20, 7–10(–20) mm, adpressed. Flower-bearing areoles slightly
to strongly modified, apical and subapical, on 1–3 ribs, often forming a true lateral cephalium sunken into
the branch, with greyish white long hairs to 4 cm. Fruit with white or magenta funicular pulp.
Rio São Francisco caatinga element: rupicolous, on outcrops of Bambuí limestone (and sandstone in
northern part of range), c. 450–1000 m, northern Minas Gerais and central-western and northern Bahia.
bahia: W Bahia, Mun. Santa Maria da Vitória, Santana and Porto Novo, on limestone, 1979, Braun 70
(ZSS, K, UFG); Mun. Santana, 16 Jan. 1991, Taylor et al. 1433 (K, HRCB, ZSS, CEPEC); l.c., 28 km
S of Santana on road to Santa Maria da Vitória, 15 Jan. 1991, Taylor et al. 1425 (K, HRCB, ZSS,
CEPEC); Mun. Santa Maria da Vitória, 1983, Braun 371 (ZSS T-06122); Mun. Cocos, W of town,
reported by Braun & Esteves Pereira (1999b). N Bahia, Mun. Sento Sé, Limoeiro, ‘Serra da Mimosa’
[Minas do Mimoso], 1985, Horst & Uebelmann 458 (ZSS); Mun. Umburanas, region of Delfino, NW of
Lagoinha, road to Minas do Mimoso, 41°20'W, 10°22'S, 950–1000 m, 4 Mar. 1974, Harley 16729 (K);
Mun. Xique-Xique/Itaguaçu da Bahia (?), Serra da Tiririca, Nov. 1917, Zehntner s.n. (US, NY); Mun.
Morro do Chapéu, APA Gruta dos Brejões, 4 Aug. 2002, Taylor (K, photo); l.c., limestone escarpment
forming E bank of Rio Salitre north of Brejão (Formosa), 5 Aug. 2002, Taylor (obs.); l.c., 21–22 km W
of the town towards América Dourada, arenitic rocks, 900 m, 13 Jan. 1991, Taylor et al. 1405 (K, HRCB,
ZSS, CEPEC); l.c., 2 km E of América Dourada, SE bank of Rio Jacaré (Rio Vereda do Romão
Gramacho), Bambuí limestone, 13 Jan. 1991, Taylor et al. 1408A (K, ZSS, photos); Mun. Ibotirama, road
BA 242, 12°45'S, 43°10'W, 2001, L. Aona 738 (UEC, photo); cent.-S Bahia, Bom Jesus da Lapa (?),
1975, Andrade-Lima 75-8165 (IPA); Bom Jesus da Lapa, Morro da Lapa, 17 Apr. 1983, Leuenberger et al.
3077 (CEPEC); Mun. Iuiú, 2 km S of town, 21 July 1989, Zappi 157 (SPF, HRCB).
minas gerais: Mun. Manga (?), Mocambinho, Serra Azul, 15°6'S, 43°59'W, 18 Mar. 1998, I. Pimenta
(K, photos); Mun. Januária, 5 km W from the town, 5 Sep. 1985, Horst & Uebelmann 716 (ZSS); l.c., 18
km E of Rio São João, 5 Sep. 1985, Horst & Uebelmann 715 (ZSS); l.c., 3 km W of the town on road
PILOSOCEREUS
MG 479 towards Pandeiros, João Menegal, 11 Aug. 1988, Eggli 1136 (ZSS, HRCB); ibid., [Pedras de]
Maria da Cruz, Faz. São Francisco, 11 Aug. 1997, I. Pimenta (K, photos); Mun. Montes Claros, 13 km
N of town towards Januária, 16°38'S, 43°55'W, 27 Jan. 1991, Taylor et al. 1455 (K, HRCB, ZSS, BHCB).
Concern, but since many populations are found on limestone it may decline due to future quarrying
activities.
Ritter (1979), having visited Montes Claros (MG) and Bom Jesus da Lapa (BA),
recognized P. zehntneri as a good species and combined it under Pilosocereus. Not taking
account of this, in 1987, E. Esteves Pereira (l.c.) described the populations from Santana
and Bom Jesus da Lapa (BA) as a new species, P. braunii, based on the presence of a true
cephalium and glaucous epidermis. Study of diverse populations of P. zehntneri and P.
braunii suggests that there is clinal variation in these characters, which become less obvious
towards the eastern limits of its distribution, the floriferous areoles being much less hairy
and modified in populations from Montes Claros (MG) and Morro do Chapéu and
América Dourada (BA). It is clear that P. braunii is represented by populations of extreme
plants that belong to a more wide-ranging and variable taxon, recognized by Ritter
(1979) as P. zehntneri, and treated by Zappi (1994) as a subspecies of P. gounellei.
The recognition of two subspecies for P. gounellei is based on the absence of absolute
discontinuities between the taxa concerned. The incomplete geographical isolation of
these taxa in regions such as west of Morro do Chapéu (BA) explains the difficulties of
delimitation between them.
In localities such as those near Montes Claros, Varzelândia, Cocos, Santana and Bom
Jesus da Lapa, P. gounellei subsp. zehntneri is sympatric with P. densiareolatus (see notes
under this species), and has been confused with it. P. superfloccosus was based on a mixture
of material (including photographs) of P. gounellei subsp. zehntneri and P. densiareolatus, and
this name has been used by some authors for the latter species.
Subg. Pilosocereus (Nos 3–20):
PILOSOCEREUS ARRABIDAE Group (Nos 3–5)

This group ranges from the coastal restinga into the caatinga via the agreste.
3. Pilosocereus catingicola (Gürke) Byles & G. D. Rowley in Cact. Succ. J. Gr. Brit.
19: 66 (1957). Lectotype (Zappi 1994): Gürke, l.c., photograph p. 55 ‘Cereus catingicola
Gürke in den sandigen Gebieten der Catinga bei Bahia [de São Salvador] nach einer von
Herrn E. Ule aufgenommenen Photographie’.
Cereus catingicola Gürke in Monatsschr. Kakt.-Kunde 18: 84 (1908). Cephalocereus catingicola (Gürke)
Britton & Rose, Cact. 2: 56 (1920). Pilocereus catingicola (Gürke) Werderm., Bras. Säulenkakt.:
104–105 (1933). Pseudopilocereus catingicola (Gürke) Buxb. in Beitr. Biol. Pflanzen 44: 252 (1968).
Pilocereus arenicola Werderm., ibid.: 109 (1933). Pilosocereus arenicola (Werderm.) Byles & G. D. Rowley in
Cact. Succ. J. Gr. Brit. 19: 66 (1957). P. catingicola subsp. arenicola (Werderm.) P. J. Braun & E. Esteves
Pereira in Succulenta 74: 134 (1995). Type: Brazil, Northern Bahia, between Sauré and Aracy, c. 350
m, Apr. 1932, Werdermann 3143 (B†).
PILOSOCEREUS
P. robustus F. Ritter, Kakt. Südamer. 1: 72, Abb. 42 & 224 (1979). Pseudopilocereus robustus (F. Ritter)
P. V. Heath in Calyx 4: 141 (1994). Pilosocereus catingicola subsp. robustus (F. Ritter) P. J. Braun & E.
Esteves Pereira in Succulenta 74: 134 (1995). Holotype: Brazil, Bahia, Ourives [Mun. Tanhaçu],
1964, Ritter s.n. (U).
[P. superbus F. Ritter, Kakt. Südamer. 1: 67 (1979), tantum quoad Abb. 35.]
Treelike or shrubby, branched above base, branching pattern 3–6-verticillate; central vascular cylinder
somewhat to strongly woody; cortex only moderately mucilaginous; epidermis olive-green to glaucous,
smooth; ribs (3–)4–12, sinuses straight, sometimes with pronounced transverse folds above the areoles.
Areoles brownish to greyish, with long hairs, presenting weak indeterminate growth. Spines yellowish
brown when young, opaque. Fertile region of stem not or slightly modified, on the apical and subapical
areoles of 1–2(–3) ribs; flower-bearing areoles with white to grey long hairs to 30 mm. Flower-buds acute
3 days before anthesis; flowers 5.5–6.7 × 4.7–7.0 cm, opening widely, somewhat compressed; pericarpel
hemiglobose, tube 38–40 × 15–22 mm at base, to 30–45 mm wide at apex, pale green, glaucous, smooth,
straight, distal half with some broad bract-scales; style 48–53 × 2–5 mm tapering to 1.8–3.5 mm, stigma-
lobes 9–12, 5 mm, exserted; ovary locule quadrangular, triangular or depressed in longitudinal section.
Fruit to 4.2–6.0 cm diam., depressed-globose, dehiscent by lateral slits, floral remnant erect or pendent,
deeply inserted; pericarp purple; funicular pulp magenta. Seeds 2–2.3 mm, with intercellular depressions.
conservation status. Least Concern [LC]; very wide-ranging, but significant reductions in its
habitats have already occurred and are likely to continue.
This species is divided in two subspecies:
1. Branches (6–)8–12 cm diam., ribs 4–6, spines stout, 10–40 mm (Bahia) . . . . . . . 3a. subsp. catingicola
1. Branches 3.5–6.0(–8.0) cm diam., ribs (5–)6–12, spines slender, 2–10 mm (Bahia northwards from
Salvador to Rio Grande do Norte and inland as far as cent.-S Pernambuco) . . . 3b. subsp. salvadorensis
3a. subsp. catingicola
vernacular names. Facheiro, Mandacaru-babão, Mandacaru-de-facho.
Tree to 6–7 m, with a well-defined trunk; vascular cylinder strongly woody; branches 8–12 cm diam.,
erect; ribs (3–)4–6, 30–35 × 15–30 mm. Areoles 7–10 mm, 14–16 mm apart, long hairs brown at base,
white at apex, to 10–15 mm. Spines (0.4–)0.6–1.6 mm diam. at base, centrals 5–6, 10–40 mm, ascending
to porrect, the largest porrect, radials 10–12, 4–13 mm, adpressed, sometimes decurved. Flower 6–6.7 ×
to 7.0 cm; inner perianth-segments sometimes with pink veins; style to 53 × 3.5–5.0 mm, stigmas c. 12,
exserted. Fruit to 6 cm diam. Seeds 2–2.3 × 1.4–1.7 mm, testa-cells domed to flat, with intercellular
depressions and fine to coarse, dense cuticular folds.
Eastern caatinga element: locally co-dominant with other arborescent plants in caatinga and ‘caatinga de
altitude’, c. 100–1100 m, eastwards from the Chapada Diamantina, northern, north-eastern, central-
eastern and southern Bahia. Endemic. Map 35.
bahia: Mun. Canudos, c. 30 km N of Euclides da Cunha on road to Canudos, 5 Jan. 1990, Taylor et al.
1358 (K, ZSS, HRCB); Mun. Jeremoabo, Serra do Xuquê [Chuquê], s.d., Luetzelburg s.n. (M, photo
filed with sheet of Melocactus zehntneri, Luetzelburg 52); Mun. Jacobina, 27–28 km W of town, towards
Lajes, 12 Jan. 1991, Taylor et al. 1398B (K, ZSS, photos); l.c., 28 km E on road to Capim Grosso, 26 Apr.
1992, Taylor & Zappi (obs.); Mun. Queimadas, 9–11 June 1915, Rose & Russell 19847 (US, NY); Mun.
Araci, 18 Jan. 1994, M.A.S. das Neves 86, 86A (HUEFS); l.c., 14.5 km N of town, 5 Jan. 1991, Taylor
et al. 1351 (K, HRCB, ZSS, CEPEC); Mun. Valente, c. 5 km SW of town towards São Domingos,
11°25'S, 39°30'W, 29 Dec. 1992, L.P. de Queiroz et al. 3038 (K, HUEFS); Mun. Conceição de Coité, 7
km NW of town on road BA 420 towards Valente, 14 July 1989, Zappi 117A (SPF, HRCB); Mun.
Morro do Chapéu, 15 km NE of town towards Várzea Nova, 12 Jan. 1991, Taylor et al. 1401 (K, HRCB,
PILOSOCEREUS
ZSS, CEPEC); l.c., 15–20 km W of town, 5 Aug. 2002, Taylor & Machado (obs.); Mun. Biritinga, 18 Jan.
1994, M.A.S. das Neves 82 (HUEFS); Mun. Riachão do Jacuípe, 3 km N on road to Conceição do Coité,
8 July 1994, F.A.R. Santos 95 (HUEFS, K, photo); Mun. Santa Bárbara, 37 km N of town on road BR
116 towards Serrinha, 13 July 1989, Zappi 117 (SPF, HRCB); l.c., at the road junction to
Juazeiro/Jacobina, 4 Jan. 1991, Taylor et al. 1348 (K, HRCB, ZSS, CEPEC); Mun. Ipirá, 19 km E of
town on road BA 052 towards Feira de Santana, 9 Feb. 1991, Taylor et al. 1610 (K, HRCB, ZSS,
CEPEC); Mun. Rafael Jambeiro, road BR 242, 4 km W of Argoim, 21 Mar. 1989, Supthut 8971 (ZSS);
Mun. Castro Alves, 13 Apr. 1995, L.P. de Queiroz 4326 (HUEFS); Mun. Milagres, 10 km towards
Itaberaba, then 2 km towards the Fazendas Morros and Antônio Romeu, 27 Jan. 1973, I. & G. Gottsberger
12-27173 (K); Mun. Marcionílio Souza, Machado Portela, 19–23 June 1915, Rose & Russell 19915 (US,
NY); Mun. Jaguaquara [Toca da Onça], 27–29 June 1915, Rose & Russell 20085 (US, NY); Mun.
Tanhaçu, Ourives, 1964, Ritter s.n. (U); l.c., 54 km on road BR 030 from Brumado towards Sussuarana,
3 Feb. 1991, Taylor et al. 1545 (K, HRCB, ZSS, CEPEC); Mun. Anagé, 1985, Horst & Uebelmann 472
(ZSS); Mun. Poções, 22 Sep. 1965, A.P. Duarte 9308 & E. Pereira 10021 (HB, RB); without precise
locality, road between Caitité and Jequié, Dec. 1912, Zehntner 645 (US).
Besides the records cited above, this taxon has also been observed in the following municípios of
Bahia: Itapicuru, Olindina, Aporá, Inhambupe, Euclides da Cunha, Filadélfia, Itiúba, Ponto Novo,
Várzea Nova, Capim Grosso, Miguel Calmon, São José do Jacuípe, Gavião, Baixa Grande, Candeal, Serra
Preta, Santa Teresinha, Iaçu, Maracás, Barra da Estiva, Contendas do Sincorá, Jequié, Planalto.
Concern, but its habitat continues to decline and its status needs to be monitored.
Subspecies catingicola is represented by populations of arborescent plants that occur inland

in the caatinga zone of Bahia and ascend into the Chapada Diamantina, presenting
branches with 4–6 ribs and strong spination. In the highland areas it is occasionally
sympatric with P. pachycladus (e.g. in various sites within Mun. Morro do Chapéu), but
hybrids have not been observed. Southern forms are often quite glaucous and the habit
of montane forms differs somewhat from those found in the lowland caatinga, but it is
clear that only one taxon is involved. This subspecies provisionally includes P. arenicola,
which was probably based on marginal populations intermediate with subsp. salvadorensis,
and thus could not be safely neotypified.
3b. subsp. salvadorensis (Werderm.) Zappi in Succ. Pl. Res. 3: 55 (1994). Holotype:
not extant (B†). Lectotype (Zappi, l.c.): Brazil, Bahia, between Bolandeiras and São
Salvador, Apr. 1932, Werdermann, l.c. infra, 37, photograph (iconotype).
Pilocereus salvadorensis Werderm., Bras. Säulenkakt.: 110 (1933). Cephalocereus salvadorensis (Werderm.) Borg,
Cacti: 107 (1937). Pilosocereus salvadorensis (Werderm.) Byles & G. D. Rowley in Cact. Succ. J. Gr.
Brit. 19: 67 (1957). Austrocephalocereus salvadorensis (Werderm.) Buxb. in Krainz, Die Kakteen, Lfg 33
(1966). Pseudopilocereus salvadorensis (Werderm.) Buxb. in Beitr. Biol. Pflanzen 44: 253 (1968).
Pilocereus hapalacanthus Werderm., Bras. Säulenkakt.: 110–111 (1933). Cephalocereus hapalacanthus
(Werderm.) Borg, Cacti: 107 (1937); E. Y. Dawson in Los Angeles County Mus. Contr. Sci. 10: 6
(1957). Pilosocereus hapalacanthus (Werderm.) Byles & G. D. Rowley in Cact. Succ. J. Gr. Brit. 19: 67
(1957). Pseudopilocereus hapalacanthus (Werderm.) Buxb. in Beitr. Biol. Pflanzen 44: 252 (1968).
Pilosocereus catingicola subsp. hapalacanthus (Werderm.) P. J. Braun & E. Esteves Pereira in Succulenta
74: 134 (1995). Holotype: not extant (B†). Lectotype (Zappi 1994): Brazil, Pernambuco, sand-dunes
N of Recife, Feb. 1932, Werdermann, l.c., 104, photograph captioned ‘Pilocer. hapalacanthus Werd.’
Pilocereus sergipensis Werderm., Bras. Säulenkakt.: 106 (1933). Cephalocereus sergipensis (Werderm.) Borg,
Cacti: 108 (1937). Pilosocereus sergipensis (Werderm.) Byles & G. D. Rowley in Cact. Succ. J. Gr. Brit.
PILOSOCEREUS
19: 67 (1957). Pseudopilocereus sergipensis (Werderm.) Buxb. in Beitr. Biol. Pflanzen 44: 253 (1968).
Type: Sergipe, Jaboatão, 1932, Werdermann, not extant (B†). Neotype (Zappi 1994): Brazil, Sergipe,
Mun. Nossa Senhora das Dores, 500 m S of the town, 110 m, 11 Feb. 1991, Taylor & Zappi 1616
(HRCB; K, isoneo.).
Pilocereus rupicola Werderm., Bras. Säulenkakt.: 109 (1933). Cephalocereus rupicola (Werderm.) Borg, Cacti:
111 (1937). Pilosocereus rupicola (Werdermann) Byles & G. D. Rowley in Cact. Succ. J. Gr. Brit. 19:
67 (1957). Pseudopilocereus rupicola (Werderm.) Buxb. in Beitr. Biol. Pflanzen 44: 252 (1968).
Holotype: Brazil, Sergipe, Serra de Itabaiana, 550 m, Mar. 1932, Werdermann 3092 (B†).
vernacular names. Facheiro, Facheiro-da-praia.
Shrubs or trees 2–10 m, with or without a definite trunk, branching at ground level and upwards; vascular
cylinder slightly woody; branches 3.5–6.0(–8.0) cm diam., erect; ribs (5–)6–12, 10–20 × 16–20 mm;
areoles 4 mm diam., 8–10 mm apart, rounded, reddish brown to grey, slightly hairy when young. Spines
0.3–0.7(–1.0) mm diam. at base, central(s) 1–11, 5–12 mm, ascending to porrect, radials 8–12, 3–4(–5)
mm, adpressed. Flowers 5.5 × 4.7 cm, inner perianth-segments white; style 48 × 1.8–2.0 mm, tapering,
stigma-lobes c. 9, exserted. Fruit to 4.2 cm diam. Seeds black, shiny, 2–2.1 × 1.1–1.3 mm, hilum-
micropylar region 6–8 mm, forming an angle of 40° with long axis of seed, testa-cells domed to flat, with
intercellular depressions and coarse, sparse cuticular folds.
Widespread humid forest (restinga) / caatinga element: in dense or sparse restinga on sand-dunes north of
Salvador (BA) to Rio Grande do Norte and (?) NE Ceará, extending up the São Francisco River valley
and westwards to the caatinga region around the borders of Bahia (Raso da Catarina), Alagoas, Sergipe
and Pernambuco, where locally co-dominant with other arborescent vegetation, near sea level to 550 m.
Endemic to North-eastern Brazil. Map 35.
rio grande do norte: Mun. Natal, Mãe Luiza, dunes, 9 Mar. 1962, Tavares 905 (US), l.c., Praia do
Meio, 1988, illustrated by Braun & Esteves Pereira (2002: 21, Abb. 34); Mun. Nísia Floresta, Praia de
Búzios, 1997, E.A. Rocha (obs.).
paraíba: Mun. Mataraca, reported by Oliveira-Filho (1993: 224); Mun. Mamanguape, 1997, E.A.
Rocha 191 (JPB); Mun. Cabedelo, Praia do Poço, 21 Feb. 1962, Castellanos 23245 (GUA), 9 Feb. 1995,
Taylor & Zappi (K, photos), Mar.–Apr. 1996, see Locatelli et al. (1997: figs 1–5); Mun. Pedras de Fogo,
Santa Emília, 24 Sep. 1962, Tavares 1053 (US).
pernambuco: Mun. Itamaracá & Mun. Abreu e Lima (Rio Doce), fide Andrade-Lima (1966: 1457);
Mun. Recife, sand-dunes, Feb. 1932, reported by Werdermann (1933: 104, photo); Mun. Vitória de
Santo Antão, 25 Apr. 1935, Pickel 3785 (IPA 4473); Mun. Inajá, Reserva Biol. Serra Negra, 17 Sep. 1995,
A. Laurênio et al. 196 (K, PEUFR); Mun. Nova Petrolândia, 3 km N of the Barragem de Itaparica, 250
m, 12 Feb. 1991, Taylor & Zappi 1619 (HRCB, K, ZSS, UFP).
alagoas: Mun. Marechal Deodoro, restinga S of Ponta do Cavalo Ruço, 14 Feb. 1995, Taylor & Zappi
(K, photos); Mun. Traipu, June 1990, R. de Lyra-Lemos (MAC, photo); Mun. (?), between Juquia and
Lagoa do Pau beach, Aug. 2001, Zappi (obs.); Mun. Igreja Nova, 6 km SE of the town on road AL 225,
14 Feb. 1991, Taylor & Zappi 1632 (HRCB, K, ZSS, MAC); Mun. (?), between Miaí de Cima and
Piaçabuçu, Aug. 2001, Zappi (obs.); Mun. Piaçabuçu, Tapera, 4 Feb. 1988, I.S. Moreira et al. 113 (MAC,
SPF); Mun. Penedo, 7 km S of town on road to Piaçabuçu, 14 Feb. 1991, Taylor & Zappi 1631 (HRCB,
K, ZSS, MAC).
sergipe: Mun. Canindé de São Francisco, border of SE/BA, to Xingozinho, 11 Feb. 1991, Taylor &
Zappi 1618A (HRCB, K); Mun. Nossa Senhora das Dores, 0.5 km S of town, 11 Feb. 1991, Taylor &
Zappi 1616 (HRCB, K, ZSS, ASE); Mun. Jaboatão, type locality of synonymous Pilocereus sergipensis
Werderm. (see above); Mun. Feira Nova, 2 km NW of town, 11 Feb. 1991, Taylor & Zappi 1617
(HRCB, K, ZSS, ASE); Mun. Nossa Senhora da Glória, 7 km NW of town, 11 Feb. 1991, Taylor &
Zappi 1617A (HRCB, K, ZSS); Mun. Itabaiana, just below western edge of highest part of Serra de
Itabaiana, c. 550 m, 3 Aug. 1998, Taylor (K, ASE, photos); Mun. Barra dos Coqueiros and Mun. Aracaju
(Praia José Sarny), restinga, 3/4 Aug. 1998, Taylor (obs.).
bahia: Mun. Glória, Brejo do Burgo, Faz. Torrão, 28 Nov. 1992, F.P. Bandeira 84 (ALCB); Mun. Paulo
Afonso, 2 km from crossing towards town, coming from Canindé de São Francisco, 11 Feb. 1991, Taylor
PILOSOCEREUS
& Zappi 1618 (HRCB, K, ZSS, CEPEC); Mun. Santa Brígida, c. 10 km W of junction for the town
with road BR 110, Raso da Catarina, c. 9°29'S, 38°12'W, 19 Dec. 1993, L.P. de Queiroz & Nascimento
3733 (K, HUEFS); Mun. Mata de São João, Aug. 2001, Zappi (obs.); Mun. Camaçari, 42 km N of
Salvador airport, towards Arembepe, restinga, 23 Mar. 1989, Supthut 8975 (ZSS); Mun. Salvador, Jardim
de Alá, 1960, A.L. Costa s.n. (ALCB 02912); l.c., Lagoa de Abaeté, 1 km N of Praia de Itapoã, 13 July
1989, Zappi 116 (SPF, HRCB), 4 Jan. 1991, Taylor et al. 1342 (K, HRCB, ZSS, CEPEC); l.c., near
Salvador, 29 May 1915, Rose & Russell 19676 (US, NY).
Although its range is extensive, its habitat continues to decline at a significant rate, especially near the
coast, and so its status needs to be monitored. For example, the once extensive population on the
peninsula of Cabedelo, Paraíba has all but disappeared since 1995, due to building developments, and
only a few plants are protected inside the local IBAMA forest reserve. The same can be said of habitats
along the road ‘Linha Verde’ on the coast of NE Bahia and of the NE coast of Alagoas.
The pollination of this subspecies by the phyllostomid bat, Glossophaga soricina Pallas, and
by hawkmoths, has been documented and photographed by Locatelli et al. (1997).
Maria Luiza Santos (Aracaju) has shown the first author photographs of what is
assumed to be dense stands of this taxon growing on the rocky banks of the lower reaches
of the Rio São Francisco, where it apparently develops into exceptionally tall specimens
with many erect branches.
4. Pilosocereus azulensis N. P. Taylor & Zappi in Cact. Consensus Initiatives 3: 8

(1997). Holotype: Brazil, Minas Gerais, Mun. Pedra Azul, 16°3'S, 41°14'W, 20 Oct.
1988, Taylor & Zappi in Harley 25220 (SPF; K, iso.).
Treelike, c. 4 m, branched above the ground, with a well-defined trunk; vascular cylinder weakly woody;
branches 8–9.5 cm diam., erect, epidermis dark green, slightly rough; ribs 8–10, 10–15 × 15 mm, sinuses
straight, and without visible transverse folds. Areoles 3–4 mm diam., 5–7 mm apart, felt greyish with
brown long hairs to 3–5 mm. Spines 0.3–0.5 mm diam. at base, opaque, brown to blackish, centrals 1–4,
10–15 mm, ascending, radials 10–14, 2–8 mm, adpressed. Fertile part of stem slightly differentiated, with
few apical or sub-apical flower-bearing areoles with bristly, black spines to 20 mm and white tufts of hair
to 10 mm. Flowers not seen. Fruits (dried) depressed-globose, floral remnant pendent, deeply inserted.
Seeds 1.5–1.6 × 1.0 mm, black, shiny, cochleariform, hilum-micropylar region 0.6–0.7 mm, forming an
angle of 25–35° with long axis of seed, testa-cells domed, with inconspicuous intercellular depressions
and hardly any cuticular folds.
South-eastern caatinga (inselberg) element: associated with gneissic inselbergs in caatinga-agreste, 650 m;
known only from the region of Pedra Azul, Minas Gerais (and from perhaps the same area recorded as a
vaguely localized collection from south of Vitória da Conquista, Bahia). Endemic to the core area within
minas gerais: Mun. Pedra Azul, ‘6 km ao sul da cidade na estrada para Jequitinhonha’, 16°3'S,
41°14'W, 20 Oct. 1988, Taylor & Zappi in Harley 25220 (K, SPF); l.c., between Pedra Azul and Salinas,
10 Sep. 1985, Horst & Uebelmann 272 (ZSS) — NB. this HU number has also often been applied to living
collections of P. pachycladus. Unlocalized: Bahia/Minas Gerais (?), S of Vitória da Conquista, 10 Apr.
1968, Castellanos 27073 (HB).
conservation status. Critically Endangered [CR D] (4); PD=2, EI=1, GD=1. Short-list score (4×4)
= 16. This species is currently known from only a small remnant of dry forest vegetation, which is being
cleared for agriculture and charcoal production. Only the above 3 collections have been recorded to date
and two of these are poorly localized and all may refer to the same small area of the type. Further study
PILOSOCEREUS
of this plant and its habitat is urgently required. In August 2002, Marlon Machado and staff from the
Univ. Estadual do Sudoeste da Bahia located c. 30 juvenile plants near the type locality (Plate 40.1), but
the adjacent forest in which the type was collected has gone.
In 1988 a single mature specimen of this taxon was observed in semi-shade of dry forest
(agreste), sympatric with Pilosocereus floccosus subsp. quadricostatus. The shape of the ribs and
spination, as well as the only slightly differentiated flowering region are reminiscent of the
P. ARRABIDAE Group, especially of some populations of the geographically distant P.
catingicola subsp. salvadorensis.
Although flowering material has yet to be examined, vegetative morphology and
seeds suggest that this species belongs to the P. ARRABIDAE Group, which is otherwise
unrepresented in the region of the Rio Jequitinhonha drainage, where all other Species
Groups in Subg. Pilosocereus are present. The alternative explanation, that it is a hybrid
involving P. floccosus subsp. quadricostatus and P. multicostatus, does not seem plausible given
the characters it displays. Furthermore, plants that are almost certainly representative of
such a hybrid have recently been seen around Pedra Azul by Marlon Machado (in Oct.
1999) and Graham Charles (K, photos, June 2002) and these are clearly not comparable
with P. azulensis (Plate 40.2).
5. Pilosocereus arrabidae (Lem.) Byles & G. D. Rowley in Cact. Succ. J. Gr. Brit. 19:
66 (1957). Lectotype (Zappi 1994): Brazil, Rio de Janeiro, Vellozo, Fl. flumin., Icones 5:
tab. 18 (1831).
Pilocereus arrabidae Lem., Rev. Hort. 34: 429 (1862). Cephalocereus arrabidae (Lem.) Britton & Rose, Cact.
2: 42 (1920). Pseudopilocereus arrabidae (Lem.) Buxb., Beitr. Biol. Pflanzen 44: 252 (1968).
Cereus warmingii K. Schum. in Martius, Fl. bras. 4(2): 204 (1890), quoad typ. Type: Brazil, Rio de Janeiro,
restingas de Copacabana, Warming s.n. (B†).
[Pilocereus exerens K. Schum. in Engler & Prantl, Pflanzenfamilien 3(6): 181 (1894), typ. excl., nom. illeg.
Art. 52.1.]
[Cephalocereus exerens (K. Schum.) Rose in Bailey, Stand. Cycl. Hort. 2: 715 (1914), typ. excl., nom. illeg.
Art. 52.1.]
[Cactus hexagonus sensu Vell., Fl. flumin.: 205 (1825, publ. 1829), non Linnaeus (1753).]
[Cactus heptagonus sensu Vell., Fl. flumin.: 205 (1825, publ. 1829), non Linnaeus (1753).]
[Cereus macrogonus sensu K. Schum. in Martius, Fl. bras. 4(2): 202, tab. 40 (1890), non Salm-Dyck (1850).]
[Pilocereus virens sensu Ule, Monatsschr. Kakt.-Kunde 13: 28 (1903), non Cereus virens DC.]
vernacular name. Facheiro-da-praia.
Shrub, 0.1–3.0(–4.0) m, or rarely with a short trunk, branching near base or above, sometimes 3–5-
verticillate; vascular cylinder weakly woody; stems 4.5–9.5 cm diam., straight, slightly upcurved,
epidermis yellow-green to dark green, sometimes slightly greyish; ribs (5–)6–8, 13–15 × 12–20 mm,
sinuses straight, with oblique transverse folds above the areoles. Areoles (5–)6–8 mm diam., (7–)15–32
mm apart, situated upon more or less conspicuous stem-projections (podaria), felt whitish to brownish or
greyish. Spines 0.6–1.1 mm diam. at base, translucent only when young, brown to greyish, centrals
(0–)2–4(–8), 15–40 mm, sometimes decurved near base, often 1 ascending and two porrect, radials 7–10,
2–20 mm, adpressed. Fertile part of stem not differentiated, flowers distributed along the branches.
Flower-buds acute 2–3 days before anthesis; flowers 6–7 × 4–5 cm; pericarpel subglobose, tube flared
from 11 to 30–32 mm wide, infundibuliform, straight, creamy green, smooth, distal half with wide,
broadly triangular bract-scales; style 45–48 × 2–3 mm, tapering, stigma-lobes c. 9, 6–7 mm, exserted;
ovary locule obovate, compressed in longitudinal section. Fruit 3–5 × 5–5.8 cm, depressed-globose,
PILOSOCEREUS
dehiscent by a lateral or abaxial slit, floral remnant pendent, deeply inserted; pericarp green to deep
reddish pink, slightly rugose; funicular pulp magenta. Seeds (1.4–)1.5–1.7(–1.8) × 1–1.2 mm, testa-cells
domed, with intercellular depressions and dense, coarse cuticular folds.
Southern humid forest (restinga) element: in dense or sparse, sandy restinga, near sea level, southern Bahia
and Espírito Santo to Rio de Janeiro. Map 35.
bahia: Mun. Santa Cruz Cabrália, 20–24 km N of Porto Seguro, 31 July 1989, Zappi 184 (SPF,
HRCB); Mun. Prado, 13 Dec. 1998, Fonseca & Guedes 1156 (ALCB, UEC).
espírito santo: Mun. Conceição da Barra, Ilha de Guriri, June 1996, M. Canal s.n. (Herb. Univ. Fed.
ES); Mun. Guarapari, on sand, [1973], Horst & Uebelmann 240 (ZSS); l.c., road ES 060, restinga close to
Lagoa Vermelha, 21 May 1990, O.J. Pereira et al. 2119 (Herb. Univ. Fed. ES, HRCB); Mun. Itapemirim,
Rodovia do Sol, S of Marataízes, 21°4'51"S, 40°50'24''W, 26 Nov. 1999, reported growing with Opuntia
monacantha (Zappi et al. 470).
rio de janeiro: Mun. São João da Barra, c. 10 km from town by the crossing towards Gruçaí, 23 Mar.
1982, Rocha et al. 893 (RB); Mun. Quiçamã (Quissamã), road to Barra do Furado, 22°6'5"S, 41°26'W,
23 Nov. 1999, Zappi et al. 397 (K, UEC).
Although its range is extensive so is the ongoing and accelerating modification of its coastal habitat.
Inhabiting a long stretch of restinga vegetation, from between Santa Cruz Cabrália and
Porto Seguro, Bahia, to west of the city of Rio de Janeiro, P. arrabidae has been rather
frequently confused with P. brasiliensis subsp. brasiliensis, with which it is sympatric, at least
along the coast of Espírito Santo (cf. Pereira 2119). There are some superficial similarities,
such as the undifferentiated flower-bearing areoles and green epidermis, but P. arrabidae
presents (5–)6–8 ribs, thicker branches, acute, straight flower-buds, large flowers, 4–5 cm
diam. at anthesis, whereas P. brasiliensis subsp. brasiliensis has 4–5 ribs, thinner branches,
4.5–5.5 cm diam. and obtuse, curved flower-buds, with narrow flowers up to 3 cm diam.
at anthesis. Their seeds also differ considerably. Contrary to the impression given in Braun
& Esteves Pereira (2003: 17), P. arrabidae is restricted to coastal rocks and dunes in Espírito
Santo, being replaced by P. brasiliensis subsp. ruschianus inland.
PILOSOCEREUS PENTAEDROPHORUS Group (Nos 6–10). Found in Mata

atlântica through to caatinga/cerrado de altitude and in the caatinga-cerrado ecotone, but avoiding
the driest areas and completely absent from the caatingas of the Rio São Francisco valley.
6. Pilosocereus brasiliensis (Britton & Rose) Backeb., Die Cact. 4: 2423 (1960). Type
(Zappi 1994): Brazil, Rio de Janeiro, Mun. Rio de Janeiro, Corcovado, 10 July 1915,
Rose & Russell 20190 (US, lecto.; NY, lectopara.).
Cephalocereus brasiliensis Britton & Rose, Cact. 2: 57, fig. 84 (1920). Pilocereus brasiliensis (Britton & Rose)
? Cereus sublanatus Salm-Dyck, Hort. Dyck. 1834: 337 (1834). Pilocereus sublanatus (Salm-Dyck) Backeb.
& F. M. Knuth, Kaktus-ABC: 333 (1935, publ. 1936). Pilosocereus sublanatus (Salm-Dyck) Byles & G.
D. Rowley in Cact. Succ. J. Gr. Brit. 19: 69 (1957). Type: a living plant of unknown provenance,
assumed not to have been preserved or lost.
Erect or semi-scandent shrub, 2–7 m, usually without a well-defined trunk, main stem little branched;
vascular cylinder moderately woody, branches 4.5–7.5 cm diam., semi-decumbent to prostrate when
PILOSOCEREUS
shaded, erect when in full light; ribs 4–6(–7), 15–17 × 8–18 mm, sinuses straight, transverse folds above
the areoles oblique. Areoles 4–7 mm diam., 6–8 mm apart, situated upon stem-projections (podaria), felt
white, brown to blackish, with white, greyish or brown long hairs to 15 mm, indeterminate growth not
observed. Spines opaque, at first reddish or yellowish brown, later greyish. Fertile part of stem not or only
slightly differentiated, flowers appearing randomly over the stem. Flower-buds obtuse 1–2 days before
anthesis; flowers 4.5 × 2.5–3.0 cm; pericarpel subglobose, tube 20–40 × 10–12 mm at base, flared to 23
mm at apex, constricted beneath and above nectar-chamber, curved, nearly cylindric, pale green, smooth,
apex with few, thick, rounded bract-scales; style 26–32 × 1–1.5 mm, tapering; stigma-lobes 7–8, 8 mm;
ovary locule hemicircular in longitudinal section. Fruit 2 × 3–3.5 cm, depressed-globose, dehiscent by a
lateral slit, floral remnant pendent, deeply inserted; pericarp wine-coloured, smooth; funicular pulp
magenta. Seeds (1.4–)1.6–1.7(–1.8) × 1.1–1.2 mm, testa-cells domed, with intercellular depressions and
dense, coarse cuticular folds.
Two subspecies are recognized as follows:
1. Branches dark green; ribs 4–5 (coastal) . . . . . . . . . . . . . . . . . . . . . . . . . . . 6a. subsp. brasiliensis

1. Branches greyish green to glaucous; ribs (4–)5–7 (inland) . . . . . . . . . . . . . 6b. subsp. ruschianus
6a. subsp. brasiliensis

Shrub. Branches semi-decumbent to prostrate when shaded, erect in exposed habitats, epidermis shiny
dark green, sometimes slightly greyish; ribs 4–5; areoles usually hairy only at the stem apex (except in
cultivation under glass), long hairs pale brown, to 10 mm, later glabrescent, brown to black. Spines
0.8–1.0 mm diam. at base, centrals (0–)1–3, 10–30 mm, ascending to porrect; radials 4–8, 3–16 mm,
adpressed. Fertile part of stem not or only slightly differentiated, flowers solitary or in groups of 2–3;
anthers 1.5–1.8 mm; stigma-lobes c. 7, not exserted much beyond anthers.
Southern humid forest (restinga) element: in restinga and gneissic inselbergs of the coast, to 100 m, Espírito
Santo; Rio de Janeiro. Endemic to South-eastern Brazil. Map 36.
espírito santo: Mun. Vitória, Campus of the Univ. Fed. do Espírito Santo, May 1990, O.J. Pereira
2120B (HRCB); Mun. Vila Velha, restinga de Jacaranema, 21 May 1990, O.J. Pereira 2120 (HRCB, Herb.
Univ. Fed. ES); Mun. Guarapari, dense restinga, 21 May 1990, O.J. Pereira 2120A (HRCB); Mun.
Anchieta, Rodovia do Sol, between Meaípe (Mun. Guarapari) and Anchieta, 20°45'13"S, 40°32'37''W,
25 Nov. 1999, Zappi et al. 451 (K, UEC); Mun. Piúma, Monte Agá, south of the town, 20°52'23"S,
40°46'24''W, 26 Nov. 1999, reported growing with Coleocephalocereus fluminensis (Zappi et al. 469, q.v.).
PD=1, EI=1, GD=2. Short-list score (2×4) = 8. Has apparently declined in the southern part of its range
(around the city of Rio de Janeiro) and will be affected by ongoing touristic developments along much
of the length of its coastal habitat.
Variable in the degree of areolar wool and long hairs developed in some populations.
6b. subsp. ruschianus (Buining & Brederoo) Zappi in Succ. Pl. Res. 3: 64 (1994).
Holotype: not extant. Lectotype (Zappi, l.c.): Brazil, Espírito Santo, Mun. Colatina,
Buining & Brederoo, l.c., 33, photograph, above right (iconotype).
Pseudopilocereus ruschianus Buining & Brederoo in Kakt. and. Sukk. 31: 33–35 (1980). Pilosocereus ruschianus
(Buining & Brederoo) P. J. Braun in Bradleya 6: 88 (1988).
PILOSOCEREUS
Sometimes treelike, to 7 metres. Branches erect or sometimes procumbent on inselbergs, epidermis grey-
green, slightly glaucous at apex; ribs (4–)5–7, variable in number between populations; areoles with white
to greyish long hairs, mainly at the apex of stems. Spines 0.5–0.8 mm diam. at base, centrals 1–3, 10–38
mm, ascending to porrect, radials 8–15, 5–12 mm, adpressed. Fertile part of stem slightly differentiated,
occupying many consecutive flowering areoles, with greyish or brownish long hairs; anthers 2.8–3.0 mm,
slightly verrucose; stigma-lobes c. 8, exserted in relation to the anthers.
Southern humid/subhumid forest (inselberg) element: on gneissic inselbergs associated with agreste and
mata seca, c. 80–700 m, southern Bahia, Espírito Santo and eastern Minas Gerais. Endemic to the core
area within Eastern Brazil. Map 36.
bahia: Mun. Vitória da Conquista, 32 km SE of town on road to Itambé, 1 km from Rio Periquito,
granitic outcrop, 16 Aug. 1988, Eggli 1175 (ZZS, HRCB); Mun. Floresta Azul, 2 km W on road BR
415, 16 April 2003, Taylor & Zappi (K, photo); Mun. Itaju do Colônia, 2 km on road between Itaju and
Pau Brasil, 24 Jan. 1969, T.S. dos Santos 349 (CEPEC); between Santa Cruz da Vitória and Itaju do
Colônia, 19 Dec. 1967, Castellanos 27058 (CEPEC).
minas gerais: Mun. Carlos Chagas, c. 25 km W of Nanuque, 1994, S. Porembski et al. (K, photo); Mun.
Galiléia, towards the border with Espírito Santo, 1982, Horst & Uebelmann 512, cult. (ZSS); l.c., road BR
259, 23 km from São Vítor towards Galiléia, 15 Dec. 1990, Taylor & Zappi 770 (HRCB, K, ZSS); Mun.
Conselheiro Pena, 7.5 km E of Goiabeira on road to Cuparaque, 15 Dec. 1990, Taylor & Zappi 777 (K,
HRCB, ZSS, BHCB); Mun. Aimorés, Terreno do Varejão, 7 July 1997, M.F. de Vasconcelos s.n. (K, BHCB).
espírito santo: E of Mantena, 1983, Braun 488 (ZSS); Mun. Nova Venézia, Pedra do Elefante, Apr.
1996, S. Porembski (K, photo); Mun. São Domingos, 15 km S of town on road to Colatina, 16 Dec. 1990,
Taylor & Zappi 783 (K, HRCB, ZSS, MBML); Mun. Pancas, Bridge over Rio Pancas, 16 Dec. 1990,
Taylor & Zappi s.n. (K, photos); Mun. (?), ‘Rio Doce’, Feb. 1917, Luetzelburg 7227 (M); Mun. Itarana,
road to Praça Oito, 19°55'34"S, 40°48'44''W, 24 Nov. 1999, Zappi et al. 417 (K, UEC); Mun. Santa
Teresa, S. João de Petrópolis, Escola Agrotécnica Federal, 8 Oct. 1985, H.Q.B. Fernandes 1554 (MBML);
l.c., Rio Cinco de Novembro, Dec. 1987, H.Q.B. Fernandes 2259 (MBML, HRCB).
Habitat is not significantly declining at present.
Variable in rib number between populations. At the northern limits of its range this
subspecies occasionally becomes a tree in semi-humid forest, such as between Itororó and
Itapetinga (BA).
7. Pilosocereus flavipulvinatus (Buining & Brederoo) F. Ritter, Kakt. Südamer. 2: 707

(1980). Holotype: not extant. Lectotype (Zappi 1994): Brazil, Piauí, Simplício Mendes,
c. 350 m, Buining & Brederoo, l.c., 138, photograph (iconotype).
Pseudopilocereus flavipulvinatus Buining & Brederoo in Succulenta 58: 137–143 (1979).
Pilosocereus carolinensis F. Ritter, Kakt. Südamer. 1: 80 (1979). P. flavipulvinatus var. carolinensis (F. Ritter)
F. Ritter, op. cit. 2: 707 (1980). Pseudopilocereus carolinensis (F. Ritter) P. V. Heath in Calyx 4: 140
(1994). Pilosocereus flavipulvinatus subsp. carolinensis (F. Ritter) P. J. Braun & E. Esteves Pereira in
Succulenta 74: 134 (1995). Holotype: Brazil, Maranhão, Carolina, 1963, Ritter 1217 (U).
P. carolinensis var. robustispinus F. Ritter, Kakt. Südamer. 1: 81 (1979). Pseudopilocereus carolinensis var.
robustipinus (F. Ritter) P. V. Heath in Calyx 4: 140 (1994). Pilosocereus flavipulvinatus var. robustispinus
(F. Ritter) P. J. Braun & E. Esteves Pereira in Succulenta 74: 134 (1995). Holotype: Brazil,
Maranhão, Barão de Grajaú, 1963, Ritter 1217A (U).
Slender, sparsely branched tree, 2–5(–8) m, with a well-defined trunk; vascular cylinder strongly woody;
branches 3–12(–15) cm diam., closely grouped and erect or sometimes semi-decumbent, epidermis olive-
green, glaucous at apex when young; ribs 6–12, 8–17 × 9–15 cm, sinuses straight, with oblique transverse
PILOSOCEREUS
folds above the areoles. Areoles 5–10 mm, 10–15 mm apart, situated upon rounded stem projections, felt
yellowish white, long hairs pale brown to grey, apical areoles sometimes unarmed. Spines 0.5–0.7 mm
diam. at base, translucent, pale yellow and ascending when young, later greyish; centrals 3–5(–7), 15–70
mm, porrect, lowermost largest; radials 12–15, 5–14 mm, adpressed. Fertile part of stem not
differentiated, subapical, on 2–4 ribs. Flower-buds acute 2 days before anthesis; flowers 5.8–6.0 × 3.5–3.7
cm, pericarpel subglobose, tube 40 × 13 mm at base to 25 mm at apex, constricted above and below the
nectar-chamber, strongly curved, slightly compressed, pale green, glaucous, smooth, distal half with acute
bract-scales; style 42–46 × 2–2.5 mm, tapering, stigma-lobes c. 9, 5 mm, exserted in relation to the
stamens; ovary locule hemicircular in longitudinal section. Fruit to 2–2.5 × 3–3.5 cm, depressed-globose,
dehiscent by a lateral slit, floral remnant pendent, deeply inserted; pericarp pale green, slightly rugose;
funicular pulp magenta. Seeds (1.6–)1.7–1.8(–1.9) × 1.1–1.2 mm, testa-cells domed to flat, with
intercellular depressions and coarse, sparse cuticular folds.
Northern caatinga element: in the caatinga/cerrado ecotone, dense, high and low shrubby caatinga, open,
seasonally flooded carnaúba (Copernicia prunifera) forest (often as epiphyte) and caatinga-mangrove ecotone,
in northern draining river valleys and at the coast, sea level to c. 350 m, northern and eastern-central
Ceará, northern to central-south-eastern Piauí and along the border regions between Piauí and
Maranhão; (?) to northern Tocantins. Endemic to North-eastern Brazil? Map 36.
maranhão: Mun. Carolina, 1963, Ritter 1217 (U); Mun. Barão de Grajaú, 1963, Ritter 1217A (U); l.c.,
just W of Rio Parnaíba near road BR 230, 3 Feb. 1990, Zappi 213 (HRCB, ZSS).
piauí: N Piauí, Mun. Luís Correia, W of Chaval [CE], 19 Feb. 1995, Taylor (K, photos); Mun. Buriti
dos Lopes, 5 km N of town, rocks by the Rio Pirangi, 19 Feb. 1995, Taylor (K, photos); Mun. Cocal, c.
17 km E of town towards border with Ceará, 19 Feb. 1995, Taylor (K, photos); Mun. Piracuruca, c. 12
& 29 km W of border with Ceará, 20 Feb. 1995, Taylor (K, photos); Cent.-SE Piauí, Mun. Picos, 23 km
from town along road BR 407 towards Jaicós, 7°13'S, 41°17'W, 6 Apr. 2000, E.A. Rocha et al. (K,
photos); Mun. Jaicós, 6 km S of town, 6 Apr. 2000, E.A. Rocha et al. (obs.); Mun. Canto do Buriti, 15
km S of town, 5 Feb. 1990, Zappi 216 (HRCB, ZSS); Mun. Simplício Mendes, see type citation above;
Mun. Paulistana, 51 km N of town on road BR 407, 6 Apr. 2000, E.A. Rocha et al. (obs.).
ceará: N Ceará, between Chaval and Camocim, various sites including mangrove, 19 Feb. 1995, Taylor
(K, photos); Mun. Granja, c. 1 km S, 20 Feb. 1995, Taylor (K, photo); Mun. Viçosa de Ceará, E of Cocal
(PI), 19 Feb. 1995, Taylor (K, photos); Mun. Coreaú, 20 Feb. 1995, Taylor (K, photos); Mun.
Maranguape, ‘Kagado’, 31 Oct. 1935, F. Drouet 2652 (F, US); E-cent. Ceará, Banabuiú [‘Banabuiá’], see
Eiten (1983: fig. 62).
Concern at present, but needs to be monitored in view of increasing habitat destruction in the areas of
higher rainfall that it inhabits.
This species ranges through the northern part of the caatinga, reaching that vegetation’s
north-western limits, and occupying the ecotones with the cerrado and forests transitional
to those of Amazônia. It is the only Pilosocereus that enters the coastal mangrove-caatinga
ecotone (NW Ceará) and is also frequently epiphytic on the trunks of carnaúba palms in
seasonally flooded palm forest at its northern limit (caatinga type No. 12 of Andrade-Lima
1981: 160). Its stems vary considerably in thickness, those of plants from the drier
vegetation of central-southern Piauí and adjacent parts of Maranhão being much more
slender than those from further north.
8. Pilosocereus pentaedrophorus (J. F. Cels) Byles & G. D. Rowley in Cact. Succ.

J. Gr. Brit. 19: 67 (1957). Type: not extant. Neotype (Zappi 1994): Brazil, Bahia, Mun.
Serrinha, 16 km N of the town on road BA 409 towards Conceição do Coité, 14 July
1989, Zappi 120 (SPF; HRCB, isoneo.).
PILOSOCEREUS
Cereus pentaedrophorus J. F. Cels, Cat. Cact. Aloées: 9 (1858); Cereastreae (subg. Cereus) pentaedrophorus
Labour., Monogr. Cact.: 365 (1853), nom. inval. (Art. 43.1). Pilocereus pentaedrophorus (J. F. Cels)
Console ex K. Schum., Gesamtbeschr. Kakt.: 174 (1897). Cephalocereus pentaedrophorus (J. F. Cels)
Britton & Rose, Cact. 2: 31 (1920). Pseudopilocereus pentaedrophorus (J. F. Cels) Buxb. in Beitr. Biol.
Pflanzen 44: 253 (1968). Pilocereus polyedrophorus Lem. in Rev. Hort. 1862: 428 (1862), nom. illeg.
(Art. 52.1). Pseudopilocereus polyedrophorus (Lem.) P. V. Heath in Calyx 2: 54 (1992), nom. illeg.
vernacular names. Facheiro, Facheiro-fino, Mandacaru-de-veado.
Shrub or treelike to 6(–15) m; vascular cylinder woody; branches blue-green, glaucous; ribs 4–8(–10),
sinuses sinuate and pronounced, with horizontal transverse folds above the areoles. Areoles 3–7(–8) mm,
12–22 mm apart, felt white to greyish black, apical young areoles frequently unarmed, growth
indeterminate, strong near base of plant. Spines 0.5–1.0 mm diam. at base, translucent, yellowish brown.
Fertile part of stem not differentiated, subapical, occupying several consecutive areoles of more than one
rib. Flower-buds curved and obtuse, constricted in the region of the nectar-chamber, varying markedly
in colour in different populations between green, reddish or purplish blue; flower 3.5–5.5 × 2.8 cm; tube
35–45 × 11 mm at base to 20–21 mm diam. at apex, constricted below and above the nectar-chamber
region, strongly curved, smooth, terete or nearly so, apex with few ovoid to spathulate bract-scales; style
35 × 1.5–2.0 mm, tapering, stigma-lobes c. 10; ovary locule hemicircular to depressed in longitudinal
section. Fruit 1.5–2.0 × 2–3 cm, depressed-globose, dehiscent by a lateral slit, with pentagonal to
hexagonal outline, floral remnant pendent, deeply inserted; pericarp pale green to wine-coloured;
funicular pulp purplish or magenta. Seeds (1.6)–1.7–1.9(–2.0) × (1–)1.1–1.3(–1.4) mm, testa-cells flat,
with intercellular depressions nearly absent and fine, sparse cuticular folds.
Two subspecies are recognized within this taxon, the typical one inhabiting forest
vegetation east of the Chapada Diamantina (Bahia) and northwards, reaching
Pernambuco, and subsp. robustus, distributed towards the southern limit of the species, in
southern Bahia and north-eastern Minas Gerais. These are distinguished as follows:
1. Branches slender, long and leaning, to 4.5(–6.0) cm diam. below apex, ribs 4–6(–7), obtuse
(Bahia to Pernambuco) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8a. subsp. pentaedrophorus
1. Branches stout, never leaning, to 7.5 cm diam. below apex, ribs (5–)6–10, acute (S Bahia
& NE Minas Gerais) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8b. subsp. robustus
8a. subsp. pentaedrophorus

Treelike, sparsely branched, to 6 m or more in high forest; branches sometimes leaning on other
vegetation for support, 3–4.5(–6.0) cm diam.; ribs 4–6(–7), 9–18 × 8–20 mm, obtuse. Central spines 1–3,
10–25 mm, porrect to deflexed, radials 3–12, 4–12 mm, adpressed, all deflexed.
Eastern caatinga element: in agreste and dense caatinga, on rocky substrates, rarely reaching into restinga
sand-dunes (N of Salvador, BA), c. 5–1000 m, north-eastern Pernambuco, western Sergipe, and north-
eastern and eastern Bahia. Endemic to the core area within North-eastern Brazil. Map 36.
pernambuco: Mun. Brejo da Madre de Deus, 5 Feb. 1965, Andrade-Lima 65-4293 (IPA), l.c., Faz. Bituri,
26 May 1995, D.C. Silva 72 (UFP); Mun. Caruaru, 9 km NE of town, 8°14'21"S, 35°54'52''W, Dec.
1991, E.A. Rocha et al. (K, photos); Mun. Garanhuns, ‘em capoeiras’, fide Andrade-Lima (1966: 1457).
sergipe: Mun. Nossa Senhora da Glória, Faz. Olhos D’agua, 13 May 1982, M.N. Almeida 80 (ASE
2384); Mun. Frei Paulo, 26 June 1981, M. Fonseca 518 (ASE 944); l.c., Faz. Serras Pretas, 16 Apr. 1985,
G. Viana 1139 (ASE 3524); Mun. Lagarto, 0.5 km S of Rio Vaza Barris, coming from São Domingos,
14 Feb. 1991, Taylor & Zappi 1633 (HRCB, K, ZSS).
bahia: Mun. Jeremoabo, Uebelmann (mss, HU 253); Mun. Jaguarari, Ritter 1221 (SGO 125489, fide Eggli
PILOSOCEREUS
et al. 1995: 497); Mun. Senhor do Bonfim, 8–9 June 1915, Rose & Russell 19921 (US, NY); Mun. Itiúba,
16 km E of Filadélfia, 25 Apr. 1992, Taylor & Zappi (obs.); Mun. Ponto Novo, 41 km N of Capim
Grosso, 25 Apr. 1992, Taylor & Zappi (obs.); Mun. Jacobina, 21 km NW of town on road BR 324
towards Lages, 16 July 1989, Zappi 131A (SPF, HRCB); l.c., 2 km S of town towards Miguel Calmon,
26 Apr. 1992, Taylor & Zappi (obs.); Mun. Capim Grosso, 16 km N of town, 25 Apr. 1992, Taylor &
Zappi (obs.); Mun. São José do Jacuípe, 10 km SE of Capim Grosso, 16 km NW of bridge over the Rio
Jacuípe, 25 Apr. 1992, Taylor & Zappi (obs.); Mun. Santaluz, c. 18 km N of Valente, 29 Dec. 1992, L.P.
de Queiroz et al. 3015 (K, CEPEC, HUEFS); Mun. Araci, 18 Jan. 1994, M.A.S. das Neves 89 (HUEFS);
Mun. Itapicuru, 11°8'S, 38°13'W, 20 Feb. 1978, R.P. Orlandi 166 (RB); Mun. Morro do Chapéu, W of
Ventura, 24 Dec. 1988, Taylor & Zappi s.n. (K, photos); Mun. Serrinha, 16 km W of town on road BA
409 towards Conceição do Coité, 14 July 1989, Zappi 120 (SPF, HRCB); Mun. Piritiba, 2 km S of
França, 26 Apr. 1992, Taylor & Zappi (obs.); l.c., 13 km S, 26 Apr. 1992, Taylor & Zappi (obs.); Mun.
Ichu (Ixu), 2 km S of town, 8 July 1994, F.A.R. Santos 91 (HUEFS); Mun. Biritinga, 18 Jan. 1994,
M.A.S. das Neves 76, 79, 80 (HUEFS); Mun. Mundo Novo, 7 & 15 km E of town on road BA 052, 26
Apr. 1992, Taylor & Zappi (obs.); Mun. Lamarão, on road between Feira de Santana and Serrinha, 4 Jan.
1991, Zappi in Taylor et al. 1348B (K, ZSS, photos); Mun. Santa Bárbara, on road between Feira de
Santana and Serrinha, 4 Jan. 1991, Taylor et al. 1348A (HRCB); Mun. Baixa Grande, 13 km SE of town,
road BA 052, 26 Apr. 1992, Taylor & Zappi (obs.); Mun. Ipirá, 26 Apr. 1992, Taylor & Zappi (obs.); Mun.
Feira de Santana, 12°15'S, 38°58'W, 5 Nov. 1983, L.R. Noblick 2751 (HUEFS); Mun. Castro Alves, 26
Jan. 1956, Andrade-Lima 56-2508 (IPA, PEUFR), 13 Apr. 1995, L.P. de Queiroz 4325 (HUEFS); Mun.
Muritiba, Faz. 8 de Dezembro, nr Itaporã, 6 Aug. 2002, Taylor (obs.); Mun. Lauro de Freitas, bairro de
Stella Maris, dunas do aeroporto, Jan. 2001, E.A. Rocha (K, photo); Mun. Boninal/Piatã, 12 km S of
Boninal on road towards Piatã, 12°48'S, 41°48'W, 22 Dec. 1988, Taylor & Zappi in Harley 25597 (K, SPF,
CEPEC); Mun. Itatim, Morro das Tocas, 12°42'S, 39°46'W, 24 Feb. 1996, E. de Melo 1444 (HUEFS);
Mun. Mucugê, 0.5 km from town, near the cemetery, 13°S, 41°28'W, 22 Dec. 1988, Taylor & Zappi in
Harley 27380 (K, SPF, CEPEC); Mun. Milagres, 26 Oct. 1978, A. Araújo 121 (RB); Mun. Marcionílio
Souza, Machado Portela, 19–23 June 1915, Rose & Russell 19921 (US, NY); Mun. Jaguaquara [Toca da
Onça], 27–29 June 1915, Rose & Russell 20081 p.p. (US, NY).
Besides the examined specimens cited above, this taxon has been observed in the following municípios
in Bahia: Tapiramutá, Andaraí, Olindina, Rui Barbosa, Itaberaba, Jequié, Vitória da Conquista.
Concern at present, but its extensive habitat continues to decline. It is rare in the north-eastern part of
its range.
8b. subsp. robustus Zappi in Succ. Pl. Res. 3: 74 (1994). Holotype: Brazil, Bahia, Mun.
Livramento do Brumado, 11 km S of town on road to Brumado, 450 m, 13°45'S,
41°49'W, 23 Nov. 1988, Taylor & Zappi in Harley 25544 (SPF; CEPEC, K, isos.).
Shrubby or treelike, densely branched, rarely more than 4 m tall; branches erect, to 7.5 cm diam.; ribs
(5–)6–10, 15–21 × 8–19 mm, acute. Central spines 0–2(–3), 20–26 mm, 1–2 ascending and sometimes 1
porrect; radials 3–8, 4–20 mm, well-developed, adpressed.
Eastern caatinga element: in dense caatinga-agreste of the Rio de Contas (Rio Gavião) and Rio Pardo
drainage systems, c. 400–900 m, southern and south-eastern Bahia and north-eastern Minas Gerais.
bahia: Mun. Ituaçu, 2 km NE of town, 18 Aug. 1988, Eggli 1197 (ZSS); Mun. Rio do Antônio, Km
43 on Brumado–Caitité road, 14 Apr. 1983, Leuenberger et al. 3068 (CEPEC, K), l.c., 2 Feb. 1991, Taylor
et al. 1523 (K, ZSS, HRCB, CEPEC); Mun. Maracás, Km 16 of Maracás–Tamburi road, 20 Apr. 1983,
Leuenberger et al. 3085 (CEPEC, K); Mun. Caitité, 8 km from Caitité towards Brumado, track to São
João, 27 Aug. 1988, Eggli 1319 (ZSS); Mun. Livramento do Brumado, 11 km S on road to Brumado,
13°45'S, 41°49'W, 23 Nov. 1988, Taylor & Zappi in Harley 25544 (K, SPF, CEPEC); Brumado–Caitité,
Horst & Uebelmann HU 121 (ZSS); Mun. Vitória da Conquista, 7 km NE of town on road BR 116, then
PILOSOCEREUS
3.5 km W, 16 April 2003, Taylor & Zappi (K, photos); l.c., c. 20 km S of town, 15°1'21"S, 40°49'51"W,
18 April 2003, Taylor et al. (obs.).
minas gerais: Mun. Taiobeiras, 35 km E of town towards road BR 116, E of bridge over Rio Atoleiro,
15°53'S, 41°57'W, 17 Oct. 1988, Taylor & Zappi in Harley 25149 (K, SPF); Mun. Águas Vermelhas, 1 km
from Curral de Dentro towards road BR 116, 31 Jan. 1991, Taylor et al. 1515D (K, HRCB).
Besides the examined specimens cited above, this taxon has been observed in the following
municípios of Bahia: Iramaia, Itaetê, Barra da Estiva, Contendas do Sincorá.
conservation status. Near Threatened [NT] (1); PD=1, EI=1, GD=2. Short-list score (1×4) = 4. At
present with an extent of occurrence of > 20000 km2, but its habitat is discontinuous and continues to
decline and regular monitoring is desirable.
Variable in rib number and stem thickness.

Intergeneric hybrids between P. pentaedrophorus and Micranthocereus purpureus, cited by
Ritter (1979), have been observed in the region of Andaraí and Ituaçu, Bahia. At the
southern limit of distribution of P. pentaedrophorus, in the drainage of the Rio Pardo (Taylor
et al. 1515D), P. p. subsp. robustus can be found sympatric with P. floccosus subsp. quadricostatus.
9. Pilosocereus glaucochrous (Werderm.) Byles & G. D. Rowley in Cact. Succ. J. Gr.

Brit. 19: 67 (1957). Type: Brazil, Bahia, near Morro do Chapéu, Serra do Espinhaço, c.
1000 m, Apr. 1932, Werdermann (B†). Lectotype (Zappi 1994): Werdermann, l.c., 102,
photograph (iconotype).
Pilocereus glaucochrous Werderm., Bras. Säulenkakt.: 106–107 (1933). Cephalocereus glaucochrous (Werderm.)
Borg, Cacti: 108 (1937). Pseudopilocereus glaucochrous (Werderm.) Buxb. in Beitr. Biol. Pflanzen 44:
252 (1968).
Treelike, to 3–5 m, sparsely branched, sometimes with a well-defined trunk; vascular cylinder woody;
branches 3–6.5(–7.0) cm diam., inclined, epidermis bluish or greyish green, glaucous; ribs (3–4[in
seedlings/juveniles]–)5–11, 7–12 × 7–18 mm, sinuses straight, transverse folds above the areoles straight.
Areoles 4–6 mm, 6–10 mm apart, situated upon rounded, low projections (podaria), with blackish felt
and white long hairs to 20 mm, mainly on the apical region of branches. Spines 0.5–1.0 mm diam. at
base, translucent when young, golden yellow to greyish, centrals 2–4(–5), ascending to porrect,
20–35(–40) mm, radials 7–14, adpressed, mainly horizontal, 8–15 mm. Fertile part of stem not or slightly
differentiated, apical to subapical, on many consecutive areoles of 2–5 ribs. Flower-buds obtuse 2 days
before anthesis; flowers 4–5.2 × 2.5 cm; pericarpel subglobose, tube 34–38 × 10–12 mm at base, flared
to 15–19 mm diam. at apex, curved, cylindric, constricted above and below the nectar-chamber, pink to
orange-red, smooth, glaucous, distal third with ovate, wine-coloured bract-scales; style 24–29 × 1.2–1.7
mm, tapering, stigma-lobes 7–8, included in relation to the stamens; ovary locule hemicircular to
depressed in longitudinal section. Fruit depressed-globose, dehiscent by a lateral slit, floral remnant
pendent, deeply inserted; pericarp olive-green to purplish, smooth; funicular pulp purplish. Seeds 1.6–1.9
× 1.1–1.2 mm, testa-cells flat with inconspicuous intercellular depressions and fine, sparse cuticular folds.
Caatinga / Northern campos rupestres (Chapada Diamantina) element: in ‘caatinga/cerrado de altitude’, c.

740–1150 m, on various substrates including sand and calcareous materials, Chapada Diamantina, central
Bahia. Endemic. Map 36.
bahia: Mun. Sento Sé, Serra São Francicsco (S of Minas do Mimoso fide G. Charles), Herm et al. (2001:
118, fig. 3); Mun. Jacobina, 28 km W of town on road BR 324 towards Lages (Lajes), 12 Jan. 1991,
Taylor et al. 1398A (K, HRCB, ZSS); Mun. Morro do Chapéu, 18 June 1994, L.P. de Queiroz 4014
(HUEFS); l.c., 15 km NW of town towards Várzea Nova, 12 Jan. 1991, Taylor et al. 1402 (K, HRCB,
ZSS, CEPEC); l.c., 5 km E of town, 3 Mar. 1973, I. & G. Gottsberger 16-3273 (K); l.c., road to Morrão,
11°35'3"S, 41°11'31''W, 5 Aug. 2001, F.R. Nonato et al. 974 (HUEFS 55719); Mun. América Dourada,
PILOSOCEREUS
Nova América, Horst & Uebelmann 218, cult. ZSS (ZSS); Mun. Bonito, 46 km S of Morro do Chapéu
on road from Utinga, 23 Dec. 1988, Taylor & Zappi in Harley 27384 (K, SPF, CEPEC); l.c., 10 km N of
Bonito, 23 Aug. 1988, Eggli 1268 (ZSS); Mun. Iraquara, 8 km S of Souto Soares towards Seabra, 13 Jan.
1991, Taylor et al. 1412 (K, HRCB, ZSS, CEPEC); Mun. Seabra, 16 km W of Seabra on road BR 242
to Ibotirama, 25 Aug. 1988, Eggli 1290 (ZSS); l.c., 27 km W, July 1989, Taylor & Zappi (K, photos).
Near Threatened, since its habitat continues to decline and the number of populations is small (c. 10),
although the plant is abundant over a wide area around the type locality.
The brightly coloured tube of the flower of this species is unusual in the genus, otherwise
occurring only in P. machrisii and P. aurisetus. The nocturnal flowers of P. glaucochrous
remain open for part of the following morning and, therefore, may be adapted for
pollination by both bats and hummingbirds. Variable in rib number between populations.
10. Pilosocereus floccosus Byles & G. D. Rowley in Cact. Succ. J. Gr. Brit. 19: 67
(1957). Holotype: Brazil, Minas Gerais, Diamantina, June 1934, A. C. Brade s.n. (RB).
Pilocereus floccosus Backeb. & Voll in Arch. Jard. Bot. Rio de Janeiro 9: 150 (1949), nom. illeg. (Art. 53)
non Lemaire in Ill. Hort. 13: tab. 470 (1866). Pseudopilocereus floccosus (Byles & G. D. Rowley) Buxb.
in Beitr. Biol. Pflanzen 44: 252 (1968).
[Cereus macrogonus sensu Warming (1892), cf. Warming (1908: 151, footnote, 254), non Salm-Dyck (1850).]
Shrub, not branched above base, or treelike, 1–4(–5) m, with or without a well-defined trunk, sparsely
or much branched; vascular cylinder weakly woody; branches erect, epidermis verrucose, greyish green;
ribs with straight sinuses and without visible transverse folds. Areoles 3–10 mm, (2–)4–10 mm apart, with
white to brownish grey felt and white or brownish long hairs to 10 mm, growth indeterminate near base
of plant. Spines 0.3–1.1 mm diam. at base, opaque, yellowish brown or reddish, becoming greyish. Fertile
part of stem strongly differentiated, flower-bearing areoles with pale brown or grey, woolly, long hairs to
15–20 mm and dark bristly spines to 35 mm, forming a ring around the subapical region of the stem.
Flower-buds obtuse 2–3 days before anthesis; flowers 4–5 × 3 cm; pericarpel subglobose, tube to 40 × 12
mm at base, flared to 25 mm diam. at apex, curved, constricted above and below nectar-chamber, green,
glaucous, smooth, distal third with broad ovate bract-scales; style 30 × 1.5–1.7 mm, tapering, stigma-lobes
9–10, exserted or included in relation to the stamens; ovary locule obtriangular in longitudinal section.
Fruit dehiscent by a lateral, abaxial or adaxial slit; pericarp wine-coloured, smooth to somewhat rugose;
funicular pulp bright red. [Seeds: see subspecies.]
conservation status. Near Threatened [NT]; see subspecies below.
When first described by Backeberg & Voll, l.c., this species was placed in the genus
Pilocereus, as P. floccosus Backeb. However, this name is illegitimate because of the prior-
published Pilocereus floccosus Lem., which relates to a Caribbean taxon (nowadays a
synonym of Pilosocereus royenii (L.) Byles & G. D. Rowley). The legitimate name is,
therefore, to be attributed to Byles & Rowley (1957), who published Pilosocereus floccosus
Byles & G. D. Rowley based on the type of Backeberg’s illegitimate name.
Easy to distinguish from all the other species of the genus by its rough, verrucose
epidermis, P. floccosus is also notable for its long-hairy flower-bearing areoles, forming a
crown or zone at the apical or subapical region of the branches. A potentially closely related
taxon occurs in north-western Minas Gerais — P. albisummus P. J. Braun & E. Esteves
Pereira in Kakt. and. Sukk. 38: 126–131 (1987). It has not been possible to study this plant
PILOSOCEREUS
at close quarters in the field, but living and herbarium material at Zürich (ZSS), photographs
of the plant, flower-buds and seeds (SEM), and its known ecological preference for Bambuí
limestone outcrops are strongly indicative of an affinity to P. floccosus subsp. floccosus.
The typical subspecies occurs only on limestone rock outcrops of the Bambuí
formation, in central Minas Gerais, where it does not normally become very tall, unless
growing in very dense forest. North-east of the distribution of subsp. floccosus, P. f. subsp.
quadricostatus lives in the caatinga associated with gneissic outcrops of the semiarid region
of the Rio Jequitinhonha and Rio Pardo drainage basins (Taylor & Zappi 1992a). This
species is subdivided as follows:
1. Branches 5–9 cm diam.; ribs 5–8; seeds shiny, testa-cells without cuticular folds (SEM)
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10a. subsp. floccosus
1. Branches 8–11 cm diam.; ribs 4–5; seeds dull, testa-cells with dense cuticular folds
(SEM) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10b. subsp. quadricostatus
10a. subsp. floccosus

Shrub, branching near base or above ground, rarely treelike, with or without a well-defined trunk;
branches 5–8 cm diam., ribs 5–8, 12–30 × 20–30 mm. Central spines 3–6, 8–25 mm, ascending to
porrect, radials 5–16, 2–25 mm, ascending or adpressed. Fruit 2.5–3.0 × 3–3.5 cm, depressed-globose.
Seeds (1.8–)1.9–2.2 × (1.2–)1.3–1.4(–1.5) mm, hilum-micropylar region forming an angle of 50° with
long-axis of seed, testa-cells flat with inconspicuous intercellular depressions and lacking cuticular folds.
Rio São Francisco (Rio das Velhas) caatinga/mata seca element: mostly on Bambuí limestone outcrops west
of the Serra do Espinhaço, c. 600–800 m, Minas Gerais. Endemic to the core area within Minas Gerais.
Map 36.
minas gerais: Mun. Buenópolis, 7 km S of town, on road BR 135 towards Curvelo, 17°56'S, 44°9'S,
11 Oct. 1988, Taylor & Zappi in Harley 24838 (K, SPF); Diamantina, June 1934, Brade s.n. (RB); Mun.
Monjolos, Horst & Uebelmann 583 (ZSS); l.c., near Rodeador, 1983, Braun 436 (ZSS T-06119); l.c. (?), W
of Conselheiro Mata, Braun 436 bis (ZSS TP-58-180); Mun. Santana do Riacho, Serra do Cipó, limestone
outcrop near Cardeal Mota, 30 Mar. 1988, Zappi in CFSC 10923 (SPF); l.c., 27 Oct. 1988, Taylor & Zappi
in Harley 25425 (K, SPF); Mun. Matozinhos, Faz. Cauaia, 31 Oct. 1996, Lombardi 1462 (BHCB, K); Mun.
Lagoa Santa, Lapa Vermelha, reported by Warming (1908), cf. Warming & Ferri (1973).
conservation status. Near Threatened [NT] (1); PD=1, EI=1, GD=1. Short-list score (1×3) = 3. A
small number of populations is currently known (c. 8) and its habitat on limestone outcrops will in future
be at risk from quarrying operations. Forest surrounding outcrops may also be cleared affecting local climate.
10b. subsp. quadricostatus (F. Ritter) Zappi in Succ. Pl. Res. 3: 86 (1994). Holotype:
Brazil, Minas Gerais, Padre Paraíso (formerly Água Vermelha), 1965, Ritter 1342 (U).
Pilosocereus quadricostatus F. Ritter, Kakt. Südamer. 1: 78, Abb. 51 (1979). Pseudopilocereus quadricostatus (F.
Ritter) P. V. Heath in Calyx 4: 140 (1994).
Shrubby or treelike, branched above ground, with a well-defined trunk; branches 8–11 cm diam., ribs
4–5(–6), 3–7 in young plants/seedlings (fide Ritter), 26–30 × 15–40 mm. Areoles either with very short
spines or central-spines 1–3, 2–50 mm, ascending to porrect, and radials 6–8, c. 2–10 mm, ascending to
adpressed. Fruit to 6 × 5 cm, depressed but often compressed due to the presence of many fruits close
together on the same rib. Seeds 1.7–1.8 × 1.1–1.3 mm, hilum-micropylar region forming an angle of 40°
with the long-axis of seed, testa-cells domed to flat with inconspicuous intercellular depressions and
dense, coarse cuticular folds.
PILOSOCEREUS
South-eastern caatinga element: in caatinga and mata seca decídua on associated gneissic inselbergs within the
drainage of the Rio Jequitinhonha and Rio Pardo, c. 250–800 m, north-eastern Minas Gerais. Endemic
to the core area within Minas Gerais. Map 36.
minas gerais: Mun. Águas Vermelhas, 1 km E of Curral de Dentro, 31 Jan. 1991, Taylor & Zappi s.n.
(K, photos); Mun. Grão Mogol, 14 km NE of town on road to Salinas, 16 Oct. 1988, Taylor & Zappi in
Harley 25146 (K, SPF) — aberrant form perhaps intermediate with subsp. floccosus; Mun. Coronel Murta,
5 km SE of town on road to Araçuaí, 21 Feb. 1988, Supthut 8861 (ZSS); Mun. Pedra Azul, 6 km S of
town on road to Jequitinhonha, 16°3'S, 41°14'W, 20 Oct. 1988, Taylor & Zappi in Harley 25221 (K, SPF);
Mun. Itaobim, road BR 367, 12–21 km W of road BR 116, 14 Dec. 1990, Taylor & Zappi 765 (K,
HRCB, ZSS, BHCB); Mun. Itinga, caatinga of Faz. Alta Viagem, 17 Oct. 1983, Rizzini & Mattos-F. s.n.
(RB 232050); Mun. Padre Paraíso [Água Vermelha], Ritter 1342 (U); l.c., ‘Km 788’ of road BR 116,
Horst & Uebelmann 172, cult. (ZSS).
conservation status. Vulnerable [VU B2ab(iii)] (2); remaining area of occupancy < 2000 km2;
PD=1, EI=1, GD=1. Short-list score (2×3) = 6. Vulnerable from the ongoing destruction of its caatinga
habitat and low number of known populations.
10b 13: Pilosocereus floccosus subsp. quadricostatus Pilosocereus magnificus

P. subsimilis Rizzini & A. Mattos in Revista Brasil. Biol. 46: 327 (1986). Holotype: Brazil, Minas Gerais,
Itinga, 18 Dec. 1984, Rizzini & Mattos 41 (RB).
Treelike or shrubby, 2–5 m, usually with a ± well-defined trunk, sparsely branched above ground;
vascular cylinder weakly woody; branches 7–10 cm diam., erect, epidermis dark green to olive-green,
more or less glaucous, slightly rough; ribs 4–7, 17–30 × 20–35 mm, sinuses straight, transverse folds not
visible. Areoles 4–6 mm diam., 2–5 mm apart, felt greyish, with white to brownish, long hairs 6–10 mm.
Spines 0.3–0.5 mm diam. at base, translucent when young, yellowish, pale brown or reddish, centrals
4–5, 7–15 mm, ascending to porrect, radials 10–14, 6–10 mm, ascending to porrect or deflexed. Fertile
part of stem slightly to strongly differentiated, flower-bearing areoles situated on all ribs of the apex of
branches forming a ring, developing black, bristly spines to 30 mm and white long hairs to 10–20 mm;
flowers 5 × 2.5 cm; tube 27 × 12 mm at base, to 20 mm diam. at apex, curved, constricted above and
below the nectar-chamber, olive-green, smooth, distal third with broad bract-scales. Fruit depressed-
globose; funicular pulp magenta. Seeds (1.5–)1.6(–1.7) × 1–1.1(–1.2) mm, hilum-micropylar region
forming an angle of 40° with the long-axis of seed, testa-cells domed to flat, with intercellular depressions
nearly inconspicuous, and coarse, sparse cuticular folds.
On gneissic inselbergs and in associated caatinga-agreste, c. 250–600 m, north-eastern Minas Gerais.

minas gerais: Mun. Padre Paraíso (formerly Água Vermelha), on road BR 116, 12 km N of town, 13
Dec. 1990, Taylor & Zappi 756 (K, HRCB, ZSS, BHCB); Mun. Itinga, 18 Oct. 1984, Rizzini & Mattos-
F. 41 (RB); l.c., 5 km E of town on road BR 367, 16°36'S, 41°43'W, 19 Nov. 1988, Taylor & Zappi in
Harley 25535 (K, SPF).
P. floccosus subsp. quadricostatus is also suspected of hybridizing with P. multicostatus, q.v.
PILOSOCEREUS ULEI Group (Nos 11–13).

Found in caatinga-agreste, caatinga and campo rupestre.
11. Pilosocereus fulvilanatus (Buining & Brederoo) F. Ritter, Kakt. Südamer. 1: 84

(1979). Holotype: Brazil, Minas Gerais, N Serra do Espinhaço, [Grão Mogol], 800–1000
m, 1968, Horst 277 (U).
Pseudopilocereus fulvilanatus Buining & Brederoo in Kakt. and. Sukk. 24: 145–147 (1973).
PILOSOCEREUS
Shrubby to tree-like, 2–3(–4) m, branching above the ground; vascular cylinder weakly woody; branches
5–12 cm diam., erect, epidermis greyish green, coated with intense sky blue wax near apex, grey-green
below; ribs 4–8, sinuses straight, transverse folds absent. Areoles almost contiguous, felt dark brown or
black with dark long hairs, paler to glabrescent with age. Spines opaque, yellowish brown to blackish,
thickened at base. Fertile part of stem strongly differentiated, flower-bearing areoles subapical, on one or
several ribs, with abundant reddish brown and golden brown long hairs to 20–40 mm. Flowers 3.6–6.0
× 3–4 cm; pericarpel subglobose, greenish; tube straight, infundibuliform, distal third with thick acute
bract-scales; nectar-chamber swollen; ovary locule hemicircular to triangular in longitudinal section. Fruit
depressed-globose, frequently compressed, dehiscent by lateral slit, floral remnant pendent, deeply
inserted; pericarp greenish, pink or dark purplish; funicular pulp magenta. Seeds with flat testa-cells,
intercellular depressions nearly absent, cuticular folds absent.
conservation status. Vulnerable [VU B1ab(iii)]; extent of occurrence estimated to be < 20000 km2
and presently known from c. 6 locations, some of which are projected to decline in the near future.
This distinctive species can be easily differentiated from the rest of the genus by the
unusual combination of intensely blue epidermis and dark reddish brown to golden brown
areolar hairs, which are more abundant when the plants are flowering. The areoles with
short dark spines are so closely arranged that it is sometimes difficult to isolate one from
the next. It is the sister species of P. ulei, from the region of Cabo Frio, Rio de Janeiro.
This taxon is divided into two subspecies, found on crystalline rock outcrops associated
with the campos rupestres, on both sides of the Serra do Espinhaço, Minas Gerais, where
the species is endemic. The subspecies are differentiated as follows:
1. Branches 8–12 cm diam.; ribs 4–7; fruit dark pink to dark purple . . . . . 11a. subsp. fulvilanatus
1. Branches to 5.5 cm diam.; ribs (5–)6–8; fruit green to brownish red . . . . . . . . . 11b. subsp. rosae
11a. subsp. fulvilanatus

Treelike or shrubby, 2–3(–4) m, branching above the ground; branches 8–12 cm diam.; ribs 4–7, 30–40
× 30–40(–45) mm. Areoles 3–7 × 2–5 mm, 1–3 mm apart, very congested. Spines 0.6–0.9 mm diam. at
base, centrals 1–7, 15–30(–45) mm, ascending to porrect, radials 8–10, 8–15 mm, adpressed. Flower-
bearing areoles extremely hairy. Flowers 3.6–5.2 × 3 cm; tube 21 × 11 mm at base, flared to 20 mm diam.
at apex, greenish; style 20 mm, stigma-lobes c. 12. Fruit 3 × 3.5–4.0 cm; pericarp dark pink to dark
purple, rugose. Seeds 1.5–1.6 × 1–1.1 mm.
South-eastern campo rupestre (Grão Mogol region) element: locally co-dominant with other woody vegetation
on quartzitic rock outcrops, campo rupestre, Serra do Espinhaço, in the drainage of the Rio Jequitinhonha,
c. 720–1000 m, northern Minas Gerais. Endemic to the core area within Minas Gerais. Map 37.
minas gerais: Mun. Grão Mogol, Serra da Bocaina, 28 km E of Caveira and 48.5 km W of Rio Vacaria
on road BR 251, 31 Jan. 1991, Taylor et al. 1513 (K, HRCB, ZSS, BHCB); Serra de Grão Mogol, 1968,
Horst 277 (U); l.c., 18 km W of town, 21 Feb. 1969, Irwin et al. 23676 & 23684 p.p. (NY); l.c., campo
rupestre NE of the town, 22 May 1982, M.C.H. Mamede et al. in CFCR 3475 (SPF); Mun. Cristália,
slopes of the Serra das Cabras, 28 May 1988, Zappi et al. in CFCR 12046 (SPF); l.c., margins of the Rio
Itacambiruçu, 16°37'S, 42°56'W, 15 Oct. 1988, Taylor & Zappi in Harley 25071 (K, SPF); Mun.
Botumirim, Ribeirão Gigante, 1991, M.G. Carvalho s.n. (K, photos).
conservation status. Vulnerable [VU B1ab(iii)] (2); extent of occurrence = 724 km2; PD=1, EI=2,
GD=1. Short-list score (2×4) = 8. Future inundation by a dam-lake and destruction of its habitat for
charcoal production make this taxon Vulnerable. It is more secure in the higher parts of its range.
PILOSOCEREUS
Known from only three populations between Grão Mogol and Botumirim (MG), subsp.
fulvilanatus presents a rather restricted distribution, but it has to be said that this region is
still rather under-explored.
11b. subsp. rosae (P. J. Braun) Zappi in Succ. Pl. Res. 3: 100 (1994). Holotype: Brazil,
Minas Gerais, Mun. Augusto de Lima, nr Santa Bárbara, 6 km from road BR 135, west
slopes of the Serra do Espinhaço, 800 m, 1982, Horst & Uebelmann 546 (ZSS; K, iso.).
Pilosocereus rosae P. J. Braun in Kakt. and. Sukk. 35(8): 178–181 (1984).
Shrubby, to 3 m, sparsely branched above ground, without a well-defined trunk; branches 5.5 cm diam.;
ribs (5–)6–8, 15–20 × 15–25 mm. Areoles 3–4 × 3–4 mm, 2–4 mm apart. Spines 0.4–0.6 mm diam. at
base; centrals 3–4, 10–20 mm; radials 8–10, to 12 mm, adpressed. Fertile part of stem differentiated, with
long hairs to 3–4 cm, and black bristly spines to 2–3 cm. Flowers 3.8–6.0 × 3–4 cm; tube 30–45 × 11
mm at base, flared to 25 mm diam. at apex, straight, greenish brown to pinkish; style 25–30 mm, stigma-
lobes c. 10. Fruit to 6.5 cm diam.; pericarp green to brownish red, striate. Seeds 1.6–1.8 × 1–1.2 mm.
[Description based on dried material and protologue.]
South-eastern campo rupestre (Rio São Francisco drainage) element: on quartzitic rock outcrops, campo
rupestre, c. 800 m, Serra do Espinhaço, in the drainage of the Rio das Velhas, central-northern Minas
minas gerais: Mun. Augusto de Lima, near Santa Bárbara, 6 km from road BR 135, west slopes of the
Serra do Espinhaço, 1982, Horst & Uebelmann 546 (ZSS, K); l.c., 6 Aug. 1988, Eggli 1104 (ZSS).
conservation status. Critically Endangered [CR B1ab(iii) + 2ab(iii)] (4); area of occupancy < 10
km2; PD=1, EI=1, GD=1. Short-list score (4×3) = 12. The single site known is now a private park of
difficult access, but needs to be visited to determine current threats.
Although it has not been possible to observe the population of subsp. rosae in the field, its
striking morphological and ecological similarities with P. fulvilanatus, sens. str., have led to
its present position (Zappi 1994). It has a less robust habit, with branches less than 6 cm
in diameter, and a tendency to develop a higher number of ribs (6–8) than in mature
stems of subsp. fulvilanatus (4–6). Otherwise, the remaining characteristics of both taxa are
very similar, and only the geographical separation on either side of the Serra do Espinhaço
— the two taxa being found c. 170 km apart — justifies the acceptance of P. rosae as a
very restricted western subspecies, which is so far known from only a single population,
found near the village of Santa Bárbara, in the Município of Augusto de Lima, Minas
Gerais. Its extent of occurrence remains uncertain.
12. Pilosocereus pachycladus F. Ritter, Kakt. Südamer. 1: 69–70, Abb. 5, 40–41

(1979). Holotype: Brazil, Bahia, Urandi, 1964, Ritter 1223 (U).
Pseudopilocereus azureus Buining & Brederoo in Kakt. and. Sukk. 26: 241–243 (1975), non P. azureus (F.
Ritter) P. V. Heath (1994) nec Pilosocereus azureus F. Ritter (1979), see below. Holotype: Brazil,
Bahia, Mun. Brumado, Umburanas, near the Rio Brumado, 450 m, s.d., Horst HU 237 (U).
Pilosocereus cyaneus F. Ritter, Kakt. Südamer. 4: 1516 (1981). Type: as above. [Nom. nov. pro
Pseudopilocereus azureus Buining & Brederoo.]
P. pachycladus (F. Ritter) P. V. Heath in Calyx 4: 140 (1994).
Pilosocereus azureus F. Ritter, Kakt. Südamer. 1: 73 (1979). Pseudopilocereus azureus (F. Ritter) P. V. Heath,
PILOSOCEREUS
l.c. (1994), nom. illeg. (Art. 53.1). Holotype: Brazil, Minas Gerais, Januária, 1959, Ritter 958 (U).
Pilosocereus atroflavispinus F. Ritter, ibid.: 68 (1979). Pseudopilocereus atroflavispinus (F. Ritter) P. V. Heath,
l.c. (1994). Holotype: Brazil, Bahia, Ituaçu, on rocks, 1964, Ritter 1349 (U).
Pilosocereus oreus F. Ritter, ibid.: 69–70 (1979). Pseudopilocereus oreus (F. Ritter) P. V. Heath, l.c. (1994).
Holotype: Brazil, Minas Gerais, Monte Azul, 1964, Ritter 1226 (U).
Pilosocereus pernambucoensis var. montealtoi F. Ritter, ibid.: 66 (1979). Pseudopilocereus pernambucoensis var.
montealtoi (F. Ritter) P. V. Heath, l.c. (1994). Holotype: Brazil, Bahia, Palmas de Monte Alto, 1964,
Ritter 1225 (U).
? Pilosocereus splendidus F. Ritter, ibid.: 69, Abb. 36 (1979). Pseudopilocereus splendidus (F. Ritter) P. V.
Heath, l.c., 141 (1994). Holotype: Brazil, Bahia, Urandi, 1964, Ritter 1224 (U).
Pilosocereus superbus F. Ritter, ibid.: 67 (1979). Pseudopilocereus superbus (F. Ritter) P. V. Heath, l.c. (1994).
Holotype: Brazil, Bahia, Anajé [Anagé], 1964, Ritter 1347 (U).
Pilosocereus superbus var. regius F. Ritter, ibid.: 67, Abb. 34 (1979). Pseudopilocereus superbus var. regius (F.
Ritter) P. V. Heath, l.c. (1994). Holotype: Brazil, Bahia, Brumado, Rio Brumado, 1964, Ritter 1343 (U).
Pilosocereus superbus var. gacapaensis F. Ritter, ibid.: 67, Abb. 36 (1979). Pseudopilocereus superbus var.
gacapaensis (F. Ritter) P. V. Heath, l.c. (1994). Holotype: Brazil, Bahia, Mun. Riacho de
Santana/Igaporã, Serra da Garapa [‘Gacapa’], Ritter 1343A (U).
Pilosocereus superbus var. lanosior F. Ritter, ibid. 67 (1979). Pseudopilocereus superbus var. lanosior (F. Ritter)
P. V. Heath, l.c. (1994). Holotype: Brazil, Bahia, Caitité, 15 km W of town, Ritter 1343B (U).
Pilosocereus cenepequei Rizzini & A. Mattos in Revista Brasil. Biol. 46: 324–327 (1986). Holotype: Brazil,
Minas Gerais, Pedra Azul, 15 Oct. 1984, Rizzini & A. Mattos-Filho 39 (RB).
P. schoebelii P. J. Braun in Cact. Succ. J. (US) 58: 150–156 (1987). Holotype: Brazil, N Minas Gerais
[probably near Janaúba], 600 m, Braun 426 (ZSS).
[Pilocereus glaucescens sensu Werderm., Bras. Säulenkakt.: 107, photo (1933).]
[Cereus ulei sensu Luetzelburg, Estud. bot. Nordéste 3: 69, 111 (1926) non (Gürke) A. Berger.]
vernacular names. Facheiro, Facheiro-azul, Mandacaru-de-facho.
Treelike or shrubby, 2–10 m or more; vascular cylinder weakly to moderately woody; branches
5.5–10.0(–15.0) cm diam., erect, epidermis glaucous, grey-green to blue, smooth; ribs 5–19, sinuses
straight, transverse folds above the areoles visible at apex of young branches. Areoles 3–10 mm diam.,
2–10 mm apart, felt white to greyish, with white long hairs. Spines translucent, golden-yellow, greyish
with age. Fertile part of stem slightly to strongly differentiated. Flower-buds acute or obtuse 1–2 days
before anthesis; flowers very variable, 4–7 × 2.2–4.5 cm, sometimes compressed; pericarpel subglobose;
tube 25–50 × (8–)12–25 mm at base, flared to 24–35 mm diam. at apex, constricted below nectar-
chamber, slightly to strongly curved, pale green to dark brown, distal third with broad bract-scales; inner
perianth-segments white to pink; style 20–58 × 1.2–3.0 mm, tapering; stigma-lobes 9–13, exserted in
relation to the anthers; ovary locule broadly hemicircular to triangular in longitudinal section. Fruit
depressed-globose, dehiscent by a lateral slit, floral remnant pendent, deeply inserted; pericarp purplish;
funicular pulp magenta. [Seed morphology: see subspecies.]
Although P. pachycladus is one of the most conspicuous species in Eastern Brazil, being
both common and widely distributed, it has an involved history of taxonomic confusion
and was first unequivocably named less than 30 years ago by Buining & Brederoo (1975),
whose chosen epithet is blocked within Pilosocereus. One of the earlier names associated
with this taxon, Pilocereus glaucescens Linke 1858, is of doubtful application, having been
based on sterile living material collected in Brazil, without a precise locality (see
‘Insufficiently known taxa’, following P. densiareolatus). Ritter (1979) argues that the
description is very ambiguous and is impossible to attribute to a single taxon with
certainty. From this description, in fact, one cannot exclude P. glaucochrous (Werderm.)
Byles & G. D. Rowley, some of the forms of P. pachycladus, or even the possibility of the
original plant being a specimen from another country. Werdermann (1933, 1942) used
PILOSOCEREUS
the name Pilocereus glaucescens for populations here included within Pilosocereus pachycladus
subsp. pachycladus. A second name incorrectly used for this species in the broad sense was
Cereus ulei by Luetzelburg (1925–26, 3: 69), while a third once applied thus is P.
piauhyensis (see P. pachycladus subsp. pernambucoensis and P. piauhyensis).
This very variable species comprises a number of heterotypic synonyms described by
Ritter, Buining & Brederoo, and Braun (see above). The species concepts utilized by
these authors have proved to be too narrow and, if applied to all the forms now known,
would lead to a new species name for each slightly different population of this complex.
P. pachycladus presents a broad range of forms, including tree-like populations, widely
distributed in the dense or sparse caatinga forests of the Nordeste, and more shrubby forms
in rupicolous populations near its southern limit, in northern Minas Gerais. To the south
of Pedra Azul, is the endemic P. magnificus, which has probably arisen by a process of
allopatric speciation following isolation from a population of common ancestry with P.
pachycladus, of which it is assumed to be the sister species.
conservation status. Least Concern [LC]; a very widespread species.
Two subspecies are recognized:
1. Ribs 5–12, high and broad; central spines long, well differentiated from radials; flower-
bearing areoles densely hairy (S of 10°S) . . . . . . . . . . . . . . . . . . . . . . . . . a. subsp. pachycladus
1. Ribs (10–)13–19, low and close together; central spines poorly differentiated from radials
equalling them or slightly longer; flower-bearing areoles scarcely hairy (N of 10°S)
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . b. subsp. pernambucoensis
12a. subsp. pachycladus

Treelike or shrubby, 1–5(–8) m, branched only at base or forming a well-defined trunk; branches
5.5–10.0(–11.0) cm diam., epidermis pale blue; ribs 5–12, 15–35 × 12–24 mm. Areoles 3–10 mm diam.,
2–12 mm apart, woolly. Spines 0.4–1.2 mm diam. at base; golden yellow to brownish or greyish, centrals
1–8, (5–)15–30(–45) mm, ascending to porrect, radials 8–18, (3–)5–15 mm, adpressed. Fertile part of
stem differentiated, with 1–4 subapical flower-bearing areoles on more than one rib, with white long
hairs and sometimes bristly spines to 30 mm. Flower-tube pale green to very dark brown, slightly to
strongly curved; inner perianth-segments white. Seeds 1.5–2.0 × 1–1.3 mm, testa-cells domed to flat,
with intercellular depressions and coarse, dense to sparse cuticular folds.
Central-southern caatinga / Northern campo rupestre element: on quartzitic outcrops and in scrub associated
with the campos rupestres, and locally co-dominant with other woody vegetation in caatinga, on stony
ground within and on either side of the Chapada Diamantina, central Bahia, south of 10°S, on limestone
outcrops in the Rio São Francisco valley (N Minas Gerais and W Bahia), in the northern part of the Serra
do Espinhaço, Minas Gerais, eastwards on gneissic inselbergs, and disjunctly in campo rupestre, northern
part of the Serra do Cabral, Minas Gerais, c. 400–1550 m. Endemic to the core area within Eastern Brazil.
Map 37.
bahia: W Bahia, Mun. Vanderlei, 3 Oct. 1985, Horst & Uebelmann 570, cult. (ZSS); l.c., 19 July 1989,
Zappi 148 (SPF); Mun. Santana, 28 km S of Santana, 15 Jan. 1991, Taylor et al. 1426 (K, HRCB, ZSS,
CEPEC); N to S Bahia, Mun. Jaguarari, Barrinha, 7–8 June 1915, Rose & Russell 19788 (US); l.c., border
with Mun. Juazeiro towards Senhor do Bonfim, between Maçaroca and Barrinha, 8 Jan. 1991, Taylor et
al. 1375 (K, HRCB, ZSS, CEPEC); l.c., 91 km S of Juazeiro on road BR 407 towards Senhor do
Bonfim, 15 July 1989, Zappi 128 (SPF, HRCB); Mun. Gentio do Ouro, Serra do Piquizeiro, 11°27'S,
42°25'W, 30 Nov. 1987, Miranda 47 (RB); l.c., Serra de Santo Inácio, Feb. 1907, reported by Ule (1908:
PILOSOCEREUS
photo II); Mun. João Dourado, APA Gruta dos Brejões, 4 Aug. 2002, Taylor (obs.); Mun. Jacobina, 28
km W of town on road BR 324 to Lages, 16 July 1989, Zappi 132 (SPF, HRCB); l.c., 27 km W of town,
12 Jan. 1991, Taylor et al. 1400 (K, HRCB, ZSS, CEPEC); Mun. Morro do Chapéu, NE of Gruta dos
Brejões, 4 Aug. 2002, Taylor (K, photos); l.c., N of Brejão (Formosa), ‘Lagedo Bordado’, 5 Aug. 2002,
Taylor (K, photo); l.c., 19–21 km W of town on road BA 052 to América Dourada, 25 Dec. 1988, Taylor
& Zappi in Harley 27391 (K, SPF, CEPEC), 13 Jan. 1991, Taylor et al. 1406 (K, HRCB, ZSS, CEPEC),
3 Aug. 2002, Taylor (K, photos); Mun. Seabra, E of town, 25 Aug. 1988, Eggli 1284 (ZSS, HRCB); Mun.
Lençois, Morro do Pai Inácio, Jan. & Mar. 1997, A. Conceição s.n. (K, photos); Mun. Mucugê, 7 km S
from the town on road towards Barra da Estiva, 21 Aug. 1988, Eggli 1248 (ZSS); l.c., 500 m from the
town, near cemetery, 22 Dec. 1988, Taylor & Zappi in Harley 27379 (K, SPF); Mun. Riacho de Santana,
55 km from Bom Jesus da Lapa on road BA 430 towards Riacho de Santana, 20 July 1989, Zappi 151
(SPF, HRCB); Mun. Rio de Contas, road to Faz. Brumadinho, 1 km from the Fazenda, 13°32'S,
41°54'W, 9 Nov. 1988, Taylor in Harley 25562 (K, SPF); Mun. Livramento do Brumado, 5 km SW of
Rio de Contas, 23 Aug. 1988, Eggli 1302 (ZSS, HRCB); l.c., 12 km on Livramento–Brumado road,
13°45'S, 41°49'W, 23 Nov. 1988, Taylor & Zappi in Harley 25546 (K, SPF); Mun. Riacho de
Santana/Igaporã, Serra da Garapa [‘Gacapa’], Ritter 1343A (U); Mun. Dom Basílio, 21 km S of
Livramento do Brumado, 12 Mar. 1991, Lewis s.n. (K, photo); Mun. Ituaçu, 1964, Ritter 1349 (U); l.c.,
2 km W of the town, 18 Aug. 1988, Eggli 1193 (ZSS); Mun. Brumado, 19 km from the town towards
Sussuarana, 2 Feb. 1991, Eggli (ZSS, photos); l.c., Umburanas, near the Rio Brumado, 1964, Ritter 1343
(U); Mun. Aracatu, 25 km from Brumado on road BR 030 towards Sussuarana, 2 Feb. 1991, Eggli (ZSS,
photo); Mun. Anagé, 1964, Ritter 1347 (U), 1985, Horst & Uebelmann 472 (ZSS); Mun. Palmas de Monte
Alto, 1964, Ritter 1225 (U); l.c., hill with TV tower, 28 Aug. 1988, Eggli 1314 (ZSS), 21 July 1989, Zappi
153 (SPF, HRCB); Mun. Malhada, 9 km on road BR 030 towards Iuiú, 28 Aug. 1988, Eggli 1318 (ZSS);
Mun. Caitité, 15 km W from the town, 1964, Ritter 1343B (U); l.c., 9 km SE from the town towards
Brumado, 27 Aug. 1988, Eggli 1310 (ZSS); l.c., Brejinho das Ametistas, Horst & Uebelmann 421 (ZSS),
26 July 1989, Zappi 177 (SPF, HRCB); Mun. Urandi, 3 km S of the town, 22 July 1989, Zappi 159 (SPF,
HRCB); Mun. Vitória da Conquista, Gameleira, 14°50'25"S, 41°0'23"W, 17 April 2003, Taylor et al.
(obs.); Mun. Licínio de Almeida, Sep. 2003, M. Machado (photos).
(photo); Mun. Espinosa, Horst & Uebelmann 402 (ZSS); l.c., 2 km N from the town on road BR 122 to
Urandi, 22 July 1989, Zappi 162 (SPF, HRCB); l.c., 8 km S from the town on road BR 122 to Monte
Azul, 22 July 1989, Zappi 162A (SPF, HRCB); Mun. Monte Azul, 1964, Ritter 1226 (U); l.c., 12 km E
from the town, E of Vila Angical, 28 Jan. 1991, Eggli (ZSS, photos), Taylor (K, photos); Mun. Januária,
1959, Ritter 958 (U); l.c., 8 km E of Rio São João, 5 Sep. 1985, Horst & Uebelmann 714 (ZSS); l.c., 15
km on road BR 135, S of the Rio São Francisco, Serra do Bom Sucesso, 11 Aug. 1988, Eggli 1141 (ZSS);
l.c., [Pedras de] Maria da Cruz, 11 Aug. 1997, I. Pimenta (K, photos); Mun. Mato Verde, 12 km E from
the town on road to Santo Antônio do Retiro, 13 Aug. 1988, Eggli 1156 (ZSS); l.c., 15°23'S, 42°45'W,
9 Nov. 1988, Taylor & Zappi in Harley 25520 (K, SPF), 22 July 1989, Zappi 164 (SPF, HRCB); Mun.
São Francisco, Faz. Daniela, 11 Aug. 1997, I. Pimenta (K, photo); Mun. Porteirinha, 28 km N from the
town towards Mato Verde on road BR 122, 27 Jan. 1991, Taylor et al. 1456 (K, HRCB, ZSS, BHCB);
Mun. Janaúba, Horst & Uebelmann 581 (ZSS); Mun. Taiobeiras, c. 35 km E from the town towards road
BR 116, after bridge over Rio Atoleiro, 15°53'S, 41°57'W, 17 Oct. 1988, Taylor & Zappi in Harley 25150
(K, SPF), 13 Aug. 1988, Eggli 1158 (ZSS, HRCB); Mun. Salinas (?), road to Pedra Azul, 1 km W from
‘Sítio’, 22 Feb. 1988, Supthut 8876 (ZSS); Mun. Pedra Azul, 15 Oct. 1984, Rizzini & Mattos-Filho 39
(RB); l.c., 10 km from the town towards road BR 116, 21 Oct. 1988, Taylor & Zappi in Harley 25401
(K, SPF); Mun. Francisco Sá/Grão Mogol, vicinity of Barrocão, 950 m, reported by Buining (1975: 26,
as Pseudopilocereus sp.); Mun. Francisco Dumont, E of the Serra do Cabral, 2 km from the town towards
Jequitaí, 7 Aug. 1988, Eggli 1118 (ZSS); N of Serra do Cabral (Francisco Dumont?), 9 Sep. 1985, Horst
& Uebelmann 147, cult. (ZSS).
Besides the examined specimens cited above, this taxon has also been observed in the following
municípios of Bahia: Caém, Jacobina, Ibitiara, Brejolândia, Igaporã/Caitité, Tanhaçu.
Concern, but the caatinga habitats continue to decline (campo rupestre habitats are less threatened).
PILOSOCEREUS
Further field study of this complex, geographically variable taxon may justify its division
into additional subspecies, especially for the distinctive regional forms from the Rio São
Francisco valley (BA/MG), Rio de Contas drainage (BA) and north-eastern Minas Gerais.
Marlon Machado has detected a population of typical subsp. pachycladus sympatric
with a similar but greener-stemmed plant, which appeared to be a different species,
perhaps Ritter’s P. splendidus, near Licínio de Almeida, Bahia.
12b. subsp. pernambucoensis (F. Ritter) Zappi in Succ. Pl. Res. 3: 109 (1994).
Holotype: Brazil, Pernambuco, Araripina, 1963, Ritter 1219 (U).
Pilosocereus pernambucoensis F. Ritter, Kakt. Südamer. 1: 65 (1979). Pseudopilocereus pernambucoensis (F. Ritter)
P. V. Heath in Calyx 4: 140 (1994) (‘pernambucensis’ ).
Pilosocereus pernambucoensis var. caesius F. Ritter, tom. cit.: 66 (1979). Pseudopilocereus pernambucoensis var.
caesius (F. Ritter) P. V. Heath, l.c. (1994). Holotype: Brazil, Pernambuco, Petrolina, Ritter 1220 (U).
[Cephalocereus piauhyensis sensu (Gürke) Britton & Rose, Cact. 2: 49 (1920), fig. 72 tantum, typ. excl.]
[Pilocereus piauhyensis sensu (Gürke) Werderm., Bras. Säulenkakt.: 111 (1933), fig. p. 107 tantum, typ. excl.]
[Pilosocereus piauhyensis subsp. piauhyensis sensu Braun & Esteves Pereira in Kaktusy 39 (special): 12, A
(2003).]
vernacular name. Calumbi.
Treelike, 2–10 m, with a well-defined trunk; branches 7–15 cm diam., epidermis grey-green to glaucous,
smooth; ribs (10–)13–19, 8–12 × 7–10 mm. Areoles 3–10 mm diam., 5–8 mm apart. Spines 0.3–0.8 mm
diam. at base, golden yellow, centrals 8–12, 12–18 mm, ascending to porrect, radials 16–18, 5–8 mm,
adpressed. Fertile part of stem slightly differentiated, with 1–3 subapical flower-bearing areoles on 1 or
more ribs, with few white long hairs and sometimes also 15–20 mm, bristly spines. Tube pale green to
brownish, curved; inner perianth-segments white to pink. Seeds (1.3–)1.4–1.6(–1.8) × (0.9–)1.0–1.1 mm,
testa-cells flat, with intercellular depressions, cuticular folds coarse, few or absent.
Northern caatinga element: locally co-dominant with other woody vegetation in dense or sparse caatinga
and agreste, on sandy or rocky substrates (including sandstones and gneissic inselbergs), c. 50–750 m,
northern Bahia (north of 10°S), Alagoas, Pernambuco, Paraíba, Rio Grande do Norte, south-easternmost
Piauí and southern and north-western Ceará (E, W & S sides of Chapada do Araripe, Chapada da
Borborema & plateau of the Serra da Ibiapaba). Endemic to North-eastern Brazil. Map 37.
piauí: SE Piauí, Mun. Jaicós, near the town, 7°13'S, 41°17'W, 6 Apr. 2000, E.A. Rocha et al. (K, photo);
Mun. Padre Marcos, 2 km along road to Simões from Marcolândia, 7°28'S, 40°41'W, 5 Apr. 2000, E.A.
Rocha et al. (K, photos); Mun. Simões, road to town from Marcolândia, W escarpment of Chapada do
Araripe, 7°36'S, 40°45'W, 5 Apr. 2000, E.A. Rocha et al. (K, photos); Mun. Paulistana, 54 km N and 24
km S of town on road BR 407, 6 Apr. 2000, E.A. Rocha et al. (obs.).
ceará: NW Ceará, Mun. Tianguá, road BR 222, 10.5 km W of access road to Tianguá, 20 Feb. 1995,
J.B. Fernandes da Silva 366 (MG), Taylor (K, photos); S Ceará, Mun. Jardim, SE escarpment of Chapada
do Araripe, 7°34'S, 39°19'W, 4 Apr. 2000, E.A. Rocha et al. (K, photos) — glaucous form with only
10–12 ribs.
rio grande do norte: ‘Sebastianópolis’ – Chapada do Apodi, 11 July 1960, Castellanos 22894 (R);
Mun. Mossoró, 11 July 1960, Castellanos 22877 (R); Mun. Tangará, W of town on road BR 226 to
Currais Novos, 6°14'S, 35°51'W, 1 Apr. 2000, E.A. Rocha et al. (obs.); Mun. São José do Campestre, 7
km S of town on road to Guarabira (PB), 1 Apr. 2000, E.A. Rocha et al. (obs.); Mun. Currais Novos, E
of town, c. 36°30'W, 1 Apr. 2000, E.A. Rocha et al. (obs.).
paraíba: Mun. Tacima, near border with Mun. Caiçara (near Logradouro), inselberg ‘Pão de Açúcar’,
6°36'S, 35°28'W, 1 Apr. 2000, E.A. Rocha et al. (K, photo); Mun. Guarabira, Serra da Jurema, 6°50'S,
35°29'W, 1 Apr. 2000, E.A. Rocha et al. (K, photos); Mun. Junco do Seridó, 22 Apr. 1978, Andrade-Lima
78-8628 (IPA); Mun. Soledade, see Eiten (1983: figs 67 & 68); Mun. Pocinhos, July 1920, Luetzelburg
PILOSOCEREUS
12594 (M); Mun. Cruz de Espírito Santo, road BR 230, near bridge over Rio Paraíba, 10 Feb. 1995,
Taylor & Zappi (obs.); Mun. São João do Cariri, road BR 412, 29–30 Oct. 1993, M.F. Agra et al. 2393
(JPB), 27–29 Apr. 1994, M.F. Agra et al. 3119, 3120, 3121 (JPB); Mun. Sumé, road BR 412, 29–30 June
1994, M.F. Agra et al. 2832 (JPB).
pernambuco: Mun. Exu, 11 km N of town, S escarpment of Chapada do Araripe, 7°25'S, 39°45'W, 5
Apr. 2000, E.A. Rocha et al. (K, photo); Mun. Araripina, 1963, Ritter 1219 (U); l.c., near the hotel
‘Pousada do Araripe’, 15 Feb. 1990, Zappi 225 (HRCB, ZSS); Mun. Taquaritinga do Norte, Gravatá do
Ibiapina, 8 Aug. 2001, Zappi (obs.); Mun. Bom Jardim, road PE 090, 8 Feb. 1993, F.A.R. Santos 62
(ALCB, HUEFS); Mun. Brejo da Madre de Deus, 9 Feb. 1993, F.A.R. Santos 66 (HUEFS, PEUFR);
Mun. Salgueiro, Morro do Cruzeiro, 17 Feb. 1990, Zappi 227A (HRCB); Mun. Gravatá, Russinha,
caatinga, 4 Feb. 1933 & 26 Apr. 1934, Pickel 1915 (F, on 2 sheets); Mun. Caruaru, 9 km NE of town,
8°14'21"S, 35°54'52''W, Dec. 1991, E.A. Rocha et al. (K, photos); Mun. Poção, 21 Oct. 1991, F.A.R.
Santos 3 (HUEFS); Mun. Pesqueira, 4 km SW from the town towards Garanhuns, 13 Feb. 1991, Taylor
& Zappi 1625 (K, ZSS, HRCB, UFP); Mun. Alagoinha, 13 Feb. 1992, F.A.R. Santos 23 (HUEFS,
PEUFR); Mun. Afrânio, 8 km S of border with PI on road BR 407, 6 Apr. 2000, E.A. Rocha et al. (obs.);
l.c., 5 km N of Rajada, 8°47'S, 40°52'W, 6 Apr. 2000, E.A. Rocha et al. (K, photos) — form with stems
15 cm diam.; Mun. Belém de São Francisco, road BR 316 near border with Mun. Floresta, 8°40'S,
38°44'W, 8 Apr. 2000, E.A. Rocha et al. (K, photos); Mun. Ibimirim, 17 Nov. 1992, F.A.R. Santos 40
(HUEFS, PEUFR); Mun. (?), Petrolina–Cabrobó, 26 Apr. 1971, Heringer et al. 366B (RB); Mun.
Nova Petrolândia, 12 km NE of crossing to old town, 12 Feb. 1991, Taylor & Zappi 1620 (K, ZSS,
HRCB, UFP); Mun. Petrolina, road BR 428, 9°11'S, 40°23'W, 8 Apr. 2000, E.A. Rocha et al. (obs.);
l.c., Ritter 1220 (U).
alagoas: Mun. Palmeira dos Índios, 6 km S of border with Pernambuco on road to Arapiraca, 13 Feb.
1991, Taylor & Zappi 1630 (K, ZSS, HRCB, MAC).
bahia: Mun. Casa Nova, 10 Feb. 1990, 50 km E of Remanso, Zappi 220 (HRCB, ZSS); Mun. Curaçá,
39 km NW of Uauá, towards Poço de Fora, 7 Jan. 1991, Taylor et al. 1368 (K, HRCB, ZSS, CEPEC);
Mun. Juazeiro, 2–6 June 1915, Rose & Russell 19744, 19758 (US, NY); Mun. Uauá, 5 km NW from the
town towards Poço de Fora, 7 Jan. 1991, Taylor et al. 1366A (K, HRCB, ZSS); l.c., 13 km SE from the
town towards Bendengó/Canudos, 5 Jan. 1991, Taylor et al. 1360 (K, HRCB, ZSS, CEPEC); Mun.
Sento Sé, near Cabeluda, 11 Jan. 1991, Taylor et al. 1394 (K, HRCB, ZSS, CEPEC).
Besides the examined specimens cited above, this taxon has also been observed in the following
municípios of Pernambuco: Lagoa Grande, Santa Maria da Boa Vista, Orocó, Floresta, Airi, Cruzeiro do
Nordeste, Arcoverde, Sertânia, Garanhuns, Caetés, Terezinha, Bom Conselho; and Bahia: Remanso.
Concern, but habitat continues to decline and has already declined very considerably.
This subspecies includes populations found from the northernmost curve of the Rio São
Francisco valley northwards (Bahia, N of 10°S), that comprise distinctly treelike,
sometimes massive plants, with usually high numbers of ribs (13–19) and fine, golden
spination, which were described by Ritter (1979) as Pilosocereus pernambucoensis. The
treatment of this taxon at subspecific level is justified by the existence of morphologically
intermediate populations in the region of Juazeiro and Sento Sé (Taylor 1375, 1394),
presenting 10–15 ribs and relatively fine spination. A 10-ribbed form has also been
observed in southern Ceará (Mun. Jardim).
P. pachycladus subsp. pernambucoensis may present either coerulescent blue (eg. Zappi
225) or greyish green epidermis (Taylor & Zappi 1630), the greenish plants predominating
in the eastern part of its range, glaucous forms being the norm to the west (eg. Plate 45.2).
There is no reason to maintain Ritter’s variety caesius (Ritter 1979), proposed on the basis
of its glaucous epidermis. However, stem thickness does differ and is somewhat correlated
with stem colour, the eastern populations, from central-eastern Pernambuco northwards
PILOSOCEREUS
to Rio Grande do Norte, being uniformly narrow-stemmed and greenish, suggesting that
these could be recognized as a further subspecies, for which there is currently no name at
any rank (Plate 45.3). This slender, greenish form obeys the distribution pattern of other
taxa characteristic of the Eastern caatingas-agrestes, eg. Tacinga palmadora.
The collection cited above for north-western Ceará seems rather disjunct on present
knowledge and underlines the need for field studies in the intervening areas of south-
western Ceará.
Britton & Rose and Werdermann confused populations of P. pachycladus subsp.
pernambucoensis with Cereus piauhyensis Gürke (1907), which they combined as Cephalocereus
piauhyensis (Gürke) Britton & Rose (1920) and Pilocereus piauhyensis (Gürke) Werdermann
(1933). The type material of Cereus piauhyensis Gürke was collected by Ule in the Serra
Branca, north of São Raimundo Nonato, Piauí. It is clear that this name belongs to another
species (see discussion under P. piauhyensis) and, to date, P. pachycladus has not been
encountered further west than Mun. Paulistana in the south-easternmost corner of Piauí.
This confused nomenclature also appears in Andrade-Lima (1989), where he follows the
concept of Britton & Rose (1920) and Werdermann (1933), and illustrates and describes P.
pachycladus subsp. pernambucoensis as P. piauhyensis (likewise Braun & Esteves Pereira 2003:
12, A). In the same work Andrade-Lima places the southern forms of P. pachycladus (ie.
subsp. pachycladus) under Pilosocereus glaucescens (Linke) Byles & G. D. Rowley, a name of
uncertain application (see notes at end of genus).
Egler (1951: 587, fig. 6) confuses this widespread northern subspecies with Facheiroa
squamosa, which in Pernambuco is restricted to the south-westernmost part of the state.
13. Pilosocereus magnificus (Buining & Brederoo) F. Ritter, Kakt. Südamer. 1: 72–73,
Abb. 1 (1979). Holotype: Brazil, Minas Gerais, Rio Jequitinhonha, 370 m, s.d. [1968],
Horst & Uebelmann 224 (U).
Pseudopilocereus magnificus Buining & Brederoo in Cact. Succ. J. (US) 44: 66–70 (1972).
[Pilosocereus glaucescens sensu Rizzini & A. Mattos, Contrib. Conhecim. Fl. NE. Minas Gerais Bahia Med.:
35, Figs 8 & 35 (1992).]
vernacular name. Facheiro.
Shrubby to treelike, 1.5–5.0 m, with or without a well-defined trunk; vascular cylinder weakly woody;
branches 4–6(–7.5) cm diam., erect, epidermis pale blue, waxy, smooth; ribs (4–)5–12(–15), 15 × 18 mm,
sinuses straight, transverse folds above the areoles only visible at apex of branches. Areoles 4 mm diam.,
2 mm apart, felt white or blackish, with white long hairs. Spines 0.2–0.4 mm diam. at base, translucent,
bristly, golden-yellow to brownish; centrals 8, to 15 mm, porrect; radials c. 16, to 10 mm, ascending.
Fertile part of stem slightly modified, flower-bearing areoles in groups of 3–6, randomly distributed along
the whole length of the branches, especially at the mid to basal part of the stems, bearing white long hairs.
Flower-buds obtuse 2–3 days before anthesis; flowers to 6 × 2.3 cm; pericarpel subglobose, pale green;
tube 35 × 10 mm at base, flared to 16 mm diam. at apex, constricted above and below the nectar-
chamber, straight to slightly curved, greenish with purplish patches, smooth, apical region with acute
bract-scales; style 41 × 1.3–2.0 mm, tapering; stigma-lobes c. 10, exserted in relation to the anthers; ovary
locule hemicircular to depressed in longitudinal section. Fruit 1.5–2.5 × 2.5–3.0 cm, depressed-globose,
dehiscent by a lateral, abaxial or adaxial slit, floral remnant erect to pendent, deeply inserted; pericarp
thick, wine-coloured to purple, smooth; funicular pulp magenta. Seeds 1.3–1.5 × 0.8–1.2 mm, testa-cells
domed, with intercellular depressions, cuticular folds lacking.
PILOSOCEREUS
South-eastern caatinga (inselberg) element: locally co-dominant with other arborescent cacti on gneissic
inselbergs in caatinga-agreste, c. 240–700 m, in the drainage of the Rio Jequitinhonha, north-eastern Minas
minas gerais: Mun. Itinga, Jenipapo, Faz. Lajeadão, 16°40'S, 41°50'W, June 2003, I. Andrade (K, photo);
l.c., 5 km E from the town on road BR 367 towards Itaobim, 16°36'S, 41°43'W, 19 Nov. 1988, Taylor
& Zappi in Harley 25533 (K, SPF); Mun. Itaobim, near the Rio Jequitinhonha, Horst & Uebelmann 224,
cult. (U, ZSS); l.c., 8 km W of town on road BR 367 towards Itinga, 14 Dec. 1990, Taylor & Zappi 763
(HRCB, K, ZSS, BHCB); ‘Rio Jequitinhonha’, s.d., Horst & Uebelmann 224 (U); Mun. Padre Paraíso,
12 km N of town, 13 Dec. 1990, recorded with P. subsimilis (Zappi 1994: 87); Mun. Caraí, on road
BR 116, c. 9.5 km S of Padre Paraíso [‘Água Vermelha’], 18 km N of Catuji, 13 Dec. 1990, Taylor &
Zappi 755 (HRCB, K, ZSS, BHCB).
Not known to be decreasing at present, but range limited (5 localities) and requiring monitoring for
habitat change.
Distinct within the genus for its approximate areoles bearing golden, bristly spines, which
contrast with its strikingly pale blue, wax-covered epidermis. P. magnificus can also be
distinguished by its small, narrow flowers, that appear randomly along the branches.
Rather variable in rib number between populations.
PILOSOCEREUS AURISETUS Group (Nos 14–16)

Stems branching only at or near base. Found on or amongst rocks in cerrado and campo rupestre.
14. Pilosocereus machrisii (E. Y. Dawson) Backeb., Die Cactaceae 4: 2419 (1960).
Holotype: Brazil, Goiás, E from Ceres, road S from Uruaçu, 3 km from the town, 26
May 1956, Dawson 15110 (R; RSA, iso.).
‘Pilocereus cuyabensis’ Backeb. in Blätt. Kakteenforsch. 1935(1): [98] (1935), nom. inval. (Art. 36.1).
‘Cephalocereus cuyabensis’ Borg, Cacti, ed. 2: 148 (1951), nom. inval. ‘Pilosocereus cuyabensis’ Byles &
G. D. Rowley in Cact. Succ. J. Gr. Brit. 19: 66 (1957), nom. inval. ‘Pseudopilocereus cuyabensis’ Buxb.
in Beitr. Biol. Pflanzen 44: 252 (1968), nom. inval.
Cephalocereus machrisii E. Y. Dawson in Los Angeles County Mus. Contr. Sci. 10: 1–8 (1957). Pseudopilocereus
machrisii (E. Y. Dawson) Buxb. in Beitr. Biol. Pflanzen 44: 252 (1968); Buining & Brederoo in
Succulenta 56: 90–94 (1977).
P. jauruensis Buining & Brederoo in Kakt. and. Sukk. 29: 153–155 (1978) (‘juaruensis’). Holotype: Brazil,
Mato Grosso do Sul, Rio Jauru [‘Juaru’], Riacho Claro, 1974, Horst & Uebelmann 454 (U). Pilosocereus
jauruensis (Buining & Brederoo) P. J. Braun in ibid. 35: 181 (1984) (‘juaruensis’).
P. saudadensis F. Ritter, Kakt. Südamer. 1: 82 (1979). Holotype: Brazil, Mato Grosso, Serra da Saudade,
1963, Ritter 1216 (U).
P. pusillibaccatus P. J. Braun & E. Esteves Pereira, Cact. Succ. J. (US) 58: 240–247 (1986). Holotype:
Brazil, Piauí, [Alto Parnaíba – Gilbués], southern border of Piauí, 500 m, Apr. 1982, E. Esteves Pereira
202 (ZSS; UFG, K, isos.).
P. cristalinensis P. J. Braun & E. Esteves Pereira in Kakt. and. Sukk. 38: 132 (1987). P. machrisii subsp.
cristalinensis (P. J. Braun & E. Esteves Pereira) P. J. Braun & E. Esteves Pereira in Schumannia 3: 188
(2002). Holotype: Brazil, Goiás, Mun. Cristalina, near the town, 1050 m, Mar. 1973, E. Esteves
Pereira 73 (ZSS; UFG, iso.).
P. lindanus P. J. Braun & E. Esteves Pereira in Kakt. and. Sukk. 38: 132 (1987) (‘lindaianus’). Holotype:
Brazil, Goiás, [Mun. Alto Paraíso de Goiás, Chapada dos Veadeiros], Oct. 1973, E. Esteves Pereira 60
(ZSS; UFG, iso.).
P. densivillosus P. J. Braun & E. Esteves Pereira in Kakt. and. Sukk. 45: 108–114 (1994). Holotype: Brazil,
PILOSOCEREUS
W Goiás, Rio Araguáia drainage, near border with Mato Grosso do Sul, 1989, E. Esteves Pereira 49
(UFG 12350, n.v.).
‘P. circinnuspetalus’ P. J. Braun & E. Esteves Pereira in Innes & Glass, The Illustrated Encyclopedia of Cacti:
246, photo. top left (1991), nom. inval. (Art. 36.1). Based on material from Goiás, Mun. Mineiros.
? P. estevesii P. J. Braun in Cact. Succ. J. (US) 71: 74–77, with illus. (1999). Holotype: Bahia, W Bahia,
‘Serra do Muquém’, 550–600 m, 1982, E. Esteves Pereira 142 (UFG 20529, n.v.).
P. goianus P. J. Braun & E. Esteves Pereira in Brit. Cact. Succ. J. 20: 93–103 (2002). Holotype: NE Goiás,
s.d., E. Esteves Pereira 89 (UFG 14873, n.v.). Synon. nov.
Shrubby, 0.4–3.2(–3.5) m, not normally branched above base; vascular cylinder weakly woody; branches
3.2–9.0 cm diam., erect, epidermis dark to bluish green, slightly greyish to glaucous, smooth; ribs 8–15,
7–15 × 13–20 mm, sinuses straight, transverse folds not visible; areoles 2–5 mm diam., 3–5 mm apart, felt
white to greyish with white to brownish long hairs to 5–10 mm. Spines 0.25–0.60 mm diam. at base,
translucent when young, reddish, pale brown or golden, centrals 3–8, 15–30(–40) mm, porrect or
ascending, the largest deflexed on old areoles, radials 9–24, 3–25 mm, adpressed, the lowermost longer.
Fertile part of stem strongly differentiated on 3 or more ribs of each branch, with yellowish to greyish
long hairs to 20–30 mm, without bristles or these to 47 mm. Flower-buds acute 2 days before anthesis;
flowers (measurements mainly based on Taylor et al. 1440 in cultivation) 3–4.5 × 3–4 cm; pericarpel
subglobose to ovoid, 10 × 14 mm; tube c. 25 × 12 mm at base, flared to 18 mm diam. at apex,
infundibuliform, straight, pinkish to brownish red, slightly striate, distal half with triangular, pinkish
bract-scales; style 32–40 × 3–1 mm, tapering, stigma-lobes 8–12, exserted or equal to anthers, to 9 mm;
ovary locule circular to compressed in longitudinal section. Fruit 1.5–2.5 × 2–3.5 cm, depressed-globose,
dehiscent by a lateral slit, floral remnant suberect to pendent, deeply inserted; pericarp exterior yellowish,
red or purplish, interior yellow, smooth; funicular pulp white or pinkish magenta. Seeds
(1.3–)1.6–1.7(–1.8) × 1–1.3 mm, testa-cells domed to flat, with intercellular depressions, cuticular folds
coarse, dense, sparse or absent.
Western cerrado element: on quartzitic, arenitic or limestone rock outcrops associated with cerrado, cerrado
de altitude or campo rupestre, c. 500–800 m, southern Piauí and western Bahia; southern Pará (Araguatins),
Goiás, Mato Grosso, Mato Grosso do Sul, western Minas Gerais (Serra da Canastra) and São Paulo
(Altinópolis & Brotas); north-eastern Paraguay. Map 38.
piauí: Alto Parnaíba–Gilbués, south-western border of Piauí, Apr. 1982, E. Esteves Pereira 202 (ZSS,
UFG, K).
bahia: W Bahia, Mun. Barreiras, Serra da Bandeira, hill with TV tower, 18 Jan. 1991, Taylor et al.
1436A (K, HRCB); l.c., road to the new airport, 2.5 km from junction of road to Brasília, 18 Jan. 1991,
Taylor et al. 1440 (K, HRCB, ZSS, CEPEC), ibid., cult. RBG Kew, July 2002 (K, photos).
conservation status. Least Concern [LC] (1); PD=2, EI=1, GD=2. Short-list score (1×5) = 5. Some
populations may be affected by loss of habitat, but the species is widespread on rock outcrops.
The recently described P. estevesii P. J. Braun is tentatively referred here. Its supposed
distinctive characteristics seem no more unusual than those of other synonyms listed above
and, given the extensive range of P. machrisii and its pronounced variability throughout this
range, it would be unwise to assume that this western Bahian plant (now said to be extinct
at its only known locality) is worthy of specific status. As Braun, l.c., comments, it is
somewhat reminiscent of the geographically proximal P. flexibilispinus P. J. Braun & E.
Esteves Pereira, from the adjacent state of Tocantins, but the flowers of the latter confirm
its proposed relationship with the P. PENTAEDROPHORUS Group (Zappi 1994), whereas
those of P. estevesii offer no differences with those of P. machrisii and its allies. P. estevesii has
fruit with magenta funicular pulp, whereas previously known populations of P. machrisii are
reported to have white fruit pulp. However, variation in this character is common in the
genus and is particularly so in the closely related P. aurisetus.
PILOSOCEREUS
Other plants that may ‘key out’ to P. machrisii are the dwarf P. parvus (L. Diers & E.
Esteves Pereira) P. J. Braun and the recently described P. bohlei Hofacker, both of which
may well prove to be a ‘good’ species. These are discussed under ‘Insufficiently known
taxa’, after P. densiareolatus. Unfortunately, the authors have not been able to study them
in the field.
15. Pilosocereus aurisetus (Werderm.) Byles & G. D. Rowley in Cact. Succ. J. Gr. Brit.
19: 66 (1957). Type: Brazil, Minas Gerais, Serra do Cipó, Werdermann 3993 (B†). Lectotype
(Zappi 1994): Werdermann, l.c., 104, photograph ‘Pilocer. aurisetus Werd.’ (iconotype).
Pilocereus aurisetus Werderm., Bras. Säulenkakt.: 103 (1933). Pseudopilocereus aurisetus (Werderm.) Buxb. in
Beitr. Biol. Pflanzen 44: 252 (1968).
P. werdermannianus Buining & Brederoo in Kakt. and. Sukk. 26: 74–77 (1975). Pilosocereus werdermannianus
(Buining & Brederoo) F. Ritter, Kakt. Südamer. 1: 76–77, Abb. 46–49 (1979). P. aurisetus subsp.
werdermannianus (Buining & Brederoo) P. J. Braun & E. Esteves Pereira in Succulenta 74: 134 (1995).
Holotype: Brazil, Minas Gerais, Mun. Conceição do Mato Dentro, 660 m, s.d., Horst & Uebelmann
227 (U).
Pilosocereus saxatilis F. Ritter, Kakt. Südamer. 1: 77 (1979), nom. inval. (Art. 34(c)). Based on: Brazil,
Minas Gerais, Conceição [do Mato Dentro], 1964, Ritter 960A (U).
Pilosocereus werdermannianus (Buining & Brederoo) F. Ritter var. diamantinensis F. Ritter, l.c. (1979).
Pseudopilocereus werdermannianus var. diamantinensis (F. Ritter) P. V. Heath in Calyx 4: 141 (1994).
Holotype: Brazil, Minas Gerais, Diamantina, 1959, Ritter 959 (U).
Pilosocereus werdermannianus (Buining & Brederoo) F. Ritter var. densilanatus F. Ritter, l.c. (1979). P. saxatilis
var. densilanatus F. Ritter, l.c. (1979), nom. inval. Pseudopilocereus werdermannianus var. densilanatus (F.
Ritter) P. V. Heath, l.c. (1994). Pilosocereus aurisetus subsp. densilanatus (F. Ritter) P. J. Braun & E.
Esteves Pereira in Succulenta 74: 134 (1995). Holotype: Brazil, Minas Gerais, Penha de França, 1959,
Ritter 960 (U).
Pilosocereus supthutianus P. J. Braun in Kakt. and. Sukk. 36: 100–103 (1985). P. aurisetus subsp. supthutianus
(P. J. Braun) P. J. Braun & E. Esteves Pereira in Succulenta 74: 134 (1995). Holotype: Brazil, Minas
Gerais, W of the Serra do Espinhaço [Mun. Engenheiro Dolabela, towards Sítio], 800 m, 1982, Horst
& Uebelmann 547 (ZSS).
[P. coerulescens sensu (Lem.) F. Ritter, Kakt. Südamer. 1:75 (1979); Zappi in Bol. Bot. Univ. São Paulo 12:
51 (1991); non Pilocereus coerulescens Lem. (1862), nom. illeg. (Art. 52.1).]
Shrubby, 1–3 m, branching only at base; vascular cylinder weakly woody; branches erect; ribs 10–17,
sinuses straight, transverse folds not visible. Areoles 2–4 mm diam., 4–6 mm apart, felt brownish to
blackish, slightly hairy to tomentose, with white long hairs to 5–10 mm. Spines 17–25, translucent, white
to golden-yellow. Fertile part of stem strongly differentiated, lateral or subapical, flower-bearing areoles
with long hairs and golden bristles. Flower-buds acute 2–3 days before anthesis; flowers 3.2–5.0 × 3–4
cm; pericarpel subglobose, greenish to brownish; tube 25–30 × 10–12 mm at base, flared to 20–28 mm
diam. at apex, infundibuliform, straight, pinkish to brownish red, slightly striate, distal half with pinkish
bract-scales; style 22–25 × 0.8–1.5 mm, tapering, stigma exserted in relation to the anthers; ovary locule
hemicircular to compressed in longitudinal section. Fruit depressed-globose, dehiscent by subapical slit
beneath the floral remnant, which is pendent, deeply inserted; pericarp thick, folded and rugose; funicular
pulp white, pink or magenta. Seeds with flat testa-cells, intercellular depressions visible to inconspicuous,
cuticular folds lacking.
conservation status. Least Concern [LC]; widespread and known from many localities that are not
known to be significantly declining.
One of the specific features of P. aurisetus is found in its fruit, which characteristically splits
across the apex, breaking the floral remnant. It also has very smooth seeds.
PILOSOCEREUS
Two subspecies are distinguished as follows:
1. Plants to 2 m; branches 2.8–5.5 cm diam.; ribs 11–13; flower-bearing areoles with white hairs
(Serra do Espinhaço) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15a. subsp. aurisetus
1. Plants to 3 m; branches 4.5–7.0 cm diam., ribs 10–17; flower-bearing areoles with golden hairs
(Serra do Cabral) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15b. subsp. aurilanatus
15a. subsp. aurisetus

vernacular names. Rabo-de-raposa, Quiabo-da-lapa, Quiabo-do-inferno, Cabeça-de-velho.
Shrub, 1–2 m; branches 2.8–5.5 cm diam., epidermis glaucous or green; ribs 11–13, 5–6 × 9–11 mm.
Areoles 2–4 mm diam., 4–6 mm apart, brown to blackish, slightly to densely hairy, hairs white. Spines
0.1–0.3(–0.35) mm diam. at base, whitish to golden yellow, centrals 5–7, 8–25(–30) mm, ascending to
porrect, radials 12–16, 7–11 mm, adpressed. Fertile part of stem lateral to subapical, flower-bearing
areoles with white long hairs to 20 mm and golden bristles to 30 mm. Stigma-lobes c. 10, 5–7 mm. Fruit
2–3 × 3.5–4.0 cm; pericarp brownish to wine-coloured; funicular pulp white, pink or magenta. Seeds
1.7–1.8 × 1.2–1.3 mm.
Widespread South-eastern campo rupestre element: quartzitic rock outcrops associated with campo rupestre,
c. 650–1300 m, Serra do Espinhaço, central Minas Gerais. Endemic to the core area within Minas Gerais.
Map 38.
minas gerais: Mun. Bocaiúva, 40 km NE of Engenheiro Dolabela, W of the Serra do Espinhaço,
Fazenda Tabual, 8 Aug. 1988, Eggli 1126 (ZSS, HRCB); l.c., Engenheiro Dolabela, towards Sítio,
1982 & 7 Sep. 1985, Horst & Uebelmann 547 (ZSS); l.c., Fazenda Laginha, near Sítio, s.d., Horst &
Uebelmann 722 (ZSS); Mun. Itamarandiba, 23 km SW of town towards Penha de França, Serra do
Mato Virgem (Serra Negra), 4 Aug. 1988, Eggli 1077 (ZSS, HRCB); Mun. Rio Vermelho, Pedra
Menina, Serra do Ambrósio, s.d., Horst & Uebelmann 109, cult. (ZSS); l.c., Faz. Sr J. Batista, 8 Mar.
1988, Zappi & Prado in CFCR 11825 (SPF); Mun. Diamantina, 1959, Ritter 959 (U); l.c., Mendanha,
18°6'S, 43°28'W, 31 Oct. 1988, Taylor & Zappi in Harley 25486 (K, SPF); l.c., road to Conselheiro
Mata, 25 Feb. 1987, Zappi et al. in CFCR 10395 (SPF); Mun. Datas, 5.5 km N of Datas on road to
Diamantina, 17 Feb. 1988, Supthut 8834 (ZSS); Mun. Gouveia, 14 Apr. 1987, Zappi et al. in CFCR
10945 (SPF); Mun. Presidente Kubitschek, Trinta Réis, 19 July 1987, Zappi et al. in CFCR 10666
(SPF); Mun. Serro, 10 Aug. 1972, Hatschbach 30133 (MBM); Mun. Conceição do Mato Dentro, 1964,
Ritter 960A (U), Horst & Uebelmann 227 (U, ZSS); l.c., road to Morro do Pilar, bridge over the Rio
Santo Antônio, 20 Nov. 1989, Zappi 195 (HRCB); Mun. Santana do Riacho, Serra do Cipó, 10 Dec.
1949, A.P. Duarte 2397 (RB), 22 July 1987, Zappi et al. in CFSC 10266 (SPF), l.c., 19°17'S, 43°34'W,
27 Oct. 1988, Taylor & Zappi in Harley 25410 (K, SPF); Mun. Cocais, W of town on road to Pedra
Pintada, 4 Aug. 2001, J.O. Rego et al. 851 (BHCB, Herb. Fund. Zoo-Bot. de Belo Horizonte 650 &
fls in alcohol [Carpoteca 762], K, photos).
conservation status. Least Concern [LC] (1); PD=2, EI=1, GD=2. Short-list score (1×5) = 5. Some
populations may be affected by excessive burning of their habitats, but threats are otherwise quite limited.
The easternmost populations from Itamarandiba and Rio Vermelho (subsp. densilanatus (F.
Ritter) P. J. Braun & E. Esteves Pereira) have densely woolly stems and at first seem rather
distinctive, but on close examination do not differ sufficiently to merit recognition as an
additional subspecies. If it were to be recognized, the logical extension of such a move
would be to give similar status to the northern- and southernmost populations cited
above, which seem to have equally divergent characteristics in terms of size, white-woolly
stems, glaucousness etc., and then other variants would also come into view.
PILOSOCEREUS
15b. subsp. aurilanatus (Ritter) Zappi in Succ. Pl. Res. 3: 123 (1994). Holotype:
Brazil, Minas Gerais, Joaquim Felício, 1964, Ritter 1325 (U).
Pilosocereus aurilanatus F. Ritter, Kakt. Südamer. 1: 77–78, Abb. 50 (1979). Pseudopilocereus aurilanatus (F.
Ritter) P. V. Heath in Calyx 4: 140 (1994).
Shrub, 1.5–3.0 m; branches 4–6 cm diam., epidermis olive-green, greyish, light blue waxy only at apex
of branches; ribs 10–17, 4.5–7.0 × 9–10 mm. Areoles 3–6 mm diam., 4–6 mm apart, felt brownish, with
white or yellowish long hairs to c. 5 mm, glabrescent when old. Spines 0.3–0.4 mm diam. at base, golden
yellow, centrals 8–9, 10–25 mm, ascending to porrect, radials 14–16, adpressed, somewhat pectinate.
Fertile part of stem lateral, with yellow to ferrugineous long hairs to 30 mm and golden bristles to 50
mm. Stigma-lobes c. 12, 5–12 mm. Fruit 3.5 × 3–4.3 cm; pericarp olive-green to wine-coloured;
funicular pulp magenta. Seeds 1.8–2.0 × (1.3–)1.4–1.5 mm.
South-eastern campo rupestre (Serra do Cabral) element: locally co-dominant with other cacti on quartzitic
rock outcrops, campo rupestre, 800–900 m, Serra do Cabral, Minas Gerais. Endemic to the core area within
Minas Gerais. Map 38.
minas gerais: Mun. Joaquim Felício, 1964, Ritter 1325 (U); l.c., road to Várzea da Palma, 17 May
1990, Zappi et al. in CFCR 13234 (SPF, HRCB); Mun. Buenópolis, 17°53'S, 44°15'W, 12 Oct. 1988,
Taylor & Zappi in Harley 24909 & 24991 (K, SPF).
conservation status. Endangered [EN B1ab(iii) + 2ab(iii)] (3); extent of occurrence < 5000 km2 and
area of occupancy estimated to be < 500 km2; PD=1, EI=2, GD=1. Short-list score (3×4) = 12. Known
only from a small region without protected areas, the surrounding habitats declining from charcoal
production activities.
Although it has a very restricted distribution and somewhat different habit, this taxon
described by Ritter (1979) from the Serra do Cabral, a disjunct mountain range west of
the Serra do Espinhaço, is not considered worthy of more than subspecific rank (Zappi
1994). It is linked to subsp. aurisetus via the form of the latter described under the
synonym, P. supthutianus. The most striking feature of subsp. aurilanatus is its stouter, taller
stems with dense, golden hairs on the flower-bearing areoles.
16. Pilosocereus aureispinus (Buining & Brederoo) F. Ritter, Kakt. Südamer. 1: 83–84
(1979). Holotype: Brazil, Bahia, E of the Rio São Francisco [Ibotirama], Serra da
Barriguda, 450 m, s.d., Horst & Uebelmann 391 (U).
Coleocephalocereus aureispinus Buining & Brederoo in Kakt. and. Sukk. 25: 73–75 (1974).
Shrubby, to 2 m, branched only at base; vascular cylinder weakly woody; branches 5–9 cm diam., erect,
epidermis dark green, smooth; ribs (18–)20–24, 3–5 × 5–7 mm, sinuses straight, transverse folds not
visible. Areoles 5 mm diam., 4–5 mm apart, felt white, with white long hairs to 10 mm, showing
indeterminate growth near base of plant. Spines 0.3–0.4 mm diam. at base, translucent, golden to
ferrugineous, sometimes bristly, centrals 8–16, 4–12(–18) mm, ascending to porrect, radials 14–16,
4–7(–12) mm, adpressed. Fertile part of stem weakly differentiated, flower-bearing areoles either lateral
or forming rings around the branches, with sparse white long hairs and golden bristles to 50 mm. Flower-
buds acute 5 days before anthesis; flowers 5 × 2–2.2 cm; pericarpel hemiglobose; tube 35 × 13 mm at
base, flared to 15 mm diam. at apex, straight, terete, narrow, pink to brownish, smooth; perianth-
segments pink to white; style 35 mm; stigma-lobes c. 10, exserted in relation to the anthers; ovary locule
hemicircular in longitudinal section. Fruit 20–26 × 20–28 mm, globose to depressed-globose, dehiscent
PILOSOCEREUS
by a lateral slit, floral remnant erect to pendent, deeply inserted; pericarp reddish to deep purple, smooth,
shiny, with folds around base of floral remnant; funicular pulp white. Seeds 1.4–1.5 × 1–1.2 mm, hilum-
micropylar region narrow, forming an angle of 60° with long axis of seed, testa-cells conic, periclinal walls
much elongated, with intercellular depressions and coarse, dense cuticular folds.
Western cerrado element: amongst arenitic rocks in cerrado, c. 450–550 m, central Bahia, east bank of the
Rio São Francisco. Endemic to Bahia. Map 38.
bahia: Mun. Ibotirama, 14 km E from the town on road BR 242, 18 July 1989, Zappi 147 (SPF,
HRCB); l.c., 13 km E from the town, 14 Jan. 1991, Taylor et al. 1422 (K, HRCB, ZSS, CEPEC); l.c.
(?), 12°13'18"S, 43°8'48''W, 2000, F.M. Sene J78 ‘sp. A’ (K, photos); E of Rio São Francisco [Ibotirama?],
‘Serra da Barriguda’, s.d., Horst & Uebelmann 391 (U); Mun. Oliveira dos Brejinhos, 30 km from
Ibotirama, ‘Rio São João’, s.d., Horst & Uebelmann 391 bis (ZSS).
conservation status. Data Deficient [DD]. Range poorly understood, but assumed to be more
extensive than the currently known habitats; if not, this taxon may merit a conservation category of
Endangered in view of its small population size.
Characteristic of this species is its small, narrow, dark-brown to pinkish flower exterior
and unusual seeds, with remarkably conic testa-cells and narrow hilum-micropylar
region. The only other species of this group whose seeds show any similarity to those of
P. aureispinus is Pilosocereus vilaboensis from Goiás. Furthermore, the seeds of a probable
synonym of the latter, P. rizzoanus P. J. Braun & E. Esteves-Pereira (1992), seem to
present intermediate characteristics. The population described as P. rizzoanus is also
geographically intermediate, occurring half way between those of P. vilaboensis and P.
aureispinus.
The peculiar testa of the seeds of P. aureispinus may be related to dispersal by ants, that
are especially abundant in the cerrado where this plant occurs, the conic testa-cells perhaps
representing an adaptation related to the transport of the seed. Indeed, some of the plants
were actually seen growing on top of anthills.
Known only from east of the Rio São Francisco, near Ibotirama, this species inhabits
arenitic rock outcrops in a phase of the cerrado and is sympatric with Facheiroa squamosa
(Gürke) P. J. Braun & E. Esteves Pereira. However, it may be expected to occur
elsewhere to the north and south in this little-botanized region.
PILOSOCEREUS PIAUHYENSIS Group (Nos 17–20)

Stems branched at base and above. Found on or amongst rocks in caatinga.
17. Pilosocereus multicostatus F. Ritter, Kakt. Südamer. 1: 79–80, Abb. 52 (1979).

Holotype: Brazil, Minas Gerais, Mun. Itaobim, 1965, Ritter 1346 (U).
Pseudopilocereus multicostatus (F. Ritter) P. V. Heath in Calyx 4: 140 (1994).
Shrubby, 1.5–5.0 m, branched at base or above, with or without a well-defined trunk; vascular cylinder
weakly woody; branches 3.8–7.5 cm diam., erect, epidermis dark green, shiny, smooth; ribs
(15–)18–25(–26), 3–6 × 6–9 mm, sinuses straight, transverse folds not visible. Areoles 2–3 mm diam., 4–7
mm apart, felt white to greyish, without long hairs. Spines 0.15–0.40 mm diam. at base, translucent,
golden yellow to brownish, centrals 3–7, 10–20 mm, ascending to porrect, radials 15–18, 5–10 mm,
adpressed. Fertile part of stem not to slightly differentiated, flower-bearing areoles randomly distributed
PILOSOCEREUS
along the branches, mostly subapical, with long, flexible, golden, to 40 mm bristles, and sparse white to
grey, long hairs. Flower-buds acute 1–3 days before anthesis; flowers 4.7 × 2.9–3.0 cm; pericarpel nearly
globose; tube 25 × 8.5 mm at base, flared to 18 mm diam. at apex, straight, terete, pale green, smooth,
apex with triangular bract-scales; style 24–26 × 1–1.5 mm, tapering; stigma-lobes c. 10, not extending
beyond the anthers; ovary locule circular to hemicircular in longitudinal section. Fruit 2.2–2.5 × 2.5–3.0
cm, depressed-globose, dehiscent by an adaxial or abaxial slit, floral remnant pendent, deeply inserted;
pericarp bright green tinged purplish, opaque, smooth; funicular pulp magenta. Seeds 1.5 × 1.2 mm,
testa-cells domed, with intercellular depressions and coarse, sparse cuticular folds.
South-eastern caatinga (inselberg) element: on gneissic inselbergs amongst caatinga-agreste in the drainage
of the Rio Jequitinhonha, c. 670–900 m, north-eastern Minas Gerais. Endemic to the core area within
minas gerais: Mun. Águas Vermelhas, 27.5 km E of Curral de Dentro towards road BR 116, 31 Jan.
1991, Taylor et al. 1517 (K, HRCB, ZSS, BHCB); Mun. Salinas, near Baixa Grande, road BR 251,
42°3'W, 31 May 2002, M. Machado (photos); Mun. Pedra Azul, 10 km SE from the town, 16°8'S,
41°12'W, 19 Oct. 1988, Taylor & Zappi in Harley 25188 (K, SPF); l.c., 10 km W from the town towards
road BR 116, 1 Feb. 1991, Taylor et al. 1519A (K, HRCB, ZSS, BHCB); Mun. Itaobim, 1965, Ritter
1346 (U).
conservation status. Near Threatened [NT] (1); extent of occurrence = < 5000 km2; PD=2, EI=1,
GD=1. Short-list score (1×4) = 4. Needs to be monitored in view of its restricted distribution and the
destruction of natural vegetation surrounding its rocky habitat.
Pilosocereus multicostatus is characterized by the high number of ribs and golden, flexible
spines, together with the slender and delicate flowers. Besides the following obviously
allied species, in its the bristly flower-bearing areoles, habit and ecology on inselbergs
in caatinga, P. multicostatus recalls P. chrysostele, which inhabits similar rock outcrops in
Pernambuco, Paraíba, Rio Grande do Norte and Ceará.
Marlon Machado has recently (June, 2002) photographed what is almost certainly a
hybrid between P. multicostatus and P. floccosus subsp. quadricostatus in the surroundings of
Pedra Azul, MG (Plate 40.2).
18. Pilosocereus piauhyensis (Gürke) Byles & G. D. Rowley in Cact. Succ. J. Gr.
Brit. 19(3): 67 (1957). Holotype: Brazil, Piauí, ‘Serra Branca’ [Mun. São Raimundo
Nonato], Jan. 1907, Ule 9 (B; K, SPF [photos ex B]).
Cereus piauhyensis Gürke in Monatsschr. Kakt.-Kunde. 18: 84–85 (1908). Cephalocereus piauhyensis
(Gürke) Britton & Rose, Cact. 2: 49 (1920), tantum quoad typ., fig. excl. Pilocereus piauhyensis
(Gürke) Werderm., Bras. Säulenkakt.: 111 (1933), tantum quoad typ. Pseudopilocereus piauhyensis
(Gürke) Buxb. in Beitr. Biol. Pflanzen 44: 253 (1968).
P. mucosiflorus Buining & Brederoo in Kakt. and. Sukk. 28: 201–203 (1977). Pilosocereus mucosiflorus
(Buining & Brederoo) F. Ritter, Kakt. Südamer. 1: 84 (1979). P. piauhyensis subsp. mucosiflorus
(Buining & Brederoo) P. J. Braun & E. Esteves Pereira in Succulenta 74: 134 (1995). Holotype:
Brazil, Piauí, São Raimundo Nonato, Serra da Capivara, 430–500 m, s.d., Horst 443 (U).
P. gaturianensis F. Ritter, Kakt. Südamer. 1: 81–82 (1979). Pseudopilocereus gaturianensis (F. Ritter) P. V.
Heath in Calyx 4: 140 (1994). Pilosocereus piauhyensis subsp. gaturianensis (F. Ritter) P. J. Braun & E.
Esteves Pereira in Succulenta 74: 134 (1995). Holotype: Brazil, Piauí, Gaturiano, 1963, Ritter 1218
(U).
? P. chrysostele subsp. cearensis P. J. Braun & E. Esteves Pereira in British Cact. Succ. J. 17: 21–27 (1999).
Holotype: Brazil, Cent./S Ceará, [apparently some 150 km or more west of the eastern border of the
state], 6 July 1981, E. Esteves Pereira 163 (UFG 13005, n.v.).
PILOSOCEREUS
vernacular names. Facheiro, Rabo-de-raposa.
Shrubby, 1.5–2.5 m, branched only at base or also above; central cylinder weakly woody; branches 5–7.5
cm diam., erect, epidermis dark green, shiny, smooth; ribs 14–24, 5 × 7 mm, sinuses straight, transverse
folds not visible. Areoles 2.5–5.0 mm diam., 5–7 mm apart, felt white to greyish, without long hairs,
presenting indeterminate growth at base of stem. Spines 0.2–0.3 mm diam. at base, translucent, golden
at first, later brownish to grey, centrals 5–9, 5–15 mm, ascending to porrect, radials 11–16, 3–8 mm,
adpressed. Fertile part of stem not or only slightly differentiated, mainly on subapical areoles, with
flexible, golden bristles and some white to grey long hairs. Flower-buds obtuse 1–3 days before anthesis;
flowers 5.2–7.5 × 2.7–4.0 cm; pericarpel hemiglobose; tube 40–45 × 11–16 mm at base, flared to 25–29
mm diam. at apex, curved, terete, pale green, glaucous, smooth, with apical, rounded, impressed bract-
scales; style 46–63 × 1.5–3.0 mm, tapering; stigma-lobes 8–11, exserted c. 1 cm in relation to the anthers;
ovary locule hemicircular in longitudinal section. Fruit 2.1 × 3.8 cm, depressed-globose, dehiscent by a
lateral slit, floral remnant pendent, deeply inserted; pericarp 4–6 mm thick, wine coloured, smooth;
funicular pulp magenta. Seeds (1.3–)1.4–1.8 × (0.9–)1.0–1.3 mm, testa-cells flat, with intercellular
depressions and coarse, sparse to absent cuticular folds.
Northern caatinga element: on arenitic/granitic rock outcrops associated with caatinga, c. 200–850 m,
south-eastern Piauí, Ceará and Rio Grande do Norte. Endemic to North-eastern Brazil. Map 39.
piauí: Mun. Dom Expedito Lopez, Gaturiano, 1963, Ritter 1218 (U), Horst & Uebelmann 520, cult.
(ZSS), 1983, Braun 465 (ZSS T-05418); l.c., Km 276 of the Petrolina–Teresina road, 5 km N of town,
1 Feb. 1990, Zappi 211 (HRCB, ZSS); Mun. São Raimundo Nonato, ‘Serra Branca’, Jan. 1907, Ule 09
(K, SPF photos ex B); l.c., Parque Nacional da Serra da Capivara, 6 Feb. 1990, Zappi 217 (HRCB), s.d.,
Horst 443, cult. (U, ZSS).
ceará: N Ceará, Serra de Meruóca, between Coreaú and Alcântaras, 20 Feb. 1995, Taylor (K, photos);
Cent. & S Ceará, unlocalized, see synonym cited above.
rio grande do norte: Mun. Açu (Assu), road to Arcoverde, Jan. 1978, Horst & Uebelmann 479, cult.
ZSS (ZSS).
Currently known from only c. 6 populations, which are not known to be declining and most plants are
± inaccessible on steep rocky slopes. The southernmost population is protected within a national park
(see above).
Described by Gürke (1908) as Cereus piauhyensis, on the basis of material collected by Ule
(Ule 09) at the Serra Branca, Piauí, its specific name was long misapplied to P. pachycladus,
first by Britton & Rose (1920), who published Cephalocereus piauhyensis (Gürke) Britton
& Rose, and later by Werdermann (1933), as Pilocereus piauhyensis (Gürke) Werderm.
Both authors had not seen true material of P. piauhyensis and confused it with what is now
known as P. pachycladus F. Ritter subsp. pernambucoensis (F. Ritter) Zappi, a very
widespread taxon from North-eastern Brazil. The same mistake is made in the illustrated
work of Andrade-Lima (1989).
The recently described, but poorly localized, P. chrysostele subsp. cearensis, is probably a
northern smaller-flowered form of this species rather than that to which it is referred by
its authors. The critical characters, which place it within P. piauhyensis rather than P.
chrysostele, are the lack of a cephalium, the slender flower-tube and the morphology of the
dehiscent fruit.
PILOSOCEREUS
19. Pilosocereus chrysostele (Vaupel) Byles & G. D. Rowley in Cact. Succ. J. Gr. Brit.
19: 66 (1957). Type: not extant. Neotype (Zappi 1994): Brazil, Ceará, ‘Serra do Cantim’,
June 1933, Luetzelburg 23755 (M; K, photo.; IPA, isoneo.).
Cereus chrysostele Vaupel in Zeitschr. Sukk.-Kunde 1: 58 (1923). Pilocereus chrysostele (Vaupel) Werderm.,
Bras. Säulenkakt.: 102–103 (1933). Cephalocereus chrysostele (Vaupel) Borg, Cacti: 104 (1937).
Pseudopilocereus chrysostele (Vaupel) Buxb. in Beitr. Biol. Pflanzen 44: 252 (1968).
vernacular names. Facheiro, Facheiro-de-serra, Rabo-de-raposa.
Shrubby, branching from the base upwards or only above and treelike with a trunk to 1.5 m or more,
1.5–6.0 m; vascular cylinder weakly woody; branches 4–7 cm diam., erect, epidermis pale green, smooth;
ribs 19–28, 5 × 5 mm, sinuses straight, transverse folds not visible. Areoles 2–3 mm diam., 4–5 mm apart,
felt white to blackish, with white long hairs to 8 mm, growth indeterminate near base of branches. Spines
0.15–0.20 mm diam. at base, translucent, pale yellow, centrals 6–8, 5–12 mm, ascending to porrect, radials
9–12, 4–10 mm, adpressed. Fertile part of stem weakly to strongly differentiated, the flower-bearing
areoles occupying the lateral and subapical region of the branches, with white hairs to 20 mm and usually
conspicuous golden, 40–60 mm bristles. Flower-buds obtuse 4 days before anthesis; flowers 4.5–6.0 × 3.7
cm; pericarpel hemiglobose; tube 33–35 × 17 mm at base, flared to 22 mm diam. at apex, constricted above
and below the nectar-chamber, straight, slightly compressed, brownish green to pinkish, distal third with
thick, broadly ovate bract-scales; style 27–28 × 1.8–2.0 mm, tapering, stigma-lobes 12–13, included in
relation to the anthers; ovary locule hemicircular to compressed in longitudinal section. Fruit 2–2.5 ×
3–3.5 cm, depressed-globose, dehiscent by a lateral or adaxial slit, floral remnant pendent, deeply inserted;
pericarp purplish, smooth; funicular pulp magenta. Seeds (1.2–)1.4–1.7 × (0.9–)1.0–1.2 mm, testa-cells flat,
with intercellular depressions or these sometimes nearly absent, cuticular folds not visible.
Northern caatinga element: on grey or whitish, granitic inselbergs with surface broken into blocks of
stone, associated with very dry, highland caatinga, 430–1190 m, central-northern Pernambuco, Paraíba,
Ceará and Rio Grande do Norte. Endemic to North-eastern Brazil. Map 39.
ceará: N Ceará, Mun. Irauçuba, 10 Jan. 1995, D. Macêdo s.n. (EAC, K, photos); W Ceará, Mun.
Tamboril, fide Uebelmann (1996): HU 939; E Ceará, Mun. Jaguaribe, Fazenda Mulungu, 11 July 1994,
D. Macêdo et al. s.n. (EAC, K, photos); unlocalized, ‘Serra do Cantim’, June 1933, Luetzelburg 23755 (M).
rio grande do norte: ‘Sebastianópolis’–Apodi, 11 July 1960, Castellanos 22633 (GUA); Mun. Angicos,
between Lajes and Fernando Pedrosa, reported by Braun & Esteves Pereira (1999a: 21); Mun. Acari,
beside road BR 427, S of Currais Novos, 6°23'S, 36°38'W, 1 Apr. 2000, E.A. Rocha et al. (K, photo).
paraíba: Mun. Brejo da Cruz, serra W of town, 6°22'S, 37°28'W, 2 Apr. 2000, E.A. Rocha et al. (K,
photos); Mun. Sousa, near border with Mun. Pombal, fide Uebelmann (1996 & mss): HU 460; Mun.
Pedra Lavrada, reported by Luetzelburg (1925–26, 3: fig. 29); Mun. Santa Luzia, 1998, E.A. Rocha 517
(JPB); Mun. Monte Horebe, 7°11'S, 38°33'W, 3 Apr. 2000, E.A. Rocha et al. (obs.); Mun. Maturéia, Serra
de Teixeira, Pico do Jabre, 1994, M.F. Agra et al. (JPB), 8 July 1994, L.P. Felix & Miranda 6583 (PEUFR);
Mun. Desterro, 1999, E.A. Rocha (JPB); Mun. Serra Branca, Coxixola, 9 km from Serra Branca, 28 Jan.
1971, Andrade-Lima et al. 1065 (IPA); Mun. Camalaú, road to Monteiro, 9 Feb. 1993, F.A.R. Santos 69
(ALCB, HUEFS, PEUFR); Mun. Monteiro, border with PE, Serra da Jabitaca, Luetzelburg, original type
locality cited in Vaupel (1923).
pernambuco: Mun. Triunfo, 8 Dec. 1991, F.A.R. Santos 16 (HUEFS); Mun. (?), Serra Talhada,
towards Carnaubeira, 22 May 1971, Heringer et al. 851 (IPA, RB, UB); Mun. Serra Talhada, 21 Nov.
1992, F.A.R. Santos 56 (HUEFS, PEUFR 13128, K, photo); l.c., rock outcrop N of the town, 13 Feb.
1990, Zappi 222 (HRCB, ZSS); Mun. Floresta, Carqueja, Serra de São Gonçalo, 23 May 1971, Heringer
et al. 858 (IPA); without precise locality, Serra da Borborema, Löfgren s.n. (US, NY, photos).
Concern, since often occuring on steep, dry rocky hillsides where destruction of vegetation is less likely.
Variable in stem thickness and rib number.

PILOSOCEREUS
20. Pilosocereus densiareolatus F. Ritter, Kakt. Südamer. 1: 73–74, Abb. 43 (1979).

Holotype: Brazil, Minas Gerais, Montes Claros, 1959, Ritter 957 (U).
Pseudopilocereus densiareolatus (F. Ritter) P. V. Heath in Calyx 4: 140 (1994).

P. densiareolatus var. brunneolanatus P. J. Braun & E. Esteves Pereira in Kakt. and. Sukk. 50: 287 (1999).
Holotype: Brazil, Bahia, limestone outcrops W of Cocos, Sep. 1984, E. Esteves Pereira 224 (UFG
12381, n.v., K, ZSS, isos. — labelled ‘P. superfloccosus var. brunneolanatus’).
‘Facheiroa Br. & R. Sp. nova’ Innes & Glass, The Illustrated Encyclopaedia of Cacti: 112 (1991).
[P. superfloccosus Buining & Brederoo in Cact. Succ. J. (US) 46: 60–63 (1974) pro parte quoad figs 1 & 2,
excl. typ. Pilosocereus superfloccosus auctt. pro parte, non (Buining & Brederoo) F. Ritter.]
vernacular names. Cabeça-de-velho, Facheiro-da-lapa.
Treelike, 2–5(–6.5) m, with a well-defined trunk; vascular cylinder weakly woody, pith wide; branches
3–7 cm diam., erect, epidermis olive-green to bluish, glaucous, smooth; ribs 13–22, 5–10 × 4–10 mm,
sinuses straight, transverse folds not visible. Areoles 2.5–5.0(–6.0) mm diam., 2–6 mm apart, felt white to
greyish with white long hairs to 20 mm at apex of branches, growth indeterminate near base of plant.
Spines to 30, 0.2–0.4 mm diam. at base, translucent, brown to golden, with pinkish or reddish shades,
greyish with age, centrals 6–15, 9–30(–35) mm, ascending to porrect, radials 14–16, 4–7 mm, adpressed.
Fertile part of stem slightly to strongly differentiated, flower-bearing areoles consecutive, occupying 1–4
ribs, sometimes somewhat immersed in the stem forming a lateral cephalium, with white, greyish or
brownish hairs to 3–8 cm and golden bristles. Flower-buds obtuse 5 days before anthesis; flowers 5–6 ×
c. 3–4 cm; pericarpel depressed-globose, olive-green; tube 3.8 × 10–17 mm at base, flared to 25 mm
diam. at apex, slightly curved, olive-green, smooth, distal half with obovate to acuminate bract-scales;
style 45 mm, tapering, stigma-lobes c. 12; ovary locule hemicircular to strongly depressed in longitudinal
section. Fruit c. 2–3 × 3 cm, depressed-globose, pinkish at base, brownish olive-green above, dehiscent
at apex and by lateral splits in the pericarp, floral remnant pendent, deeply inserted; funicular pulp white.
Seeds brown-black, shiny, (1.7–)1.8–2.5 × 1.3–1.9 × 1.4–1.6 mm, testa-cells flat, intercellular depressions
and cuticular folds absent.
Southern Rio São Francisco caatinga element: sometimes locally co-dominant with other arborescent
cacti and other woody plants on Bambuí limestone outcrops in caatinga, c. 450–800 m, central-northern
Minas Gerais and western Bahia. Endemic to the core area within Eastern Brazil. Map 39.
bahia: Mun. Santana, 8 km from Porto Novo, 1975, Horst 394 bis, cult. p.p., veg. (U, ZSS); l.c., 8 km
N of Porto Novo, 2 Oct. 1985, Horst & Uebelmann 394 ter, cult. (ZSS), 16 Jan. 1991, Taylor et al. 1432
(K, HRCB, ZSS, CEPEC); l.c., 28 km S from Santana towards Santa Maria da Vitória, 15 Jan. 1991,
Taylor et al. 1424 (K, HRCB, ZSS, CEPEC); Mun. Bom Jesus da Lapa, Morro da Lapa, 17 Apr. 1983,
Leuenberger et al. 3077A (CEPEC); l.c., W from the town, 1972, Horst & Uebelmann 394, pro parte (ZSS
T-06002); Mun. Cocos, limestone outcrops W from the town, Sep. 1984, E. Esteves Pereira 224 (ZSS T-
05836, UFG, K).
minas gerais: Mun. Manga, region of Jaíba, Mar. 1998, I. Pimenta (K, photo); Mun. Varzelândia, 2 km
E from the town, 23 July 1989, Zappi 168 (SPF, HRCB); l.c., 10 km N of Varzelândia on road to Jaíba,
12 Aug. 1988, Eggli 1149 (ZSS); Mun. Capitão Enéas, limestone outcrop visible from road BR 122, 9
June 2002, G. Charles (K, photo); Mun. Montes Claros, 13 km N from the town on road BR 135
towards Januária, 16°38'S, 43°55'W, 27 Jan. 1991, Taylor et al. 1454 (K, HRCB, ZSS, BHCB); 30 km
NW of Engenheiro Dolabela towards Sítio, 8 Aug. 1988, Eggli 1127 (ZSS); Mun. Bocaiúva, Granjas
Reunidas/Eng. Dolabela, 3 May 1936, A.P. Duarte 7756 (RB); l.c., access road to Engenheiro Dolabela,
17°28'S, 44°1'W, 4 Nov. 1988, Taylor & Zappi in Harley 25503 (K, SPF), 17 June 1990, Zappi 236
(HRCB); l.c., E of Engenheiro Dolabela, 7 Aug. 1988, Eggli 1109 (ZSS).
Concern, but habitat liable to decline through limestone quarrying, so monitoring is desirable.
Variable in stature and extent of cephalium wool developed.

PILOSOCEREUS
Probable hybrids with P. pachycladus subsp. pachycladus:

? P. occultiflorus P. J. Braun & E. Esteves Pereira in Cact. Succ. J. (US) 71: 310–315, figs 1–5, 7–9, tab. 1
& 2 (1999). Type: Minas Gerais, W of Rio São Francisco, ‘400 m’, E. Esteves Pereira 223 (UFG, holo.
n.v., BONN, ZSS, iso.).
Shrubby to treelike; branches to 8 cm diam., epidermis bluish green; ribs 8–11, 10 × 8 mm. Areoles c. 5
mm diam., 3 mm apart. Spines 0.3–0.7 mm diam. at base, centrals 10(–35) mm. Fertile part of stem with
brownish hairs to 8 cm.
bahia: Mun. Santana, 8 km from Porto Novo, 460 m, 16 Jan. 1991, Taylor et al. 1434 (K, HRCB, ZSS,
CEPEC); 28 km S from Santana towards Santa Maria da Vitória, 530 m, 15 Jan. 1991, Taylor et al. 1427
(K, HRCB, ZSS, CEPEC).
minas gerais: unlocalized (W of Rio São Francisco), E. Esteves Pereira 223 (see synonym tentatively cited
above); Mun. Varzelândia, 10 km N from the town, on road to Jaíba, 12 Aug. 1988, Eggli 1150 (ZSS).
Braun & Esteves Pereira’s recently described species is referred here as a putative hybrid
with only little hesitation, since its characters are strongly suggestive of hybrid origin (N.B.
stem morphology, cephalium, flower form, fruit, seed size etc.), involving P. pachycladus
and P. densiareolatus, the latter being reported as sympatric, l.c. Pilosocereus pachycladus
certainly also occurs in this region (see above) and obvious hybrids involving these two
species have been studied by the present authors in the other habitats cited here.
Pilosocereus densiareolatus F. Ritter (1979) was described on the basis of plants inhabiting
Bambuí limestone outcrops in central-northern Minas Gerais, with flower-bearing areoles
only moderately differentiated.
Northwards from Minas Gerais, the same species presents an increasingly well-
developed lateral cephalium, with flower-bearing areoles immersed in the branches, and
those populations have become known as Pilosocereus superfloccosus (Buining & Brederoo)
F. Ritter, described from western Bahia. This is, in fact, a confused and misapplied name,
since its protologue (cf. Pseudopilocereus superfloccosus Buining & Brederoo 1974a) clearly
illustrates reproductive parts of Pilosocereus gounellei subsp. zehntneri, and the positively
identifiable elements amongst its type material (U, holo.) consist of fragments of flower
and fruit of Pilosocereus gounellei subsp. zehntneri, and none of ‘P. superfloccosus’ as
interpreted by most authors.
The problem of confusion between P. densiareolatus and P. gounellei subsp. zehntneri is
not an unusual one. Most material examined from these species was found to be mixed
(Zappi 1994). Field study of such sympatric populations indicates that the arborescent P.
densiareolatus only flowers when the branches are far away from the ground, ie. 2.5–4.0
metres high, and, furthermore, the flowers are hidden in a hairy lateral cephalium. P.
gounellei subsp. zehntneri has a shrubby to treelike habit, but produces flowers when less
than 1 metre tall, its reproductive parts being much more accessible and obvious to
collectors. Young plants of P. densiareolatus frequently look like P. gounellei subsp. zehntneri,
but subsequently the branching pattern is completely different, being candelabriform for
P. gounellei subsp. zehntneri and erect for P. densiareolatus, whose spination is also denser and
finer. The examination of the apex of the fruits has proved to be the best character to
differentiate them: P. densiareolatus has a deeply sunken, pendent floral remnant, typical of
PILOSOCEREUS
Subgenus Pilosocereus, while P. gounellei subsp. zehntneri has fruits with rounded and
superficially inserted floral remnants, characteristic of Subgenus Gounellea.
Insufficiently known taxa

1) The following names refer to a taxon treated as incertae sedis here (cf. Zappi 1994: 103;
Ritter 1979: 64–65): Pilocereus glaucescens Linke (1858); Cephalocereus glaucescens (Linke)
Borg; Pilosocereus glaucescens (Linke) Byles & G. D. Rowley; Pseudopilocereus glaucescens
(Linke) Buxb.; ‘Cereastreae glaucescens’ Labour. (1853), nom. inval. (Art. 43.1); Pilocereus
coerulescens Lem. (1862), nom. illeg. (type as for above); Pilosocereus coerulescens (Lem.) F.
Ritter, nom. illeg. (The plant described by Lemaire under this illegitimate name originated
from the Serra do Cipó, MG and is identifiable as P. aurisetus, q.v.) Being guided only by
Labouret’s original description it is clear that the small plant he had before him could have
represented any one of at least 3 Brazilian species, even assuming that his statement that
it came from Brazil was correct. This plant was not preserved and it is futile to speculate
further on its identity in the absence of a definite locality.
2) The recently described Pierrebraunia brauniorum E. Esteves Pereira in Kakt. and. Sukk.
50: 311–314 (1999) will likely remain a botanical mystery until the secrecy about its
geographical origin is overcome. Esteves Pereira gives away no more than to say that it
emanates from high mountains in the Serra do Espinhaço of Minas Gerais (a mountain
range more than 1000 km in extent) and inhabits an area where a decidedly improbable
list of other cactus genera are said to grow (improbable in the sense that more than 15
years’ study of the cacti of Eastern Brazil by the present authors has so far failed to reveal
any instance where the genera Facheiroa and Cipocereus occur in proximity, especially since
in Minas Gerais the former genus is restricted to limestone outcrops close to the Rio São
Francisco, very far from the Serra do Espinhaço, where Cipocereus is found). The plant’s
extraordinary combination of characters is unequalled and its stated rarity lends support
to the idea, ventured in Taylor (2000), that this is some sort of bizarre intergeneric hybrid,
the likes of which are not unknown elsewhere in the family, eg. Bergerocactus with
Pachycereus (Pachgerocactus) and, separately, with Myrtillocactus (Myrtgerocactus), from
northern Baja California, Mexico. In the case of the vaingloriously named Pierrebraunia
brauniorum, its combination of few-ribbed stems covered by a visibly roughened epidermis
and small, deep pink, hummingbird syndrome flowers tempts the suggestion that this is a
hybrid between two genera, the most plausible being a Pilosocereus (eg. P. floccosus subsp.
quadricostatus) and either an Arrojadoa or a Micranthocereus, both of which are said to occur
in the vicinity. There is a good illustration of the plant in Braun & Esteves Pereira (2002:
6, Abb. 1).
3) The following may be a distinct species of dwarf habit, but has not been studied in
vivo by the authors, either in habitat or in cultivation. It was tentatively treated as a
possible synonym of P. machrisii by Zappi (1994), whose distribution it overlaps in western
Bahia (eg. it is reported from Mun. São Desidério, Sítio Grande, road to Barreiras, s.d.,
Horst & Uebelmann HU 203, cult. [ZSS], with P. machrisii being known from just north
of Barreiras; Map 38). While it should ‘key out’ under P. machrisii, a photograph of the
above collection in cultivation has been kindly supplied by Graham Charles as an
additional aid to identifying this little-known taxon (see Plate 49.2): Pilosocereus parvus (L.
MICRANTHOCEREUS
Diers & E. Esteves Pereira) P. J. Braun in Bradleya 6: 88 (1988); Pseudopilocereus parvus L.

Diers & E. Esteves Pereira in Kakt. and. Sukk. 33: 100–104 (1982). Holotype: Brazil,
Goiás, [Mun. Posse], Serra Geral de Goiás, border with Bahia, 900–1000 m, Apr. 1978,
E. Esteves Pereira 94 (KOELN [‘Succulentarium’], n.v.; UFG, iso.).
4) As the manuscript of the present work was being finalized, Andreas Hofacker published
a new taxon, Pilosocereus bohlei Hofacker in Kakt. and. Sukk. 52: 253–257 (2001). This
plant is known from a small population in the Serra São Francisco, northern Bahia, on
quartzite (Map 38), where it is sympatric with P. pachycladus, Micranthocereus flaviflorus and
Melocactus zehntneri (the holotype is cited, l.c. 255, as A. Hofacker 442, UFG 24356, n.v.).
It will probably prove to be a ‘good’ species, but its relationship seems rather unclear at
present for lack of adequate data on the morphology and dehiscence of its fruit. In the
protologue, Hofacker compares it with Stephanocereus luetzelburgii and Pilosocereus gounellei
— the swollen stem bases are reminiscent of the former, while the narrower apical fertile
part of the stem strongly resembles flower-bearing growth of the latter. However, in
response to a request for more information Hofacker (in litt., 15.10.01) communicated
that the fruit of his new taxon had deeply inserted perianth remains, which tends to rule
out any affinity with P. gounellei. Besides, the flower of P. bohlei is different from either of
the above, being short and S-shaped. In its basally branching habit and general stem
morphology it could belong to the P. AURISETUS Group and may be ‘keyed out’ here
under P. machrisii (q.v.). A brief summary of its morphology follows, based on the
protologue: to 1.8 m, branched at base only, stems blue-green to strongly glaucous, to 12
cm diam. in the basal 35–40 cm, but narrowing to only 5 cm diam. in the fertile region
above; ribs 9–12, tuberculate, to 15 × 15 mm; areoles oval, to 7 × 4 mm, long hairs
(vegetative part) to 30 mm, white; spines red-brown to light brown or more golden in
the fertile region, central 1, to 25 mm, radials 30–40, to 20 mm; areoles in fertile region
bearing dense tufts of long white hairs to 70 mm, developing on up to 7 of the ribs;
flower to 55 × 35 mm, white, tube curved; fruit 35 × 40 mm, greenish to bluish green,
funicular pulp white; seed 1.0 × 0.7 × 0.4 mm, matt black, with strongly developed
cuticula sculpturing (seed SEMs kindly supplied by A. Hofacker). See Plate 49.3.
21. MICRANTHOCEREUS Backeb.
in Blätt. Kakteenf. 1938(6): [22] (1938). Type: M. polyanthus (Werderm.) Backeb.
Including Austrocephalocereus Backeb. (1938) and Siccobaccatus P. J. Braun & E. Esteves Pereira
(1990); Micranthocereus subg. Austrocephalocereus (Backeb.) P. J. Braun & E. Esteves Pereira
(1991a); M. subg. Siccobaccatus (P. J. Braun & E. Esteves Pereira) N. P. Taylor (1991b).
Columnar, erect, unbranched or more commonly shrublike but branched only at base; vascular cylinder
rather to scarcely woody, cortical tissues mostly very mucilaginous; seedlings frequently densely hairy and
juvenile to adult stages also with long, flexible, bristly hypertrophic spines at stem base (except species nos.
1–3 and in populations of other species growing in sand). Branches erect or decumbent, 3–10 cm diam. or
more, not constricted, epidermis dull to bright green or glaucous; ribs 14–33, low, crenate, sinuses straight.
Areoles with felt and long hairs; spines weak, flexible to brittle, not very pungent, central spines porrect to
deflexed, longer than the numerous radials; indeterminate growth at basal areoles present, sometimes
MICRANTHOCEREUS
forming long curled bristles. Flower-bearing region of stem lateral or subapical, differentiated and sometimes
forming a cephalium modifying the stem’s morphology. Flowers nocturnal or diurnal, when nocturnal
strongly odoriferous, 15–42 × 5–32 mm; pericarpel smooth, naked, hemiglobose and often very clearly
delimited from the tube; tube broadened below, then cylindric, white, greenish, yellowish, orange-red, pink
or magenta, with broad bract-scales only near the apex; innermost filaments curved towards the style;
perianth-segments triangular to lanceolate or spathulate, innermost white, yellowish, pinkish or purplish.
Fruit obovate to turbinate, naked, indehiscent except in M. albicephalus and perhaps M. auriazureus, flower
remnant drying brownish or black, erect, sometimes tardily deciduous, forming a broad scar at the fruit
apex, but ± superficial; pericarp smooth, funicular pulp white or pinkish, solid. Seeds 1–1.7 mm,
cochleariform or elongate; testa-cells flat or convex; cuticular folds present or absent.
The relationships of this endemic Brazilian genus within tribe Cereeae are unclear, but fruits
with non-impressed apex, ± superficial floral remnants and stems bearing lateral cephalia are
found in Pilosocereus subg. Gounellea (cf. P. gounellei subsp. zehntneri) and Coleocephalocereus.
Some species in the latter genus display hypertrophic spine growth at the base of stems as
in the majority of Micranthocereus species, but their seeds differ in shape and in the position
of the hilum region. A hybrid between Pilosocereus (subg. Pilosocereus) pentaedrophorus and M.
purpureus has been recorded from two distant sites at the eastern margins of the Chapada
Diamantina, Bahia and implies that these two genera may be closely related.
The 8 species treated here seem distinct, though Nos 2 & 3, 4 & 5, and 6 & 7 may
represent vicariant species-pairs, the last being difficult to distinguish, yet rather disjunct.
Nos 1–7 are campo rupestre taxa (see Map 5), while No. 8 is found amidst caatinga forest on
limestone outcrops. A ninth species closely related to No. 8 occurs on Bambuí limestone
outcrops around the Serra Geral de Goiás, from southern Tocantins and Goiás to north-
western Minas Gerais (M. estevesii (Buining & Brederoo) F. Ritter). Two of the species are
single-site endemics (Nos 4 & 5) and two more are known from very small areas (Nos 1
& 6) and thus their conservation status needs to be carefully monitored. Plates 49.4–52.4.
1. Plants 0.3–2.5(–3.0) m, suffrutescent or with solitary inclined stems, vascular cylinder not
woody, or, if woody, ribs 14–17; seeds cochleariform (crystalline rocks and sandstones,
campos rupestres, Serra do Espinhaço and Chapada Diamantina) . . . . . . . . . . . . . . . . . . . . . . . . . . 2
1. Plants to > 3 m, maturing when > 1.2 m, erect with very woody vascular cylinder; ribs
21–30 or more; fruits drying inside the cephalium; seeds elongate, ± twisted (Bambuí
limestone, SW Bahia and adjacent Minas Gerais) . . . . . . . . . . . . . . . . . 8. dolichospermaticus
2. Flowers > 30 × 20 mm, anthesis predominantly nocturnal . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
2. Flowers slender, < 25 × 11 mm, anthesis diurnal . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
3. Ribs 23–29(–32); cephalium wool white to yellowish; epidermis bright green; flowers
greenish or pinkish white outside (Serra do Espinhaço: N Minas Gerais and southernmost
Bahia) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2. albicephalus
3. Ribs 10–26; cephalium wool pale brown, with pinkish or grey shades; epidermis grey-green
or glaucous; flowers deep magenta outside (Chapada Diamantina, Bahia) . . . . . . . . 3. purpureus
4. Stem solitary; floral remnants strongly blackened; central spines and bristles of flower-bearing
areoles dark red to brown; ripe fruits green; seeds black (N Minas Gerais: Serra da Bocaina
& Serranópolis) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1. violaciflorus
4. Stems branched at base; floral remnants pale brown, not blackened; spines and bristles of
flower-bearing areoles mostly golden or pale yellow; ripe fruits red or pinkish; seeds brownish . . . 5
5. Flowers 15–18 mm, outer and inner perianth-segments of contrasting colours; stems 3–5.5 cm
diam., erect to ± inclined or decumbent . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
MICRANTHOCEREUS
5. Flowers 20–25 mm, perianth-segments ± concolorous or innermost paler; stems 5.5–7.0 cm

diam., erect . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
6. Flowers orange-red/magenta-pink, with white to yellowish inner perianth-segments (N Bahia:
Mun. Morro do Chapéu to Mun. Sento Sé) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7. flaviflorus
6. Flowers pale purplish with pale cream or white inner perianth-segments (S Bahia: Mun.
Caitité) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6. polyanthus
7. Fertile part of stem not sunken, bristles and wool white or greyish (nr Grão Mogol, Minas
Gerais) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4. auriazureus
7. Cephalium sunken, bristles and wool golden or brownish (W of Seabra, Bahia) . . . . . 5. streckeri
Subg. Austrocephalocereus (Backeb.) P. J. Braun & E. Esteves Pereira (Nos 1–3): Stems
woody or lacking well-developed wood, erect and branched at base or solitary and then
often inclined, 0.7–2.0(–3.0) m high, with a ± sunken, and sometimes discontinuous
cephalium; lacking hypertrophic spine development at stem base; flowers diurnal and/or
nocturnal; fruit with persistent, blackened perianth remains (inconsistently blackened in M.
purpureus), pericarp fleshy; seeds black, testa-cells with convex periclinal walls and
cuticular sculpturing. Endemic to the core area within Eastern Brazil and characteristic of
the Northern campos rupestres.
1. Micranthocereus violaciflorus Buining in Kakt. and. Sukk. 20: 129–130 (1969).

Holotype: Brazil, Minas Gerais, ‘Chapada Diamantina’ [Mun. Grão Mogol, Serra da
Bocaina], 24 June 1968, Buining in Horst 275 (U).
Solitary, rarely branched at base, to 1 m high; vascular cylinder very woody. Stems erect to inclined,
3.5–4.0 cm diam.; ribs 14–17, 2 × 3–5 mm; epidermis green. Areoles 2 mm, to 3 mm apart, long hairs
white to grey. Spines flexible, pale golden to golden brown, reddish or dark brown, central spines 6–8,
longest one to 25 mm, radials 25–30, 5–15 mm. Fertile part of stem lateral, sunken, flower-bearing
areoles with abundant, loose white hairs and long reddish or dark brown bristles to 3 cm. Flowers poorly
known, 23 × 7 mm, diurnal, tube violet or pinkish red, perianth-segments deep pink, spreading to erect.
Fruit obovate, truncate at apex, 11 × 9 mm, flower remnants black, pericarp smooth, dull, green. Seeds
c. 1 mm, cochleariform, black; testa-cells convex, with cuticular folds (Barthlott & Hunt 2000: pl. 50.4).
Northern campo rupestre (N Serra do Espinhaço) element: amongst rocks at c. 900–1100 m, Serra da
Bocaina and Serranópolis, northern Minas Gerais. Endemic to the core area within Minas Gerais. Map 40.
minas gerais: Mun. Porteirinha, Serranópolis, 15°50'S, 42°49'W, Mar. 2000, fide M. Machado (in litt.,
11.04.00 and photo on Machado’s website); Mun. Grão Mogol, 48 km W of bridge over Rio Vacaria,
28.5 km E of Caveira (road junction to Riacho dos Machados and Porteirinha) on road BR 251, 31 Jan.
1991, Taylor et al. 1514 (K, HRCB, ZSS, BHCB).
conservation status. Vulnerable [VU D2] (2); area of occupancy estimated to be < 20 km2; PD=2,
EI=1, GD=1. Short-list score (2×4) = 8. Will be under threat if charcoal producers move into the two
very restricted areas where it occurs. Needs a population survey and regular monitoring.
This rare species is characterized by a suite of presumably plesiomorphic character states

(a rather woody vascular axis, only moderately mucilaginous stem tissue, absence of
hypertrophic spines at stem base, green pericarpel, strongly blackening, persistent perianth
remnants and black seeds with convex testa-cells), which suggest that its phylogenetic
position within Micranthocereus is basal. However, it has most characters in common with
MICRANTHOCEREUS
Subg. Austrocephalocereus. It seems to be a relictual species, occupying a restricted habitat

in a small area in the northern half of the Serra do Espinhaço (MG).
2. Micranthocereus albicephalus (Buining & Brederoo) F. Ritter, Kakt. Südamer. 1:

108 (1979). Holotype: Brazil, Minas Gerais, Mato Verde, 950 m, Aug. 1972, Buining in
Horst 348 (U).
Austrocephalocereus albicephalus Buining & Brederoo in Kakt. and. Sukk. 24: 73–75 (1973).
Coleocephalocereus albicephalus (Buining & Brederoo) F. H. Brandt in Kakteen Orchid. Rundschau 6:
124 (1981).
Micranthocereus monteazulensis F. Ritter, Kakt. Südamer. 1: 105–106 (1979). Holotype: Brazil, Minas
Gerais, Serra de Monte Azul, 1964, Ritter 1214 (U).
M. aureispinus F. Ritter, tom. cit.: 107–108 (1979). Holotype: Brazil, Minas Gerais, SW of Monte Azul,
1971, Ritter 1482 (U).
Shrubby, to 0.5–2.5(–3.0) m, branched mainly at base; vascular cylinder not woody. Stems erect, 6–8.5
cm diam.; ribs 23–29(–32), 5 × 6.5 mm; epidermis bright green. Areoles 3–4 mm apart, long hairs white
to grey. Spines flexible to brittle, golden, showing only weak indeterminate growth at basal areoles, central
spines to 12, to 10 mm, radials 15–18, 5–15 mm, the lowermost slightly longer. Fertile part of stem lateral,
sunken into the stem, flower-bearing areoles with abundant, white, compact hairs and long golden bristles
to 6.5 cm. Flowers 45–50 × 26 mm, nocturnal; pericarpel 4–5 × 9 mm, whitish to cream; tube 27–30 ×
10–25 mm, broadened at base, constricted at the middle, white or pale green with pink shades, with broad
bract-scales only at apex; perianth-segments 10 × 3.5–5.5 mm, triangular to lanceolate, outer segments
spreading to recurved, white, brown to magenta-pink at tips, inner segments spreading, white, pinkish to
reddish brown at tips; stamens included in relation to the perianth-segments; nectar-chamber 7 × 8 mm;
style 32 mm, stigma-lobes c. 8, exserted; ovary locule 3–4 × 4–5 mm, obtriangular in longitudinal section.
Fruit broadly turbinate, 3.5 cm diam., flower remnants strongly blackened; pericarp smooth, shiny, pale to
dark green or wine-coloured, apically dehiscent. Seeds 1.5–1.7 mm, cochleariform, black; testa-cells
convex, with cuticular folds (Taylor 1991a: pl. 1.1).
Northern campo rupestre element: between crystalline rocks and on cliffs in campo rupestre, c. 800–1000 m,
Serra do Espinhaço (Serra Geral) northern Minas Gerais and adjacent southernmost Bahia. Endemic to
the core area of Eastern Brazil. Map 40.
bahia: Mun. Licínio de Almeida, Ritter 1415 (U); ibid. (?), 12 km from town towards Brejinho das
Ametistas, 14°32'4"S, 42°31'51''W, 12 Mar. 1994, N. Roque et al. in CFRC 15039 (K, SPF), 10 June
2002, G. Charles (photos).
minas gerais: Mun. Monte Azul, mountains E of town (Serra Geral), 1964, Ritter 1214 (U); l.c.,
mountains SW of town, 1971, Ritter 1482 (U); Mun. Mato Verde/Rio Pardo de Minas, 12 km E of Mato
Verde on road to Santo Antônio do Retiro, 15°23'S, 42°45'W, 9 Nov. 1988, Taylor & Zappi in Harley
25519 (K, SPF), 13 Aug. 1988, Eggli 1155 (ZSS).
PD=2, EI=1, GD=1. Short-list score (1×4) = 4. Potentially threatened due to its limited range and
generally small size of its 4 known populations.
This is the southern sister species of the following.
3. Micranthocereus purpureus (Gürke) F. Ritter in Kakt. and. Sukk. 19: 157 (1968);
Buining in Ashingtonia 2: 28–29 (1975). Holotype: Brazil, Bahia, Serra do Sincorá, 1906,
Ule 4 (B, n.v.).
MICRANTHOCEREUS
Cephalocereus purpureus Gürke in Monatsschr. Kakt.-Kunde 18: 86 (1908). Austrocephalocereus purpureus

(Gürke) Backeb. in Jahrb. Deutsch. Kakteen-Ges. 1941(2): 53 (1942).
Cephalocereus lehmannianus Werderm. in Repert. Spec. Nov. Regni Veg. Sonderbeih. C: tab. 46 (1932);
Bras. Säulenkakt.: 118–119 (1933). Austrocephalocereus lehmannianus (Werderm.) Backeb. in Cact.
Succ. J. (US) 23: 149 (1951). Micranthocereus lehmannianus (Werderm.) F. Ritter in Kakt. and. Sukk.
19: 157 (1968), nom. inval. (Art. 33.2). Coleocephalocereus lehmannianus (Werderm.) F. H. Brandt in
Kakteen Orchid. Rundschau 6: 124 (1981), nom. inval. (Art. 33.2). Holotype: Brazil, Bahia, Serra
do Espinhaço, 1100 m, Apr. 1932, Werdermann 3295 (B†). Lectotype (designated here):
Werdermann, l.c., tab. 46 (1932).
M. haematocarpus F. Ritter, Kakt. Südamer. 1: 105 (1979). Holotype: Brazil, Bahia, Ituaçu, 1964, Ritter
1328 (U).
M. ruficeps F. Ritter, Kakt. Südamer. 1: 106–107 (1979). Holotype: Brazil, Bahia, Jacobina, 1963, Ritter
1212 (U).
Shrubby, 0.6–2.5(–3.0) m, branched at base; vascular cylinder not woody. Stems erect, 4.5–10.0 cm
diam.; ribs 12–26, rarely only 10 when immature, 4–5 × 8–9 mm; epidermis blue-green to strongly
glaucous. Areoles 3–4 mm, to 6–8 mm apart, long hairs pale brown to grey. Spines flexible to brittle,
golden or reddish, brown when old, showing only weak indeterminate growth at basal areoles, central
spines 4–7, to 20(–25) mm, radials 20–25, to 10 mm. Fertile part of stem lateral, sunken into the stem,
flower-bearing areoles with pale brown, pinkish or greyish, abundant, loose or compact hairs and few
long golden or brownish bristles to 4 cm. Flowers 42 × 30–32 mm, partially opening at late afternoon,
fully expanded at night; pericarpel 8 × 12 mm, whitish to pale pink; tube 20–25 × 20–25 mm, broadened
at base, slightly constricted at the middle, deep pink-magenta, with broad bract-scales only at apex;
perianth-segments 9–11 × 6–9 mm, triangular to lanceolate, outer segments recurved, deep pink to
reddish, inner segments spreading to recurved, white; stamens included in relation to the perianth-
segments; nectar-chamber 7 × 8 mm; style 25–28 mm, stigma-lobes c. 10, exserted; ovary locule 4–6 ×
5–6 mm, obtriangular in longitudinal section. Fruit broadly turbinate, 1.7–2.0 cm diam., flower remnants
brown, usually not blackened, pericarp smooth, shiny, purplish pink. Seeds 1.5–1.7 mm, cochleariform,
black; testa-cells convex, with cuticular folds (Barthlott & Hunt 2000: pl. 50.2–3).
Northern campo rupestre (Chapada Diamantina) element: on crystalline rocks in campo rupestre and its
ecotones with caatinga and cerrado, c. 350–1900 m, eastern flanks and highest peaks of the Chapada
Diamantina. Endemic to Bahia. Map 40.
bahia: Mun. Jacobina, 2.5 km E of bridge to E of town, 16 July 1989, Zappi 129A (SPF); Mun. Morro
do Chapéu, 19 June 1994, L.P. de Queiroz 4038 (HUEFS); l.c., 9 km NE of town towards Jacobina, 23
Aug. 1988, Eggli 1271 (ZSS); l.c., 19.5 km SE of town on road BA 052, Rio do Ferro Doido, 31 May
1980, Harley et al. 22900 (K, CEPEC); l.c., 11°49'S, 41°12'W, 11 Mar. 1996, A. Conceição 2248 (HUEFS);
Mun. Seabra, c. 28 km W of town on road BA 242, 14 Jan. 1991, Taylor et al. (obs.); ibid., 16 km E of
town, 18 July 1989, Taylor & Zappi (K, photos); Mun. Lençóis, 24 Sep. 1965, A.P. Duarte 9363 & E.
Pereira 10076 (RB, HB); l.c., 14 km NW of town, Serra do Brejão, 12°27'S, 41°21'W, Harley et al. 22378
(K, CEPEC); Mun. Lençois, Morro do Pai Inácio, Mar. 1997, A. Conceição s.n. (K, photos); Mun. Andaraí,
Rio Paraguaçu, 12°48'S, 41°20'W, 27 Dec. 1988, Taylor & Zappi in Harley 27429 (K, SPF, CEPEC), 20
Aug. 1988, Eggli 1223 (ZSS); Mun. Mucugê, 4 km WNW of town on the road to Boninal, 13°1'S,
41°26'W, 22 Dec. 1988, Taylor & Zappi in Harley 25599 (K, SPF, CEPEC); l.c., surroundings of town, 22
Dec. 1988, Taylor & Zappi in Harley 27381 (K, SPF, CEPEC); l.c., 3 km towards Igatu, 41°21'W, 22 Aug.
1986, J.D.C. Arouck Ferreira et al. 368 (HRB, RB), 10 Oct. 1987, M.L. Guedes et al. 1539 (ALCB); Mun.
Piatã, road Piatã–Inúbia, c. 25 km NW of Piatã, 13°4'19"S, 41°55'24''W, 24 Feb. 1994, Sano et al. in
CFCR 14539 (K, SPF); Mun. Abaíra, Serra do Rei, 13°17'S, 41°54'W, 12 Jan. 1994, W. Ganev 2791 (K,
HUEFS), Feb. 1992, Sano & Laessøe in Harley 50897 (SPF, CEPEC); Mun. Rio de Contas, Pico das Almas,
‘Campo do Queiroz’, 13°32'S, 41°57'W, 11 Dec. 1988, Taylor in Harley 25564 (K, SPF, CEPEC); l.c.,
Cachoeira do Fraga, 13°35'S, 41°50'W, 24 Nov. 1988, Taylor & Zappi in Harley 25547 (K, SPF, CEPEC);
Mun. Ituaçu, c. 2 km NE and 10 km N of town, 19 Aug. 1988, Eggli 1203, 1214 (ZSS).
conservation status. Least Concern [LC] (1); PD=2, EI=1, GD=2. Short-list score (1×5) = 5. Its
numerous habitats are not significantly affected by change at present.
MICRANTHOCEREUS
The name Cephalocereus purpureus was misapplied to what is now known as

Coleocephalocereus goebelianus by Britton & Rose (1920) and Werdermann (1933), the latter
redescribing the true Cephalocereus purpureus as C. lehmannianus. Luetzelburg (1925–26, 1:
fig. 38) illustrated the true Micranthocereus purpureus as ‘Pilocereus na caatinga’ and
commented in the caption that it was always accompanied by Cereus leucostele (=
Stephanocereus leucostele), which is not the case. He was evidently confusing it with C.
goebelianus at this stage (however, cf. Luetzelburg 1925–26, 3: 69). The true identity of
M. (Austrocephalocereus) purpureus was recognized by Ritter (1968) and Buining (1975b).
M. purpureus is the most wide-ranging and variable species in the genus and a
characteristic element of the campo rupestre flora in the higher and eastern parts of the
Chapada Diamantina, where it is constantly associated with Stephanocereus luetzelburgii. It
hybridizes with Pilosocereus pentaedrophorus near Andaraí and Ituaçu.
Subg. Micranthocereus (Nos 4–7): Stems only rarely woody (M. cf. flaviflorus), branched at
base, erect or semi-sprawling, 0.3–1.2 m high, with a superficial to sunken, and sometimes
discontinuous cephalium and greater or lesser development of hypertrophic spines at stem
base; seedlings densely woolly; flowers diurnal; fruit with tardily deciduous, non-blackening
perianth remains, pericarp fleshy; seeds brown to brown-black, testa-cells with nearly flat
periclinal walls lacking cuticular sculpturing. Endemic to the core area within Eastern
Brazil and characteristic of the Northern and South-eastern campos rupestres.
4. Micranthocereus auriazureus Buining & Brederoo in Cact. & Succ. J. (US) 45:
120–123 (1973) (‘auri-azureus’). Holotype: Brazil, Minas Gerais, Grão Mogol, 900–1000
m, 17–18 Aug. 1972, Buining in Horst & Uebelmann 346 [‘348’] (U).
Shrubby, to 1.2 m, much branched at base; vascular cylinder not woody. Stems erect, 6–7 cm diam.; ribs
15–19, 5 × 8–9 mm; epidermis blue-green, glaucous. Areoles 3–7 mm, to 4 mm apart, long hairs white
to grey. Spines flexible to brittle, golden, showing indeterminate hypertrophic growth at basal areoles,
forming long curled bristles, central spines to 6, 2–3(–8) cm, radials 25–30, 5–15 mm. Fertile part of stem
lateral, superficial, flower-bearing areoles with white hairs and long bristles to 4(–8) cm. Flowers 20–25 ×
10–11 mm, diurnal; pericarpel 2.5 × 5 mm, brownish to pale pink; tube to 16 × 7–9 mm, broadened at
base, constricted at the middle, deep magenta pink, with broad bract-scales only at apex; perianth-segments
4–7 × 2.5–3.5 mm, triangular to lanceolate, outer segments spreading, magenta-pink, inner segments erect
to spreading, lilac-pink; stamens included in relation to the perianth-segments; nectar-chamber 8 × 5–6
mm; style 15–19 mm, stigma-lobes c. 8, exserted; ovary locule 3–4 × 2–3 mm, obtriangular in longitudinal
section. Fruit obovate, truncate at apex, 1.8 × 1.4–1.6 cm, flower remnants pale brown, deciduous when
fruit ripe, pericarp smooth, dull, pale pink to reddish, sometimes splitting open transversely below apex.
Seeds c. 1.5 mm, cochleariform, brown, shiny; testa-cells flat to convex, without cuticular folds.
South-eastern campo rupestre (Grão Mogol) element: between crystalline rocks and in quartz sand, c.
750–1000 m, Serra do Barão and vicinity, northern Minas Gerais. Endemic to the core area within Minas
Gerais. Map 40.
minas gerais: Mun. Grão Mogol, 21 May 1982, Giulietti et al. in CFCR 3416 (SPF, K), 24 Aug. 1986,
Zappi et al. in CFCR 9925 (SPF, MBM), 21 May 1987, Pirani & Mello-Silva in CFCR 10773 (SPF), 26
May 1988, Zappi et al. in CFCR 11944 (SPF), 15 Oct. 1988, Taylor & Zappi in Harley 25072 (SPF, K).
MICRANTHOCEREUS
conservation status. Endangered [EN B1ab(iii) + 2ab(iii)] (3); extent of occurrence < 5000 km2; area
of occupancy estimated to be < 100 km2; PD=2, EI=1, GD=1. Short-list score (3×4) = 12. Endangered
due to its very restricted distribution, part of which may be subject to inundation by a dam-lake in future.
However, another part of its area is inside a recently created reserve (Parque Estadual da Serra do Barão).
According to the list in Uebelmann & Braun (1984) and the label on the type specimen at
Utrecht, the type collection of M. auriazureus is HU 346, and the name should be
considered legitimate, since it is clear that HU 348 was cited erroneously as its holotype in
the protologue (HU 348 is the type number for the prior-published M. albicephalus, q.v.).
5. Micranthocereus streckeri Van Heek & Van Criek. in Kakt. and. Sukk. 37: 102–105,
with illus. (1986). Holotype: Brazil, Bahia, W of Seabra, Van Heek & Van Criekinge 85/250
(KOELN [‘Succulentarium’], n.v.).
Shrubby, to 0.8 m, densely branched at base. Stems erect, 3–5(–5.5) cm diam., vascular cylinder not
woody; ribs 17–24(–25), 2–3 × 6–8 mm; epidermis grey-green, slightly glaucous. Areoles 2–3 mm, to 5
mm apart, with white to grey trichomes. Spines flexible, pale golden, hypertrophic growth of spines at
basal areoles present, central spines 6–8, spreading, lowermost longest, to 25 mm, radials 20–30, 5–10
mm. Fertile part of stem lateral, superficial to sunken, flower-bearing areoles with compact white to
brownish hairs and long pale golden to reddish brown bristles to 30 mm. Flowers 40–50 × 22 mm,
diurnal; pericarpel 3.5 × 3.5 mm, pale pink; tube to 18 × 3–5 mm, broadened at base, cylindric above,
deep pink, with bract-scales only at apex; perianth-segments 4–5 × 2.5 mm, triangular to lanceolate, outer
segments spreading, deep pink, inner segments erect, purplish pink; stamens included in relation to the
perianth; style 13–15 mm, stigma-lobes 5–6, included; ovary locule 2 × 2.5 mm, obtriangular in
longitudinal section. Fruit obovoid, 11 × 10 mm, floral remnants pale brown, deciduous; pericarp
smooth, dull, purplish. Seeds c. 1.2 mm, cochleariform, blackish, shiny; testa-cells flat to convex, without
cuticular folds.
Northern campo rupestre (Chapada Diamantina) element: in the campo rupestre/cerrado de altitude ecotone,
c. 1100 m, west of Seabra, central Bahia. Endemic. Map 40.
bahia: Mun. Seabra, 27–28.5 km W of town towards Ibotirama, 25 Aug. 1988, Eggli 1289 (ZSS), 17
July 1989, Zappi 145 (SPF), 14 Jan. 1991, Taylor et al. 1415 (K, HRCB, ZSS, CEPEC).
conservation status. Critically Endangered [CR B1ab(iii) + 2ab(iii); C2a(ii); D] (4); PD=2, EI=1,
GD=1. Short-list score (4×4) = 16. Critical due to its very restricted distribution, small population size
(< 50 individuals seen) and potential for alteration of its habitat, which is above a main highway.
This species is poorly understood, being known from only the locality cited above where,
as recently illustrated by Heek & Strecker (2002: 116), it appears to be undergoing
introgression with the sympatric M. purpureus, making assessment of its typical
morphological state rather difficult. Specimens which, in 1991, the authors interpreted to
be least influenced by this introgression showed a certain resemblance to the
geographically distant M. auriazureus (Grão Mogol, MG), but have ± sunken cephalia and
darker seeds.
6. Micranthocereus polyanthus (Werderm.) Backeb. in Jahrb. Deutsch. Kakteen-Ges.

1941(2): 51 (1942). Holotype: Bahia, near Caitité, May 1932, Werdermann 3457 (B†).
Lectotype (designated here): Werdermann, l.c. infra, tab. 43 (1932).
MICRANTHOCEREUS
Cephalocereus polyanthus Werderm. in Repert. Spec. Nov. Regni Veg. Sonderbeih. C, Lfg 11, tab. 43
(1932); Bras. Säulenkakt.: 114–116 (1933). Arrojadoa polyantha (Werderm.) D. R. Hunt in Bradleya
5: 92 (1987).
Shrubby, to 1.2 m, much branched at base; vascular cylinder not woody. Seedlings with abundant white
hairs and long curled, hypertrophic bristles at base. Stems erect, 3.5–5.0 cm diam.; ribs 15–20, 4–5 × 3–6
mm; epidermis grey-green, glaucous. Areoles 2–4 mm, to 10 mm apart, long hairs white to grey. Spines
flexible, pale golden, showing indeterminate growth at basal areoles, forming long curled bristles, central
spines 4–8, to 15 mm, radials 20–28, 5–10 mm. Fertile part of stem lateral, superficial to slightly sunken,
flower-bearing areoles with abundant, loose white hairs and long, pale golden bristles to 20 mm. Flowers
18 × 7–8 mm, diurnal; pericarpel 2 × 3 mm, pale pink; tube to 16 × 5–6 mm, broadened at base,
constricted at the middle, coral red to pale pink, with broad bract-scales only at apex; perianth-segments
4–5 × 1–2 mm, triangular to lanceolate, outer segments spreading to erect, coral red to reddish pink, inner
segments opening very slightly, pale cream to white; stamens included in relation to the perianth-
segments; style 13–15 mm, stigma-lobes c. 8, included; ovary locule 1–2 × 1.5–2.0 mm, obtriangular in
longitudinal section. Fruit obovate, truncate at apex, 5–7 × 5 mm, flower remnants pale brown; pericarp
smooth, dull, pale pink. Seeds c. 1 mm, cochleariform, dark brown, shiny; testa-cells flat to convex,
without cuticular folds.
Northern campo rupestre element: in quartz sand amongst crystalline rocks, c. 900–1000 m, Mun. Caitité,
southern Bahia. Endemic. Map 40.
bahia: Mun. Caitité, c. 9 km ESE from town on road to Brumado, then c. 2 km S on footpath towards
‘São João’, 27 Aug. 1988, Eggli 1309 (ZSS); l.c., 25 km S of town, 1 km SW of Brejinho das Ametistas,
26 July 1989, Zappi 176 (SPF), 2 Feb. 1991, Taylor et al. 1537 (K, HRCB, ZSS, CEPEC).
conservation status. Endangered [EN B2ab(iii)] (3); area of occupancy estimated to be < 100 km2;
PD=2, EI=1, GD=1. Short-list score (3×4) = 12. Endangered due to its restricted distribution and
projected decline due to mining activities.
In its overall appearance, but especially in habit, soft stems, superficial cephalium and
bicoloured flowers, it is assumed to be the southern sister species of the following.
7. Micranthocereus flaviflorus Buining & Brederoo in Kakt. and. Sukk. 25: 25–27
(1974). Holotype: Brazil, Bahia, Mun. Sento Sé / Umburanas, Serra do Curral Feio, 850
m, 22 July 1972, Buining in Horst 389 (U).
M. densiflorus Buining & Brederoo in Cact. & Succ. J. (US) 46: 113–116 (1974). M. flaviflorus subsp.
densiflorus (Buining & Brederoo) P. J. Braun & E. Esteves Pereira in Succulenta 74: 132 (1995).
Holotype: Brazil, Morro do Chapéu, 900 m, Horst 221 (U).
M. uilianus Brederoo & C. A. L. Bercht in Succulenta 63: 178–183 (1984). M. flaviflorus var. uilianus
(Brederoo & C. A. L. Bercht) P. J. Braun & E. Esteves Pereira in Succulenta 74: 132 (1995), nom.
inval. (Art. 33.2?). Holotype: Brazil, Bahia, Mun. Sento Sé, ‘Serra da Mimosa, near Limoeiro, 1130
m’, 4 July 1974, Buining in Horst & Uebelmann 439 (U; ZSS, iso.).
Shrubby, 30–100 cm, much branched at base; vascular cylinder not woody (except in the sand-dwelling
populations from E side of Mun. Morro do Chapéu). Stems partly decumbent, 4–5 cm diam.; ribs 12–16,
3–5 × 5–7 mm; epidermis grey-green, glaucous. Areoles 2–3.5 mm, to 3 mm apart, long hairs white or
cream to grey. Spines flexible, golden, sometimes reddish, showing indeterminate hypertrophic growth
at basal areoles in some populations, forming long curled bristles, central spines 4–8, 10–25 mm, radials
20–25, 10–12 mm. Fertile part of stem lateral to apical, superficial, flower-bearing areoles with loose
white to yellowish hairs and long pale to brownish golden bristles to 2.5 cm. Flowers 14–18 × 5–8 mm,
diurnal; pericarpel 2 × 2.5–3.0 mm, pale pink; tube to 12 × 4–6 mm, broadened at base, constricted at
the middle, orange-red to pink, with bract-scales only at apex; perianth-segments 2–4 × 1–3 mm,
MICRANTHOCEREUS
triangular to lanceolate, outer segments spreading to erect, orange-red to reddish or pink/magenta; inner
segments opening very slightly, yellow to pale cream; stamens included in relation to the perianth-
segments; style 10–12 mm, stigma-lobes 4–6, included; ovary locule 1.5 × 1.8–2.0 mm, obtriangular in
longitudinal section. Fruit obovoid to globose, truncate at apex, 5–8 × 5–8 mm, flower remnants pale
brown, deciduous; pericarp smooth, dull, red to reddish pink. Seeds c. 1–1.5 mm, cochleariform,
blackish, shiny; testa-cells flat to convex, without cuticular folds (Barthlott & Hunt 2000: pl. 50.5–8).
Northern campo rupestre (Chapada Diamantina) element: in the campo rupestre/caatinga ecotone on quartz
sand or sandstone between low shrubs, c. 700–1200 m, northern and western flanks of the Chapada
Diamantina (northwards from the region of Morro do Chapéu), northern Bahia. Endemic. Map 40.
bahia: Mun. Sento Sé, ‘Serra do Tinga’, Aug. 1912, Zehntner 274 (US), ibid., s.n. (M); Serra do
Mimoso, near Limoeiro, 4 July 1974, Buining in Horst & Uebelmann 439 (U, ZSS); l.c., 10°19'S, 41°24'W,
11 July 2000, G. Charles (photos); Mun. Sento Sé / Umburanas, Serra do Curral Feio [S of Minas do
Mimoso], s.d., Horst 389 (U); l.c., 10°22'S, 41°19'W, 9 Apr. 1999, L.P. de Queiroz et al. 5120 (HUEFS
36691); Mun. Umburanas, ‘estrada velha Delfino – Mimoso de Minas, Serra do Curral Frio’, near
Delfino, 10°24'7"S, 41°18'43''W, 9 Mar. 1997, Harley et al. (ALCB, UEC, K, photos), ibid., 10°24'14"S,
41°18'41''W, 3 Apr. 2002, Oliveira et al. 118 (HUEFS 59063); Mun. Várzea Nova, 11°16'S, 40°56'W,
Hofacker (fide Machado, pers. comm.); Mun. Morro do Chapéu, north of town, W bank of the Rio
Salitre, c. 8 km from Icó, N of Brejão (Formosa), ‘lagedo bordado’, 5 Aug. 2002, Taylor & Machado (K,
photos); l.c., W of town, S of road BA 052 towards ‘Pedra Duas Irmãs’, c. 11°33'40"S, 41°17'50''W, June
2002, M. Machado (photos); l.c., SE of the town, 2 Aug. 2002, Taylor (K, photos); l.c., S of town, terreno
de M. Machado, 3 Aug. 2002, Taylor (K, photos); l.c., Boqueirão dos Lages, 17 May 1975, A.L. Costa
& G.M. Barroso s.n. (HBR, ALCB); l.c., 19–20 km W of town on road BA 052, 11°28'S, 41°19'W, 25
Dec. 1988, Taylor & Zappi in Harley 27392 (K, SPF, CEPEC), 14 Jan. 1977, Hatschbach 39565 (MBM);
l.c., 22 km W, 17 July 1989, Zappi 135 (SPF, K), l.c., c. 26–27 km W, then 1 km S on road to Cafarnaum,
26 Dec. 1988, Taylor & Zappi (obs.).
conservation status. Least Concern [LC] (1); extent of occurrence = 1024 km2; PD=2, EI=1, GD=2.
Short-list score (1×5) = 5. Area of occupancy unknown, but assumed to be < 10% of extent of occurrence.
The distribution of this species seems somewhat disjunct, but the region between its
northern and southern sites has been little explored. However, this includes extensive areas
of limestone and derived calcareous caatinga soils, which would not suit it and thus it is
almost certain that its distribution is highly fragmented. However, populations observed
contain abundant plants and numerous seedlings and some are rather extensive. It is highly
variable between populations, and much more so than the above synonymy indicates.
More importantly, its eastern and south-eastern populations (on more sandy terrain, Plate
52.1) may prove to be indistinguishable from the preceding species, when both are better
understood. At present it is largely the remarkable geographical disjunction between these
two taxa that causes the writers to hesitate in proposing their merger or re-shaping.
In August 2002, Taylor & Machado observed hummingbirds and a yellow butterfly
actively visiting the flowers in two populations of the species in Mun. Morro do Chapéu.
Subg. Siccobaccatus (P. J. Braun & E. Esteves Pereira) N. P. Taylor (No. 8): Stems normally
unbranched above ground unless damaged, forming erect, very woody columns to 5 m
or more, with hypertrophic spine development at base when young; cephalium deeply
sunken, continuous; flowers nocturnal; fruit drying out to become paper-thin, perianth
remains deciduous; seeds slender elongate or with many, small testa-cells. Limestone west
of the Rio São Francisco.
MICRANTHOCEREUS
8. Micranthocereus dolichospermaticus (Buining & Brederoo) F. Ritter, Kakt.

Südamer. 1: 108 (1979). Holotype: Brazil, Bahia, W of Bom Jesus da Lapa, 460 m, Horst &
Uebelmann 395 (U)
Austrocephalocereus dolichospermaticus Buining & Brederoo in Kakt. and. Sukk. 25: 76–79 (1974). Siccobaccatus
dolichospermaticus (Buining & Brederoo) P. J. Braun & E. Esteves Pereira in Succulenta 69: 7 (1990).
Erect, fertile from c. 1.2 m, reaching more than 3 m, unbranched above ground, but often sprouting at
the base and forming a compact group of erect stems; vascular cylinder very woody. Seedlings with spines
curved upwards and with long, curled, hypertrophic bristles at base. Stems 8–12 cm diam.; ribs
21–30(–32), 10–20 × 10–22 mm, in cephalium region 12 × 13 mm; epidermis blue-green, strongly
glaucous. Areoles 3–5 mm, to 2–6 mm apart, long hairs white to grey, to 3 cm. Spines hard, brittle,
golden to brown and black when old, central spines 4–7, to 15 mm, radials 8–12, to 5 mm. Fertile part
of stem lateral, sunken into the stem, flower-bearing areoles with white to pale yellow, abundant,
compact hairs and long brownish or reddish bristles to 4 cm. Flowers 4 × 2.5 cm, nocturnal, all white or
cream-coloured; pericarpel 5 × 11–12 mm; tube 35–38 × 12–18 mm, slightly constricted at the middle,
with broad bract-scales only at apex; perianth-segments 9–11 × 6–9 mm, triangular to lanceolate, outer
segments recurved, inner segments spreading; stamens included in relation to the perianth-segments;
nectar-chamber 10 × 9–10 mm; style 32–36 mm, stigma-lobes 8–9, exserted; ovary locule 4–5 × 2 mm,
depressed in longitudinal section. Fruit broadly turbinate, 9–10 mm diam., flower remnants pale brown,
deciduous; pericarp dry, brownish, wrinkled. Seeds 2 mm, elongate, brownish; testa-cells flat, hilum
depressed (Barthlott & Hunt 2000: pl. 51.1–2).
Southern Rio São Francisco caatinga element: on Bambuí limestone outcrops surrounded by caatinga or
forest with caatinga elements, c. 450–650 m, west of the Rio São Francisco, south-western Bahia and
adjacent Minas Gerais. Endemic to Eastern Brazil. Map 40.
bahia: SW Bahia, Mun. Santana, 30 km from town towards Porto Novo, 16 Jan. 1991, Taylor et al. 1431
(K, HRCB, ZSS, CEPEC); ‘Mun. Bom Jesus da Lapa, Faz. Serra Solta’ (unlocalised), 17 July 1975,
Andrade-Lima 75-8193 (IPA); l.c., W of Rio São Francisco, reported by Andrade-Lima (1977: figs 2, 6
& 9, as ‘Austrocephalocereus purpureus’); l.c., Mun. Serra do Ramalho, Serra do Ramalho, 13°31'S,
43°45'W, 2001, photographed by J. Jardim growing with Facheiroa cephaliomelana (J. Jardim 3486, CEPEC);
near the Rio Carinhanha, 1986, illustrated by Braun & Esteves Pereira (2002: 178, Abb. 185).
minas gerais: Braun & Esteves Pereira, l.c., report this species from adjacent to the Bahian border.
conservation status. Least Concern [LC]; PD=2, EI=1, GD=1. Short-list score (1×4) = 4. At one
time thought to be threatened from the destructive collection of seeds and also potentially from the
quarrying of limestone, but range and abundance now known to be much greater.
The peculiar, slender-elongate seeds of this species are assumed to be an adaptation for
wind dispersal (W. Barthlott, pers. comm.). It is closely related to M. estevesii (Buining &
Brederoo) F. Ritter, the only member of this genus occurring outside the geographical
area treated here (NW Minas Gerais to S Tocantins).
22. COLEOCEPHALOCEREUS Backeb.
Blätt. Kakteenf. 1938(6): unpaged [22] (1938). Type: C. fluminensis (Miq.) Backeb.
Including Buiningia Buxb. (1971); Coleocephalocereus subg. Buiningia (Buxb.) P. J. Braun

(1988).
Literature: Taylor (1991a: 17–19).

COLEOCEPHALOCEREUS
Tall columnar, sometimes unbranched (unless damaged), or low-growing and sometimes caespitose,
with most branches arising near the rootstock or from the decumbent stem bases; stems depressed-
globose to cylindric, often asymmetric at apex once cephalium has developed, vascular cylinder rather
to scarcely woody, cortical tissues with or without mucilage, pith sometimes chlorophyllous; ribs 9–30
or more, low and rounded to higher and triangular; areoles large to rather small, close-set. Spination
various, sometimes poorly developed; hypertrophic spines developed at stem base in some species.
Cephalium lateral, always deeply sunken, composed of bristles and wool in variable proportions. Flowers
rather small, c. 2–7.5 cm, diurnal (hummingbird-pollination syndrome) or nocturnal and odoriferous
(bat syndrome), white, greenish or magenta, tubular, pericarpel naked, narrower than tube at anthesis,
tube with few minute, naked bract-scales; perianth-segments spreading or the innermost remaining ±
erect in the diurnal species. Fruit expelled from cephalium, obovoid-clavate, red or deep pink, opening
by a small basal pore. Seeds small, 0.8–1.8 mm (pyriform in C. goebelianus), black, testa-cells flat and
smooth to strongly convex and with considerable cuticular sculpturing; hilum ± perpendicular to long-
axis, large (small and sunken in C. goebelianus).
A genus of 6 well-defined species (plus 2 heterotypic subspecies), all native to Eastern

Brazil (5 endemic) and ranging between the caatinga (3 taxa) and Mata atlântica (5 taxa)
regions, almost exclusively on or closely associated with gneiss/granite inselbergs (C.
goebelianus also on other substrates). Recent chloroplast gene sequence studies (Cowan &
Soffiatti unpubl.) have confirmed its monophyletic status and the distinctness of the
subgenera Simplex and Buiningia, the former being basal, the latter the most derived. The
3 subgenera are allopatric and can be recognized on the basis of seed-morphology,
presence/absence of stem mucilage, spination and floral pollination syndrome. Although
there is currently little disagreement over the circumscription of the genus,
Coleocephalocereus names have also been published for species here referred to Cipocereus,
Stephanocereus, Pilosocereus, Micranthocereus and Espostoopsis. Plates 53.1–55.4.
1. Flowers nocturnal, expanding fully, white within; seeds verrucose, testa-cells domed; ribs
with (rarely without) transverse folds above the areoles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
1. Flowers diurnal (often a.m.), inner perianth-segments scarcely expanding, yellow-green or
magenta; seeds smooth, testa-cells flat; ribs lacking transverse folds (Rio Jequitinhonha
drainage, NE Minas Gerais) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
2. Spines > 16 per areole, some strongly hooked in seedlings, lacking hypertrophic spination
at ground level; stem tissues non-mucilaginous (cent.-N Minas Gerais & S Bahia) . . . 4. goebelianus
2. Spines < 17 in areoles remote from the cephalium and stem base, not or scarcely hooked in
seedlings, or hypertrophic spination developed at ground level; stem tissues mucilaginous
(NE Minas Gerais & Espírito Santo southwards) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
3. Stems with long, hypertrophic spines near base . . . . . . . . . . . . . . . . . . . . . . . . 1. buxbaumianus
3. Stems lacking hypertrophic spines near base . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
4. Flowers c. 30–60 mm; cephalium bristles mostly yellowish or porrect, intermixed with
abundant whitish wool; stem 6–19-ribbed . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2. fluminensis
4. Flowers c. 19–35 mm; cephalium bristles dark brown and adpressed, not intermixed with
wool; stem 12–34-ribbed . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3. pluricostatus
5. Flowers yellow-green; seeds c. 1.35 mm; spines yellowish . . . . . . . . . . . . . . . . . . . . . . 5. aureus
5. Flowers magenta; seeds c. 1.75 mm; spines reddish brown . . . . . . . . . . . . . . . . . . . 6. purpureus
Subg. Coleocephalocereus (Nos 1–3): stems short to tall (0.5–5.0 m), branched at or above
base or solitary, tissues mucilaginous; ribs often with transverse epidermal folds above the
COLEOCEPHALOCEREUS
areoles; spines finely needle-like; flowers nocturnal, whitish at least within; seeds with
strongly convex testa-cells, these ornamented with characteristic, crown-like, cuticular
folds encircling each convex periclinal wall (SEM), hilum ± broad. NE to S Minas Gerais
and adjacent Espírito Santo (drainage of Rios Mucuri, Doce and others to the south); Rio
de Janeiro and off-shore islands of São Paulo.
1. Coleocephalocereus buxbaumianus Buining in Succulenta 53: 28–33 (1974);

Taylor in Bradleya 9: 17, adnot. (1991). Holotype: Minas Gerais, Mun. (?) Itambacuri,
1972, Horst 379 (U).
C. braunii L. Diers & E. Esteves Pereira in Kakt. and. Sukk. 36: 28–35 (1985). Holotype: Espírito Santo,
Mun. (?) Afonso Cláudio, 500–600 m, 1983, Horst, Braun (470) & E. Esteves Pereira (172) HBE 4
(KOELN, n.v.); Braun 470 (ZSS, iso.), E. Esteves Pereira 172 (UFG, iso.).
vernacular names. Mandacaru-de-topete, Facheiro-das-pedras.
Stems solitary or branched at or near base, erect from the rootstock or the basal parts decumbent on the
rock surface and the apical parts slightly inclined, to 220 × 5–13 cm. Ribs 14–36, 5–10 × 10–15 mm, ±
tuberculate, with prominent folds above the areoles, epidermis dark green, shiny. Areoles 1.5–7.0 ×
1.5–5.0 mm, 6–13 mm apart, with pale yellowish to brownish, later greyish felt, displaying very strong
indeterminate growth on basal half of stem. Spines finely needle-like, yellow, golden or partly brownish
at first, later pale greyish to black, c. 10–25 per areole above the stem base, central and radial spines
weakly differentiated, centrals (0–)3–6(–9), 8–40 mm, radials 7–18, to 28 mm, basal areoles with
abundant, sinuate, hypertrophic spines to 400 mm or more. Cephalium rather broad, sometimes as wide
as the stem, with dense, usually golden bristle-spines to 60 mm, wool inconspicuous except near stem
apex, bristle-spines blackish when old. Flowers (25–)30–75 × 23–40 mm, brownish to greenish white
when in bud; pericarpel 3–6 × 5–9 mm, with or without small bract-scales, tube 20–50 × 10–13 mm at
apex, bearing few, 2–11 × 0.2–2.0 mm bract-scales; nectar-chamber c. 7–12 × 6–7 mm; outer perianth-
segments 7–17 × 3.5–6.0 mm, whitish with reddish, brownish or greenish apex and mid-stripe, inner
segments 8–20 × 3.5–6.0 mm, acuminate, white; stamens 8–20 mm, white; style 18–60 mm, stigma-lobes
c. 8–10, to 5 mm, pale yellow to whitish; ovary locule 1–5 × 2–6 mm. Fruit 18–35 × 10–22 mm, often
somewhat laterally compressed or channelled, blood-red to shocking pink at apex, deep or paler pink
below, whitish near the basal pore, floral remnants whitish near base, darker to blackish above. Seed c.
1.3–1.8 × 1–1.4 mm.
conservation status. Least Concern [LC]; > 10 locations exist, the majority declining only rather
slowly (except in the west of its range).
Two subspecies are recognized, subsp. buxbaumianus replacing subsp. flavisetus at the
eastern edge of the species’ range, which is poorly recorded due to the inaccessibility of
many of its inselberg habitats:
1. Stems 5–8 cm diam., branching freely from the decumbent bases, forming loose, sprawling
clumps; flowers c. 25–42 × 25 mm (NE/E Minas Gerais, E of 42°W, and Espírito Santo)
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1a. subsp. buxbaumianus
1. Stems 7–13 cm diam., solitary and erect or forming small compact clusters; flowers c. 50–75 ×
30–40 mm (SE & SW Minas Gerais, W from 42°W) . . . . . . . . . . . . . . . . . 1b. subsp. flavisetus
1a. subsp. buxbaumianus

Ribs 14–22. Cephalium 4–7(–8) cm wide. Fruit quite naked, without bract-scales.
COLEOCEPHALOCEREUS
Southern humid/subhumid forest (inselberg) element: locally co-dominant with other cacti on gneissic
inselbergs/lajedos, 100–700 m, eastern Minas Gerais and western Espírito Santo (Rio Doce drainage).
Endemic to the core area within South-eastern Brazil. Map 41.
minas gerais: NE & E Minas Gerais, (?) Mun. Santa Maria do Salto, Sep. 2003, J.R. Stehmann (photos);
Mun. Carlos Chagas, c. 25 km W of Nanuque on road to Teófilo Otoni, 1994, S. Porembski et al. (K,
photos); Mun. Itambacuri, 20 km S of Teófilo Otoni on road BR 116, 14 Dec. 1990, Taylor & Zappi
(obs.); l.c. (?), July 1974, Buining & Horst in Horst 379 (U); Mun. Galiléia, 23 km along road BR 259
from São Vítor towards Galiléia, 15 Dec. 1990, Taylor & Zappi 771 (K, ZSS, HRCB, BHCB).
espírito santo: Mun. (?) Afonso Cláudio, (?) Pontões [‘Mun. Três Pontões’], July 1983, E. Esteves
Pereira 172 (UFG), Braun 470 (ZSS).
conservation status. Least Concern [LC] (1); PD=1, EI=2, GD=2. Short-list score (1×5) = 5. Five
populations are recorded above, but this is almost certainly less than the total that could be found if more
intensive field studies were undertaken. Although the vegetation around the inselbergs where this taxon
occurs has largely been destroyed, the actual habitat of the plant is less affected. However, disturbance is
likely to increase and so monitoring of its status is essential.
The presence of abundant, long, basal hypertrophic spines in this taxon may have a
moderating influence on the plant’s temperature near to the sun-baked rock surface
(Porembski et al. 1998: 115, fig. 6). It is quite variable in stature and spination, the first
collection cited above having a distinctive habit and unusual dark cephalium bristles. It
remains to be seen whether it belongs here.
1b. subsp. flavisetus (F. Ritter) N. P. Taylor & Zappi in Cactaceae Consensus Initiatives
3: 7 (1997). Holotype: Minas Gerais, Mun. Engenheiro Caldas, 1965, Ritter 1339 (U).
C. flavisetus F. Ritter, Kakt. Südamer. 1: 127–128 (1979).
C. estevesii L. Diers in Kakt. and. Sukk. 29: 201–205 (1978). Holotype: Minas Gerais, Mun. Pedra do
Indaiá, 1974, E. Esteves Pereira 66 (KOELN, n.v.; UFG, iso.).
Ribs 18–36. Cephalium to 13 cm wide. Fruit bearing minute bract-scales.
Southern humid/subhumid forest (inselberg) element: on gneissic inselbergs/lajedos, 100–1000 m,

south-eastern and south-western Minas Gerais (disjunct). Endemic to the core area within Minas Gerais.
Map 41.
minas gerais: SE Minas Gerais, Mun. Engenheiro Caldas, 10 km S of the town, 1965, Ritter 1339 (U);
Mun. São Joao do Oriente, 6 km E of the turning to Iapú on road BR 458, 13 Dec. 1990, Taylor & Zappi
754 (K, ZSS, HRCB, BHCB); Mun. (?) Caratinga, SW of the town, 1982, Horst & Uebelmann 511 (ZSS);
SW Minas Gerais, Mun. Pedra do Indaiá [‘Mun. Itapecerica, granito ao norte da cidade’], road MG 050,
June 1974, E. Esteves Pereira 66 (UFG); Mun. Carmo da Mata, F. Campos, plant cultivated at Jardim Bot.,
Fund. Zoo-Bot. de Belo Horizonte, 31 July 2001, Taylor (obs.); Mun. (?), between São Joao del Rei and
Morro do Ferro, Horst & Uebelmann 593 (Uebelmann, mss.).
PD=1, EI=1, GD=2. Short-list score (2×4) = 8. More secure in the eastern sector of its range, especially
where it occurs on larger, steep-sided inselbergs, but probably threatened in the west, where mining
operations have destroyed much of its habitat. By no means a common plant.
The eastern part of the range of this species is mostly within the drainage of the Rio
Doce in a region of relatively low rainfall (< 1000–1250 mm/yr, cf. map ‘isoietas anuais
1914–1938’ in Azevedo 1972; Nimer 1973: 40, fig. 18), where there is a mixture of
COLEOCEPHALOCEREUS
Mata atlântica and caatinga-like vegetation (Luetzelburg 1925–26, 2: 112–115). The

following two species are also found in this region, but range further south into wetter
areas. They have not yet been found truly sympatric with C. buxbaumianus, but occur in
very close proximity.
2. Coleocephalocereus fluminensis (Miq.) Backeb. in Jahrb. Deutsch. Kakteen Ges.

1941(2): 53 (1942); Taylor in Bradleya 9: 17, adnot. (1991). Type: not extant. Lectotype
(designated here): Brazil, Rio de Janeiro, Vellozo, l.c. infra, tab. 20 (1831).
Cactus melocactus Vell., Fl. flumin.: 205 (1829), Icones 5: tab. 20 (1831) non L. (1753). Cereus fluminensis
Miq. in Bull. Sci. Phys. Nat. Neérl. 1838: 48 (1838). C. vellozoi Lem. in Rev. Hort. 1862: 427
(1862), nom. illeg. (Art. 52.1). Cephalocereus melocactus K. Schum. in Martius, Fl. bras. 4(2): 215
(1890), nom. illeg. (Art. 52.1). Pilocereus melocactus K. Schum. in Monatsschr. Kakt.-Kunde 3: 20
(1893), nom. illeg. (Art. 52.1). Cereus melocactus A. Berger in Annual Rep. Missouri Bot. Gard. 16:
62 (1905), nom. illeg. (Art. 52.1). Cephalocereus fluminensis (Miq.) Britton & Rose, Cact. 2: 29 (1920).
Austrocephalocereus fluminensis (Miq.) Buxb. in Beitr. Biol. Pflanzen 44: 240 (1968).
? Leocereus paulensis Speg. in Anales Soc. Ci. Argent. 99: 116 (1925), pro parte. Type: see Zappi & Taylor
(1992b).
Coleocephalocereus paulensis F. Ritter, Kakt. Südamer. 1: 161 (1979). C. fluminensis subsp. paulensis (F. Ritter)
P. J. Braun & E. Esteves Pereira in Succulenta 74: 84 (1995). Holotype: São Paulo, Ilhabela, Ritter
1352 (U).
C. fluminensis var. braamhaarii P. J. Braun in Kakt. and. Sukk. 33: 118–121, with illus. (1982). C.
fluminensis subsp. braamhaarii (P. J. Braun) P. J. Braun & E. Esteves Pereira in Succulenta 74: 84
(1995). Holotype: Brazil, Espírito Santo, coast near Vitória, Braamhaar 1/1980 (KOELN
[‘Succulentarium’], in spirit, n.v.).
C. diersianus P. J. Braun & E. Esteves Pereira in Kakt. and. Sukk. 39: 48–53 (1988). Holotype: Brazil,
Minas Gerais, Mun. (?) Lajinha, c. 500 m, 1983, Horst, Braun & E. Esteves Pereira HBE 3 (KOELN,
n.v.), E. Esteves Pereira 171 (UFG, iso.).
Stems solitary or more commonly branched from the decumbent basal parts, erect above or inclined towards
the rock surface, to 2 m high and, including decumbent parts, > 300 × (3.5–)4.0–12.0 cm; ribs
(5–)6–16(–19), ± triangular in cross-section, to 16 × 10–30 mm, with ± well-defined folds above the areoles,
epidermis light to dark, shiny green or greyish green. Areoles 1–4 mm, to c. 8 mm apart, with white to
greyish felt at first, indeterminate growth weak. Spines mostly few, 1–16 per areole, to 40 mm, finely
needle-like, rarely stouter, whitish, yellowish or darker, central and radial spines poorly differentiated, the
former 0–3, the latter 0–13. Cephalium mostly no more than half the stem’s diameter, composed of
abundant white wool interspersed with yellow or more rarely brownish bristle-spines, these straight or
curved, to 30 mm. Flowers 30–60 × 20–40 mm; pericarpel c. 4 × 5–7 mm, tube 25–40 × 8–15 mm, with
few narrow bract-scales, the uppermost to 7 × 2.5 mm; nectar-chamber 8–17 × 4–7.5 mm; outer perianth-
segments c. 12–20 × 3.5–6.0 mm, greenish brown, purplish pink or creamy yellow, inner segments to 15
× 2.5–6.0 mm, white; stamens to 25 mm; style to 53 mm; stigma-lobes 8–15, to 6 mm, pale yellow. Fruit
15–26 × 13–20 mm, red, paler towards base. Seed 1.25–1.50 × 1.0 mm (Taylor 1991a: pl. 1.2).
conservation status. Least Concern [LC]; widespread on usually inaccessible inselbergs.
The following subspecies are distinguished:

1. Stem erect except near base, to 12 cm diam. (southwards from border region between
E Minas Gerais and SE Bahia) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2a. subsp. fluminensis
1. Stem decumbent except at the inclined apex, to 6 cm diam. (NE Minas Gerais)
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2b. subsp. decumbens
COLEOCEPHALOCEREUS
2a. subsp. fluminensis
Ribs to 16(–19), epidermis light to dark shiny green. Spines all finely needle-like, mostly pale yellow to
greyish white, rarely partly brownish. Cephalium bristle-spines slender hairlike, mostly pale.
Southern humid/subhumid forest (inselberg) element: locally dominant on gneissic inselbergs or lajedos
within the Mata atlântica and restinga zones, near sea level to c. 900 m, north-eastern (MG/BA border
region) to south-eastern Minas Gerais (border region between MG and ES/RJ) and Espírito Santo; Rio
de Janeiro and off-shore islands of São Paulo (to Ilha Queimada Grande). Map 41.
minas gerais: Mun. Nanuque, 1994, S. Porembski et al. (K, photos); Mun. Serra dos Aimorés, c. 8 km
E of Nanuque on road to state border (BA), 1994, S. Porembski et al. (K, photo); (?) Mun. Mantena, near
frontier with ES, Horst & Uebelmann 335 (Uebelmann, mss.); Mun. Conselheiro Pena, 7.5 km E of
Goiabeira on road to Cuparaque, with C. pluricostatus, 15 Dec. 1990, Taylor & Zappi 775 (K, ZSS,
HRCB) — aberrant dwarf form; Mun. Aimorés, Pedra Lorena, 5 July 1997, M.F. de Vasconcelos s.n. (K,
BHCB); Mun. Lajinha [‘ES, Lagedinho’], arredores do cemitério, July 1983, E. Esteves Pereira 171 (UFG);
Mun. Juiz de Fora, ‘perto de Pedreira’, 1999, L. Aona (UEC, K, photo); Mun. Além Paraíba, road BR
116, 5 km N of Rio Aventureiro, 18 Dec. 1990, Taylor & Zappi 793B (K, photo).
espírito santo: Mun. Nova Venécia, 1989, H.B.Q. Fernandes s.n., cult. at MBML, May 1990 (HRCB)
— somewhat intermediate with subsp. decumbens; ibid., Pedra do Elefante, Apr. 1996, S. Porembski (K,
photo.); Mun. São Gabriel da Palha, 29.5 km SE of Barra de São Francisco near road ES 080, 16 Dec.
1990, Taylor & Zappi 780 (K, ZSS, HRCB, MBML); Mun. Pancas, 11 km NW of Ângelo Frechiani on
road ES 341, 0.7 km W of bridge over Rio Pancas, 16 Dec. 1990, Taylor & Zappi 786 (K, ZSS, HRCB,
MBML) — somewhat intermediate with subsp. decumbens; Mun. Colatina, 4 km W of Ângelo Frechiani
on road ES 341, 16 Dec. 1990, Taylor & Zappi 785A (K, photo); Mun. Santa Teresa, Rio Cinco de
Novembro, H.B.Q. Fernandes 2259, cult. MBML, May 1990 (HRCB); Mun. Vitória, May 1990, Taylor
& Zappi (obs.); Mun. Conceição do Castelo, near road BR 262, 5 km E of Venda Nova, 850–900 m, 17
Dec. 1990, Taylor & Zappi (obs.); Mun. Guarapari, road ES 060, Km 30–35, Setiba–Guaraparí, near
Lagoa Corais, 28 Aug. 1987, O.J. Pereira 1013 & J.M.L. Gomes 236 (SPF); Mun. Piúma, Monte Agá,
south of the town, 20°52'23"S, 40°46'24''W, 26 Nov. 1999, Zappi et al. 469 (UEC, K).
rio de janeiro: Mun. (?), between Muriaé (MG) and Itaperuna, Uebelmann (mss.).
This subspecies is very variable in size, habit, rib number, cephalium colour etc., and
especially so near the northern limits of its range, but apart from this the variation does
not seem to show any obvious geographical pattern that would allow its division into
additional subspecies, nor does there seem to be any point in naming every local form. It
is sometimes sympatric with C. pluricostatus.
2b. subsp. decumbens (F. Ritter) N. P. Taylor & Zappi in Cact. Consensus Initiatives 3:
7 (1997). Holotype: Minas Gerais, Padre Paraíso [‘Água Vermelha’], Ritter 1340 (U).
Coleocephalocereus decumbens F. Ritter in Kakt. and. Sukk. 19: 160–161 (1968).
Stem to c. 6 cm diam.; ribs to 13, epidermis dark or grey-green. Spines stout, stiff, dark brownish at first.
Cephalium bristle-spines stiff, dark brown.
Southern humid/subhumid forest (inselberg) element: locally dominant on gneissic inselbergs and lajedos,
in the agreste/Mata atlântica transition, c. 650 m, north-eastern Minas Gerais. Endemic to the core area
within Minas Gerais. Map 41.
minas gerais: Mun. Padre Paraíso [Água Vermelha], 1965, Ritter 1340 (U); ibid., 2 km N of town on
road BR 116, 23 Feb. 1988, Supthut 8890 (ZSS); Mun. Caraí, N of Catuji, 21 Oct. 1988, Taylor & Zappi
COLEOCEPHALOCEREUS
in Harley 25403 (K, SPF); Mun. Teófilo Otoni, reported by W. Uebelmann (mss.); l.c., 1999, L. Aona,
cult. UNICAMP, Campinas, São Paulo.
conservation status. Endangered [EN B1ab(iii) + 2ab(iii)] (3); extent of occurrence = 91 km2 and
area of occupancy < 10 km2; PD=1, EI=2, GD=1. Short-list score (3×4) = 12. Endangered from urban
expansion at its extensive type locality, but range probably under-recorded, although the 3 known
populations are fragmented.
The range of this taxon in north-eastern Minas Gerais is not well understood and more
field studies are needed in the area between Padre Paraíso (MG) and the Rio Doce in
Espírito Santo (eg. municípios Nova Venécia and Pancas), where plants intermediate with
subsp. fluminensis have been observed and collected. At the type locality it is represented
by a relatively uniform population occupying an area some kilometres in extent. It
appears sufficiently distinct to warrant subspecific status at present, but may well prove to
be of lesser significance once the variation of the species as a whole is better known.
3. Coleocephalocereus pluricostatus Buining & Brederoo in Krainz, Die Kakteen,

Lfg. 46–47 (1971). Holotype: Minas Gerais, Horst & Uebelmann 245 (ZSS).
C. pluricostatus subsp. uebelmanniorum P. J. Braun & E. Esteves Pereira in Kakt. and. Sukk. 44: 150–154,
with illus. (1993). Holotype: Brazil, S Espírito Santo, c. 250 km S of the type locality of the species,
[Jerônimo Monteiro fide Uebelmann (1996)], 1991, W. & R. Uebelmann HU 1502 (UFG, n.v.).
Stem an unbranched, erect, solitary column or branched from the decumbent base, 100–500 × 7.5–11.0
cm; ribs (in seedlings 9–)12–34, 8–18 × 12–26 mm, epidermis dark green. Areoles to 2.5 mm, to c. 9
mm apart, at first with whitish to pale grey felt, soon almost naked, indeterminate growth almost lacking.
Spines (0–)4–8, to 11 mm, brownish, not very conspicuous, sometimes lacking on older stems, central
spine 0–1, radials to 8, slender. Cephalium becoming nearly as broad as the stem, composed of densely
packed, dark, shiny brown, curved-adpressed bristle-spines, wool sparse, inconspicuous. Flowers c. 19–35
× 10–22 mm; pericarpel to 2.5–3.0 × 4.5–6.0 mm; tube to c. 24 × 12 mm, bearing few, greenish, to 1.5
mm bract-scales; nectar-chamber to 9 × 12 mm; outer perianth-segments 6–7 × 2.5 mm, pale green or
yellowish, inner segments c. 7 × 2–2.5 mm, white; stamens to c. 12 mm, bearing pale yellow anthers;
style 15–19 mm, stigma-lobes 6–10, 2–3 mm, greenish white; ovary locule 1–1.5 × 3 mm. Fruit 14–25
× 8–13 mm, red, glossy. Seed 1–1.2 × 0.8–1.1 mm, testa-cells strongly convex.
Southern humid/subhumid forest (inselberg) element: on gneissic inselbergs within the Mata atlântica
zone, 100–300 m, near the eastern border of Minas Gerais and in adjacent Espírito Santo, from the region
of Barra de São Francisco southwards for some 250 km. Endemic to the core area within South-eastern
Brazil. Map 41.
minas gerais: Mun. (?) Mantena, ‘im Osten des Staates Minas Gerais’ [10 km from frontier with
Espírito Santo (Uebelmann, mss.)], June 1968, Horst & Uebelmann 245 (ZSS); Mun. Conselheiro Pena,
7.5 km E of Goiabeira on road to Cuparaque, with dwarf form of C. fluminensis, 15 Dec. 1990, Taylor &
Zappi 776 (K, ZSS, HRCB, BHCB).
unlocalized (mg/es): [probably near the Rio Doce, c. 1917], Luetzelburg ‘32’ (M).
espírito santo: Mun. Barra de São Francisco, 1 km from border of Minas Gerais on road ES 080, 16
Dec. 1990, Taylor & Zappi 778A (K, photo); Mun. Itarana, Jatibocas, May 1946, Brade 18537 (RB); S
Espírito Santo, Mun. Jerônimo Monteiro, 27 Oct. 1991, W. & R. Uebelmann 1502, cult. A. Hofacker,
June 2003, Taylor (obs.).
Concern, since its habitats are mostly inaccessible, steep slopes. Range and frequency of occurrence
probably under-recorded.
COLEOCEPHALOCEREUS
The caespitose, southern form described by Braun & Esteves Pereira as subsp.
uebelmanniorum, and published without precise locality information, is neither so
geographically remote as its authors claim, nor the first collection from Espírito Santo,
since Brade collected the species in the intervening area of the state in May 1946 (see
above). Their use of subspecific status, which may be justified, needs to be evaluated in
the light of studies of such intervening populations, but Herm et al. (2001: 100) illustrate
a plant with slender yellowish flowers, which seem different from those observed further
north by the present authors and might indicate specialization towards hummingbird
pollination. The first collection of the species is even earlier than Brade’s, dating from
around 1917, by Luetzelburg, who is known to have visited the border region between
Minas Gerais and Espírito Santo while travelling up the Rio Doce (Luetzelburg
1925–26). It should also be noted that the phrase ‘Holotypus: Brasilia, Espirito Santo, in
regione septentrionali ...’ following the Latin diagnosis published by Braun & Esteves Pereira
(l.c.: 152) is evidently an error (the locality is south, not north of the locus classicus).
Subg. Simplex N. P. Taylor (No. 4): stem tall (to 6.5 m), normally solitary unless damaged,
tissues non-mucilaginous; ribs with transverse epidermal folds above the areoles; spines
stout, hooked in seedlings; flowers nocturnal, white; seeds with strongly convex testa-
cells, hilum narrow, sunken. S Bahia & cent.-N Minas Gerais. Type and only species:
4. Coleocephalocereus goebelianus (Vaupel) Buining in Kakt. and. Sukk. 21:

202–206 (1970); F. Ritter ex Backeb., Kakteenlex.: 92 (1966), nom. inval. (Art. 33.2).
Type: Bahia, Serra das Almas, Luetzelburg 32 (B†). Neotype (designated here): Bahia,
Mun. Ituaçu, c. 10 km S of town towards Tanhaçu, 18 Aug. 1988, Eggli 1195 (ZSS).
Cereus goebelianus Vaupel in Z. Sukkulentenk. 1: 58 (1923).
Coleocephalocereus pachystele F. Ritter in Kakt. and. Sukk. 19: 157 (1968). Holotype: Bahia, Urandi, 1964,
Ritter 1234 (U, not found Oct. 1991; SGO 122062, lecto. designated here, cf. Eggli et al. 1995: 500).
[Cephalocereus purpureus Britton & Rose, Cact. 2: figs 25, 27 & 28 (1920); Werderm., Bras. Säulenkakt.:
58, 117 (1933), non Gürke (1908).]
[Austrocephalocereus purpureus sensu (Gürke) Backeb. in Blätt. Kakteenf. 1938(6): [22] (1938) non
Cephalocereus purpureus Gürke (1908).]
[C. lehmannianus sensu Werderm., quoad sem., ibid., 120, fig. 6 (1933), non Werderm. (1932).]
Stem solitary, normally branching only if damaged, but occasionally forming short branches at the point
where the stem develops the cephalium, columnar-erect, 150–500(–650) × 7–12 cm, tissues non-
mucilaginous; ribs c. 10 in seedlings, 14–30 or more in mature individuals, to 10 × 15 mm, somewhat
tuberculate, with transverse folds above the areoles, epidermis mid- to dark green. Areoles to 5 mm
diam., 6–12 mm apart, with sparse whitish felt. Spination variable depending on age and maturity, spines
mostly curved, pale with dark tips and bases, mostly rather stiff, not flexible: seedlings with strongly
recurved to hooked central spines 10–30 mm, juvenile plants with straight central spines 60–100 mm,
adult stems with shorter and finer spination of c. 4–8 central spines to 60 mm and 12–18 radial spines to
15 mm. Cephalium developed when immature stem overtops surrounding vegetation (ie. when the plant
is between 0.4 and 6.0 m high), becoming as broad as the stem, comprising creamy white wool
interspersed with abundant dark, curved bristle-spines to 25–50 mm. Flower c. 40–55 × 25–28 mm;
pericarpel white, c. 3–6 × 6–7 mm; tube c. 35–45, with few, 1 mm, pinkish bract-scales; nectar-chamber
to 18 × 10 mm; perianth-segments c. 10–12 × 4–5 mm, the outermost red-brown to greenish, the
remainder white; stamens to 20 mm, white; style to c. 45 mm, stigma-lobes c. 10, to 4 mm, whitish.
COLEOCEPHALOCEREUS
Fruit broadly clavate, c. 20–30 × 17–22 mm, magenta-red at apex, paler below, to white at base. Seeds
1.5 × 1.2 mm, rather narrow at the hilum; testa-cells strongly convex (Taylor 1991a: pl. 1.3).
Southern caatinga element: on quartz, gneiss/granite inselbergs (rarely on limestone outcrops) and stony
soil of the caatinga and caatinga/campo rupestre ecotone, c. 300–1000 m, southern margins of the Chapada
Diamantina and planalto de Maracás and western flank of the Serra do Espinhaço, central-southern Bahia
to central-northern Minas Gerais. Endemic to the core area of Eastern Brazil. Map 41.
bahia: Mun. Ibipitanga, 5–10 km N of Ibiajara towards Novo Horizonte, 12°56'S, 42°10'W, Aug. 2003,
M. Machado (obs.); Mun. Iramaia, 40 km W of Pé-de-Serra on road to Iramaia from Maracás via Placas,
5 Feb. 1991, Eggli (obs.); Mun. Riacho de Santana, fide Uebelmann (1996): HU 1496; Mun. Rio de
Contas, by road at start of descent to Livramento, Serra das Almas, c. 1000 m, 23 Nov. 1988, Taylor
(obs.); Mun. Livramento do Brumado, 26 Oct. 1978, A. Araújo 044 (RB); Mun. Dom Basílio, 21 km S
of Livramento do Brumado, 21 Mar. 1991, G.P. Lewis (K, photos); Mun. Maracás, 5 km E of Porto
Alegre above N bank of the Rio de Contas, 300 m, 4 Feb. 1991, Taylor et al. (obs.); Mun. Ituaçu, c. 10
km S of town towards Tanhaçu, 18 Aug. 1988, Eggli 1195 (ZSS); Mun. Tanhaçu, Ourives, reported by
Ritter (1979: 123, Abb. 71–73); Mun. Brumado, 7 km N of town towards Livramento, 22 Nov. 1988,
Taylor (K, photos); Mun. Aracatu, 25 km E of Brumado on road BR 030 towards Sussuarana, 3 Feb.
1991, Taylor (K, photos); Mun. Guanambi, 9.5 km from town on road to Caitité, 25 July 1989, Zappi
171A (K, fr. & seeds); Mun. Palmas de Monte Alto, Ritter 1234 (SGO 122060, fide Eggli et al. 1995: 500);
Mun. Urandi, c. 3 km S of town above railway line, 22 July 1989, Taylor & Zappi (obs.); Mun. Vitória
da Conquista, Pradoso, Gameleira, 17 April 2003, Taylor et al. (obs.); Mun. Cordeiros, 15°4'32"S,
41°54'W, Sep. 2003, M. Machado (obs.).
minas gerais: Mun. Espinosa, 2 km N towards MG/BA border, 22 July 1989, Taylor & Zappi (obs.);
Mun. Monte Azul, Serra Geral, ‘Serra Gillette’, c. 10 km E of town, 28 Jan. 1991, Taylor et al. 1458
(HRCB, BHCB, K, ZSS); Mun. Porteirinha, 8 km S of town on road to Riacho dos Machados, 7 Nov.
1988, Taylor & Zappi in Harley 25515 (SPF, K).
Concern at present, but its habitat continues to be lost and monitoring is desirable.
A specimen labelled as Luetzelburg 32 at Munich (M) and annotated as being the type
number of Cereus goebelianus Vaupel by Werdermann cannot be considered a true
duplicate of the collection described by Vaupel, since it represents Coleocephalocereus
pluricostatus (E Minas Gerais and Espírito Santo), which is unknown from Bahia, where
the type of the former was collected, and has much smaller flowers than those described
by Vaupel, who does not appear to have studied it. As an examination of his materials has
clearly shown, Luetzelburg’s numbering and labelling of his collections was rather chaotic
and this Munich specimen should not be allowed to further destabilize the nomenclature
of the Bahian species (see below).
Until the late 1960s, this unmistakable plant was known by the misapplied name
Cephalocereus (Austrocephalocereus) purpureus (= Micranthocereus purpureus (Gürke) F. Ritter).
Vaupel’s description of the stem, ribs and spination are not representative and probably
referred to a juvenile plant or juvenile base of the stem, but the original details of
cephalium and flower clearly refer to this species. The confusion with M. purpureus began
with Britton & Rose (1920: fig. 25) and was compounded by Werdermann (1933, 1942),
who believed C. goebelianus to be synonymous, having redescribed the true M. purpureus
as C. lehmannianus in 1932. Ritter (1968) recognized the problem, but not being sure of
the precise identity of Vaupel’s name, redescribed the plant treated here as
Coleocephalocereus pachystele.
COLEOCEPHALOCEREUS
Subg. Buiningia (Buxb.) P. J. Braun (Nos 5 & 6): plants low (< 1.2 m), caespitose; stems
non-mucilaginous; ribs lacking transverse epidermal folds above the areoles; spines very
long, finely to stoutly needle-like; flowers diurnal, coloured; seeds with flat testa-cells,
hilum broad. NE Minas Gerais (drainage of Rio Jequitinhonha).
5. Coleocephalocereus aureus F. Ritter in Kakt. and. Sukk. 19: 158–160 (1968).

Holotype: Minas Gerais, Itaobim, 1964, Ritter 1341 (U).
Buiningia aurea (F. Ritter) Buxb. in Krainz, Die Kakteen, Lfg. 46–47 (1971).
B. brevicylindrica Buining, l.c. (1971). Coleocephalocereus brevicylindricus (Buining) F. Ritter, Kakt. Südamer.
1: 122 (1979). Holotype: Minas Gerais, Horst & Uebelmann HU 167 (ZSS; ZSS, iso.).
Buiningia brevicylindrica var. elongata Buining, l.c. (1971). Coleocephalocereus brevicylindricus var. elongatus
(Buining) F. Ritter, tom. cit. 122 (1979). C. elongatus (Buining) P. J. Braun in Bradleya 6: 92 (1988).
Holotype: Minas Gerais, Horst & Uebelmann HU 271 (ZSS?).
Buiningia brevicylindrica var. longispina Buining, l.c. (1971). Coleocephalocereus brevicylindricus var. longispinus
(Buining) F. Ritter, tom. cit. 122 (1979). Holotype: Minas Gerais, Horst & Uebelmann HU 167a
(ZSS?).
Stem solitary or caespitose, mostly erect, 15–115 × 6–22 cm, ovoid to cylindric, rather variable in form;
ribs 10–20, low and blunt-edged, epidermis yellow-green to slightly glaucous. Areoles 3–5 mm, white
felted, 4–11 mm apart, showing indeterminate growth near stem base. Spines golden yellow (blackish in
age and darker in seedlings), flexible, central spine(s) 1–4, to 50 × 1 mm at base, radials c. 10–15, to 15
mm, strong hypertrophic spine growth near stem base in some populations. Cephalium making the stem
strongly assymetric in low-growing populations, reaching c. half the width of the stem, composed of
abundant white wool mixed with curved, golden bristle-spines to 30 mm. Flowers 30–42 × 17–20 mm,
greenish yellow to green; pericarpel to 5 × 5 mm; tube 20–25 mm; nectar-chamber c. 5 mm; outer
perianth-segments to 10 × 3 mm, inner segments c. 8 × 2 mm, erect, expanding to give an opening of
only 2–4 mm; stamens c. 15 mm; stigma-lobes c. 5–7, whitish. Fruit red to pinkish red, 16–24 × 13–17
mm. Seed c. 1.35 mm, smooth (Barthlott & Hunt 2000: pl. 51. 3–4).
South-eastern caatinga (inselberg) element: locally dominant on gneissic inselbergs/lajedos, c. 280–910 m,

north-eastern Minas Gerais (Rio Jequitinhonha drainage and watershed with Rio Pardo). Endemic to the
core area within Minas Gerais. Map 41.
minas gerais: Mun. Águas Vermelhas, c. 15 km E of Curral de Dentro on road BR 251, 31 Jan. 1991,
Taylor et al. 1516A (K, ZSS, photos); Mun. Pedra Azul, 16 Jan. 1965, G. Pabst 8336 & E. Pereira 9447
(HB); ibid., 5 km do aeroporto, 21 Sep. 1965, A.P. Duarte 9287 (HB, RB); ibid., 3 km from town, 11
Dec. 1984, Harley et al. in CFCR 6692 (K, SPF); ibid., 8 km towards road BR 116, 18 Oct. 1988, Taylor
& Zappi in Harley 25186 (K, SPF); ibid., 10 km along road to Almenara, 19 Oct. 1988, Taylor (K, photos);
Mun. Salinas, near Baixa Grande, road BR 251, 1 km W of junction to [Amparo do] Sítio, 22 Feb. 1988,
Supthut 8874 (ZSS); l.c., a few km further E, 42°3'W, 31 May 2002, M. Machado & G. Charles (K,
photos); Mun. Medina, road BR 116, Km 87, 21 Oct. 1988, Taylor & Zappi in Harley 25402 (K, SPF);
Mun. Itaobim, 500 m W of town on N bank of Rio Jequitinhonha, 18 Nov. 1988, Taylor (K, photos);
ibid., S bank, 8 km W of road BR 116 on road BR 357 towards Itinga, 14 Dec. 1990, Taylor & Zappi
764 (K, HRCB, BHCB).
Some habitats significantly declining due to quarrying of gneiss/granite.
Very variable in habit and stem morphology. According to Marlon Machado the first and
second localities along road BR 251 cited above have already all but disappeared through
quarrying activities.
COLEOCEPHALOCEREUS
6. Coleocephalocereus purpureus (Buining & Brederoo) F. Ritter, Kakt. Südamer. 1:

128 (1979). Holotype: Brazil, Minas Gerais, 1972, Horst & Uebelmann 359 (U).
Buiningia purpurea Buining & Brederoo in Kakt. and. Sukk. 24: 121–123, with illus. (1973).
Like C. aureus, but stems to c. 80(–87) × 11.5 cm; ribs 12–18, to c. 15 × 23 mm, epidermis mid- to dark
green; areoles to 5 mm, to 15 mm apart; spines yellow and reddish brown to red, the largest central
directed downwards, to 70 × 1 mm at base; cephalium with wool and bristle-spines coloured like those
of the stem; flowers magenta, 37–48 × 20–27 mm; style reddish above, stigma-lobes 4–5, 2.5 mm,
reddish; fruit red, 16–25 × 11–17 mm; seed c. 1.75 mm (Taylor 1991a: pl. 1.4).
South-eastern caatinga (inselberg) element: on gneissic inselbergs and lajedos in caatinga, c. 240–300 m,
near the Rio Jequitinhonha, north-eastern Minas Gerais. Endemic to the core area within Minas Gerais.
Map 41.
minas gerais: Mun. Itinga, 4 km E [corrected from ‘W’ by W. Wayt Thomas, in litt., 29.02.2000] of
town on road BR 367, 15 Feb. 1988, W.W. Thomas et al. 5972 (SPF, MBM, NY), 19 Nov. 1988, Taylor
& Zappi in Harley 25532 (K, SPF).
conservation status. Critically Endangered [CR C2b] (4); area of occupancy < 10 km2; PD=2,
EI=1, GD=1. Short-list score (4×4) = 16. Critically Endangered, since the principal habitat is close to a
road and may be visited by collectors for plants and seed. It is on privately owned agricultural land whose
future must be regarded as uncertain and it is reasonable to project decline in habitat quality and numbers
of individuals. The owner of the property has recently begun cutting the forest at the top of the inselberg
(M. Machado, pers. comm.).
The range of this species deserves further investigation, but is clearly much more
restricted than that of its sister species, C. aureus, and is currently known only from the
original locality cited above and from 3 very small, recently discovered subpopulations
within a few kilometres of it (M. Machado and I. Ribeiro de Andrade, pers. comms.).
Uebelmann (1996) states it is from Itaobim, which is here assumed to refer to the Itinga
locality, since the former is the closest town of any size and is the likely starting point for
the short journey to Itinga. However, it should not be difficult to determine the extent
of the species’ range more precisely, working eastwards from the known localities along
the Rio Jequitinhonha valley.
23. MELOCACTUS (L.) Link & Otto
in Verh. Vereins Beförd. Gartenbaues Königl. Preuss. Staaten 3: 417 (1827), nom. cons.
Type: Cactus melocactus L., typ. cons. Cactus [unranked] Melocactus L., Gen. Pl., ed. 5: 210
(1754).
vernacular names. Cabeça-de-frade, Coroa-de-frade.
Literature: Taylor (1991a).

Stem unbranched (unless damaged), highly succulent, scarcely woody except at extreme base, depressed-
globose to cylindric, vegetative part rarely to 50 cm, with 7–21 vertical ribs, mucilage absent to abundant;
areoles small to large, indeterminate growth absent. Spines (3–)4–21 or more per areole, usually only
weakly differentiated into central and radial series, sometimes strongly curved to hooked at apex in
seedlings. At reproductive maturity stem apex converted into a terminal cephalium of bristles and dense
MELOCACTUS
trichomes, vegetative growth ceasing; cephalium remaining small or sometimes attaining 30 cm in length,
to 12 cm diam. Flowers short lived, expanding mid- to late afternoon, small (15–29 × 4–18 mm), tubular,
naked, only the pink, magenta or red perianth and sometimes uppermost part of tube exserted from the
cephalium (rarely cleistogamous); pericarpel very small, conspicuously narrower than the swollen nectar-
chamber at base of tube; perianth-segments tapered to linear, arranged in 2–4 series; lowermost stamens
with filaments enlarged at base to protect nectar-chamber, anthers minute, < 1 mm; style and 4–8 stigma-
lobes very slender, equalling stamens or long-exserted and conspicuous. Fruit short- to elongate-clavate
(c. 10–45 × 5–12 mm), naked except for the pale floral remains attached at apex, white, pink, magenta
or reddish, expressed from the cephalium and sometimes falling down from the plant, funicular pulp often
becoming liquid, translucent. Seeds small (0.85–1.75 mm), blackish, globose to ovoid; testa-cells few,
large, flat to strongly convex or almost pointed (for SEM micrographs of all species treated below, except
M. lanssensianus, see Taylor 1991a).
One of the most widely distributed genera of Cactaceae, comprising at least 34 species,
ranging from South-eastern and North-eastern Brazil (16–17 spp.) and the Llanos-
Amazonian region (4 spp.), northwards to the Caribbean (to N Cuba) (10 spp.), and W
to the Andes and Central America (S Peru to W Mexico) (5 spp.). Braun & Esteves
Pereira (2002: 79) indicate that Melocactus also occurs in the Northern Brazilian state of
Tocantins, but it is unclear which species this could be. The greatest concentration of taxa
and centre of diversity is in Eastern Brazil (especially Bahia), and 18 out of the total of 22
species and heterotypic subspecies recognized here are endemic to the core area. The
genus is characteristic of the caatingas-agrestes and Northern campos rupestres, only Melocactus
violaceus occurring in sand-dunes of the Mata atlântica and M. ernestii being occasional on
rocks above brejo forest (cf. Porembski et al. 1998: 116).
Karyotype information has been published in two papers by Das et al. (1998a & b), but
is not included below, since no permanently preserved voucher materials were cited and
considerable doubt must exist concerning the identities of the taxa studied, which were of
unstated origin from a living collection (the unreliable identity of most living plants offered
from commercial sources is well known and compounded by the widespread occurrence
of man-made hybrids). A more recent study (Resende 2002, Assis et al. 2003), based on
wild-collected material identified by Marlon Machado, indicates that at least two of the
species exist in both diploid and tetraploid races and that tetraploidy is the norm amongst
the 11 Brazilian taxa counted. Such studies, if expanded, may be valuable to enable an
understanding of why some sympatric taxa produce hybrids, while others do not.
Various species occur sympatrically and sometimes form hybrid swarms. The key below
does not attempt to account for plants of hybrid origin, which include (cf. Taylor 1991a):
No. 1a No. 10, 1b 14, 1b 15, 2a 10, 2a 11, 2a 13, 8 11, 9a 11, 11
14, 13 14, 14 15.
For all taxa, fruit, seed and edaphic data are essential for precise identification. Plates
56.1–64.4.
1. Fruit entirely white, white but very pale pink at apex, or pale lilac-pink to pink and only
10–20 mm long . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
1. Fruit red or pinkish magenta, at least at apex, 10–45 mm long . . . . . . . . . . . . . . . . . . . . . . . . . 13
2. Stem lacking mucilage; fruit length 1.5–2.0 × diam. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
2. Stem with at least some mucilage in the green cortical tissues or highly mucilaginous; fruit
length 2 or more × diam. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
MELOCACTUS
3. Lowermost radial spine markedly longer than longest central spine; flower pinkish
magenta; seed to 1.35 mm, testa-cells strongly convex (granite/gneiss inselbergs & lajedos,
cent.-S Bahia) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5. deinacanthus
3. Lowermost radial spine ± equal to or shorter than longest central spine; flower red, at
least without; seed 1.35–1.75 mm, testa-cells almost flat (limestone, N Minas Gerais,
W & cent.-S Bahia) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6. levitestatus
4. Fruit white, or white and very pale pink only at apex, or pale pink and seed with almost
flat testa-cells . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
4. Fruit lilac-pink to pink; seed with testa-cells strongly convex at end opposite hilum . . . . . . . . . . 8
5. Central spine(s) > 20 × 1.5 mm (inland Bahia, limestone) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
5. Central spine lacking or < 20 × 1.5 mm (coastal sand dunes)
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16c. violaceus subsp. margaritaceus
6. Perianth-segments 0.7–1.7 mm wide; fruit to 6 mm diam. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
6. Perianth-segments 1.4–2.2 mm wide; fruit 7–9.5 mm diam. . . . . . . . . . . . . . 9. pachyacanthus
7. Stem glaucous, at least when young; cephalium apex with brownish tufts of wool, bristles
not or scarcely exserted; seeds (1.3–)1.4–1.75 mm . . . . . . . . . . . . . . . . . . . . . . . . . . . 7. azureus
7. Stem never glaucous; cephalium apex lacking brownish wool tufts, bristles usually
well-exserted; seeds 1.05–1.30 mm . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8. ferreophilus
8. Flowers cleistogamous (N & SE Pernambuco & E Paraíba) . . . . . . . . . . . . . . 12. lanssensianus
8. Flowers opening prior to fruit development . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
9. Stem pale bluish waxy-glaucous, at least when young . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
9. Stem always plain green . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
10. Stem 11–16 cm diam., pith chlorophyllous; ribs 8–12; areoles 13–20 mm apart on the ribs
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14. concinnus
10. Stem to 25 cm diam., pith white; ribs 10–22; areoles 20–40 mm apart on the ribs . . 11. zehntneri
11. Spines (3–)4–6 per areole, largest 1.2–2.0 mm thick (cent. & S Bahia, 900–1500 m)
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15. paucispinus
11. Spines > 6 per areole or only 0.5–1.0 mm thick . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
12. Radial spines 5–10(–11), almost straight, 0.5–1.5 mm thick (restinga and sandy places inland,
0–1100 m) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16. violaceus
12. Radial spines (6–)7–11, ± curved, to 2.5 mm thick (inland, especially in the caatinga)
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11. zehntneri
13. Fruit to 16 × 7 mm, dark red to base; cephalium white-woolly throughout, bristles
hidden (Bahia: Mun. Ourolândia, Mun. Morro do Chapéu & Mun. Barra) . . . 13. glaucescens
13. Fruit to 45 × 12 mm, reddish to magenta at apex, paler below; cephalium with conspicuous
bristles or white-woolly only at apex . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
14. Ribs ± rounded in cross-section (but edge often acute) or very low; stems depressed-globose
to taller than broad . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
14. Ribs triangular in cross-section; stems broader than tall . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
15. Lowermost radial spine recurved at apex, to 35 mm, central spine 1, to 22 mm; stem ±
hemispheric; ribs very low (Bahia: Mun. Vitória da Conquista) . . . . . . . . . . . . . . . 4. conoideus
15. Lowermost radial spine straight or outcurved at apex, or > 35 mm, central spines > 1,
> 22 mm, or stem and ribs not as above . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
16. Lowermost radial spine > 40 mm and < 1.5 mm thick, or stem taller than broad or lacking
mucilage and/or flowers with pinkish, long-exserted stigma-lobes; ribs 9–16 . . . . . . . . . . . . . . 17
16. Lowermost radial spine < 40 mm or > 1.5 mm thick; stem depressed to globose, with
mucilage in the green cortical tissues; stigma-lobes white, scarcely exserted; ribs 8–10
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3. bahiensis (inconcinnus)
MELOCACTUS
17. Spines 9–14 per areole, lowermost radial 40–80 mm; stem to 15 × 18 cm . . . . . . . . . . . 1. oreas
17. Spines (11–)14–21 per areole, lowermost radial 50–150 mm; stem to 45 × 35 cm . . . . 2. ernestii
18. Stem light greyish blue-green or quite glaucous; flower with c. 23 perianth-segments visible
from above; spines usually < 40 mm (S & E Bahia) . . . . . . . . . . . . . . . . . . . . 10. salvadorensis
18. Stem pale to dark green; flower with 25–33 perianth-segments visible from above; spines
to 60 mm (Pernambuco, Bahia & N Minas Gerais) . . . . . . . . . . . . . . . . . . . . . . . . . 3. bahiensis
MELOCACTUS OREAS Group (Nos 1–4)

Characteristic of exposed outcrops of gneiss/granite and sandstones or crystalline rocks/
gravels (especially quartz) between low shrubs, only very rarely occurring on limestone.
1. Melocactus oreas Miq., Monogr. Melocacti: 113 (1840). Type: ‘Habitat circa
Bahiam’ [Salvador] (assumed not to have been preserved). Neotype (Taylor 1991a:
24–25, illus.): Bahia, Mun. Itatim, Zappi 181A (K, photo.).
Cactus oreas (Miq.) Britton & Rose, Cact. 3: 227 (1922).

Melocactus rubrisaetosus Buining et al. in Succulenta 56: 161 (1977). M. oreas subsp. rubrisaetosus (Buining et
al.) P. J. Braun in Bradleya 6: 95 (1988). M. oreas var. rubrisaetosus (Buining et al.) Rizzini, Melocactus
no Brasil: 52–53 (1982), nom. inval. (Art. 33.2). Holotype: Bahia, Mun. Milagres, Horst 137 (U
531282?).
M. oreas var. submunitis Rizzini, tom. cit.: 54, fig. 22 (1982), nom. inval. (Art. 37.1).
[M. oreas subsp. bahiensis (Britton & Rose) Rizzini, tom. cit.: 52–53 (1982), excl. typ. M. oreas var.
bahiensis (Britton & Rose) Rizzini, l.c., fig. 23 (1982), excl. typ.]
[M. conoideus sensu Rizzini, tom. cit.: figs 22 & 34, above right (1982) non Buining & Brederoo (1973).]
Stem depressed-globose to elongate, 8.5–15.0(–35.0) × 10–18 cm, mid- to dark green; ribs 10–16, rather
variable in height, to 40 mm broad near stem base, always ± rounded in cross-section, but edge often
acute; areoles 6–9 × 4–6 mm, 10–18 mm apart on the ribs. Spines 9–14, horn-coloured, pale brown or
dull reddish brown, acicular, central(s) 1–4, lower 27–45 mm, radials 8–11, lowermost 40–80 mm,
slender and flexuous, rarely stout and stiff, outcurved or straight. Cephalium to 12 × 4–8 cm, bristles dark
red-brown. Flowers light to dark pink-magenta, 17–22 × (7–)9–10 mm. Fruit red at apex, magenta to
pinkish below, fading to almost or quite white at base, elongate clavate, 14–28 × 5–9 mm. Seed nearly
globose, 0.85–1.20 × 0.8–1.15 mm, testa-cells convex.
conservation status. Least Concern [LC]; known from > 10 locations, but requires regular
monitoring in view of its limited range.
This endemic NE Brazilian species is divisible into the following subspecies:
1. Ribs 10–13; stem usually depressed (N Chapada Diamantina, Bahia, 700–1000 m)

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1b. subsp. cremnophilus
1. Ribs 12–16; stem depressed to elongate (E Bahia < 500 m) . . . . . . . . . . . . . . . . 1a. subsp. oreas
1a. subsp. oreas

Stem subglobose to elongate, 10–15(–35) × 10–18 cm; ribs (11–)12–16, to 15–25 mm high; areoles 9 ×
4–6 mm. Spines 9–14, central(s) 1–4, lower 35–45 mm, radials 8–10, lowermost to 40–80 mm.
Chromosome number: 2n = 44 (Assis et al. 2003).
Eastern caatinga element: on more or less exposed, horizontal, granite/gneiss lajedos and arenitic rocks
MELOCACTUS
in caatinga, within the drainage of the Rio Paraguaçu, at 300–500 m, eastern Bahia. Endemic to Bahia.
Map 42.
bahia: Mun. Macajuba, 2.5 km N of town towards Baixa Grande, 9 Feb. 1991, Taylor et al. 1605A (K,
photos); Mun. Ipirá, 42.5 km E of Baixa Grande, 9 Feb. 1991, Taylor et al. 1609C (K, photos); Mun. Rui
Barbosa, quarry 1.5 km SW of town by road to Morro das Flores, 7 Feb. 1991, Taylor et al. 1565A (K,
photo); Mun. Itaberaba, 9.5 km from Rui Barbosa on road to village of Alagoas, 7 Feb. 1991, Taylor et
al. 1578C (K, photo); Mun. Iaçu, 5.5 km W of the town, 7 Feb. 1991, Taylor et al. 1580A (K, photos);
Mun. Itatim, 7 km N of Milagres beside road BR 116, 27 June 1989, Zappi 181A (K, photo); Mun.
Milagres, Horst 137 (U 531282, ZSS), K. Preston-Mafham s.n. (K, photo); Mun. Nova Itarana, 34 km S
of Milagres on road BR 116, 27 July 1989, Zappi 180A (K, photo).
Concern at present, but further modification of its habitats (eg. quarrying) and collection (sold for
medicinal purposes) is likely and it requires monitoring in view of its relatively restricted distribution.
The name M. oreas Miq. is here employed (subsp. oreas) for a plant having a relatively
limited, lowland distribution and stems with up to 16 ribs, the high rib-count being the
only detail in Miquel’s original description that positively excludes its application to the
more common and widespread M. ernestii Vaupel, which has up to 13(–15) ribs only and
ranges into higher and more humid areas (including outcrops in brejo forest), but is
otherwise much more variable.
Subspecies oreas has been observed sympatric with M. ernestii and M. salvadorensis (all 3
were found together at Iaçu). It will occasionally hybridize with the latter. It may also
hybridize or intergrade with M. bahiensis in north-eastern Bahia.
1b. subsp. cremnophilus (Buining & Brederoo) P. J. Braun in Bradleya 6: 95 (1988).

Holotype: Bahia, Mun. Morro do Chapéu, Horst 223 (ZSS, apparently never deposited;
U 531296, lecto. designated here).
Melocactus cremnophilus Buining & Brederoo in Cact. Succ. J. (US) 44: 3, with illus. (1972).
Like subsp. oreas, but stem globose to strongly depressed-globose, usually rather small, 8.5–12.0 ×
13.5–14.5 cm; ribs 10–13, often very low; areoles to c. 6 × 5 mm. Spines 9–13, central(s) 1–4, 27–40 ×
1 mm, lower slightly upcurved, radials 8–9, straight to slightly outcurved, lowermost to 55–70 × 1 mm,
more strongly outcurved. Chromosome number: 2n = 44 (Assis et al. 2003).
Caatinga/Northern campo rupestre (Chapada Diamantina) element: on ± exposed, usually horizontal,

crystalline/sandstone and granitic rocks, 700–1000 m, northern part of the Chapada Diamantina, Bahia.
Endemic to Bahia. Map 42.
bahia: Mun. Jacobina, 21 km W of Jacobina on road BR 324 to Lajes (Lages), 16 July 1989, Zappi 131B
(K, frs in spirit, photos); Mun. Morro do Chapéu, c. 8 km NW of town on road to Várzea Nova, M.
Machado (obs.); l.c., c. 5 km SE of town, 2 Aug. 2002, Taylor (K, photos); l.c., 19 km E of town beside
road BA 052, 24 Dec. 1988, Taylor & Zappi in Harley 27385 (K, SPF, CEPEC), Horst 223 (U 531296);
Mun. Seabra, Serra da Água de Rega, 25 km N of Seabra on road to Água de Rega, near waterfall, 25
Feb. 1971, Harley s.n. (K, photo); l.c., 11.5 km W of Seabra on road BR 242, 18 July 1989, Zappi 143B
(K, photo).
conservation status. Least Concern [LC] (1); area of occupancy estimated to be < 2000 km2; PD=1,
EI=1, GD=1. Short-list score (1×3) = 3. No serious deterioration to its habitats is known to be occurring.
The habitat to the east and north-east of Morro do Chapéu is crossed by roads and may be visited by
tourists, but the rock outcrops involved are sufficiently extensive to ensure its survival.
MELOCACTUS
Subspecies cremnophilus is isolated from subsp. oreas by a zone of Atlantic Forest on the
eastern flank of the Chapada Diamantina, Bahia. It has been found sympatric with M.
bahiensis, M. concinnus and M. paucispinus, and will hybridize with the last two taxa. The
population previously identified as belonging to this subspecies from Pernambuco (Mun.
Caetés, 4 km NW of the town, then 0.5 km NE towards a fazenda, 13 Feb. 1991, Taylor
& Zappi 1627B, K, photos) requires further study (cf. Taylor 1991a).
2. Melocactus ernestii Vaupel in Monatsschr. Deutsch. Kakteen-Ges. 30: 8 (1920);

Buining in Krainz, Die Kakteen, Lfg 58 (1974). Type: Bahia, Mun. Barra da Estiva (?),
Ule (photos, see Vaupel, l.c. and Ule 1908).
M. oreas Miq. subsp. ernestii (Vaupel) P. J. Braun in Bradleya 6: 95 (1988).
M. erythracanthus Buining & Brederoo in Cact. Succ. J. (US) 45: 223 (1973). M. oreas f. erythracanthus
(Buining & Brederoo) P. J. Braun in Bradleya 6: 95 (1988). Holotype: Bahia, Mun. Morro do
Chapéu, Horst 220 (U 531270).
M. azulensis Buining, Brederoo & S. Theun. in Kakt. and. Sukk. 28: 154 (1977). M. oreas f. azulensis
(Buining et al.) P. J. Braun in Bradleya 6: 95 (1988). Holotype: Minas Gerais, Mun. Pedra Azul, Horst
168 (U 531281?).
M. longispinus Buining, Brederoo & S. Theun. in Succulenta 56(6): 137, with illus. on journal cover
(1977). M. oreas var. longispinus (Buining et al.) P. J. Braun in Bradleya 6: 95 (1988). Holotype: Bahia,
Mun. Iaçu, Horst 435 (U, not found Apr. 1989; K, lecto. designated here).
M. nitidus F. Ritter, Kakt. Südamer. 1: 139 (1979). Holotype: Minas Gerais, Mun. Itaobim, Ritter 1210
(U 531257).
M. interpositus F. Ritter, Kakt. Südamer. 1: 140 (1979). Holotype: Bahia, Mun. Iaçu, Ritter 1207 (U
531258).
[M. oreas sensu auctt. pro parte et Buining in Krainz, Die Kakteen, Lfg. 55–56 (1973), non Miq. (1840).]
Stem with or without mucilage, subglobose to shortly cylindric, 9–45 × 7–22(–35) cm, light yellow-
green to dark green; ribs 9–13(–14), to c. 30 mm high and 70 mm broad near stem base, ± rounded in
cross-section but sometimes with an acute edge; areoles to 13 × 9.5 mm, to 28 mm apart. Spines
(11–)14–21, red and yellow banded or dull reddish to brownish, centrals (3–)4–8, lower somewhat
upcurved or straight, to 32–90 × 1.3 mm, radials 7–13, lowermost to 45–150 × 1.3 mm, straight or
outcurved. Cephalium to 18 × 8 cm, composed of pale to bright pinkish red bristles, sometimes covered
by white wool at apex. Flower 19.5–29.0 × 9–18 mm; stigma-lobes white to reddish, sometimes exserted.
Fruit deep pink to red at apex, rather variable in size, 14–45 × 5–12 mm. Seed 1.05–1.35 × 0.8–1.1 mm,
testa-cells convex.
1. Stigma-lobes not or scarcely exserted, white; green stem tissues mucilaginous, ribs 10–13(–15);
central spines 1–4(–6) (NE Brazil & NE Minas Gerais) . . . . . . . . . . . . . . . . . . 2a. subsp. ernestii
1. Stigma-lobes exserted, often pinkish red; stem lacking mucilage or with traces in the green
cortical tissues, ribs 9–11; central spines 4–8 (S Bahia & N Minas Gerais: between the Rio São
Francisco and 42°W) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2b. subsp. longicarpus
2a. subsp. ernestii

Stem mucilaginous at least in the green cortical tissues, subglobose to shortly cylindric, 10–45 × 14–22(–35)
cm; ribs 10–13(–15), to c. 30 mm high and 40–70 mm broad near stem base; areoles to 10 × 9 mm. Spines
MELOCACTUS
(11–)14–18, centrals usually 4(–6), lower somewhat upcurved, to 40–90 × 1.3 mm, radials 7–13,
lowermost to 50–150 × 1.3 mm, outcurved. Flower 19.5–25.0 × 9–15 mm; stigma-lobes white, not
exserted. Chromosome numbers: 2n = 22, 44 (Assis et al. 2003).
Eastern caatinga element: locally dominant on exposed crystalline/sandstone rocks and especially gneissic
inselbergs, including outcrops in brejo forest, 50–1190 m, southern Paraíba, eastern Pernambuco, western
Alagoas, western Sergipe, Bahia (E of 42°W) and north-eastern Minas Gerais. Reported from Ceará and
Rio Grande do Norte by Luetzelburg (1925–26) and likely to occur there. Endemic to Eastern Brazil.
Map 42.
paraíba: Mun. Esperança, 16 Nov. 1920, Luetzelburg 15000 (M); Mun. Pocinhos, July 1921, Luetzelburg
12602 (M); between Soledade and Campina Grande, Nov. 1920, Luetzelburg 51, 57B (M); Mun.
Maturéia, Serra de Teixeira, Pico do Jabre, 1994, M.F. Agra et al. (JPB).
pernambuco: Mun. Sertânia, Serra do Pinheiro, reported by Andrade-Lima (1966: 1458, as M. oreas);
l.c., 13 km W of Arcoverde, 12 Feb. 1991, Taylor & Zappi 1624 (K, ZSS, HRCB, UFP); Mun. Brejo da
Madre de Deus, 9 Feb. 1993, F.A.R. Santos 65 (HUEFS); Mun. Pesqueira, E of Cimbres, ‘alto da serra’
on road between Pesqueira and border with Paraíba state, 11 Feb. 1995, Taylor (K, photos); Mun.
Caruaru, 9 km NE of town, 8°14'21"S, 35°54'52''W, IPA reserve, Dec. 1999, Zappi (K, photos); Mun.
Alagoinha, 1992, F.A.R. Santos 25 (HUEFS); l.c., 4.5 km SW of the town on road to Garanhuns, 13
Feb. 1991, Taylor & Zappi 1626B (K, photos); Mun. Caetés, 4 km NW of town, 13 Feb. 1991 (observed
growing with M. bahiensis, see Taylor & Zappi 1627).
alagoas: reported from Piranhas etc. by Luetzelburg (1925–26, 1: 107).
sergipe: Mun. São Francisco de Canindé, reported as ‘M. oreas’ by Rizzini (1982: 53, figs 8 & 19).
bahia: Mun. Jaguarari, Flamengo, Ritter 1210 (SGO 122126, fide Eggli et al. 1995: 493); Mun. Paulo
Afonso, rock beside waterfall of the Rio São Francisco, 1919, Luetzelburg 92 (M); Mun. Morro do Chapéu,
21 km W of Morro do Chapéu beside road BA 052, 25 Dec. 1988, Taylor & Zappi in Harley 27394 (K,
SPF, CEPEC), 1971, Horst 220 (U 531270); l.c., c. 2 km W of Ventura, 24 Dec. 1988, Taylor & Zappi in
Harley 27388 (K, SPF, CEPEC); Mun. Candeal, beside road BR 324, 20.5 km NW of Tanquinho, 25 Apr.
1992, Taylor & Zappi (K, photo); Mun. Ipirá, 30 km E of Ipirá along road BA 052 to Feira de Santana, 2
Apr. 1976, Davidse et al. 11832 (SP, MO), 9 Feb. 1991, Taylor et al. 1611D (K, photos); Mun. Feira de
Santana, 74.5 km E of Ipirá on road BA 052, 9 Feb. 1991, Taylor et al. 1611E (K, photos); l.c., N of town,
Serra de São José, Nov. 2001, Zappi (K, photos); Mun. Cachoeira/São Félix, Barragem de Bananeiras,
12°32'S, 39°5'W, May 1980, Iscardino et al. in Grupo Pedra do Cavalo 7 (ALCB, RB); Mun. Iaçu, 1964,
Ritter s.n. (U 531258); l.c., 5.5 km W, 7 Feb. 1991, Taylor et al. 1580B (K, photos); Mun. Itatim, Morro
das Tocas, ápice, 12°43'S, 39°42'W, 31 Aug. 1996, F. França 1791, 1792, 1793, 1794 (HUEFS, K, photos)
& F. França s.n. (K, photos); l.c., 12°47'S, 39°47'W, L.P. de Queiroz & T.S.N. Sena 3193 (K, HUEFS);
Mun. Santa Teresinha, Serra da Jibóia, 7 Aug. 2002, Taylor (K, photos); Mun. Marcionílio Sousa, Machado
Portella, reported by Buining (1974b); Mun. Maracás, 6–7 km E of Porto Alegre, immediately N of the
Rio de Contas, 4 Feb. 1991, Taylor et al. 1552A (K, photos); Mun. Barra da Estiva, ‘vom Rio de Contas’
[Rio Sincorá, cf. Ule (1908)], photos in Vaupel, l.c. (1920); Mun. Iramaia, 38 km from Contendas do
Sincorá towards Maracás, 4 Feb. 1991, Taylor et al. 1550A (K, photos); Mun. Dom Basílio, 21 km SE of
Livramento do Brumado, 13°48'S, 41°47'W, 21 Mar. 1991, G.P. Lewis s.n. (K, photos); Mun. Ituaçu, c.
2 km NE of the town, 19 Aug. 1988, Eggli s.n. (ZSS, photo); Mun. Jequié, 41.5 km E of Porto Alegre,
4 Feb. 1991, Taylor et al. 1555A (K, photos); l.c., c. 30 km E of town, along the Rio de Contas, reported
by Buining (1974b); Mun. Anajé, 11 km SE of town, 19 Dec. 1988, Taylor & Zappi in Harley 25594 (K,
SPF, CEPEC); l.c., 14 km from Vitória da Conquista, 11 Apr. 1980, Chalet 32 (B).
minas gerais: NE Minas Gerais, Mun. Pedra Azul, near the town, Horst 168 (U 531281); Mun.
Itaobim, 1 km W of town, 18 Nov. 1988, Taylor & Zappi in Harley 25527 (K, SPF); l.c., 8 km from road
BR 116 towards Itinga, 9 Apr. 1983, Martinelli & Leuenberger 9228 (B).
This subspecies can be found sympatric with M. oreas, M. bahiensis, M. salvadorensis, M.

zehntneri, M. glaucescens and M. concinnus. It can be distinguished from the first-named
with difficulty, when they are not growing sympatrically, mainly because subsp. ernestii
MELOCACTUS
remains a highly variable and complex taxon. It requires much further study, which
should also embrace M. oreas, M. bahiensis and M. conoideus.
Although disputed by Braun & Esteves Pereira (2002: 132), the following may represent
a name for hybrids between M. ernestii and M. zehntneri, which have been observed in
the municípios of Poção/Jataúba (F. A. R. Santos 5 & 6, PEUFR, K, photos) and
Alagoinha (Taylor & Zappi, K, photos), Pernambuco: M. horridus Werderm. in Notizbl.
Bot. Gart. Berlin-Dahlem 12: 227 (1934). Holotype: Pernambuco, Mun. Serra Talhada,
1932, Werdermann 2934a (B, in spirit; K, photos). It has yet to be recollected at the type
locality, but the plants observed match the type well. Although the ranges of M. ernestii
and M. zehntneri overlap, they are usually separated ecologically, M. ernestii preferring
more humid rocks at higher altitudes.
2b. subsp. longicarpus (Buining & Brederoo) N. P. Taylor in Bradleya 9: 26 (1991).

Holotype: Minas Gerais, Mun. Porteirinha, Horst 149 (U 531269).
Melocactus longicarpus Buining & Brederoo in Cact. Succ. J. (US) 46: 191 (1974). M. deinacanthus subsp.
longicarpus (Buining & Brederoo) P. J. Braun in Bradleya 6: 94 (1988).
M. florschuetzianus Buining & Brederoo in Ashingtonia 2: 25 (1975). M. deinacanthus subsp. florschuetzianus
(Buining & Brederoo) P. J. Braun in Bradleya 6: 94 (1988). Holotype: Minas Gerais, Mun. Francisco
Sá / Grão Mogol, Horst 148 (U 531286).
M. mulequensis Buining & Brederoo in Succulenta 55: 46 (1976). M. deinacanthus f. mulequensis (Buining
& Brederoo) P. J. Braun in Bradleya 6: 94 (1988). Holotype: Bahia, Mun. Rio do Antônio (?), Horst
122 (U 531299).
M. montanus F. Ritter, Kakt. Südamer. 1: 141 (1979). Holotype: Bahia, Mun. Urandi, Ritter 1358 (U
531254).
M. neomontanus Van Heek & Hovens in Succulenta 63: 78, illus. (1984). Holotype: Bahia, van Heek &
Hovens 81-135 (U, not found Apr. 1989). Lectotype (designated here): van Heek & Hovens, l.c., illus.
Stem lacking mucilage or with a little in the green cortical tissues, epidermis light to dark green,
occasionally somewhat glaucous, 9–20 × 7–20 cm; ribs 9–11(–12), to 25–50 mm broad near stem base;
areoles to 13 × 9.5 mm. Spines (13–)14–21, sometimes very thinly overlaid with grey, nearly straight, but
sometimes strongly curved to hooked at apex in seedlings, centrals (3–)4–8, lower to 32–53 × 1–1.8 mm,
radials (8–)10–13, lowermost to 45–90 × 0.8–1.8 mm. Cephalium bristly or densely white-woolly at
apex. Flowers often large, 22–29 × 9–18 mm, with whitish, pink or reddish, ± exserted stigma-lobes.
Fruit pink to red at apex, elongate-clavate, 16–45 × 6.5–12.0 mm.
Southern caatinga element: on gneissic inselbergs and in other rocky places in caatinga, c. 450–950 m,
between the Rio São Francisco and 42°W in southern Bahia, and on west side of the Serra do Espinhaço
in northern Minas Gerais. Endemic to the core area of Eastern Brazil. Map 42.
bahia: Mun. Guanambi, 7 km W of Guanambi beside road BR 030, 20 July 1989, Zappi 152 (K, SPF);
Mun. Rio do Antônio, 43.5 km W of Brumado on road BR 030, 2 Feb. 1991, Taylor et al. 1523B (K,
photos), l.c. (?), 1972, Horst 122 (U 531299); Mun. Urandi, 3 km S of town, above railroad, 22 July 1989,
Zappi 160 (K, SPF).
minas gerais: Mun. Espinosa, 8 km S of town on road BR 122 to Mato Verde, 22 July 1989, Zappi
163 (K, SPF); Mun. Monte Azul, c. 8 km E of the town, Serra Geral, ‘Serra Gillette’, 28 Jan. 1991, Taylor
et al. 1460 (K, ZSS, HRCB, BHCB); Mun. Porteirinha, 8 km S of town on road to Riacho dos
Machados, 7 Nov. 1988, Taylor & Zappi in Harley 25514 (K, SPF); Mun. Francisco Sá / Grão Mogol,
vicinity of Barrocão, 950 m, Horst 148 (U 531286).
Concern at present, but habitats are being disturbed in certain parts of its range; needs to be monitored.
MELOCACTUS
Subspecies longicarpus is sympatric with M. salvadorensis in Mun. Rio do Antônio, Bahia.
3. Melocactus bahiensis (Britton & Rose) Luetzelb., Estud. bot. Nordéste 3: 111 (1926).
Type: Bahia, Mun. Marcionílio Sousa, Rose & Russell 19935 (US, lecto. designated here;
NY, lectopara.).
Cactus bahiensis Britton & Rose., Cact. 3: 234 (1922). Melocactus oreas subsp. bahiensis (Britton & Rose)
Rizzini, Melocactus no Brasil: 52–53 (1982), tantum quoad typ. M. oreas var. bahiensis (Britton &
Rose) Rizzini, l.c. (1982), tantum quoad typ.
? M. brederooianus Buining in Succulenta 51: 28 (1972). M. inconcinnus var. brederooianus (Buining) F.
Ritter, Kakt. Südamer. 1: 137 (1979). Holotype: Bahia, Mun. Senhor do Bomfim, Buining 1001 (U
531252).
M. acispinosus Buining & Brederoo in Krainz, Die Kakteen, Lfg 62 (1975), excl. photograph of plant apex
(= M. azureus Buining & Brederoo). Holotype: Bahia, Mun. Jacobina, Horst 258A (U 531308).
? M. inconcinnus Buining & Brederoo in Kakt. and. Sukk. 26: 193 (1975). Holotype: Bahia, Mun.
Brumado (?), Buining 1003 (U 531295).
Stem globose, depressed-globose or pyramidal, 9.5–21 × 11–21 cm, pale to dark green, with mucilage at
least in the green cortical tissues; ribs 8–14, very low or up to 30 mm high, very variable in shape; areoles
6–14 × 4.5–8.0 mm, to 24 mm apart on the ribs. Spines 9–16, mostly straight, central(s) 1–4, 17–50 mm,
radials (7–)8–12, lowermost to 22–60 mm. Cephalium usually small, rarely more than 5 × 6.5–8.5 cm.
Flower pinkish magenta, 20–23 × 10–12.5 mm. Fruit reddish to magenta at apex, paler below, 17–25 ×
6.5–9.0(–10.0) mm. Seed 1.05–1.35 × 0.85–1.30 mm, testa-cells convex.
Until recently the name M. bahiensis was commonly misapplied in Brazilian literature to
the more widely ranging M. zehntneri. This misuse can be attributed to Werdermann,
whose incorrect determinations of herbarium material preserved at Recife (IPA) seem to
have misled two generations of botanists and ecologists.
The true M. bahiensis is highly variable and its taxonomy is poorly understood, as are its
relationships with other species in the M. OREAS Group. The following infraspecific taxa
are provisionally recognized:
1. Central spine(s) 1–4, the lower and largest usually > 25 mm, lowermost radial spine 24–60 mm . . . 2
1. Central spine 1, to 25 mm, lowermost radial spine 22–32 mm (N Bahia)
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3a(ii). subsp. bahiensis f. acispinosus
2. Ribs rounded to somewhat acute but scarcely triangular in cross-section, or lowermost radial
spine > 40 × 1.5 mm (Pernambuco & NE to S Bahia) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
2. Ribs acute and triangular in cross-section; lowermost radial spine to 40 × 1.5 mm (S edge of
Bahia to central Minas Gerais: Diamantina) . . . . . . . . . . . . . . . . . . . . . 3b. subsp. amethystinus
3. Ribs c. 8–10, sharply acute at edge, to 60 mm diam.; areoles 8–14 mm; spines to 2 mm
diam. (S Bahia) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3a(iii). subsp. bahiensis f. inconcinnus
3. Ribs 10–13, ± rounded, to 45 mm diam.; areoles to c. 8 mm; spines to c. 1 mm diam.
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3a(i). subsp. bahiensis f. bahiensis
3a. subsp. bahiensis

Stem globose to depressed-globose or pyramidal, 9.5–21 × 11–21 cm; ribs 8–13, high and acute to low
and rounded, to 60 mm broad; areoles to 8–14 × 6.5–8.0 mm (to only 6 × 4.5 mm in f. acispinosus). Spines
MELOCACTUS
9–16, pale reddish brown overlaid with grey, central(s) 1–4, lower 17–50 × 1–2 mm, porrect, straight or
upcurved, radials (7–)8–12, lowermost 22–60 × 1–2 mm. Seed 1.05–1.20 × 0.8–1.05 mm (to 1.35 × 1.3
mm in f. acispinosus).
Eastern caatinga/Northern campo rupestre element: on more or less exposed crystalline rock formations
(quartzitic-arenitic, granite/gneiss), quartz gravel etc., rarely on limestone, campo rupestre /caatinga,
380–1300 m, northern and eastern Pernambuco, to c. 14°50'S in Bahia. Endemic to the core area within
North-eastern Brazil. Map 43.
pernambuco: Mun. Triunfo, Lajedo do Piripiri, 1140 m, 22 Nov. 1992, F.A.R. Santos 59 (HUEFS,
PEUFR); Mun. Taquaritinga do Norte, road BR 104, c. 3 km S of turning to the town, 8 Feb. 1993,
F.A.R. Santos 63 (PEUFR 13225, K, photo); Mun. Serra Talhada, ‘Serra da Carnaubeira’ [Serra do
Morcego ou da Carnaúba?], on rock at top of the Serra, 22 May 1971, Heringer, D. Andrade-Lima et al.
842 (IPA 19729, UB); Mun. Caetés, 4 km NW of the town, then 0.5 km NE towards a fazenda, 13 Feb.
1991, Taylor & Zappi 1627 (K, ZSS, HRCB, PEUFR).
bahia: Mun. Senhor do Bomfim, Andorinha, Buining s.n. (U 531252); Mun. Monte Santo, above
Monte Santo by path to santuário, 14 July 1989, Zappi 123 (K, SPF); l.c., near village of Pedra
Vermelha, 14 July 1989, Zappi 123A (K, fruit in spirit, photos); Mun. Itiúba/Cansanção, 15 km E of
Itiúba on road to Cansanção, 19 Feb. 1974, Erskine 91, cult. R.B.G. Kew, 1980 (K, photo); Mun.
Jacobina, 2 km E of Jacobina, valley slope N of river, 16 July 1989, Zappi 131 (K, SPF), 1972, Horst
258A [‘258’] (U 531308, ZSS AA-20-255); (?) Mun. Santa Bárbara, 43.5 km from Feira de Santana on
road BR 116 to Serrinha, 13 July 1989, Zappi 119 (K, SPF); Mun. Rui Barbosa, Serra de Rui Barbosa,
Taylor et al. 1582 (K, ZSS, HRCB, CEPEC); Mun. Marcionílio Sousa, Machado Portella, on tops of
nearly barren hills, 1915, Rose & Russell 19935 (US, NY), Horst 388 (U 531261); Mun. Rio de Contas,
between Rio de Contas and the Pico das Almas, Fazenda Brumadinho, 27 Nov. 1988, Taylor in Harley
25553 (K, SPF, CEPEC); l.c., between Faz. Brumadinho and Faz. Silvino, near the Pico das Almas,
1300 m, Dec. 1988, Taylor (K, photos); Mun. Paramirim, between Caraíbas and Paramirim, 30 Nov.
1988, Taylor in Harley 27029A (K, frs in spirit, photos); Mun. Livramento do Brumado, just N of
Livramento, rocks by waterfall of Rio Brumado, 13°38'S, 41°50'W, Jan. 1974, Harley et al. 15345 (K,
U, CEPEC, IPA 22990); Mun. Jussiape, 25 km from Rio de Contas towards the town, 25 Nov. 1988,
Taylor & Zappi in Harley 25549 (K, SPF, CEPEC); Mun. Brumado (?), Rio São João, between Brumado
and Livramento, 1973, Buining 1003 (U 531295); Mun. Ituaçu, c. 2 km SW of the town, exposed
limestone, 18 Aug. 1988, Eggli s.n. (ZSS, photo); Mun. Tanhaçu, E of Sussuarana, Serra Escura, 18 April
2003, Taylor et al. (K, photos); Mun. Vitória da Conquista, Pradoso, Gameleira, 17 April 2003, Taylor
et al. (K, photos).
Two distinctive local Bahian taxa of this complex are distinguished in the key above.
Forma acispinosus (Buining & Brederoo) N. P. Taylor (1991a: 28) is represented by the
collections cited from the municípios of Senhor do Bomfim, Itiúba and Jacobina, and f.
inconcinnus (Buining & Brederoo) N. P. Taylor (1991a: 30) by those from Paramirim,
Livramento do Brumado, Brumado and Ituaçu. However, recently (April 2003), the last-
named has been found by the authors sympatric with typical M. bahiensis f. bahiensis near
Vitória da Conquista and east of Sussuarana (Serra Escura) and would appear to be a
distinct species, distinguishable by the characters given in the key, above.
Subspecies bahiensis can be found sympatric with M. oreas, M. ernestii, M. zehntneri, M.
concinnus and M. violaceus subsp. ritteri.
3b. subsp. amethystinus (Buining & Brederoo) N. P. Taylor in Bradleya 9: 30 (1991).

Holotype: Bahia, Mun. Caitité, Horst 270 (U 531291).
MELOCACTUS
Melocactus amethystinus Buining & Brederoo in Krainz, Die Kakteen, Lfg. 50–51 (1972).
M. lensselinkianus Buining & Brederoo in Succulenta 53: 68 (1974). Holotype: Minas Gerais, Mun.
Itaobim, Horst 381 (U 531275).
M. griseoloviridis Buining & Brederoo in Kakt. and. Sukk. 25: 98 (1974) (‘grisoleoviridis’). Holotype: Minas
Gerais, Mun. Itamarandiba, Horst 405 (U 531274).
M. glauxianus Brederoo & C. A. L. Bercht in Succulenta 63: 55, illus. (1984). Holotype: Minas Gerais,
Horst 382 (U, not found Apr. 1989). Lectotype (designated here): Brederoo & Bercht, l.c., illus.
M. ammotrophus Buining, Brederoo & C. A. L. Bercht in Succulenta 63: 33 (1984). Holotype: Minas
Gerais, Mun. Grão Mogol, Horst 353 (U 531306).
Stem depressed-globose to pyramidal, 10–13 × 13–19 cm; ribs 9–14, usually triangular in cross-section,
edge acute; areoles 6–10 × 5–7 mm. Spines 9–13, all ± straight, brown to reddish or horn yellow, thinly
overlaid with grey, central(s) 1–4, lower 15–33 × 1–1.8 mm, radials 8–10, lowermost 24–40 × 1.2–1.5
mm. Seed 1.1–1.3 × 0.95–1.15 mm. Chromosome number: 2n = 44 (Assis et al. 2003).
Eastern caatinga/Northern campo rupestre element: under and between shrubs on mainly crystalline
(quarzitic/arenitic) rock formations in the Serra do Espinhaço region and on exposed granite/gneiss
further east, campo rupestre/caatinga, 300–1000 m, southern Bahia to central Minas Gerais. Endemic to the
core area of Eastern Brazil. Map 43.
bahia: Mun. Caetité, Brejinho das Ametistas, 26 July 1989, Zappi 175 (K, SPF), Horst 270 (U 531291);
Mun. Licínio de Almeida, 10 June 2002, G. Charles (obs.); l.c. (?), E of Urandi, reported by Braun
(1988a: 208–209, as ‘HU 535’); Mun. Presidente Jânio Quadros, reported by Rizzini (1982: 75, fig. 39
above left).
minas gerais: Mun. Taiobeiras, between Taiobeiras and Curral de Dentro, by bridge over Rio Atoleiro,
17 Oct. 1988, Taylor & Zappi in Harley 25148 (K, SPF); Mun. (?), c. 15 km W of Curral de Dentro on
road BR 251, 13 Aug. 1988, Eggli s.n. (ZSS, photo), 31 Jan. 1991, Taylor et al. 1516C (K, photos); Mun.
Salinas, Fruta de Leite, reported by Vandecaveye & Keirse in Succulenta 63: 213 (1984); Mun. Medina,
reported by Rizzini (1982: 74); Mun. Grão Mogol, Córrego da Bonita, 16°35'S, 42°54'W, 29 May 1988,
Zappi et al. in CFCR 12095 (SPF 50304); Mun. Cristália, S bank of Rio Itacambiruçú near Grão Mogol,
15 Oct. 1988, Taylor & Zappi in Harley 25073 (K, SPF); Mun. Coronel Murta, 5 km SE of town, 21 Feb.
1988, Supthut 8860 (ZSS); Mun. Itinga, c. 5 km E of town, S of road to Itaobim, 19 Nov. 1988, Taylor &
Zappi in Harley 25532A (K, frs in spirit, photo); Mun. Itaobim, c. 1 km W of town on N side of Rio
Jequitinhonha, 18 Nov. 1988, Taylor & Zappi in Harley 25526 (K, SPF), 1972, Horst 381 (U 531275); Mun.
Itinga, 1999, Machado et al. (obs.); Mun. Araçuaí, reported by Rizzini (1982: 74); Mun. Itacambira, N of
the town, 1987, Braun 844 (ZSS); Mun. Itamarandiba, near Itamarandiba, 1972, Horst 405 (U 531274);
Mun. Diamantina, Mercês, Horst & Uebelmann HU 174 (Uebelmann, mss; Eerkens 1983: 277).
This subspecies is sometimes sympatric with M. ernestii, M. zehntneri and M. concinnus, but
hybrids between them have not been observed.
4. Melocactus conoideus Buining & Brederoo in Krainz, Die Kakteen, Lfg. 55–56
(1973). Holotype: Bahia, Mun. Vitória da Conquista, Horst 183 (U 531247).
vernacular name. Cabeça-de-frade-do-Periperi.
Stem mucilaginous in the green cortical tissues, strongly depressed-globose to hemispheric, to 10 × 17

cm; ribs 11–15, very low and rounded, to 2.5 × 4 cm; areoles 6.5–7.5 × 6.5 mm. Spines dark brown
overlaid with grey, central 1, 20–22 × 1.5 mm, radials 8–11, straight to slightly recurved, lowermost
20–35 × 1.5–1.7 mm, recurved near apex. Flower pinkish magenta, c. 22 × 10 mm. Fruit lilac-magenta,
18–21 × 5–6 mm. Seed 1.05–1.25 × 0.9–1.05 mm, testa-cells strongly convex at periphery. Chromosome
number: 2n = 44 (Assis et al. 2003).
MELOCACTUS
Eastern caatinga/Northern campo rupestre element: under and between shrubs in quartz gravel, campo
sujo/cerrado de altitude, 900–1050 m, Serra do Periperi, Mun. Vitória da Conquista, south-eastern Bahia.
bahia: Mun. Vitória da Conquista, Morro do Cruzeiro above Vitória da Conquista, 26 July 1989, Zappi
179 (K, SPF), Horst 183 (K, in spirit [ex cult. Kew, Accn. No. 413-83.04910], U 531247, ZSS AA-20-73);
l.c., NW of the above, 16 April 2003, Taylor et al. (K, photos).
conservation status. Critically Endangered [CR B1ab(iii, iv, v) + 2ab(iii, iv, v)] (4); extent of
occurrence < 100 km2 and area of occupancy estimated to be < 10 km2; PD=2, EI=1, GD=1. Short-list
score (4×4) = 16. Regrettably, survival of M. conoideus in the wild continues to be severely threatened by
extraction of the quartz gravel in which it grows, and it was formerly impacted by commercial collection
for the European horticultural market (Taylor 1992b). Since June 1992 it has been listed on Appendix I
of C.I.T.E.S. In 1989, M. conoideus appeared close to extinction at the type locality in the Serra do
Periperi above Vitória da Conquista, BA, where it is sympatric with the widespread M. concinnus.
However, from the original locality other adjacent areas of similar habitat could be seen and subsequently
Brazilian cactus enthusiasts have confirmed that healthy extensions of the population exist, some of these
including 1000s of individuals. These areas are more distant from the BR 116 highway and one is now
officially protected, although the extraction of gravel and wood continues throughout. Staff at the
Universidade Estadual do Sudoeste da Bahia (Vitória da Conquista) have successfully propagated a
substantial number of plants from wild seed and hope to reintroduce the species into a securely protected
area of its original habitat in the near future. One encouraging piece of news is that it appears M. conoideus
is capable of recolonizing areas from which gravel has been extracted and may even be able to take
advantage of the reduced competition that such disturbance temporarily creates. A substantial grant to
fence the officially protected area was made available in 2003 by the British Cactus & Succulent Society.
This should improve the chances of the species’ survival considerably.
The strongly depressed habit of this species is, as in the case of M. paucispinus, assumed to
be an adaptation to minimize the effects of fire. It is very close to forms of M. bahiensis
and may ultimately be best treated as a subspecies of it, once the latter has been better
studied in the field.
A plant that could represent a southern subspecies of M. conoideus has recently been
found in Mun. Itinga, Minas Gerais, by Inês Ribeiro de Andrade. Its habitat, at 1160 m
altitude, is comparable and it differs from its Bahian relative principally in having stouter,
strongly recurved spines.
MELOCACTUS DEINACANTHUS Group (No. 5)

Restricted to gneiss outcrops in central-southern Bahia.
5. Melocactus deinacanthus Buining & Brederoo in Kakt. and. Sukk. 24: 217 (1973).
Holotype: Bahia, Horst 153 (U 531251).
Stem not mucilaginous, globose to elongate, 15–35 × 12–25 cm; ribs 10–12, acute, to 40 × 60 mm broad
near stem base; areoles to 15 × 10 mm, to 33 mm apart on the ribs. Spines 15–21, reddish brown, centrals
4–7, lower to 53 × 3 mm, radials 11–14, lowermost to 80 × 2.5 mm, somewhat recurved. Cephalium to
25 × 9 cm, with exposed bristles at apex. Flowers pinkish magenta, to 26 × 11.5 mm; stigma-lobes
whitish, not exserted. Fruit entirely white, rarely pinkish, shortly clavate, 12–22 × 6–12 mm. Seed
1.15–1.35 × 1.05–1.25 mm, scarcely narrowed or broadest at the hilum, testa-cells strongly convex.
Southern Rio São Francisco caatinga element: on gneissic inselbergs and lajedos in caatinga east of the Rio
São Francisco, 500–540 m, central-southern Bahia. Endemic to Bahia. Map 44.
MELOCACTUS
bahia: Mun. Bom Jesus da Lapa, Juá, 21 Apr. 1992, Hatschbach & E. Barbosa 56810 (MBM); l.c., 34.5
km SE of Bom Jesus on road BR 430 to Riacho de Santana, south side of road, Morro da Barriguda, 20
July 1989, Zappi 150 (K, SPF); l.c., 22 km from Riacho de Santana, 16 Apr. 1983, Leuenberger et al. 3075
(B, CEPEC); l.c., 1971, Horst 153 (U 531251); l.c., 2–10 km N of Juá towards Baixada Grande and
Favelândia, incl. Faz. Barauninha, Jan. 2003, M. Machado (K, photos).
conservation status. Critically Endangered [CR B1ab(iii) + 2ab(iii)] (4); area of occupancy estimated
to be < 10 km2; PD=3, EI=2, GD=1. Short-list score (4×6) = 24. It seems to be known from only c. 5
sites within 10 km of each other and it may be hybridising with M. zehntneri at one of these (M.
Machado, pers. comm.). Its rarity and vulnerability to commercial exploitation has caused it to be placed
on Appendix I of C.I.T.E.S., since June 1992. Part of the original locality is now a quarry.
M. deinacanthus is readily distinguished from all other members of the genus by the
combination of shortly clavate, pure white fruits and uniquely shaped seeds, which are
very broad at the hilum and with strongly convex testa-cells. It appears to have a very
limited distribution east of the Rio São Francisco and is presently known from only the
locality cited above, where it is a co-dominant element of the rupicolous vegetation.
MELOCACTUS LEVITESTATUS Group (No. 6)

Restricted to outcrops of Bambuí limestone in central-northern Minas Gerais and western
Bahia.
6. Melocactus levitestatus Buining & Brederoo in Cact. Succ. J. (US) 45: 271 (1973).
Holotype: Bahia, Mun. Bom Jesus da Lapa, Horst 397 (U 531285).
M. diersianus Buining & Brederoo in Kakt. and. Sukk. 26: 169 (1975). Holotype: Minas Gerais, Mun.
Bocaiúva, Horst 404 (U 531287).
M. securituberculatus Buining & Brederoo in Cact. Succ. J. (US) 48: 38 (1976). M. levitestatus f.
securituberculatus (Buining & Brederoo) P. J. Braun & E. Esteves Pereira in Schumannia 3: 188 (2002).
Holotype: Bahia, Mun. Iuiú, Horst 446 (U 531298).
M. warasii E. Pereira & Bueneker in Bradea 2(30): 213, with illus. (1977). Holotype: Bahia, Mun. Bom Jesus
da Lapa, R. Buenecker (HB 1827, in spirit, †). Lectotype (designated here): Pereira & Bueneker, l.c., illus.
M. rubrispinus F. Ritter, Kakt. Südamer. 1: 135 (1979). M. diersianus f. rubrispinus (F. Ritter) P. J. Braun
in Bradleya 6: 94 (1988). Holotype: Minas Gerais, Mun. Bocaiúva, Ritter 1330 (U 531267).
M. uebelmannii P. J. Braun in Kakt. and. Sukk. 36: 232 (1985). Holotype: Bahia, Mun. Brejolândia, Horst
528 (ZSS).
[M. aff. zehntneri sensu Andrade-Lima in Revista Brasil. Biol. 37: 182–191, fig. 7 (1977).]
[M. azureus sensu Rizzini, Melocactus no Brasil: 60, fig. 12 above (1982) non Buining & Brederoo (1971).]
Stem without mucilage, light grey-green to dark green, or sometimes rather glaucous and with reddened
edges to the ribs, slightly depressed or globose to cylindric, 15–68 × 14–30 cm; ribs 9–12(–15), to 5 × 8
cm; areoles to 11 × 10 mm, with abundant creamy wool at first, to 30–42 mm apart on the ribs. Spines
8–15, brownish red thinly overlaid with grey, central(s) 1–4(–6), 17–33 × 1–2.5 mm, mostly ascending,
lower up- or decurved towards apex, radials 7–10, all strongly recurved, lower 3 21–33 × 1.5–2.0 mm.
Cephalium to 18 × 7–12 cm. Flower entirely red or deep magenta within, red without, 20–27 × 6–9
mm, expanding fully or the inner perianth-segments remaining ± erect and forming a tube around the
exserted stigma-lobes. Fruit pure white or faintly pinkish at apex, shortly clavate, 12–20(–24) × 7–12 mm.
Seed 1.35–1.75 × (1.2–)1.3–1.65 mm, testa-cells flat and smooth.
Southern Rio São Francisco caatinga element: usually on elevated outcrops of Bambuí limestone amidst
high caatinga forest, c. 450–780 m, western and central-southern Bahia and central-northern Minas
MELOCACTUS
Gerais. Endemic to the core area of Eastern Brazil. Map 44.

bahia: Mun. Vanderlei, near Vanderlei, 25 km from road BR 242, 19 July 1989, Zappi 149 (K, SPF);
Mun. Brejolândia, 1985, Horst 528 (ZSS AA-58-25/26/27, AV-20-72); Mun. Santana, 28 km S of the
town, Faz. São Geraldo, path to Gruta do Padre, 15 Jan. 1991, Taylor et al. 1424A (K, photos); l.c., c. 8
km N of Porto Novo, Jan. 1991, Eggli (K, photo); Mun. Bom Jesus da Lapa, W of Rio São Francisco,
1978, Horst 456 (ZSS AV-20-36), 1972, Horst 397 (U 531285, ZSS AV-20-89/132); Mun. Iuiú, 2 km S
of town, 21 July 1989, Zappi 155 (K, SPF), 1974, Horst 446 (U 531298).
minas gerais: Mun. (?) Manga, ‘Mocambinho, Morro Solto’, 15°6'S, 43°59'W, 18 Mar. 1998, I.
Pimenta (photos); l.c., ‘Serra Azul’ [between ‘Mocambinho’ and Jaíba], 18 Mar. 1998, I. Pimenta (K,
photos); Mun. Itacarambi/Januária, vale do Rio Peruaçu, illustrated in Costa et al. (1998: Fig. 29); Mun.
Januária, c. 15 km SE of the Rio São Francisco on road BR 135, 11 Aug. 1988, Eggli s.n. (ZSS, photo);
Mun. Varzelândia, c. 10 km N of the town, 12 Aug. 1988, Eggli s.n. (ZSS, photo); Mun. Capitão Enéas,
reported by Rizzini (1982: 60, fig. 24 above); Mun. Montes Claros, 1964, Ritter s.n. (U 531250); Mun.
Bocaiúva, Eng. Dolabela, limestone outcrop in sugarcane plantation near road BR 135, 4 Nov. 1988,
Taylor & Zappi in Harley 25505 (K, SPF); Mun. Monjolos, between Rodeador and Conselheiro Mata,
18°17'27"S, 44°2'W, Aug. 2003, M. Machado (obs.).
Concern at present, but in need of monitoring in view of the potential threat from limestone quarrying,
though this is unlikely to affect the plant over the length of its considerable range.
MELOCACTUS AZUREUS Group (Nos 7–9)

Restricted to limestone outcrops (Bambuí) in northern Bahia.
7. Melocactus azureus Buining & Brederoo in Kakt. and. Sukk. 22: 101 (1971).
Holotype: Bahia, [Mun. Juçara / ‘Juassara’ fide Uebelmann (1996)], 1968, Horst 256 (U,
not found Apr. 1989). Lectotype (Taylor 1991a): colour photograph in Buining &
Brederoo, l.c.
M. krainzianus Buining & Brederoo in Krainz, Die Kakteen, Lfg 62 (1975), excl. photo of plant apex (=
M. bahiensis (Britton & Rose) Luetzelb.). M. azureus var. krainzianus (Buining & Brederoo) P. J. Braun
in Bradleya 6: 94 (1988). Holotype: Bahia, Mun. Irecê, Horst 264 (U 531300).
Stem mucilaginous in the chlorophyllous cortical tissues, depressed-globose to shortly cylindric, usually
intensely glaucous, less so when very old, 13–26 × 14–19 cm; ribs (9–)10, triangular in cross-section, to 35
× 40 mm; areoles to 11 × 7.5 mm, to 28–35 mm apart on the ribs. Spines 9–13, blackish or dark brown to
reddish overlaid with grey, some hooked in seedlings, all somewhat recurved to decurved or almost straight,
lower central 21–43 × 1.5–2.0 mm, radials 8–10, lowermost 21–40 × 1–2 mm. Cephalium to c. 12 × 7–9
cm, composed of reddish bristles and white wool, apex with brownish tufts of wool, bristles not or scarcely
exserted. Flowers pinkish magenta, 19–23 × 8–11.5 mm. Fruit entirely white, slightly pinkish at apex or
pale pink throughout, to 17 × 6 mm. Seed (1.3–)1.4–1.75 × 1.2–1.5 mm, testa-cells flat and smooth.
Northern Rio São Francisco caatinga element: on flat, exposed Bambuí limestone in caatinga at c. 450–750
m, in the region of Irecê, drainage of the Rio Jacaré and Rio Verde, central-northern Bahia. Endemic to
Bahia. Map 44.
bahia: Mun. São Gabriel, Sítio Arqueológico, 10°57'54"S, 41°41'38''W, 27 Aug. 2002, A. Cardoso
(photos); Mun. Itaguaçu da Bahia, road BA 052, just S of the Rio Verde, 14 July 2000, G. Charles
(photos); Mun. Juçara, near Juçara, 25.5 km from road BA 052, 17 July 1989, Zappi 139 (K, SPF); Mun.
Irecê, 9 km W of cross-roads with road BA 052, 18 June 1994, L.P. de Queiroz 3996 (HUEFS, K); l.c.,
beside road to Seabra, 1 km from traffic island immediately S of Irecê, 17 July 1989, Zappi 140A (K, fls
in spirit); 12 km E of Irecê, S of road BA 052 to Morro do Chapéu, 25 Dec. 1988, Taylor & Zappi in
Harley 27396 (K, SPF, CEPEC), 1972, Horst 264 (U 531300).
MELOCACTUS
conservation status. Endangered [EN B1ab(i, ii, iii, iv) + 2ab(i, ii, iii, iv)] (3); extent of occurrence
< 2000 km2 and area of occupancy < 50 km2; PD=2, EI=1, GD=2. Short-list score (3×5) = 15.
Endangered by actual or potential habitat destruction (agricultural development) at its few known
localities and not included within any protected area.
This species was observed sympatric with M. pachyacanthus subsp. viridis, q.v., but it is
doubtful whether this habitat still holds plants of either species. Near the type locality of
M. azureus there is a large population of plants that on their vegetative appearance seem to
be somewhat intermediate between this species and M. pachyacanthus (G. Charles, photos).
8. Melocactus ferreophilus Buining & Brederoo in Krainz, Die Kakteen, Lfg 52

(1973). Holotype: Bahia, Mun. Barro Alto/Souto Soares, Horst 217 (U 531260).
M. azureus subsp. ferreophilus (Buining & Brederoo) N.P. Taylor in Bradleya 9: 40 (1991).
Like M. azureus, but usually cylindric, dark to grey-green, 15–45 × 13–19 cm; areoles to 13 × 11 mm.
Spines 11–15, lower central to 30–53 × 2–2.2 mm, radials 7–11, lowermost to 40–53 × 2 mm, often
strongly decurved towards its apex. Cephalium lacking brownish woolly tufts at apex, scarcely woolly in
old plants, bristles conspicuous, ± well-exserted. Seed 1.05–1.30 × 0.95–1.10 mm.
Northern Rio São Francisco caatinga element: on raised, exposed, karstic Bambuí limestone in caatinga at
c. 520–850 m, in the drainage of the Rio Jacaré (Rio Vereda do Romão Gramacho) and tributaries,
between São Gabriel (APA de Gruta dos Brejões) and Mulungo do Morro, central Bahia. Endemic to
Bahia. Map 44.
bahia: Mun. São Gabriel, Sítio Arqueológico, 10°57'54"S, 41°41'38''W, 27 Aug. 2002, A. Cardoso
(photos); l.c., APA de Gruta dos Brejões, escarpment above Rio Jacaré, 4 Aug. 2002, Taylor (K, photos);
Mun. Morro do Chapéu, 2.5 km E of América Dourada, south bank of Rio Jacaré (Rio Vereda do
Romão Gramacho), N of road BA 052, 17 July 1989, Zappi 138 (K, SPF), 13 Jan. 1991, Taylor et al.
1408B (K, frs in spirit, photos); Mun. Barro Alto, Gameleira [near Lagoa do Boi], Horst 217 (U 531260,
ZSS TP-20-236); Mun. Mulungu do Morro, Horst & Uebelmann 606 (ZSS TP-58-86).
conservation status. Endangered [EN B1ab(iii) + 2ab(iii)] (3); area of occupancy < 50 km2. PD=2,
EI=1, GD=1. Short-list score (3×4) = 12. Endangered from potential limestone quarrying, actual
modification of the environment surrounding its habitat and in view of its very limited known area of
occupancy at 5 localities.
This taxon is here elevated to specific rank following the discovery by Aloísio Cardoso
(Administrador, APA de Gruta dos Brejões, BA) of the first population cited above,
where it grows sympatric with M. azureus without evidence of intermediates or hybrids.
At the second and third localities cited this species was found sympatric at the edge of its
habitat with M. zehntneri and there was some evidence of introgression between them. Its
epithet was given in the mistaken belief that the limestone upon which it occurs was an
iron-rich rock of volcanic origin.
9. Melocactus pachyacanthus Buining & Brederoo in Kakt. and. Sukk. 27(1): 1, with
illus. (1975) [vol. for 1976 but Heft 1 publ. Dec. 1975]. Holotype: Bahia, Mun.
Ourolândia, Horst 407 (U 531288).
Stem mucilaginous, depressed-globose to ovoid, pale green to grey-green, glaucous when young, 15–30
× 20 cm; ribs 9–11, low, to 25–37 × 50–65 mm; areoles to 13–15 × 9–11 mm, to 25 mm apart on the
MELOCACTUS
ribs. Spines 10–12, all very stout, reddish brown overlaid with grey, central(s) 1–3, lower 28–46 × 2–3(–4)
mm, porrect or slightly ascending, radials 8–9, straight to slightly recurved, lower 3 25–49 × 2–2.5 mm.
Cephalium to 30 × 10 cm, with dull pinkish red bristles, wool conspicuous or sparse. Flowers reddish to
deep pinkish magenta, 22–25 × 7–10 mm, sometimes only half-expanded or scarcely exserted from
cephalium. Fruit whitish or rather pale pink becoming very pale in lower half, 16.5–20.0 × 7–9.5 mm,
somewhat flattened. Seed 1.2–1.4 × 0.95–1.15 mm, testa-cells flat and smooth, some elongate.
conservation status. Endangered [EN B2ab(i, ii, iii, iv)]; area of occupancy estimated to be < 100
km2 and numbers of plants at 5 of the c. 10 known locations are undergoing severe decline or may have
already been lost.
An endemic Bahian species divisible into two subspecies:

1. Stem globose to elongate-ovoid, strongly glaucous except when old; cephalium with bristles
and wool visible . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9a. subsp. pachyacanthus
1. Stem depressed-globose, plain green or slightly glaucous when young; cephalium of very
dense red bristles only . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9b. subsp. viridis
9a. subsp. pachyacanthus

Stem glaucous. Cephalium to c. 9 cm diam., with bristles and conspicuous wool.
Northern Rio São Francisco caatinga element: on flat, ± exposed Bambuí limestone in caatinga at 400–645
m, near the Rios Salitre and Jacaré, northern Bahia. Endemic to Bahia. Map 44.
bahia: Mun. Juazeiro, 19 km W from Maçaroca on road to Fazenda Lagoa do Angico, 8 Jan. 1991,
Taylor (K, photo); Mun. Juazeiro/Campo Formoso, 2.5 km W of Curral Velho, 9 Jan. 1991, Taylor et
al. 1386 (K, ZSS, HRCB, CEPEC); Mun. Ourolândia, 65 km W of Jacobina, 1972, Horst 407 (U
531288); Mun. Morro do Chapéu, Área de Proteção Ambiental (APA) de Gruta dos Brejões, 4 Aug.
2002, Taylor (K, photos); Mun. São Gabriel, 57 km from Umburanas on road to Gameleira, 14 km NE
of the latter, 13 July 2000, G. Charles (photos).
conservation status. Endangered [EN B1ab(i, ii, iii, iv)] (3); area of occupancy = < 500 km2; PD=1,
EI=1, GD=2. Short-list score (3×4) = 12. Endangered by ongoing habitat destruction (urban/agricultural
development, quarrying) at 3 out of its 5 known, widely separated localities, but the largest population
at the APA Gruta dos Brejões is safe and numbers a few 1000s of mature individuals and seedlings.
This subspecies is sometimes partially sympatric with M. zehntneri, with which it may
exchange genes, but can be distinguished by its darker coloured flowers, smooth seeds and
stronger spination.
9b. subsp. viridis N. P. Taylor in Bradleya 9: 40, pl. 16 (1991). Holotype: Bahia, Mun.
América Dourada, Taylor & Zappi in Harley 27400 (CEPEC; SPF, K, isos.).
Stem never or only slightly glaucous when young. Cephalium c. 10 cm diam. or wider, with very dense
bristles and little or no wool.
Northern Rio São Francisco caatinga element: on flat, ± exposed Bambuí limestone in caatinga at 700–800
m, in the region of Irecê, Gruta dos Brejões and to the west of Várzea Nova, central-northern Bahia.
bahia: Mun. Morro do Chapéu, 10°57'53"S, 41°23'57"W, Sep. 2003, A. Cardoso (photos); l.c., E bank
of the Rio Salitre, nr ‘lagedo bordado’, N of the village of Brejão (Formosa), W of Várzea Nova and c.
MELOCACTUS
8 km from Icó, 5 Aug. 2002, Taylor (K, photos); Mun. Irecê, beside road BA 432 to Seabra, 1 km beyond
traffic island S of Irecê, 17 July 1989, Zappi 140 (K, SPF); Mun. América Dourada, c. 17 km S of road
BA 052, 2 km E of Belo Campo, 26 Dec. 1988, Taylor & Zappi in Harley 27400 (K, SPF, CEPEC).
conservation status. Critically Endangered [CR B2ab(i, ii, iii, iv); C2a(ii); D] (4); area of
occupancy < 10 km2. PD=1, EI=1, GD=1. Short-list score (4×3) = 12. Almost certainly extirpated by
habitat destruction (agricultural development) at two of its localities, apart from the second cited above,
where a total of less than 50 individuals are known to survive. It has recently been reported that
additional populations occur on top of limestone escarpments along the course of the Rio Salitre (M.
Machado, in litt.).
Near Irecê, Bahia, subsp. viridis has been found sympatric with M. azureus, but can be
distinguished by its greener epidermis, stouter spination, massive cephalium of very dense
bristles and scant wool, less expanded flowers with much broader perianth-segments,
larger, more pinkish fruits and smaller seeds. The habitats of these taxa in this region have
mostly been destroyed by agricultural development. The second population cited above
is somewhat morphologically intermediate with subsp. pachyacanthus.
MELOCACTUS VIOLACEUS Group (Nos 10–16)

Species Nos 13, 15 & 16 are restricted to sand or very sandy/gravelly substrates amongst
low, sparse shrubs and No. 14 is also commonly found on similar materials, but more
often associated with taller or denser vegetation. Nos 10 & 12 are most commonly found
on crystalline rocks, gneiss etc. or very stony substrates, whereas No. 11 is found on any
of the above as well as on limestone and soils derived from it.
10. Melocactus salvadorensis Werderm. in Notizbl. Bot. Gart. Berlin-Dahlem 12:

228 (1934). Type (syntypes): Bahia, Mun. Ipirá (Camisão) and Mun. São Félix/Muritiba,
Bananeiras, 1932, Werdermann 3391, 3392 (both B†). Neotype (Taylor 1991a): Bahia,
Bananeiras, Horst 301 (U 531290), illustrated in Krainz, Die Kakteen, Lfg 54 (1973).
[M. macrodiscus sensu Rizzini, Melocactus no Brasil: fig. 32 below (1982) non Werderm. (1932).]
[M. inconcinnus sensu F. Ritter, Kakt. Südamer. 1: 136–137, Abb. 78 (1979) non Buining & Brederoo (1975).]
Stem with mucilage in the green cortical tissues, pyramidal-globose to depressed-globose, greyish blue-green
or quite glaucous, 12–20 × 12–25 cm; ribs 8–14, triangular-acute, 25–30 × 30–50 mm; areoles to 7–8.5 ×
7–8 mm, 20–30 mm apart on the ribs. Spines horn yellow to reddish brown, overlaid with grey at first,
central(s) 1–4, 15–30 × 1.5–2.5 mm, radials 7–10, lowermost 20–46 × 1.8–2.2 mm. Cephalium to 15 × 6–10
cm, composed of dense reddish bristles and sparse grey-white wool. Flowers pinkish magenta, opening c. 4–5
p.m., to 25 × 12 mm, but sometimes scarcely exserted from cephalium and less expanded. Fruit deep lilac-
magenta, clavate, c. 17 × 6.5–9.0 mm. Seed 1.1–1.3 × 0.9–1.15 mm, some testa-cells markedly elongated,
others strongly convex, especially at periphery. Chromosome number: 2n = 44 (Assis et al. 2003).
Eastern caatinga element: usually on or adjacent to exposed gneiss/granitic rocks/inselbergs and in stony
soil of the caatinga, low elevations to 1040 m, within the Rio Paraguaçu (Rio Jacuípe) and Rio de Contas
(Rio Gavião) drainage systems, eastern and southern Bahia. Endemic to Bahia. Map 45.
bahia: Mun. Riachão do Jacuípe, near road BR 324, 9 km NW of bridge over Rio Jacuípe, 25 Apr.
1992, Taylor & Zappi (K, photo); Mun. Candeal, beside road BR 324, 20.5 km NW of Tanquinho, 25
Apr. 1992, Taylor & Zappi (K, photos); Mun. Ipirá, 42.5 km E of Baixa Grande on road BA 052, 9 Feb.
1991, Taylor et al. 1609D (K, photos); Mun. São Félix/Muritiba/Conceição da Feira, Bananeiras, 1971,
MELOCACTUS
Horst 301 (U 531290); Mun. Boa Vista do Tupim, c. 14 km da balsa para travessia do Rio Paraguaçú para
João Amaro, na estrada para Boa Vista do Tupim, 27 Apr. 1994, L.P. de Queiroz & Nascimento 3886
(HUEFS, K); Mun. Iaçu, 5.5 km W of the town, 7 Feb. 1991, Taylor et al. 1580C (K, photos); Mun.
Itatim, Morro do Agenor, 12°42'S, 39°46'W, 1 Sep. 1996, F. França 1815 (HUEFS, K, photo); l.c., 7 km
N of Milagres beside road BR 116 to Feira de Santana, 27 July 1989, Zappi 181 (K, SPF); Mun. Rio de
Contas, 8 km S of Jussiape (Juçiape) beside road to town of Rio de Contas, just W of the Rio de Contas,
22 Dec. 1988, Taylor & Zappi in Harley 25595 (K, SPF, CEPEC); Mun. Iramaia, 34 km from Contendas
do Sincorá towards Maracás, 4 Feb. 1991, Taylor et al. 1549A (K, photo); Mun. Maracás, 6–7 km E of
Porto Alegre, 4 Feb. 1991, Taylor et al. 1553 (K, ZSS, HRCB, CEPEC); Mun. Jequié, c. 1 km S of Jequié
beside road BR 116, 0.5 km S of the Rio de Contas, 27 July 1989, Zappi 180 (K, SPF); l.c., 10 km S, 21
Aug. 1996, Hatschbach 63224 (MBM, K); Mun. Aracatu, 25 km along road BR 030 towards Sussuarana,
3 Feb. 1991, Taylor et al. 1542A (K, photos); Mun. Tanhaçu, 54 km E of Brumado along road BR 030
towards Sussuarana, 3 Feb. 1991, Taylor et al. 1545A (ZSS, photo); Mun. Rio do Antônio, 36 km from
Brumado on road BR 030 to Caetité, 14 Apr. 1983, Leuenberger et al. 3064 (B, CEPEC), 2 Feb. 1991,
Taylor et al. 1523A (K, fr. in spirit, photos); Mun. Brumado, 7 km N of Brumado beside road to
Livramento, 22 Nov. 1988, Taylor & Zappi in Harley 25536 (K, SPF, CEPEC); Mun. Jacaraci, 2 km NW
of Jacaraci, Horst & Uebelmann 535 pro parte (ZSS AV-20-37); Mun. Condeúba and Mun. Piripá, reported
by Rizzini (1982: 71, figs 36 & 37); Mun. Cordeiros, 15°4'S, 41°53.5'W, Sep. 2003, M. Machado (obs.).
Concern at present in view of its extensive range and number of locations, but its habitat continues to
decline.
This species can be found sympatric with M. oreas, M. ernestii (both subspp.), M. zehntneri
and, probably, M. bahiensis and M. concinnus. It occasionally hybridizes with M. oreas and
M. ernestii subsp. ernestii. It is easily confused with M. zehntneri, but has much darker,
magenta fruit. Its pollination biology has been studied by Raw (1996).
11. Melocactus zehntneri (Britton & Rose) Luetzelb., Estud. bot. Nordéste 3: 111
(1926). Type: Bahia, Mun. Juazeiro, Rose & Russell 19728 (US, lectotype designated here;
NY, lectopara.).
Cactus zehntneri Britton & Rose, Cact. 3: 236, with illus. of type, fig. 248 (1922).
Melocactus macrodiscus Werderm., Blühende Kakteen (Repert. Spec. Nov. Regni Veg. Sonderbeih. C), Lfg
12 (1932). Holotype: Bahia, Mun. Caitité, Werdermann s.n. (B†). Lectotype (Taylor 1991a):
Werdermann, l.c. tab. 47 (1932).
M. curvicornis Buining & Brederoo in Kakt. and. Sukk. 23: 33 (1972). M. zehntneri var. curvicornis (Buining &
Brederoo) P. J. Braun in Bradleya 6: 98 (1988). Holotype: Bahia, Mun. Macaúbas, Horst 128 (U 531266).
M. giganteus Buining & Brederoo in Cact. Succ. J. (US) 45: 227 (1973). Holotype: Bahia, Mun. Gentio
do Ouro, Horst 266 (U 531294).
M. helvolilanatus Buining & Brederoo in Succulenta 55: 261 (1976). Holotype: Piauí, Santo Antônio de
Lisboa, Horst 444 (U 531289).
M. canescens F. Ritter, Kakt. Südamer. 1: 134 (1979). M. zehntneri subsp. canescens (F. Ritter) P. J. Braun
in Bradleya 6: 98 (1988). Holotype: Bahia, Mun. Tanhaçu, Ritter 1333 (U 531248).
M. canescens var. montealtoi F. Ritter, l.c. (1979). Holotype: Bahia, Mun. Palmas de Monte Alto, Ritter
1437 (U 531249).
M. zehntneri var. viridis F. Ritter, l.c., 132 (1979). Holotype: Piauí, Dom Expedito Lopes, Ritter 1206a (U
531268).
M. zehntneri var. ananas Rizzini, Melocactus no Brasil: 64 (1982), nom. inval. (Art. 37.1).
M. arcuatispinus Brederoo & Eerkens in Succulenta 62: 97, with illus. (1983). M. zehntneri subsp.
arcuatispinus (Brederoo & Eerkens) P. J. Braun & E. Esteves Pereira in Schumannia 3: 188 (2002).
Holotype: Bahia, Mun. (?), Horst 424 (U, not found Apr. 1989).
MELOCACTUS
M. douradaensis Hovens & W. Strecker in ibid. 63: 3, with illus. (1984). Holotype: Bahia, Mun. América
Dourada, Hovens et al. 81/172 (U, not found Apr. 1989).
M. saxicola L. Diers & E. Esteves Pereira in Kakt. and. Sukk. 35: 196, with illus. (1984). Holotype: Bahia,
Mun. Barreiras, E. Esteves Pereira 119 (KOELN [‘Succulentarium’], n.v.; UFG, iso.).
[M. bahiensis sensu Werderm., Bras. Säulenkakt.: 40 (1933) et auctt. mult. bras., non (Britton & Rose)
Luetzelb. (1926).]
[M. salvadorensis sensu Rizzini, Melocactus no Brasil: 72, fig. 36 (1982) non Werderm.]
Stem highly mucilaginous, at least in the non-chlorophyllous tissues, dark to grey- or light green, often
glaucous, sometimes strongly so when growing on limestone, but sometimes plain dark green (when on
sand), hemispheric to cylindric, very variable in shape, 11–48 × (9–)13–25 cm; ribs 10–22, 22–35 × 22–57
mm, with a sharply acute edge; areoles to 12 × 9 mm, set into notches and c. 20–40 mm apart on the
ribs. Spines 8–13(–15), horn yellow, brown or dull pink, but usually thickly overlaid with grey except at
the dark tips, at least some hooked at apex in seedlings, centrals (0–)1–2(–4), upcurved, 15–40(–45) ×
1.3–3.0 mm, radials 7–11, weakly to strongly recurved, lower 1–3 largest, 19–39(–45) × 1.5–2.5 mm.
Cephalium to 11(–30) × 6–10 cm, composed of fine, dense, pale pinkish red bristles and sparse to
abundant, white to creamy wool. Flowers pale to deep pink, not at all to well-exserted from cephalium,
sometimes with the tube exserted also, ± expanded, to 25 × 4–13 mm. Fruit very pale (rarely almost
white) to deep lilac-pink, rarely magenta (SE Pernambuco), 12–20 × 5–9(–10) mm. Seed 1.0–1.4 ×
0.8–1.35 mm, testa-cells strongly convex at end opposite hilum. Chromosome numbers: 2n = 22, 44
(Assis et al. 2003).
Widespread caatinga / Northern campo rupestre element: in soil or sand and on rocks of various types,
including limestone, gneiss/granite (inselbergs), sandstones, quartzitic and other crystalline formations, in the
caatinga (rarely on rocks in cerrado, W Bahia, or in campo rupestre, S Bahia), c. 200–1000 m, northern Piauí,
northern Ceará and Rio Grande do Norte to southern Bahia. Endemic to North-eastern Brazil. Map 45.
piauí: N Piauí, Mun. Buriti dos Lopes, 5 km NE of town on road BR 343, rocks by the Rio Pirangi, 19
Feb. 1995, Taylor (K, photos); Mun. Cocal, c. 17 km E of town, 19 Feb. 1995, Taylor (K, photos); SE Piauí,
Mun. Dom Expedito Lopes, Gaturiano, Ritter 1206a (U 531268); Mun. Santo Antônio de Lisboa, Horst 444
(U 531289); Mun. Simões, foot of Chapada do Araripe on road from Marcolândia, 7°36'S, 40°45'W, 5 Apr.
2000, E.A. Rocha et al. (K, photos); Mun. Jaicós, 28 km S of town on road BR 407, 6 Apr. 2000, E.A.
Rocha et al. (obs.); Mun. Canto do Buriti, c. 15 km S of town, partly exposed granite, 5 Feb. 1990, Zappi
215 (HRCB, ZSS); Mun. Paulistana, 25 km S of Paulistana, 1982, Horst 519 (ZSS AV-20-48).
ceará: Mun. Tauá, 6 km along road to Marrecas/Cocotá from junction with road BR 020, S of Tauá,
30 Jan. 1990, Zappi 209 (HRCB, K, photos).
rio grande do norte: Mun. Açu, 1978, Horst 480 (ZSS); Mun. Tangará, road BR 226, W of town
towards Santa Cruz, 6°14'S, 35°51'W, 1 Apr. 2000, E.A. Rocha et al. (K, photos).
paraíba: Mun. Pocinhos, Parque das Pedras, 28 June 1998, E.A. Rocha 420 (JPB, PEUFR); Mun.
Taperoá, Nov. 1920, Luetzelburg 54 (M); Mun. Campina Grande, Nov. 1920, Luetzelburg 14001 (M);
Mun. Itaporanga, Serra Água Branca, 7–10 Jan. 1994, M.F. Agra et al. 2559 (JPB); Mun. São João do
Cariri, c. 1–2 km from the town, on road BR 412, 29 June 1998, E.A. Rocha 426 (JPB); Mun. Sumé,
29–30 June 1994, M.F. Agra et al. 2827 (JPB); Mun. Monteiro, s.d., Luetzelburg 86 (M), Nov. 1920,
Luetzelburg 12321 (M).
pernambuco: Mun. Araripina, 15 Feb. 1990, Zappi 226B (HRCB); Mun. Taquaritinga do Norte, 8 Feb.
1993, F.A.R. Santos 63 (HUEFS); Mun. Ouricuri, Faz. Barrinha, 12 July 1984, G. Costa Lima 19 (IPA
42153); Mun. Salgueiro, 1.5 km N of city, granite outcrop, 17 Feb. 1990, Zappi 227 (HRCB, ZSS, K,
photos); Mun. Serra Talhada, serra N of city, exposed granite, 13 Feb. 1990, Zappi 224 & 224A (HRCB,
ZSS, K, photos); Mun. Brejo da Madre de Deus, Faz. Nova, 23 Jan. 1934, B. Pickel 3503 (IPA 4478), Jan.
1938, V. Sobrino s.n. (IPA 657); Mun. Ibimirim, 18 Nov. 1991, F.A.R. Santos 41 (HUEFS); Mun.
Alagoinha, 9 Dec. 1991, F.A.R. Santos 20 (HUEFS, PEUFR 12412, K, photo); l.c., Faz. Cajueiro Seca, 6
km from town, near the quarry, 31 Mar. 1998, E.A. Rocha & F. Lins 360 (IPA); Mun. Afrânio, road BR
407, 5 km NW of Rajada, 8°47'S, 40°52'W, 6 Apr. 2000, E.A. Rocha et al. (obs.); Mun. Floresta, Serra de
São Gonçalo, 23 May 1971, Heringer, Andrade-Lima et al. 864 (IPA 19753); l.c., road PE 360, N of the Serra
Negra, 8°33'S, 38°1'W, 8 Apr. 2000, E.A. Rocha et al. (K, photos); Mun. Garanhuns, reported by Rizzini
MELOCACTUS
(1982: 68, fig. 30); Mun. Santa Maria da Boa Vista, 18 Nov. 1992, F.A.R. Santos 47 (PEUFR, HUEFS);
Mun. Petrolina, 19 Nov. 1992, F.A.R. Santos 50 (HUEFS); l.c., 20 km from Petrolina towards Recife,
granite serra, 13 Aug. 1966, Hunt 6494 (K, photo); Mun. Nova Petrolândia, 12 km NE of the turning to
the old town on road BR 110, 12 Feb. 1991 (observed with Pilosocereus tuberculatus, Taylor & Zappi 1621).
alagoas: Mun. Ouro Branco, fide Uebelmann (1996): HU 475; Mun. Piranhas, Oct. 1993, R. de Lyra-
Lemos s.n. (MAC); Mun. Dois Riachos and Mun. Santana do Ipanema, reported by Rizzini (1982: 67,
fig. 29 left).
sergipe: Mun. Canindé de São Francisco, 23 Apr. 1981, M. Fonseca et al. 481 (ASE 896); Mun. Porto
da Folha, 12.5 km from Monte Alegre de Sergipe towards Paulo Afonso (Bahia), 11 Feb. 1991, Taylor &
Zappi 1617C (K, photo); Mun. Nossa Senhora da Glória and Mun. Tobias Barreto, reported by Rizzini
(1982: 67, figs 28 above & 29 right).
bahia: W Bahia, Mun. Barreiras, NW of the city, Serra da Bandeira, cliff edges, road to new airport
and near Telebahia tower, 18 Jan. 1991, Taylor et al. 1441 (K, ZSS, HRCB, CEPEC) & 1441A (K,
HRCB); N Bahia, Mun. Glória, Brejo do Burgo, 27 Aug. 1995, F.P. Bandeira 267 (HUEFS); Mun.
Remanso, 11 km E of town, 9°33'S, 42°W, 7 Apr. 2000, E.A. Rocha et al. (K, photos); l.c., c. 30 km
from Remanso towards Casa Nova, 10 Feb. 1990, Zappi 220B (K, seeds); Mun. Juazeiro, 6.5 km along
dirt road to Rodeodouro from road BA 210, 10 Jan. 1991, Taylor et al. 1389 (K, ZSS, HRCB, CEPEC);
Mun. Paulo Afonso, reported by Rizzini (1982: 67, fig. 34 below left); Mun. Sento Sé, 10 km E of
turning to Bazuá on road BA 210, Jan. 1991, Taylor et al. 1390A (K, photo); Mun. Curaçá, 4 km SE of
Poço de Fora, 7 Jan. 1991, Taylor et al. 1368A (ZSS, photo); Mun. Uauá, 12.5 km N towards Patamuté,
6 Jan. 1991, Taylor et al. s.n. (ZSS, photo); Mun. Campo Formoso, 2.5 km W of Curral Velho, 9 Jan.
1991, Taylor et al. 1386C (K, photo); Mun. Jaguarari, 10 km N of Flamengo on road BR 407 to
Juazeiro, 15 July 1989, Zappi 125A (K, frs in spirit, photo); Mun. Xique-Xique, 26 km SE of the town
on road BA 052, 13 Jan. 1991, Taylor et al. (obs.); Mun. Queimadas, 3 June 2000, C.T.S. Andrade
(photos); Mun. Jacobina, 10 km from Lajes (Lages) on road to Catinga do Moura, on limestone, 16 July
1989, Zappi 133A (K, photo); Mun. Gentio do Ouro, 16 June 1994, L.P. de Queiroz 3951 (HUEFS,
K); l.c., S of Xique-Xique, Serra de Santo Inácio, Horst 266 (U 531294); Mun. Morro do Chapéu, 2.5
km E of América Dourada, south (east) bank of Rio Jacaré (Rio Vereda do Romão Gramacho), N of
road BA 052, exposed Bambuí limestone, 17 July 1989, Zappi 138A (K, SPF); Mun. Valente, 6 km N
of Valente beside road to Santa Luz, on fine quartz sand, 14 July 1989, Zappi 121 (K, SPF); Mun. Araci,
14.5 km N towards Tucano, 5 Jan. 1991, Taylor et al. 1354A (K, photos); Mun. Ibotirama, few km E
of the town, 1986, P.J. Braun 604 (ZSS); S Bahia, Mun. Bom Jesus da Lapa, c. 32 km NE of Bom Jesus
da Lapa, just beyond Caldeirão, 43°13'W, 13°10'S, caatinga with damp sandy area, 18 Apr. 1980, Harley
21525 (K, CEPEC); Mun. Macaúbas, Horst 128 (U 531266, 531280); Mun. Paramirim, 43 km W of
Livramento do Brumado beside road to Paramirim, 4 km from Caraíbas, 28 Nov. 1988, Taylor in Harley
25555 (K, SPF, CEPEC); Mun. Livramento do Brumado, 12 km S of Livramento beside road to
Brumado, 23 Nov. 1988, Taylor & Zappi in Harley 25545A (K, fls in spirit); Mun. Dom Basílio, 21 km
S of Livramento do Brumado, 21 Mar. 1991, G.P. Lewis s.n. (K, photos); Mun. Brumado, reported by
Rizzini (1982: 67, fig. 28 below); Mun. Aracatu, 25 km E along road BR 030 towards Sussuarana, 3
Feb. 1991, Taylor et al. 1542B (K, photo); Mun. Tanhaçú, Ourives, 1964, Ritter s.n. (U 531248); Mun.
Palmas de Monte Alto, 1964, Ritter s.n. (U 531249); Mun. Caetité, near Brejinho das Ametistas, c. 1000
m, 26 July 1989, Zappi 177 (K, SPF).
Concern, with an enormous distribution over much of NE Brazil, but exploited in many parts of its range
as cattle fodder, for cactus candy and even for horticultural plantings, as seen in Pernambuco (Taylor
1991a) and elsewhere.
With the exception of one disjunct occurrence in the cerrado of western Bahia (near
Barreiras), the range of M. zehntneri corresponds very closely to the limits of the caatinga.
It is absent, however, from the caatingas of northern Minas Gerais and adjacent
southernmost Bahia, and is replaced by the related and similar M. salvadorensis (q.v.) in the
dry valleys of the Rio Paraguaçu and Rio de Contas in eastern Bahia (E of 41°W). It can
MELOCACTUS
be found sympatric with M. ernestii, M. bahiensis, M. ferreophilus, M. pachyacanthus, M.

salvadorensis, M. lanssensianus (fide Braun) and M. concinnus. Braun & Esteves Pereira (2002:
79) indicate that Melocactus occurs in Maranhão, which could relate to M. zehntneri, but
this remains to be backed up by published evidence.
The name M. horridus Werderm. is here applied to hybrids between M. zehntneri and M.
ernestii (q.v.). It does not appear to be a synonym of M. zehntneri, as suggested by Braun &
Esteves Pereira (2002: 132). Introgression between M. zehntneri and M. concinnus is evident
at São Rafael and Tareco, Mun. Morro do Chapéu (M. Machado & Taylor, obs.).
M. zehntneri varies greatly in stem size, shape and colour, in spine number, length,
thickness and colour, and in the degree to which its flowers are exserted from the
cephalium and thus able to expand. Its pollination biology has been studied by Locatelli
& Machado (1999a).
12. Melocactus lanssensianus P. J. Braun in Succulenta 65: 26, 61–63 (1986); P. J.

Braun & E. Esteves Pereira in Schumannia 3: 135, Abb. 150 (2002). Holotype:
Pernambuco, Mun. Caetés, 1977, Horst 474 (ZSS).
Stem highly mucilaginous, depressed-globose, to 8 × 14 cm, bluish grey-green; ribs c. 12, acute, 25–30
mm high, to 40 mm wide near stem base; areoles circular, 5–6 mm diam., c. 13 mm apart on the ribs.
Spines 8–9, pinkish grey to beige, yellowish to brownish at apex, stiff, central 1, upcurved, to 30–35 ×
1.5–2.0 mm, radials 7–8, recurved, lowermost longest, to 35–40 × 1.5 mm. Cephalium to 2.5 × 7 cm,
with pale red to salmon-coloured bristles. Flowers apparently cleistogamous, drying up without
expanding to form 6 mm long floral remains attached to the fruit. Fruit pink, c. 17 × 6 mm, containing
18–32 seeds. Seed 1.1 × 0.9 mm, testa-cells strongly convex.
Eastern caatinga element: on exposed granitic outcrops of serras in caatinga, c. 900 m, Mun. Caetés, south-
eastern Pernambuco. Map 45.
pernambuco: Mun. Caetés, Horst 474 (ZSS AV-20-33/44, AA-58-55, TP-58-35).
conservation status. Data Deficient [DD]; (material from the type locality is cultivated at the Royal
Botanic Gardens, Kew).
At its type locality said by Braun, l.c., to occur sympatrically with M. zehntneri, but
neither species could be observed by the present authors in the município cited above
during field studies in 1991. However, it is not the only cleistogamous taxon of this
relationship known from the region — the following cleistogamous entities may also
belong here, the first being most similar to Horst 474 (see Map 45):
paraíba: Mun. Tacima, near border of Mun. Caiçara, inselberg known as ‘Pão de Açucar’, 6°36'S,
35°28'W, 1 Apr. 2000, E.A. Rocha et al. (K, photos) — like M. zehntneri, but stem rather small and spines
longer and more slender; Plate 61.
pernambuco: Mun. Serra Talhada, rocky mountain slope c. 1 km north of the town, 13 Feb. 1990,
Zappi 223 (HRCB, ZSS, K) — not dissimilar from M. bahiensis in some respects and possibly worthy of
description as a new taxon (see Taylor 1991a: pl. 14).
13. Melocactus glaucescens Buining & Brederoo in Cact. Succ. J. (US) 44: 159, with
illus. (1972). Holotype: Bahia, Mun. Morro do Chapéu, [W of town], Horst 219 (U, not
found Apr. 1989). Lectotype (Taylor 1991a): Buining & Brederoo, l.c. fig. 2.
MELOCACTUS
[M. pruinosus sensu P. J. Braun in Succulenta 67: 114–117 (1988) and in Bradleya 6: 95 (1988) pro parte,
non Werderm. (1934).]
Stem highly mucilaginous, depressed-globose to pyramidal, intensely light blue-glaucous at first, light
grey-green in age, 13–18 × 12.5–24.0 cm; ribs 7–15, to 40 mm high and 60 mm broad near stem base;
areoles to 8 × 5 mm, to 21 mm apart on the ribs. Spines 6–10, brown thickly overlaid with grey, dark
brown to blackish at tip, centrals (0–)1–2, lower 11–20 × 0.8–1.7 mm, ascending and upcurved, radials
5–8, nearly straight to somewhat recurved, lower 3 11–55 × 1–1.8 mm. Cephalium to 10 × 6–7.5 cm,
bristles hidden beneath dense, creamy white wool. Flowers lilac-magenta, c. 25 × 15.5 mm. Fruit entirely
deep magenta-red, 9.5–16.0 × 5–7 mm, terete or somewhat flattened. Seed 1.1–1.35 × 0.9–1.15 mm,
testa-cells strongly convex. Chromosome number: 2n = 44 (Assis et al. 2003).
Caatinga/Northern campo rupestre (Chapada Diamantina) element: in the open and between low shrubs
of the caatinga on very sandy soil between small stones or in pure sand, sometimes occuring on adjacent,
flat sandstone rocks, 650–1000 m, Chapada Diamantina, Bahia. Endemic to Bahia. Map 46.
bahia: Mun. Ourolândia, N of Olho d’água do Fagundes, 10°54'41"S, 41°15'4"W, Aug. 2003, M.
Machado (obs.); Mun. Morro do Chapéu, NE of Gruta dos Brejões, between ‘Vermelho’ and ‘Mulungu
da Gruta’, 4 Aug. 2002, Taylor (K, photos); l.c., north of town towards Várzea Nova, W bank of the Rio
Salitre, c. 8 km from Icó, N of Brejão (Formosa), ‘lagedo bordado’, 5 Aug. 2002, Taylor (K, photos); l.c.,
12 km N of Tareco, along the Rio Jacaré, after village of ‘Canabrava’, ‘Cavalo Morto’, 10 April 2003,
M. Machado (obs.); l.c., 20–21 km W of Morro do Chapéu near road BA 052 to Irecê, 25 Dec. 1988,
Taylor & Zappi in Harley 27393 (K, SPF, CEPEC), 3 Aug. 2002, Taylor (K, photos); c. 25 km W of Morro
do Chapéu, 1 km S of road BA 052 towards Cafarnaum, slope E of road, 26 Dec. 1988, Taylor & Zappi
in Harley 27397 (K, SPF, CEPEC).
conservation status. Critically Endangered [CR B1ab(iii) + 2ab(iii)] (4); area of occupancy
estimated to be < 10 km2; PD=2, EI=1, GD=2. Short-list score (4×5) = 20. In view of its rarity and the
threats from commercial collection M. glaucescens has been placed on Appendix I of C.I.T.E.S. since June
1992. A local reserve to protect this and other rare species, incorporating the locality of Harley 27393
(above), has recently been implemented, but plants can still be damaged by fire. Its conservation status
may soon need to be revised in view of the new discoveries made by Marlon Machado and colleagues
over the past 4 years.
On account of its striking white-woolly cephalium, lilac-magenta flowers and small, deep
magenta-red fruits, this is one of the most distinctive species and cannot be confused with
any other member of the genus. The plant differs significantly at each of its sites of
occurrence. At the northernmost site in Mun. Morro do Chapéu the plants have much
longer spines (to 55 mm), whereas at the third site (listed above) the stems are smaller and
have consistently fewer ribs (7–9).
M. glaucescens is restricted in distribution, being mainly known from 6 small areas north
and west of Morro do Chapéu, Bahia, the latter two south-western sites being close to a
main road (BA 052) and therefore easily accessed. At these two sites there is also evidence
of hybridization with contiguous populations of other members of the genus and it is
disturbing that ‘pure’ M. glaucescens is rapidly disappearing or restricted to very small areas
of fine sand. At the fifth locality cited above (Harley 27393) M. glaucescens hybridizes with
M. ernestii and the product of these two very different taxa can be identified as M.
albicephalus Buining & Brederoo in Krainz, Die Kakteen, Lfg 52 (1973). At the final
locality (Harley 27397) M. glaucescens had formed a hybrid swarm with M. concinnus,
although this may not be the only taxon involved (Taylor 1991a: plate 17, below).
According to Marlon Machado (pers. comm.) pure M. glaucescens can no longer be found
MELOCACTUS
at this site. In the limited area of pure white sand at the previous locality, where the plant
is found in pure form, it is at risk from fires set by the local farmer and various badly
damaged specimens were seen (2002). At its new northern sites, which are c. 80 and 40
km distant from the type locality, there are no other Melocactus taxa close enough for gene
exchange to occur and the species seems more secure from other forms of disturbance.
Further investigation may reveal yet more localities for this exceptionally beautiful
species and, remarkably, Alex Braga, a contact of Marlon Machado, says he has recently
found the species in the dunes of the Rio São Francisco, in Mun. Barra, on the west bank
of the river, far from the Chapada Diamantina.
14. Melocactus concinnus Buining & Brederoo in Kakt. and. Sukk. 23: 5–7 (1972).
Holotype: Bahia, Mun. Seabra, Horst 214 (U 531262).
? M. pruinosus Werderm. in Notizbl. Bot. Gart. Berlin-Dahlem 12: 228 (1934). Type: Bahia, Mun. Morro
do Chapéu, Werdermann 3285 (B†).
M. pruinosus var. concinnus (Buining & Brederoo) P. J. Braun in Bradleya 6: 95 (1988).
M. axiniphorus Buining & Brederoo in Succulenta 55: 193 (1976). M. concinnus subsp. axiniphorus (Buin.
& Brederoo) P. J. Braun & E. Esteves Pereira in Schumannia 3: 188 (2002). Holotype: Bahia, Mun.
Vitória da Conquista, Horst 450 (U 531307).
M. robustispinus Buining, Brederoo & S. Theun. in ibid. 56: 116–119 (1977). M. zehntneri subsp.
robustispinus (Buining et al.) P. J. Braun in Bradleya 6: 98 (1988). Holotype: Minas Gerais, Mun. Mato
Verde, Horst 403 (U 531293).
[M. macrodiscus var. macrodiscus sensu F. Ritter, Kakt. Südamer. 1: 133 (1979) non Werderm. (1932).]
Stem highly mucilaginous, with green, chlorophyllous pith, grey-green and glaucous, at least when
young, intensely so in seedlings, depressed-globose, 8–12(–13) × 11–16 cm; ribs 8–12, to 20(–30) mm
high and 35–60 mm broad near stem base, often laterally creased; areoles to 6.5 × 6 mm, 13–20 mm apart
on the ribs. Spines 6–9, red then blackish when developing, later horn yellow or pale reddish brown to
brown, thickly overlaid with grey except at the dark tips, central (0–)1, 10–19 × 1 mm, upcurved, radials
6–8, slightly to strongly recurved, lower 1–3 largest, 15–26(–28) × 1–2 mm, sometimes hooked at apex.
Cephalium to 5.5 × 4–9 cm, composed of fine, dense, pale pinkish red bristles and creamy wool. Flowers
deep pink, scarcely to well-exserted from cephalium, 20–23 × 6–12 mm. Fruit pale lilac-pink to pink,
13–18(–22.5) × 5–8.5 mm, terete or somewhat flattened. Seed 1.05–1.35 × 0.95–1.3 mm, testa-cells
strongly convex. Chromosome number: 2n = 44 (Assis et al. 2003).
Eastern caatinga/Northern campo rupestre element: in the open or beneath low to tall shrubs and trees in
caatinga, carrasco, cerrado de altitude and campo rupestre, in stony soil, quartz sand or gravel, or between
crystalline rocks, 550–1300 m, Chapada Diamantina, Serra do Espinhaço and Serra do Periperi (Vitória
da Conquista), central Bahia to northern Minas Gerais. Endemic to core area of Eastern Brazil. Map 46.
bahia: Mun. Morro do Chapéu, at Tareco and São Rafael, 25 Dec. 1988 & 5 Aug. 2002, Taylor (obs.);
l.c., 19 km W of town on road BA 052 to Irecê, dense caatinga, 25 Dec. 1988, Taylor & Zappi in Harley
27389 (K, SPF, CEPEC); c. 25 km W of Morro do Chapéu, 1 km S of road BA 052 en route to
Cafarnaum, E of road, crystalline rocks and gravel, 26 Dec. 1988, Taylor & Zappi in Harley 27398 (K,
SPF, CEPEC); l.c., c. 2 km S of town, terreno de M. Machado, 3 Aug. 2002, Taylor (K, photos); Mun.
Seabra, 29 km N of town on road to Água de Rega, 27 Feb. 1971, Irwin et al. 31221 (K, UB); l.c., 11.5
km W of Seabra on road BR 242 to Ibotirama, caatinga on granite with sand between the rocks, 18 July
1989, Zappi 143 (K, SPF); Mun. Piatã, 12 km S of Boninal, Taylor & Zappi in Harley 25596 (K, SPF,
CEPEC); Mun. Abaíra, between Catolés and Abaíra, 13°18'S, 43°49'W, 31 Jan. 1992, Pirani et al. in
Harley 51396 (SPF, CEPEC, HUEFS, K); Mun. Érico Cardoso [Água Quente], 2 hours’ walk W of
‘Campo Queiroz’, Serra do Pau Queimado (region of Pico das Almas), gravel from mine workings, 11
Dec. 1988, Taylor in Harley 25563 (K, SPF, CEPEC); Mun. Jussiape (Juçiape), 25 km from Rio de Contas
PLATE 46
1 2
3 4
Plate 46. 46.1 Pilosocereus magnificus. NE Minas Gerais, Taylor & Zappi 763 (NT). 46.2 P. pachycladus subsp.
pachycladus. Bahia, Taylor et al. 1406 (UE). 46.3 P. machrisii. W Bahia, Taylor et al. 1440, cult. Kew (K). 46.4 P.
aurisetus subsp. aurisetus. Minas Gerais, S of Diamantina, 2002 (GC).
PLATE 47
1 2 3
4 5
Plate 47. 47.1 Pilosocereus aurisetus subsp. aurilanatus. Minas Gerais, Joaquim Felício, 2002 (MM). 47.2 P. aurisetus
subsp. aurisetus (densilanatus). Minas Gerais, S of Itamarandiba, 2002 (GC). 47.3 Ibid., in fruit. Minas Gerais, Zappi
& Prado in CFCR 11825 (DZ). 47.4 P. aureispinus. Bahia, Taylor et al. 1422 (NT). 47.5 P. multicostatus. NE Minas
Gerais, Harley et al. 25188 (NT).
PLATE 48
2 3
Plate 48. 48.1 Pilosocereus multicostatus. NE Minas Gerais, Taylor et al. 1517 (NT). 48.2 P. piauhyensis, flower.
Piauí, Zappi 217 (NT). 48.3 P. chrysostele. N Pernambuco, Zappi 222 (NT).
PLATE 49
1 2
3 4
Plate 49. 49.1. Pilosocereus densiareolatus. W Bahia, Porto Novo, 2002 (MM). 49.2 P. parvus. Bahia, Horst &
Uebelmann HU 203, cult. G. Charles (GC). 49.3 P. bohlei. Type locality, 2000 (GC). 49.4 Micranthocereus
violaciflorus. Minas Gerais, Taylor et al. 1514 (NT).
PLATE 50
1 2
3 4
Plate 50. 50.1 Micranthocereus albicephalus. Bahia, N of Licínio de Almeida, 2002 (GC). 50.2 Ibid. Minas Gerais,
Harley et al. 25519 (NT). 50.3 M. purpureus. Bahia, Harley et al. 27429 (NT). 50.4 M. auriazureus. Minas Gerais,
Grão Mogol, 2002 (GC).
PLATE 51
1 2
3 4
Plate 51. 51.1 Micranthocereus auriazureus. Minas Gerais, Grão Mogol, 2002 (GC). 51.2 M. streckeri. Type
locality, 2002 (MM). 51.3 Ibid. Ibid. (MM). 51.4 M. polyanthus. S Bahia, Brejinho das Ametistas, 2002 (GC).
PLATE 52
1 2
3 4
Plate 52. 52.1 Micranthocereus flaviflorus, sens. lat. Bahia, SE of Morro do Chapéu, 2002 (NT). 52.2 M. flaviflorus.
Bahia, Zappi 135 (NT). 52.3 Ibid., in fruit. Ibid. (NT). 52.4 M. dolichospermaticus. W Bahia, Porto Novo,
2002 (GC).
PLATE 53
1 2
Plate 53. 53.1 Coleocephalocereus buxbaumianus subsp. buxbaumianus. Minas Gerais, Taylor & Zappi 771 (NT).
53.2 C. buxbaumianus subsp. flavisetus. E Minas Gerais, Taylor & Zappi 754 (NT). 53.3 C. fluminensis subsp.
decumbens. NE Minas Gerais, S of Padre Paraíso, 2002 (GC).
PLATE 54
1 2
3 4
Plate 54. 54.1 Coleocephalocereus fluminensis subsp. decumbens. NE Minas Gerais, S of Padre Paraíso, 2002 (GC).
54.2 C. fluminensis subsp. fluminensis. Espírito Santo, Taylor & Zappi 786 (NT). 54.3 C. pluricostatus. E Minas
Gerais, Taylor & Zappi 776 (NT). 54.4 C. goebelianus. Bahia, Mun. Maracás, 2002 (MM).
PLATE 55
1 2
3 4
Plate 55. 55.1 Coleocephalocereus aureus. NE Minas Gerais, Pedra Azul, 2002 (MM). 55.2 Ibid. NE Minas
Gerais, Mun. Salinas, 2002 (GC). 55.3 C. purpureus, NE Minas Gerais, Harley et al. 25532 (NT). 55.4 Ibid., in
fruit. Ibid. (NT).
PLATE 56
1 2
3 4
Plate 56. 56.1 Melocactus oreas subsp. oreas. Bahia, Taylor et al. 1578C (NT). 56.2 M. oreas subsp. cremnophilus.
Bahia, SE of Morro do Chapéu, 2002 (NT). 56.3 M. ernestii subsp. ernestii. E Bahia, Jequié, 2002 (GC). 56.4
M. ernestii subsp. longicarpus. Minas Gerais, Harley et al. 25514 (NT).
PLATE 57
1 2
3 4
Plate 57. 57.1 Melocactus bahiensis subsp. bahiensis. Bahia, nr Vitória da Conquista, 2003 (NT). 57.2 M. bahiensis
subsp. amethystinus. Minas Gerais, Harley et al. 25073 (NT). 57.3 M. conoideus. Type locality, 2002 (MM). 57.4
M. bahiensis (M. inconcinnus). Bahia, Harley et al. 27029A (NT).
PLATE 58
1 2
Plate 58. 58.1 Melocactus deinacanthus. Bahia, type locality, 2003 (MM). 58.2. Ibid. Ibid., Zappi 150 (NT).
58.3 Ibid. Bahia, new locality 8–10 km from type, 2003 (MM).
PLATE 59
1 2
3 4
Plate 59. 59.1 Melocactus levitestatus. W Bahia, Porto Novo, 2002 (GC). 59.2 M. azureus. N Bahia, Mun.
Itaguaçu da Bahia, 2000 (GC). 59.3 Ibid. N Bahia, Harley et al. 27396 (NT). 59.4 M. ferreophilus. N Bahia, Taylor
et al. 1408B (NT).
PLATE 60
1 2
Plate 60. 60.1 Melocactus pachyacanthus subsp. pachyacanthus. Bahia, APA Gruta dos Brejões, 2002 (NT). 60.2
M. pachyacanthus subsp. viridis. Type collection (NT). 60.3 M. salvadorensis. Bahia, Harley et al. 25536 (NT).
PLATE 61
1 2
3 3
Plate 61. 61.1 Melocactus zehntneri. Bahia, Zappi 177 (NT). 61.2 Ibid. Bahia, Harley et al. 25555 (NT). 61.3 M.
cf. lanssensianus. Paraíba, Tacima (ER).
PLATE 62
1 2
3 4
Plate 62. 62.1 Melocactus glaucescens. Bahia, NE of Gruta dos Brejões, 2002 (NT). 62.2 Ibid. Bahia, W bank of
the Rio Salitre, ‘lagedo bordado’, 2002 (NT). 62.3 Ibid. Bahia, W of Morro do Chapéu, Harley et al. 27393
(NT). 62.4 Ibid. Ibid., Harley et al. 27397 (NT).
PLATE 63
Plate 63. 63.1 Melocactus concinnus. Bahia, Morro do Chapéu, 2002 (NT). 63.2 M. paucispinus. Ibid. (NT).
PLATE 64
1 2
3 4
Plate 64. 64.1 Melocactus concinnus × M. paucispinus. Bahia, Morro do Chapéu, 2002 (NT). 64.2 M. violaceus
subsp. violaceus. Paraíba, NE of Pedras de Fogo, 2000 (NT). 64.3 M. violaceus subsp. ritteri. Bahia, Taylor et al.
1583 (NT). 64.4 M. violaceus subsp. margaritaceus. Bahia, Itapoã, 2002 (GC).
PLATE 65
Plate 65. 65.1 Harrisia adscendens, in bud. Bahia, Taylor et al. 1388 (NT). 65.2 Ibid., in fruit. Bahia, S of Feira
de Santana, 1989 (NT).
PLATE 66
1 2
3 4
Plate 66. 66.1 Leocereus bahiensis, in fruit. Bahia, W bank of the Rio Salitre, ‘lagedo bordado’, 2002 (NT). 66.2
Ibid., flower sectioned. Bahia, Harley et al. 25552 (NT). 66.3 Facheiroa ulei. Bahia, near Santo Inácio, 2000 (GC).
66.4 Ibid. Ibid. (GC).
PLATE 67
1 2
3 4
Plate 67. 67.1 Facheiroa cephaliomelana subsp. cephaliomelana. Cult. Univ. Bonn (WR). 67.2 F. cephaliomelana
subsp. estevesii. S Bahia, Iuiú, 2002 (GC). 67.3 F. squamosa. Type locality, 1990 (NT). 67.4 Ibid. N Bahia, Taylor
et al. 1372 (UE).
PLATE 68
1 2
3 4
Plate 68. 68.1 Espostoopsis dybowskii. N Bahia, Jaguarari, 2000 (GC). 68.2 Ibid. S Bahia, Taylor et al. 1551 (NT).
68.3 Arthrocereus melanurus subsp. melanurus. Minas Gerais, Horst & Uebelmann HU 594, cult. G. Charles (GC).
68.4 A. melanurus subsp. magnus. Type locality, Aug. 2003 (NT).
PLATE 69
2 3
Plate 69. 69.1 Arthrocereus melanurus subsp. odorus. Minas Gerais, Horst & Uebelmann HU 1555, cult. G. Charles
(GC). 69.2 A. rondonianus. Minas Gerais, Joaquim Felício, 2002 (GC). 69.3 Ibid. Loc. cit., Horst & Uebelmann
HU 145, cult. Univ. Heidelberg (WR).
PLATE 70
1 2
Plate 70. 70.1 Arthrocereus glaziovii. Minas Gerais, Horst & Uebelmann HU 330, cult. G. Charles (GC). 70.2 Ibid.,
in bud. Minas Gerais, Harley et al. 25500 (NT). 70.3 Ibid. Ibid. (NT).
PLATE 71
1 2
3 4
5 6
Plate 71. 71.1 Discocactus zehntneri subsp. zehntneri. N Bahia, Serra do Francisco, S of Piçarrão, 2000 (GC).
71.2 D. zehntneri subsp. boomianus. Bahia, W of Morro do Chapéu, 2002 (NT). 71.3 Ibid. (D. horstiorum). Bahia,
Serra do Curral Feio, 2000 (GC). 71.4 D. bahiensis. Bahia, Rodeadouro, 2000 (NT). 71.5 Ibid. Bahia, Horst &
Uebelmann HU 438, cult. G. Charles (GC). 71.6 Ibid. Bahia, São Rafael, 2000 (GC).
PLATE 72
2 3
Plate 72. 72.1 Discocactus heptacanthus subsp. catingicola. Bahia, Harley 25558 (NT). 72.2 D. placentiformis, with
spent flowers and fire-scorched stem. Minas Gerais, Harley et al. 24988 (NT). 72.3 Ibid., in fruit. Minas Gerais,
Taylor et al. 1450 (NT).
PLATE 73
Plate 73. 73.1 Discocactus pseudoinsignis. Minas Gerais, Grão Mogol, 2002 (GC). 73.2 D. horstii. Ibid., 1981
(WR).
PLATE 74
Plate 74. 74.1 Uebelmannia buiningii. Type locality, 2002 (GC). 74.2 U. gummifera (meninensis). Minas Gerais,
Pedra Menina, 2002 (GC).
PLATE 75
1 2
Plate 75. 75.1 Uebelmannia gummifera. Minas Gerais, SW of Itamarandiba, 2002 (GC). 75.2 Ibid. Minas Gerais,
Penha da França, 2002 (GC). 75.3 Ibid. Minas Gerais, Tromba D’Anta, 2002 (MM).
PLATE 76
1 2
Plate 76. 76.1 Uebelmannia pectinifera subsp. pectinifera. Minas Gerais, W of Mendanha, 2002 (GC). 76.2 Ibid.
(green form). Minas Gerais, Inhaí, 2002 (MM). 76.3 U. pectinifera subsp. flavispina. Minas Gerais, N of Datas, 2002
(GC).
PLATE 77
Plate 77. Uebelmannia pectinifera subsp. horrida. Minas Gerais, Mun. Bocaiúva, 2002 (GC).
MELOCACTUS
on road to town, 25 Nov. 1988, Taylor & Zappi in Harley 25550 (K, SPF, CEPEC); Mun. Ituaçu, 13°48'S,
41°16'W, 22 June 1987, L.P. de Queiroz 1632A (HUEFS); l.c., c. 2 km NE of the town, 18 Aug. 1988,
Eggli s.n. (ZSS, photo); Mun. Caetité, Brejinho das Ametistas, campo rupestre on quartzitic rocks, 26 July
1989, Zappi 174 (K, SPF); Mun. Licínio de Almeida, Braun 402 (ZSS TP58-116); Mun. Jacaraci, Horst
468 (ZSS AV-20-32); Mun. Vitória da Conquista, Morro do Cruzeiro above Vitória da Conquista, low
shrub vegetation on quartz gravel, 26 July 1989, Zappi 178 (K, SPF), Horst 450 (U 531307); Mun.
Cordeiros, c. 15°4'S, 41°53.5'W, Sep. 2003, M. Machado (obs.).
minas gerais: Mun. Monte Azul, Serra Geral, c. 9 km E of the town by path to village of Gerais, 28
Jan. 1991, Taylor et al. 1474 (K, ZSS, HRCB, BHCB); Mun. Mato Verde, 12 km E of Mato Verde on
road to Santo Antônio do Retiro, cerrado de altitude and campo rupestre on quartz sand and gravel, 9 Nov.
1988, Taylor & Zappi in Harley 25517 (K, SPF), 22 July 1989, Zappi 165A (K, fls in spirit, photos), Aug.
1972, Horst 403 (U 531293, K); Mun. Grão Mogol, Serra da Bocaina, 48.5 km W of bridge over Rio
Vacaria on road BR 251, 31 Jan. 1991, Taylor et al. 1515A (K, photo); Mun. Turmalina, Peixe Cru, by
the Rio Jequitinhonha, July 1991, M.G. Carvalho (K, photo).
Concern, but some populations affected by burning of vegetation started by man (campos rupestres) and
the southernmost population is likely to be submerged under an hydro-electric dam-lake.
When found sympatric or contiguous with populations of other species, M. concinnus not
infrequently forms hybrid swarms. These include allied species, such as M. glaucescens, M.
zehntneri and M. paucispinus, as well as the unrelated M. oreas subsp. cremnophilus. It is also
found with M. bahiensis, M. conoideus and M. salvadorensis.
15. Melocactus paucispinus Heimen & R. J. Paul in Kakt. and. Sukk. 34: 227–229,
with illus. (1983). Holotype: Bahia, Mun. Seabra, Heimen et al. 81/149 (KOELN, n.v.).
Stem highly mucilaginous, strongly depressed, hemispheric or disc-shaped, usually partly buried in the
sand, light grey-green, never glaucous even when young, 7–11.5 × 15–20 cm; ribs 9–10, to 2.5–4.0 cm
high and 3.7–5.0 cm broad near stem base; areoles c. 3.5–7.0 × 2.5–5.5 mm, to 24 mm apart on the ribs.
Spines (3–)4–6, recurved, pale grey with dark brown tips, all radial, lowermost largest, 20–32 × 1.2–2.0
mm, uppermost 1–3 much smaller, to 10 × 1 mm. Cephalium to 3–6 × 7–8 cm, composed of dense, fine,
pinkish red bristles and creamy white wool. Flowers deep pink, (18–)24–26 × (9–)12.5–14.5 mm. Fruit
pale lilac-pink, (14–)16.5–19.0 × 5.5–7.5 mm. Seed 1.3–1.6 × 1.2–1.5 mm, testa-cells strongly convex.
Chromosome number: 2n = 44 (Assis et al. 2003).
Northern campo rupestre element: in sand or more rarely quartz gravel, caatinga/cerrado de altitude, c.
900–1500 m, in the Chapada Diamantina and northern Serra do Espinhaço, northern to southern Bahia.
bahia: Mun. Morro do Chapéu, W of town, S of road BA 052 towards ‘Pedra Duas Irmãs’, c.
11°33'40"S, 41°17'50''W, June 2002, M. Machado (K, photos); l.c., E & S of town, in various sites within
the area delimited by 11°33–37'S & 41°0–10'W, 2 Aug. 2002, Taylor (K, photos); Mun. Seabra, 28.5 km
W of Seabra near road BR 242 to Ibotirama, 18 July 1989, Zappi 144 (K, SPF); Mun. Piatã, between
Piatã and Inûbia, R. Harley (K, photos); Mun. Abaíra, 9 km N of Catolés, ‘caminho de Ribeirão de Baixo
a Piatã, gerais entre Serra do Atalho e a Serra da Tromba’, 13°14'S, 41°55'W, 10 July 1995, L.P. de
Queiroz et al. 4420 (HUEFS, K); Mun. Érico Cardoso [Água Quente], Pico das Almas, valley near Serra
do Pau Queimado, near old road between Mato Grosso do Norte and Água Quente, quartz gravel, 23
Dec. 1988, Hind in Harley 27430 (K, SPF, CEPEC); Mun. Rio de Contas, Pico das Almas, between Rio
de Contas and Faz. Brumadinho, 0.5 km before Faz. Brumadinho, 27 Nov. 1988, Taylor in Harley 25554
(K, SPF, CEPEC); Mun. (?), E of Urandi, reported by Braun (1988a: 208, 171, illus. ‘HU 536’).
MELOCACTUS
conservation status. Endangered [EN B2ab(v)] (3); extent of occurrence > 10,000 km2, but area of
occupancy assumed to be < 50 km2; PD=2, EI=1, GD=1. Short-list score (3×4) = 12. In view of its
erratic distribution, rarity and desirability to collectors, this species has been listed in Appendix I of
C.I.T.E.S. since June 1992. Only the population to the E & S of Morro do Chapéu is known to be large,
the remaining 6 being much smaller (some with < 50 individuals) and severely fragmented/disjunct.
The strongly depressed stem of M. paucispinus seems to be an adaptation to minimize

damage caused by fire, which periodically sweeps through its cerrado/campo rupestre habitat.
By remaining partly buried in sand and exposing only the upper half of its flattened stem,
it benefits from the cooler air drawn in at ground level as the fire passes, though the edges
of its ribs may still get scorched. This adaptation is identical to that displayed by members
of the ecologically comparable genus Discocactus, with which juvenile plants of M.
paucispinus are readily confounded. Besides hummingbirds, a small lizard, c. 5 cm long,
was seen visiting the flowers at Morro do Chapéu.
A population photographed by Ray Harley (K, photos), between Piatã and Inûbia
(Mun. Piatã, Bahia), appears to represent plants showing introgression with M. concinnus,
with bluish grey-green epidermis, 11 ribs and mostly 5, well-developed spines per areole.
Introgression has also been observed by Taylor (K, photos) in a population south of
Morro do Chapéu, in which the hybrid plants were larger than either parent and strongly
resembled M. zehntneri. Similar plants were seen south of Rio de Contas. It also
hybridizes with M. oreas subsp. cremnophilus.
16. Melocactus violaceus Pfeiff. in Allg. Gartenzeitung 3: 313 (1835); Enum. Cact.:
45–46 (1837). Type (see Schumann 1897–98): Brazil, cult. Schelhase (B†). Neotype
(Taylor 1991a): Brazil, Rio de Janeiro, D. Sucre 9186 (RB 192529).
? Cactus melocactoides Hoffmanns., Verz. Pfl.-Kult. Nachtr. 3: 24 (1826). Type not known to have been
preserved; see Taylor (1980: 67). ? Melocactus melocactoides (Hoffmanns.) DC., Prodr. 3: 461 (1828).
M. depressus Hook. in Bot. Mag. 65: tab. 3691 (1838). M. melocactoides var. depressus (Hook.) Rizzini,
Melocactus no Brasil: 82 (1982). M. melocactoides f. depressus (Hook.) Rizzini, tom. cit. 82, 84, fig. 49
(1982). Type: Pernambuco, Gardner (apparently not preserved). Lectotype (Taylor 1991a): Hooker,
l.c., tab. 3691 (1838).
M. melocactoides var. violaceus (Pfeiff.) Rizzini, tom. cit., 81, 83 (1982).
M. violaceus subsp. natalensis P. J. Braun & E. Esteves Pereira in Cact. Succ. J. (US) 69: 71–74 (1997).
Type: Rio Grande do Norte, Esteves Pereira 340 (UFG 17657, n.v.). M. melocactoides var. natalensis
Rizzini, tom. cit. (1982), nom. inval. (Art. 37.1).
Stem highly mucilaginous, dark green, never glaucous even when young, usually broader than tall, often
very small, 5–18(–20) × 6–17(–20) cm; ribs 8–15; areoles to 5 × 5 mm, 6–18 mm apart on the ribs.
Spines 5–10(–12), brownish but thickly overlaid with grey except at the dark brown to blackish tips,
very slender, 0.5–1.5 mm thick, centrals absent or 1, to 19 mm, ascending and often slightly upcurved,
radials almost straight or slightly recurved, lower 3–5 largest and nearly equal, 14–24(–26) mm.
Cephalium to 5.5(–11) × 3.7–8.5 cm, with abundant white wool and sparse to numerous, ± exserted,
pale red, very fine bristles. Flowers deep pink, 15–25 × 6–13.5 mm, scarcely to well-exserted from the
cephalium. Fruit pale pink to lilac-pink or white, 12.5–19.0 × 5.5–7.5 mm. Seed 1.2–1.5 × 1–1.4 mm,
testa-cells strongly convex.
conservation status. Vulnerable [VU A3c]; see assessments of subspecies below.

MELOCACTUS
Two of the 3 subspecies recognized here are endemic to the core area of Eastern Brazil:
1. Fruit lilac- to pale pink . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2

1. Fruit white to very pale pink (NE Bahia, Sergipe & Alagoas) . . . . . . 16c. subsp. margaritaceus
2. Flower to 25 × 13.5 mm; spines 6–12; ribs 9–15 (NE Minas Gerais; and coastal regions
of E Brazil up to 35 km inland, from Rio Grande do Norte to Rio de Janeiro) . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16a. subsp. violaceus
2. Flower c. 18–22 × 7–10 mm; spines 5–6; ribs 8–10 (Bahia inland: Jacobina & Rui Barbosa)
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16b. subsp. ritteri
16a. subsp. violaceus

Stem depressed-globose, hemispheric or disc-shaped, 6–16 cm diam.; ribs 9–15, to 10–25 × 40 mm broad
near stem base; areoles 6–18 mm apart on the ribs. Spines 6–12. Cephalium to 5.5(–7.0) × 3.7–8.5 cm.
Flower ± exserted from cephalium, with 17–25 perianth-segments visible from above.
Widespread humid forest (Mata atlântica/restinga) and Northern campo rupestre (cerrado) element: between
shrubs in sand of the coastal restinga, riverine sand-dunes and similar habitats further inland, 0–150 m (in
sandy cerrado de altitude at 1100 m, NE Minas Gerais only), Rio Grande do Norte to Rio de Janeiro, but
apparently rather discontinuous, perhaps for lack of suitable habitats and due to extensive habitat
destruction. Map 46.
rio grande do norte: Mun. Genipabu, reported by Braun & Esteves Pereira (2002: 133, Abb. 146);
Mun. Natal, c. 1902, F.H. Rodier Heath s.n., cult. R.B.G. Kew, 1908 (K), cf. also Cactus J. (Croyden) 6:
63–65, figs 1 & 2 (1938).
paraíba: Mun. Mamanguape/Mun. Rio Tinto, Reserva de Guaribas, Aug. 2001, J. Jardim (obs.); Mun.
Santa Rita, M.F. Agra (obs.); l.c. (?), plants on sale in market in João Pessoa, assumed to be from Santa
Rita site, 1992, McRobb (K, photo ref. BZ 92-151); Mun. Pedras de Fogo, beside sugarcane plantation E
of dirt road leading N from Destilaria Giasa, 7°20'S, 35°4'W, c. 150 m, 31 Mar. 2000, E.A. Rocha et al.
(K, photos) — c. 30 km from the coast.
pernambuco: Mun. Jaboatão dos Guararapes, Prazeres (Estação), Oct. 1932, B. Pickel 3224 (IPA 4479);
Mun. Ipojuca, restinga near baía de Suape, 1978, Andrade-Lima & Medeiros-Costa 241 (IPA 22705); l.c.,
Porto das Galinhas, 1988, reported by Braun & Esteves Pereira (2002: 21).
sergipe: between Penedo (AL) and Pindoba, c. 30–35 km from the ocean, c. 6 km from the Rio São
Francisco, 100 m, 1978, Horst 482 (ZSS AV-20-24/41/94).
bahia: Mun. Maraú, Jan. 2000, E.A. Rocha (CEPEC); Mun. Uruçuca/Ilhéus, west of Serra Grande,
‘Campo Cheiroso’, Aug. 2001, Zappi (obs.); Mun. Canavieiras, A.M. de Carvalho (living plant seen at
Ilhéus, Feb. 1991); Mun. Santa Cruz Cabrália, 2–6 km S of Santa Cruz Cabrália beside road BR 367, 31
July 1989, Zappi 185 (K, SPF); Mun. Porto Seguro, c. 11 km N of Porto Seguro on road BR 367 to
Santa Cruz Cabrália, beneath shrublets of Chamaecrista, Cuphea and Melastomataceae, 30 July 1989, Zappi
183 (K, SPF); Mun. Prado, near Hotel Praia do Prado, 23 Sep. 1989, M.C. Vianna 1994 (GUA).
minas gerais: Mun. Jequitinhonha, c. 47 km S of Pedra Azul on road to Jequitinhonha, ‘Serra da Areia’
[Serra da Sapucaia?], 1100 m, 20 Oct. 1988, Taylor & Zappi in Harley 25265 (K, SPF).
espírito santo: Mun. Itaúnas, 15 May 1987, G.P. Lewis (K, photo); Mun. Linhares, Reserva CVRD,
near Parajú road, 19°6–18'S, 39°45'–40°19'W, reported by Figueira et al. (1994: 295 & fig. 1); Mun.
Guarapari, KM 29–30 on road ES 060, Lagoa Vermelha, 21 May 1990, O.J. Pereira 2124 & Zappi (HRCB).
conservation status. Vulnerable [VU A3c] (2); PD=1, EI=1, GD=2. Short-list score (2×4) = 8.
Vulnerable from habitat modification by development for the tourist industry and for pineapple and
sugarcane plantations. In Paraíba, the locality at Santa Rita has already been destroyed by a pineapple
plantation, while that north-east of Pedras de Fogo has all but disappeared due to the cultivation of
sugarcane. However, it may be protected in the northern part of this state in the Guaribas reserve (see
above). Bahian populations are similarly affected, including habitat destruction due to the planting of
coconut groves (eg. at ‘Campo Cheiroso’, Uruçuca/Ilhéus).
MELOCACTUS
16b. subsp. ritteri N. P. Taylor in Bradleya 9: 57 (1991). Holotype: Bahia, Mun. Rui
Barbosa, Ritter 1209a (U 531256).
Melocactus macrodiscus var. minor Ritter, Kakt. Südamer. 1: 133 (1979). Type: as above.
Stem depressed-globose, 8–13.5 cm diam.; ribs 8–10, to c. 15 × 50 mm; areoles 10–14 mm apart on the
ribs. Spines 5–6. Cephalium to c. 5 × 5 cm. Flower with up to 24 perianth-segments visible from above,
sometimes scarcely exserted from cephalium.
Eastern caatinga / Northern campo rupestre element: between Vellozia shrubs in fine quartz sand or gravel,
campo rupestre, c. 450–860 m, near Jacobina and above Rui Barbosa, central-eastern Bahia. Endemic to
Bahia. Map 46.
bahia: Mun. Jacobina, 2 km E of Jacobina town centre, valley slope N of Rio Itapicurumirim, 16 July
1989, Zappi 130 (K, SPF); Mun. Rui Barbosa, 1964, Ritter s.n. (U 531256); l.c., Serra de Rui Barbosa,
8 Feb. 1991, Taylor et al. 1583 (K, ZSS, HRCB, CEPEC).
conservation status. Critically Endangered [CR B2ab(iii, v)] (4); area of occupancy < 10 km2;
PD=1, EI=1, GD=1. Short-list score (4×3) = 12. Known from only two small locations, both near to
towns, giving a projected decline in habitat and numbers of individuals (through collection).
This very local taxon occurs sympatrically with different forms of M. bahiensis at both
localities cited above. It is very similar to the following, but has pink fruits like subsp. violaceus.
16c. subsp. margaritaceus N. P. Taylor, l.c., pl. 19 (1991). Holotype: Sergipe, Mun.
Santo Amaro das Brotas, Rizzini & Mattos (RB 215018).
? Melocactus pentacentrus Lemaire, Cact. Nov. Gen. Sp.: 108 (1839). Type: Bahia [= Salvador], not known
to have been preserved.
M. ellemeetii Miquel in Nederl. Kruidk. Arch. 4(3): 336, tab. 1 (1858) [a separate print from Miquel filed
in the herbarium at BR is annotated 1857]. Type: Bahia, Salvador, a living plant in the garden of
Ellemeet. Lectotype (designated here): Miquel, l.c., tab. 1 (1858).
M. margaritaceus Rizzini, Melocactus no Brasil: 76, figs 42–45 (1982), nom. inval. (Art. 37.1).
M. margaritaceus var. disciformis Rizzini, tom. cit.: 78 (1982), nom. inval. (Art. 43.1).
M. margaritaceus var. salvadoranus Rizzini, tom. cit.: 78 (1982), nom. inval. (Art. 43.1).
[M. depressus sensu Schumann, Gesamtb. Kakt. Nachtr.: 129 (1903) pro parte.]
Stem depressed-globose, rarely shortly cylindric, 5–15(–25) × 7–13(–20) cm; ribs 8–15, to 20 × 40 mm;
areoles 7–10 mm apart on the ribs. Spines 5–10. Cephalium to 5(–11) × 3.7–6.5 cm. Flower exserted
from cephalium, with c. 19 perianth-segments visible from above. Chromosome number: 2n = 22 (Assis
et al. 2003).
Northern humid/subhumid forest (restinga) element: on coastal dunes of fine, white sand and inland
(Serra de Itabaiana, SE), near sea level to c. 400 m, Alagoas, Sergipe and eastern Bahia (south to Salvador).
Endemic to the core area within North-eastern Brazil. Map 46.
alagoas: (?) Mun. Maragogi, coast N of the town, 1988, Braun 1006 (ZSS); Mun. Marechal Deodoro,
16 km S of Maceió, 9°47'S, 35°52'W, 2 Feb. 1982, J.H. Kirkbride jr 4620 (MAC, UB); l.c., restinga S of
Ponta do Cavalo Ruço, 14 Feb. 1995, Taylor & Zappi (K, photos); Mun. Feliz Deserto, 30 Sep. 1981,
R.F. Rocha 145 (MAC).
sergipe: Mun. Itabaiana, on rocks and sand of east slope (to c. 400 m) and at base of Serra da Itabaiana,
3 Aug. 1998, Taylor (ASE, K, photos); Mun. Santo Amaro das Brotas, 10 Mar. 1981, Rizzini & Mattos,
cult. Jard. Bot. Rio de Janeiro (RB 215018); Mun. São Cristóvão, reported by Rizzini (1982: 77).
bahia: Mun. Camaçari, near Arembepe, Rod. Linha Verde, 18 Aug. 1995, Hatschbach et al. 63363
HARRISIA
(MBM, K); Mun. Lauro de Freitas, 8 km N of Itapoã, Mar. 1989, Supthut 89-72 (ZSS); Mun. Salvador,
1 km N of praia of Itapoã, Lagoa de Abaeté, 13 July 1989, Zappi 115 (K, SPF); l.c., Dunas de Itapoã,
12°56'S, 38°21'W, 13 Mar. 1996, M. Gomes 12 (ALCB).
conservation status. Vulnerable [VU A3c] (2); PD=1, EI=1, GD=2. Short-list score (2×4) = 8.
Vulnerable from habitat modification by the rapidly expanding tourist industry and agricultural
development (sugar cane, coconut), especially along the extensive coastal area in Bahia served by the
‘Linha Verde’ road and along the coast of NE Alagoas. Away from the coast it is most abundant and
currently best protected in the state ecological reserve at the Serra de Itabaiana, Sergipe, managed by the
local office of IBAMA. A decline of the populations of this taxon by at least 30% can be projected within
20 years (c. 3 generations).
If this taxon should one day prove to be worthy of specific rank, as Rizzini believed, then
the earliest, clearly typified species name for it is M. ellemeetii Miquel, l.c. (1858).
However, at least one of its northern populations has flowers and very pale pink fruits,
which are somewhat intermediate with subsp. violaceus, further weakening the differences
between them.
Tribe TRICHOCEREEAE F. Buxbaum
As noted earlier, the distinction between this tribe and Cereeae Salm-Dyck is currently
unclear when analysed using DNA gene-sequence data (Wallace unpubl., Soffiatti
unpubl.). Their traditional circumscriptions, based primarily on the presence or absence
of hair-spines/spines on the pericarpel and flower-tube, is maintained here, although
Espostoopsis F. Buxb. (Trichocereeae) and some species of Cipocereus Ritter (Cereeae)
represent exceptions in each case. In the gene sequence study by Soffiatti (2002, unpubl.)
the enigmatic Uebelmannia is placed amongst the Brazilian Cereeae, in a position close to
Cipocereus, but here it is left at the end of the Cereeae/Trichocereeae pending further
DNA data.
24. HARRISIA Britton
in Bull. Torr. Bot. Club. 35: 561 (1908). Type: H. gracilis (Mill.) Britton (Cereus gracilis
Mill.).
Including Eriocereus (A. Berger) Riccob. (1909); Harrisia subg. Eriocereus (A. Berger)
Britton & Rose (1920).
A genus of c. 10 species, with a disjunct distribution between the Caribbean (Subg.

Harrisia, fruits indehiscent) and central South America (Subg. Eriocereus, fruits dehiscent).
Two species are native of Brazil: H. balansae (K. Schum.) N. P. Taylor & Zappi (Mato
Grosso & Mato Grosso do Sul, first recorded by Hoehne 1915; ‘H. pomanensis’ misapplied
by Braun & Esteves Pereira 2002: 127, Abb. 140) and that treated here, which is endemic
to the Nordeste and isolated from its nearest congeners by some 1800 km. In its gross
morphology it is most similar to the east Andean H. pomanensis (F.A.C. Weber) Britton
HARRISIA
& Rose (syn. Eriocereus tarijensis F. Ritter; Argentina & Bolivia). However, gene sequence
data (Wallace 1997: 11) indicate that H. adscendens is sister taxon to the Caribbean Subg.
Harrisia, linking the two subgenera and indicating the path of radiation of the genus from
its presumed origin in the eastern Andes of Bolivia, where it is believed to have common
ancestry with Samaipaticereus Cárdenas. Plate 65.
1. Harrisia adscendens (Gürke) Britton & Rose, Cact. 2: 155 (1920). Type: Brazil,
Bahia, ‘Tambury’, Oct. 1906, Ule 7072 (B, lecto. designated here; HBG, lectopara. [K,
photos ex B]).
Cereus adscendens Gürke in Monatsschr. Kakt.-Kunde 18: 66 (1908). Eriocereus adscendens (Gürke) A.
Berger, Kakteen: 341 (1929).
vernacular names. Rabo-de-raposa, Passa-prá-lá.
Shrubby, to 2 m or more, free-standing or semi-scandent and partly pendulous, sometimes forming

dense masses, with basitonic or mesotonic branching, ± segmented. Juvenile growth of markedly
different appearance to the adult, the ribs scarcely visible and bearing very small areoles and fine spines.
Adult stems erect to decumbent, 2–7 cm diam.; ribs 6–10, low, rounded, tuberculate at the areoles,
sinuses sinuous, marked by a dark line (as in H. pomanensis); epidermis pale grey-green, sometimes
reddish. Areoles 6–10 mm diam., 25–50 mm apart (much closer in juvenile forms), with white, greyish
or brownish, early deciduous hairs, subtended by red scale leaves when first emerging from apical
meristem. Central spines 1–3, 15–30 mm, radials 4–7, 3–15 mm. Flower-bearing region not
differentiated; young flower-buds covered in golden hair-spines to 15 mm long, flowers nocturnal,
15–22 × 12–17 cm; pericarpel greenish, covered in tubercles bearing areoles with bristles and triangular
bract-scales; flower-tube 5–7 cm, infundibuliform, 1–2 cm diam., bearing scattered areoles with
triangular bract-scales and hairs to 15 mm; perianth-segments to 6 cm, outer segments reflexed,
lanceolate to linear, fleshy, dark red with green shades, inner segments spreading, lanceolate and
fimbriate, delicate, white; stamens exerted in relation to the perianth-segments, curved, anthers linear;
style 13–20 × 0.3–0.4 cm, stigma-lobes 12–15, exserted; ovary locule ovoid to oblong in longitudinal
section. Fruit globose, to 10 cm diam., dehiscent by splitting irregularly, flower remnants deciduous;
pericarp greenish then bright orange-red, covered in podaria bearing woolly areoles; funicular pulp
white. Seeds c. 3–4 mm, cochleariform, black, shiny; testa-cells flat, smooth.
Widespread central-southern caatinga element: common along roadsides, often in farm hedges, amongst
semi-open vegetation or scrambling over rocks (inselbergs), caatinga-agreste, especially on soils containing
clay, c. 50–700 m, from north-western (Xique-Xique), central-northern and -eastern Bahia (drainage of
Rio Paraguaçu), from 13°S, northwards to southern Ceará and southern Paraíba (c. 7°S). Endemic to the
core area within North-eastern Brazil. Map 47.
piauí: reported by Luetzelburg (1925–26, 3: 111), but see comments below.
ceará: S Ceará, Mun. Barro, road CE 292, between Cuncas and border with PB, 7°6'S, 38°42'W, 4
Apr. 2000, E.A. Rocha et al. (K, photos); Mun. Milagres, 2 km E of town on road BR 116, 4 Apr. 2000,
E.A. Rocha et al. (obs.); Mun. Mauriti, ‘Mamalu’, road CE 296, shortly after the border with PB, 7°32'S,
38°40'W, 3 Apr. 2000, E.A. Rocha et al. (K, photos); Mun. Jardim, 1.5 km S of town towards Cedro
(PE), 4 Apr. 2000, E.A. Rocha et al. (obs.).
paraíba: Mun. Itaporanga, Serra Água Branca, 7–10 Jan. 1994, M.F. Agra et al. 2515, 2561 & 2562
(JPB); Mun. Desterro, 1998, E.A. Rocha (JPB); Mun. Conceição, planted in the town ex Serra do
Cardoso, 3 Apr. 2000, E.A. Rocha et al. (obs.); Mun (?), between Sumé and Monteiro, 22 Feb. 1962,
Castellanos 23269 (GUA); Mun. Monteiro, very common around the town, 11 Feb. 1995, Taylor (obs.).
pernambuco: Mun. Araripina, behind hotel ‘Pousada do Araripe’, 15 Feb. 1990, Zappi 226A (K, photos);
Mun. Exu, 3 km S of town, 7°32'S, 39°43'W, 5 Apr. 2000, E.A. Rocha et al. (obs.); Mun. Serrita, Fazenda
HARRISIA
Pirapora, 7°48'S, 39°14'W, 4 Apr. 2000, E.A. Rocha et al. (obs.); Mun. Triunfo, 8 Dec. 1991, F.A.R. Santos
15 (PEUFR, HUEFS); Mun. Serra Talhada, Estação experimental do IPA, 7°59'S, 38°19'16''W, 18 Jan.
1996, M.L. Gomes 125 (IPA); Mun. Parnamirim, road BR 122, Poço do Fumo, 8 Dec. 1971, Andrade-Lima
et al. 1219 (IPA, PEUFR); l.c., 18 km from the town, road BR 316, 0.5 km after the Rio Favela, 11 Feb.
1993, F.A.R. Santos 71 (ALCB, HUEFS, PEUFR); Mun. Custódia, Rio Salgado, road BR 232, 9 Dec.
1991, F.A.R. Santos 19 (PEUFR, HUEFS); Mun. Sertânia, Moderna, 8 Apr. 2000, E.A. Rocha et al. (obs.);
Mun. Arcoverde, W of town, 8 Apr. 2000, E.A. Rocha et al. (obs.); Mun. Brejo da Madre de Deus, Fazenda
Nova, 23 Jan. 1934, Pickel 3500 (IPA); Mun. Pesqueira, Ipanema, 12 Jan. 1956, Andrade-Lima 56-2488
(IPA); Mun. Alagoinha, 13/14 Feb. 1992, F.A.R. Santos 21, 24, 27 (PEUFR, HUEFS); Mun. Afrânio, 5
km NW of Rajada on road BR 407, 8°47'S, 40°52'W, 6 Apr. 2000, E.A. Rocha et al. (K, photos); Mun.
Belém do São Francisco, 12 km E of Cabrobó on road BR 428, 8 Apr. 2000, E.A. Rocha et al. (obs.); Mun.
Floresta, 8°36'S, 38°27'W, 8 Apr. 2000, E.A. Rocha et al. (obs.); Mun. Inajá, Reserva Biol. Serra Negra, 9
Dec. 1995, A. Laurênio et al. 265 (K, PEUFR); Mun. Petrolina, Dec. 1914, Luetzelburg 729 (M), l.c., 5 km
norte da CPATSA, 16 Aug. 1983, Fotius 3552 (IPA); l.c., road W of Petrolina towards village of Tapera,
immediately N of the Rio São Francisco, 9°27'S, 40°41'W, 7 Apr. 2000, E.A. Rocha et al. (obs.).
alagoas: Mun. Arapiraca, near Craíbas, Tingui, 22 Nov. 1983, Staviski 669 (MAC).
sergipe: Mun. Porto da Folha, 9.5 km from Monte Alegre de Sergipe towards Poço Redondo, 12 Feb.
1991, Taylor & Zappi (obs.); Mun. Nossa Senhora da Glória, Faz. Olhos D’agua, 21 Dec. 1981, G. Viana
331 (ASE 2006); Mun. Frei Paulo, 5 km após o povoado de Mocambo, 26 Mar. 1981, M. Fonseca et al.
450 & 451 (ASE 852 & 853).
bahia: NW Bahia, Mun. Barra, dunes by the Rio São Francisco, 10°47'16"S, 42°46'22''W, 23 June 1996,
Hind s.n. (K, photos); N & NE Bahia, Mun. Juazeiro, near Rodeadouro, 10 Jan. 1991, Taylor et al. 1388
(HRCB, K, ZSS, CEPEC); Mun. Uauá, 1.7 km ESE of town towards Bendengó, 6 Jan. 1991, Taylor et al.
1362 (HRCB, K, ZSS, CEPEC); Mun. Jaguarari, Barrinha, 8 June 1915, Rose & Russell s.n. (US); Mun.
Monte Santo, 14 July 1989, Taylor & Zappi (obs.); Mun. Itiúba, road to Filadélfia, 10°39'S, 39°44'W, 11 May
2002, Nascimento & Nunes 103 (HUEFS); Mun. Queimadas, 3 June 2000, C.T.S. Andrade (photos); Mun.
Jacobina, ‘margem esquerda do Rio Jacuípe’, 21 Jan. 1985, Passos 25 (BAH 6022); l.c., 27–28 km WNW of
town towards Lajes, 12 Jan. 1991, Taylor et al. (obs.); Mun. Morro do Chapéu, NE of Gruta dos Brejões, 4
Aug. 2002, Taylor (obs.); l.c., W of Icó, Brejão (Formosa), ‘Lagedo Bordado’, 5 Aug. 2002, Taylor (obs.); l.c.,
26 km E of town, near Ventura, 11°42'S, 40°58'W, 24 Dec. 1988, Taylor & Zappi in Harley 27387 (K, SPF,
CEPEC); Mun. Miguel Calmon, 30 km S of Jacobina, 26 Apr. 1992, Taylor & Zappi (obs.); Mun. São José
do Jacuípe, 20 km SE of Capim Grosso, 6 km NW of bridge over the Rio Jacuípe, 25 Apr. 1992, Taylor &
Zappi (obs.); Mun. Teofilândia, 2.5 km S of town, 16 km N of Serrinha towards Araci, 5 Jan. 1991, Taylor
et al. 1349 (HRCB, K, ZSS, CEPEC); Mun. Riachão do Jacuípe, 33 km NW of Tanquinho, 25 Apr. 1992,
Taylor & Zappi (K, photo); l.c., 24.5 km NW of Tanquinho, 25 Apr. 1992, Taylor & Zappi (obs.); Mun.
Lamarão, 15 July 1959, Castellanos 22441 (R); Mun. Tanquinho, 8 July 1994, F.A.R. Santos 93 (HUEFS);
Mun. Ipirá, 24 km E of town on road BA 052, 9 Feb. 1992, Taylor et al. (K, photos); Mun. Iaçu, 5.5 km W
of town towards Marcionílio Souza, 7 Feb. 1991, Taylor et al. 1580 (HRCB, K, ZSS, CEPEC); Mun. Rafael
Jambeiro, Km 16 on road BR 242, 29 Sep. 1973, A.L. Costa s.n. (ALCB); Mun. Castro Alves, May & Oct.
1972, G.C.P. Pinto s.n. (ALCB 02903, 02905); Mun. Itatim, 1997, F. França s.n. (K, photo.); Mun. Milagres,
10 km towards Itaberaba and then 2 km towards the Fazendas Morros and Antônio Romeu, 27 Jan. 1973,
I. & G. Gottsberger 11-27173 (K); Mun. Brejões, 3 km N of the junction to the town on road BR 116, Aug.
2001, Zappi (obs.); unlocalized, ‘Mina Poço, Mina Angico’, 27 Mar. 1966, Castellanos 25943 (HB).
Although reported from Piauí by Luetzelburg (1925–26, 3: 111), this species has not been
seen there by the authors. However, its occurrence at Araripina, Pernambuco is close to
the border with that state, although it is not known if suitable edaphic conditions exist
(ie. soils with clay content). Braun & Esteves Pereira (2002: 79, 128) indicate that it
occurs in Rio Grande do Norte and western Minas Gerais, but concrete evidence is
awaited and their distribution map (l.c. 78) is contradictory. If it does occur in Minas
Gerais, it will be important to determine whether it is truly native or introduced there.
HARRISIA
The places in which this species is commonly found today suggest that its dispersal and
establishment is assisted by man and his animals. Its fruits are edible and it is often seen in
farmhouse hedges.
25. LEOCEREUS Britton & Rose
Cact. 2: 108 (1920). Type: L. bahiensis Britton & Rose.
An isolated genus comprising a single species endemic to the core area of Eastern Brazil.
Its placement in Trichocereeae is provisional and awaits confirmation via DNA gene
sequence data.
Other species referred to Leocereus by Britton & Rose (l.c.) belong in Arthrocereus, except
for Cereus oligolepis Vaupel (Britton & Rose, l.c., 225), which is a Pilosocereus (Zappi 1994).
L. squamosus (Gürke) Werderm. (1933) is Facheiroa squamosa (Gürke) P. J. Braun & E.
Esteves Pereira. For L. paulensis Speg., see Coleocephalocereus fluminensis (Zappi & Taylor
1992b). Plates 66.1–66.2.
1. Leocereus bahiensis Britton & Rose, Cact. 2: 108 (1920); N. P. Taylor & Zappi in
Bradleya 8: 107–108 (1990). Type: Bahia, Mun. Sento Sé / Xique-Xique, Serra do Tinga
[‘Tingga’], Zehntner 266 (US, lecto. designated here; NY, K, lectoparas.).
Cereus bahiensis (Britton & Rose) Luetzelb., Estud. Bot. Nordéste 3: 111 (1926).
Leocereus urandianus F. Ritter, Kakt. Südamer. 1: 222 (1979). L. bahiensis var. urandianus (F. Ritter) P. J. Braun
& E. Esteves Pereira in Kakt. and. Sukk. 41: 205 (1990). L. bahiensis subsp. urandianus (F. Ritter) P. J.
Braun & E. Esteves Pereira in Succulenta 74: 132 (1990). Holotype: Bahia, Mun. Urandi, Ritter 1231
(U).
L. estevesii P. J. Braun in Kakt. and. Sukk. 41: 204 (1990). Holotype: SW Piauí, Mun. Corrente/Gilbués,
E. Esteves Pereira 207 (UFG 12380; ZSS, iso.).
L. bahiensis var. barreirensis P. J. Braun & E. Esteves Pereira in ibid. 41: 205 (1990). L. bahiensis subsp.
barreirensis (P. J. Braun & E. Esteves Pereira) P. J. Braun & E. Esteves Pereira in Succulenta 74: 132
(1995). Holotype: Bahia, Mun. Barreiras, E. Esteves Pereira 118 (UFG 12360; ZSS, iso.).
L. bahiensis var. exiguospinus P. J. Braun & E. Esteves Pereira in Kakt. and. Sukk. 41: 205 (1990). L.
bahiensis subsp. exiguospinus (P. J. Braun & E. Esteves Pereira) P. J. Braun & E. Esteves Pereira in
Succulenta 74: 132 (1995). Holotype: Bahia, Mun. (?), ‘Canabrava’, E. Esteves Pereira 135 (UFG [not
found Jan. 1991]; ZSS, iso.).
L. bahiensis var. robustispinus P. J. Braun & E. Esteves Pereira in Kakt. and. Sukk. 41: 205 (1990). L.
bahiensis subsp. robustispinus (P. J. Braun & E. Esteves Pereira) P. J. Braun & E. Esteves Pereira in
Succulenta 74: 132 (1995). Holotype: Bahia, Mun. Gentio do Ouro, P. J. Braun 702 (ZSS).
Erect or semi-scandent, to 2(–3) m high, with one or more seldom-branched stems arising from a woody
rootstock; stems cylindrical, very woody, lacking mucilage, bright or olive-green, to 200 × 1–2.5 cm; ribs
(10–)12–20, rounded, obtuse, low, c. 1–2 × 2–4.5 mm, sinuses sinuate; areoles circular, 1–2.8 mm, 4–7
mm apart. Spines 8–21, not clearly differentiated into centrals and radials, slender acicular, yellowish to
dark brownish or blackish, central(s) 1–9, ± erect, 6–20 mm, radials 7–12, adpressed, 3–5 mm. Flowers
subapical, nocturnal, tubular, 40–73 × 20–34 mm, white; pericarpel and tube to c. 50 mm, green, clothed
with 2–3 × 0.8–1.0 mm, brownish, triangular-acuminate bract-scales, pericarpel c. 9–10 × 6.5–7.5 mm,
with 4 mm, porrect, pale spines in the bract-scale axils, the tube with dark hairs and bristles to 12 mm;
nectar-chamber 12–17 × 8–12 mm, narrowed to 6–8.5 mm at base and to 7–11 mm above, apex of tube
12–18 mm diam.; perianth-segments 4–17 × 1.8–4.5 mm, outer series patent to reflexed, green to dull
FACHEIROA
brownish red, oblong-apiculate, inner series suberect, pure white, lanceolate; lowermost stamens with
thickened filaments protecting the nectar-chamber, 15–18 mm, anthers cream, 1.6–2.5 mm; style 30–40
× 0.7–1.0 mm, stigma-lobes 6–11, white, 1.5–4.5 × 0.6 mm. Fruit globose to ovoid, 23–36 × 16–32 mm,
red or purplish, pericarp very shiny, indehiscent, at maturity the spine-clusters becoming detached, floral
remains persisting, erect, funicular pulp magenta. Seeds 1.2–1.8 × 1–1.2 mm, black, glossy, hilum oblique,
testa-cells flat to slightly convex, with intercellular pits (Barthlott & Hunt 2000: pl. 27.6–8).
Widespread central-southern caatinga element: growing between and through shrubs, rocky and sandy
places in ‘caatinga de altitude’, campo rupestre and their ecotone, and amongst rocks and cliffs in cerrado (W
Bahia), 550–1500 m, central-northern Minas Gerais to northern Bahia (especially in the Serra do
Espinhaço and Chapada Diamantina), westwards to southern Piauí and in the Chapadão da Bahia (Espigão
Mestre) towards the border with Goiás. On present evidence, endemic to Eastern Brazil. Map 47.
piauí: Mun. Canto do Buriti, fide Uebelmann (1996): HU 663; Mun. Corrente/Gilbués, E. Esteves
Pereira 207 (UFG 12380, ZSS).
bahia: W Bahia, Mun. Roda Velha, fide Uebelmann (1996): HU 950; Mun. Barreiras, Serra da
Bandeira, edge of cliff overlooking Barreiras with Telebahia installations, 18 Jan. 1991, Taylor et al. 1436
(K, ZSS, HRCB, CEPEC); l.c. (?), Apr. 1981, E. Esteves Pereira 118 (UFG, ZSS); N Bahia, Mun. Sento
Sé, 10°7'S, 41°25'W & 10°19'S, 41°24'W, 11 July 2000, G. Charles (photos); l.c., Serra do Tinga
[‘Tingga’], Aug. 1912, Zehntner 266 (US, NY, K), Zehntner in IFOCS 2040 (M); Mun. Jaguarari,
Barrinha, 7–8 June 1915, Rose & Russell 19820 (K); Mun. Gentio do Ouro, nr Gameleira, July 1986,
Braun 702 (ZSS); Mun. Morro do Chapéu, W of Icó, Brejão (Formosa), ‘Lagedo Bordado’, 5 Aug. 2002,
Taylor (K, photos); l.c., c. 20 & 25 km W of Morro do Chapéu, S of road BA 052, 26 Dec. 1988, Taylor
& Zappi (obs.); Mun. Cafarnaum, 12 Nov. 1956, E. Pereira 2152 (RB); Mun. Barra do Mendes/Seabra,
c. 20 km W of Seabra, 1986, Braun 721 (ZSS); Mun. Ibitiara, Serra Malhada, c. 52 km W of Seabra on
road BR 242, 14 Jan. 1991, Taylor et al. 1416 (K, ZSS, HRCB, CEPEC); Mun. Seabra, 6 km E of Seabra
beside road BR 242, 25 Aug. 1988, Eggli 1283 (ZSS); S Bahia, Mun. Piatã/Boninal, 12 km S of Boninal
on road to Piatã, 12°48'S, 41°48'W, 22 Dec. 1988, Taylor & Zappi in Harley 25596A (SPF, K, photos);
Mun. Érico Cardoso [Água Quente], NW slopes of Pico das Almas, 13°29'S, 42°W, 11 Dec. 1988, Taylor
& Hind in Harley 25552B (K, photos); Mun. Rio de Contas, 3 km from the town towards Livramento,
13°36'S, 41°50'W, 23 Nov. 1988, Taylor & Zappi in Harley 25540 (K, SPF, CEPEC); Mun. Ituaçu, c. 2
km NE of the town, 18 Aug. 1988, Eggli s.n. (ZSS, photo), Ritter 1470 (U); Mun. Urandi, 1964, Ritter
1231 (U).
minas gerais: Mun. Monte Azul, mountains c. 9 km E of the town, 15°10'S, 42°48'W, 28 Jan. 1991,
Taylor et al. 1473 (K, ZSS, HRCB); Mun. Mato Verde, 12 km from Mato Verde on road to S. Antônio
do Retiro, 15°23'S, 42°45'W, 9 Nov. 1988, Taylor & Zappi in Harley 25521 (K, SPF).
Braun, l.c. and Braun & Esteves Pereira (2002: 79, 85, Abb. 95) report the synonymous
L. estevesii from southern Maranhão, but no corroborating material has been seen by the
present authors.
26. FACHEIROA Britton & Rose
Cact. 2: 173 (1920); P. J. Braun & E. Esteves Pereira (1986–89). Type: Facheiroa pubiflora
Britton & Rose (= Facheiroa ulei (Gürke) Werderm.).
Including Zehntnerella Britton & Rose (1920); Facheiroa subg. Zehntnerella (Britton &
Rose) P. J. Braun & E. Esteves Pereira (1986).
Columnar, erect, treelike, sometimes shrubby, to 8 m; central cylinder woody, solid. Branches 2.4–7.0
cm diam., erect, not constricted, epidermis dull to bright green, not glaucous; ribs 15–20, low, crenate,
FACHEIROA
sinuses straight. Areoles with felt and long hairs; spines weak, flexible to brittle, not very pungent, central
spines porrect to deflexed, longer than the numerous radials. Flower-bearing region of stem lateral,
subapical or flowers randomly disposed, when differentiated sometimes forming a cephalium that
transforms the structure of the stem. Flowers nocturnal, to 5 × 2.5 cm; pericarpel and tube covered in
triangular bract-scales subtending tufts of hair; tube cylindric, stout; sometimes with a tuft of hairs
between the nectar-chamber and the filament bases; innermost filaments curved towards the style;
perianth-segments short, triangular to lanceolate or spathulate, innermost white, pinkish or purplish. Fruit
globose to depressed-globose or pear-shaped, indehiscent, covered in scales and hairs, flower remnant
drying brownish, erect. Seeds c. 1–1.5 mm, cochleariform; testa-cells flat or convex; cuticular folds
present or absent.
A genus of 3 allopatric species, endemic within the core area of Eastern Brazil (Rio São
Francisco drainage), mainly in caatinga vegetation, and of uncertain relationship within the
Trichocereeae (but cf. Yungasocereus Ritter (1980) and Vatricania Backeberg (1950), both
from the eastern edge of the Andes in Bolivia, the latter genus nowadays sometimes
included within Espostoa Britton & Rose). Buxbaum (1959) published Espostoa subg.
Facheiroa (Britton & Rose) F. Buxbaum. Plates 66.3–67.4.
1. Flower-bearing part of stem strongly modified, rarely only poorly differentiated . . . . . . . . . . . . . 2

1. Flower-bearing part of stem not at all modified (cent.-S Bahia northwards, non-calcareous rocks)
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3. F. squamosa
2. Flowers to 3 × 2.8 cm; perianth-segments sometimes pinkish (Bambuí limestone outcrops of
SW Bahia and cent.-N Minas Gerais) . . . . . . . . . . . . . . . . . . . . . . . . . . . 2. F. cephaliomelana
2. Flowers 4–4.7 × 2 cm; perianth-segments white (cent.-N Bahia, non-calcareous rocks) . . . 1. F. ulei
1. Facheiroa ulei (Gürke) Werderm., Bras. Säulenkakt.: 113 (1933); P. J. Braun & E.
Esteves Pereira in Kakt. and. Sukk. 38: 26, Abb. 1 (1987); ibid. 39: 64 (1988). Holotype:
Brazil, Bahia, Mun. Gentio do Ouro, Serra de Santo Inácio, 1907, Ule 12 (B, n.v.).
Cephalocereus ulei Gürke in Monatsschr. Kakt.-Kunde 18: 85 (1908). Espostoa ulei (Gürke) Buxb. in
Oesterr. Bot. Zeitschr. 106: 155 (1959), nom. inval. (Art. 33.2).
Facheiroa pubiflora Britton & Rose, Cact. 2: 173 (1920) (‘publiflora’). Type: Brazil, Bahia, Mun. Xique-
Xique, Serra da Cana-brava, Oct. 1917, Zehntner in Rose s.n. (US, lecto. designated here; K,
lectopara.)
vernacular name. Facheiro-preto.
Treelike or often shrubby, to 6 m or more, much branched. Stems erect, to 7 cm diam.; ribs 18–20, 8–9
× 7–8 mm, epidermis dark green to grey-green. Areoles 4–5 mm, 10–12 mm apart, long hairs grey.
Spines flexible, brown, central spines 2–3, to 18 mm, radials 13–15, 3–15 mm. Flower-bearing region
strongly differentiated, lateral, sunken into stem, with greyish wool and bristles to 10 mm. Flowers 4–4.7
× 1.7–2.0 cm; pericarpel and tube wine-coloured, covered in lanceolate to triangular, 2–5 mm bract-
scales, bearing tufts of short, reddish to golden brown hairs; flower tube cylindric, to 30 × 22 mm;
perianth-segments 8–10 × 3.5–5.0 mm, lanceolate-spathulate, whitish; stamens included in relation to the
perianth-segments; nectar-chamber 10 × 12 mm; style 27 mm, stigma-lobes 10–15, exserted; ovary locule
saucer-shaped in longitudinal section, 3 × 8 mm. Fruit depressed-globose to pear-shaped, 3.5–6.0 × 4 cm.
Seeds c. 1 × 0.7 mm, broadly oval, brown (Barthlott & Hunt 2000: pl. 29.2–3).
Northern Rio São Francisco caatinga element: in caatinga at c. 500 m or more, north-western edges of the
Chapada Diamantina, from the region of Xique-Xique to the Serra da Chapada, central-northern Bahia.
FACHEIROA
bahia: Mun. Xique-Xique/Itaguaçu da Bahia, Serra da Cana-brava, Oct. 1917, Zehntner in Rose s.n. (K,
US); Mun. Xique-Xique/Gentio do Ouro, 11°6'S, 42°43'W, 24 June 1996, Guedes et al. in PCD 3025
(K, ALCB); Mun. Gentio do Ouro, Serra do Santo Inácio, 19 July 1972, Horst & Uebelmann 265 (U);
Mun. Brotas de Macaúbas / Barra do Mendes, S of Minas de Espírito Santo and Serra da Chapada,
reported by Braun & Esteves Pereira, l.c. (1988).
conservation status. Data Deficient [DD]. No population data and indications of threats to its
habitats are currently available for this species.
This taxon has not been studied in the field by the authors and little is known about its
ecology and relationship with other members of the genus, although it is clearly a
distinct species.
2. Facheiroa cephaliomelana Buining & Brederoo in Kakt. and. Sukk. 26: 121–124
(1975). Holotype: Brazil, Bahia, Santa Maria da Vitória, 25 July 1974, Horst & Uebelmann
447 (U).
F. pilosa F. Ritter, Kakt. Südamer. 1: 219 (1979). Holotype: Brazil, Minas Gerais, Mun. Januária, 1959,
Ritter 1000 (U).
F. tenebrosa P. J. Braun & E. Esteves Pereira in Kakt. and. Sukk. 39: 126–131 (1988). Holotype: Brazil,
Bahia, W of the Rio São Francisco [Mun. Cocos], E. Esteves Pereira 183 (KOELN
[‘Succulentarium’], n.v.; ZSS, UFG, K, isos.).
Treelike or usually shrubby, sometimes to 4 m or more, much branched. Stems erect, 2.4–6.0 cm
diam.; ribs 18–32, 4–8 × 4–9 mm; epidermis green to grey-green. Areoles 1.5–3.0 mm, 1–8 mm apart,
long hairs grey to brownish. Spines flexible, brown or golden, central spine(s) 1–5, to 40 mm, radials
10–12, 3–15 mm. Flower-bearing region differentiated, lateral, with brownish to greyish or white
wool and bristles to 10 mm. Flowers 2.8–3.5 × 1.9–2.8 cm; pericarpel and tube wine-coloured,
covered in lanceolate to triangular, 4.5–6.0 mm bract-scales, bearing tufts of short or long, white to
golden brown hairs in their axils; flower-tube 19–20 × 18 mm; perianth-segments 8–10 × 3.5–5.0 mm,
lanceolate-spathulate, pink; stamens included in relation to the perianth-segments; nectar-chamber 5–6
× 10–13 mm; style 23–27 mm, stigma-lobes 10–12; ovary locule saucer-shaped in longitudinal section,
3 × 8–10 mm. Fruit depressed-globose to turbinate, 2.5 × 1.8–2.5 cm. Seeds c. 1.2–1.5 mm,
cochleariform, brown, dull or shiny.
conservation status. Vulnerable [VU B2ab(iii)]; area of occupancy estimated to be < 2000 km2 and
decline projected in extent of habitat at its < 10 known locations (see below).
The distribution of F. cephaliomelana, as broadly circumscribed here, is paralleled or somewhat

exceeded by that of other similarly variable cactus taxa exclusive to Bambuí limestone
outcrops from the same region, eg. Tacinga saxatilis, Pilosocereus densiareolatus and Melocactus
levitestatus. It is divisible into two subspecies, the first of which is somewhat heterogeneous
and may merit further subdivision (cf. Braun & Esteves Pereira 2002: 120, Abb. 133):
1. Cephalium somewhat sunken into stem, conspicuous (SW Bahia & N Minas Gerais) . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2a. subsp. cephaliomelana
1. Cephalium superficial or only weakly developed (cent.-S Bahia, E of the Rio São
Francisco) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2b. subsp. estevesii
FACHEIROA
2a. subsp. cephaliomelana
vernacular names. Facheiro-preto, Xique-xique-preto.
Southern Rio São Francisco caatinga element: locally co-dominant on outcrops of raised Bambuí limestone
within caatinga/cerradão, 460–750 m, south-western Bahia (W of the Rio São Francisco) and central-
northern Minas Gerais. Endemic to the core area of Eastern Brazil. Map 48.
bahia: Mun. Santa Maria da Vitória, 25 July 1974, Horst & Uebelmann 447 (U); Mun. Serra do Ramalho,
Serra do Ramalho, 13°31'S, 43°45'W, 2001, J. Jardim 3486 (CEPEC); l.c. (?), ‘Mun. Bom Jesus da Lapa
Faz. Serra Solta’, 17 July 1975, Andrade-Lima 75-8203 (IPA); Mun. Cocos, July 1984, Esteves Pereira 183
(UFG, K).
minas gerais: Mun. Januária, 1959, Ritter 1000 (U); l.c., 3 km W on road MG 479 to Pandeiros, 11
Aug. 1988, Eggli 1135 (ZSS), l.c., 5 km towards Serra dos Araras, 5 Sep. 1985, Horst & Uebelmann 718
(K); Mun. Varzelândia, W of town, 12 Aug. 1988, Eggli 1147 (ZSS).
GD=2. Short-list score (2×7) = 14. Relatively rare, with only c.5 populations currently known, and
restricted to isolated limestone outcrops, which may be quarried in future leading to rapid decline in some
local populations.
This taxon has not been studied in the field by the present authors, but the characteristics
of the populations treated here have been recorded in some detail by Braun & Esteves
Pereira (1986–89).
2b. subsp. estevesii (P. J. Braun) N. P. Taylor & Zappi in Cact. Consensus Initiatives 3:
7 (1997). Holotype: Brazil, Bahia, E of the Rio São Francisco, [E of Malhada, Iuiú], July
1984, E. Esteves Pereira 186 (KOELN [‘Succulentarium’], n.v.; UFG, K, isos.).
Facheiroa estevesii P. J. Braun in Kakt. and. Sukk. 37: 74–79 (1986).
vernacular names. Facheiro-da-serra, Xique-xique-amarelo.
Southern Rio São Francisco caatinga element: on outcrops of raised Bambuí limestone amidst high caatinga
forest, 600–650 m, central-southern Bahia (E of the Rio São Francisco). Endemic to the locality cited
below. Map 48.
bahia: Mun. Iuiú, July 1984, E. Esteves Pereira 186 (UFG, K); l.c., 2 km S of Iuiú, July 1989, Taylor &
Zappi (K, photos), 28 Aug. 1988, Eggli 1315 (ZSS).
Short-list score (2×5) = 10. Restricted to a single, albeit extensive locality, near a town, where there is
the potential for limestone quarrying.
3. Facheiroa squamosa (Gürke) P. J. Braun & E. Esteves Pereira in Kakt. and. Sukk.
40: 202 (1989). Holotype: Brazil, Piauí, Serra Branca [Mun. São Raimundo Nonato],
Jan. 1907, Ule 10 (B; K, SPF photos ex B).
Cereus squamosus Gürke in Monatsschr. Kakt.-Kunde 18: 70 (1908). Leocereus squamosus (Gürke)
Zehntnerella squamulosa Britton & Rose, Cact. 2: 177 (1920). Type: Brazil, Bahia, Juazeiro, Serra do
Atoleiro, 4 June 1915, Rose & Russell 19760 (US, lecto. designated here; NY, lectopara.).
FACHEIROA
Z. polygona F. Ritter, Kakt. Südamer. 1: 214 (1979). Facheiroa squamosa var. polygona (F. Ritter) P. J. Braun
& E. Esteves Pereira in Kakt. and. Sukk. 40: 202 (1989). F. squamosa subsp. polygona (F. Ritter) P. J.
Braun & E. Esteves Pereira in Succulenta 74: 130 (1995). Holotype: Brazil, Bahia, Flamengo, 1963,
Ritter 1228 (U).
Zehntnerella chaetacantha F. Ritter, Kakt. Südamer. 1: 215 (1979). Facheiroa chaetacantha (F. Ritter) P. J.
Braun & E. Esteves Pereira in Kakt. and. Sukk. 40: 202 (1989). Holotype: Bahia, Guanambi, 1964,
Ritter 1229 (U).
Zehntnerella chaetacantha var. montealtoi F. Ritter, Kakt. Südamer. 1: 215–216 (1979). Facheiroa chaetacantha
var. montealtoi (F. Ritter) P. J. Braun & E. Esteves Pereira in Kakt. and. Sukk. 40: 202 (1989).
Holotype: Brazil, Bahia, Palmas de Monte Alto, 1964, Ritter 1229a (U).
F. deinacanthus Rauh ex Innes & Glass, Illustrated Encyclopaedia of Cacti: 112 (1991), nom. inval. (Arts
36 & 37).
vernacular names. Facheiro, Facheiro-preto.
Shrubby to treelike, to 6 m or more, much branched. Stems erect, old branches weak, becoming curved,
4–8 cm diam.; ribs 13–24, 3–5 × 3–4 mm; epidermis green to grey-green. Areoles to 2 mm, 4–7 mm apart,
long hairs grey to brownish. Spines flexible, brown, central spines 1–8, to 30 mm, radials 8–18, 3–15 mm.
Flower-bearing region not differentiated, flowers lateral, subapical or randomly distributed. Flowers 3–5 ×
2–3 cm; pericarpel and tube green, covered in lanceolate to triangular, green 6–8 mm bract-scales, bearing
tufts of short, white to golden or reddish brown hairs; flower-tube 20–40 × 17–22 mm, cylindric or
infundibuliform; perianth-segments 8–9 × 3–4 mm, lanceolate spathulate, white; stamens included in
relation to the perianth-segments; a tuft of hairs between the nectar-chamber and the filaments bases
appearing randomly; nectar-chamber 8–9 × 8 mm; style 23–27 mm, stigma-lobes 10–12; ovary locule
saucer-shaped in longitudinal section, 2–4 × 8–10 mm. Fruit globose to depressed-globose, 2–3.5 cm diam.
Seeds c. 1–1.2 mm, broadly oval, dark reddish brown, dull (Barthlott & Hunt 2000: pl. 29.5–6).
Rio São Francisco caatinga element: locally co-dominant with other arborescent cacti on non-calcareous
(mostly granite/gneiss) inselbergs or lajedos and in very stony ground of the caatinga (rarely on arenitic
rocks in cerrado or at the margins of campo rupestre), 390–1020 m, south-eastern Piauí, western
Pernambuco and northern Bahia, and disjunctly in central to southern Bahia, eastwards to the watershed
with the Rio de Contas drainage system. Endemic to the core area within North-eastern Brazil. Map 48.
piauí: Mun. São Raimundo Nonato, Serra de Capivara, Feb. 1990, Taylor & Zappi (obs.); l.c., Serra
Branca, Jan. 1907, Ule 10 (K, SPF [photos ex B]), Feb. 1990, Taylor & Zappi (K, photos).
pernambuco: Mun. Petrolina, road BR 122 [428], Serra [de] Santa, 18 Nov. 1992, F.A.R. Santos 49
(HUEFS, PEUFR); l.c., 20 km NE of Petrolina on road to Recife, 13 Aug. 1966, Hunt 6491 (K).
bahia: N Bahia, Mun. Juazeiro, 14 km E of town towards Curaçá, 7 Jan. 1991, Taylor et al. 1372 (K,
HRCB, ZSS, CEPEC); l.c., Serra do Atoleiro, 4 June 1915, Rose & Russell 19760 (US, NY); Mun. Casa
Nova, Santana, Crista da Serra, 4 Jan. 1984, Fotius 3672 (IPA); Mun. Sento Sé, 13.5 km S of main road
to Juazeiro in the direction of Cabeluda, 11 Jan. 1991, Taylor et al. 1395 (K, HRCB, ZSS, CEPEC); l.c.,
10°7'S, 41°25'W, 11 July 2000, G. Charles (photos); Mun. Sento Sé / Campo Formoso, Serra São
Francisco, reported by Buining & Brederoo (1974b: 257, fig. 8); Mun. Jaguarari, Flamengo, 1963, Ritter
1228 (U); cent. & S Bahia, Mun. Ibotirama, 14 km E of town on road BR 242, 18 July 1989, Taylor &
Zappi (obs.); l.c. (?), 12°13'18"S, 43°8'48''W, 2000, F.M. Sene J78 ‘sp. B’ (K, photos); Mun. Ibitiara, Serra
Malhada, 52 km W of Seabra on road to Ibotirama, 1020 m, 14 Jan. 1991, Taylor et al. 1418 (K, HRCB,
ZSS, CEPEC); Mun. ‘Seabra’, July 1987, Esteves Pereira 238 (UFG); Mun. Rio do Pires, Ibiajara, fide
Uebelmann (1996): HU 268; Mun. Macaúbas, fide Uebelmann (1996): HU 127; Mun. Paramirim, near
Rio Paramirim, 13°26'S, 42°15'W, 28 Nov. 1988, Taylor in Harley 25557 (K, SPF, CECEC), ibid.,
between Caraíbas and Paramirim, north of the road from Livramento, 30 Nov. 1988, Taylor (K, photos);
Mun. Riacho de Santana, 22 km from Bom Jesus da Lapa, 16 Apr. 1983, Leuenberger et al. 3074 (CEPEC,
B), 20 July 1989, Taylor & Zappi (obs.); Mun. Rio de Contas, southern limits of município bordering on
Mun. Brumado, 13°49'39"S, 41°34'25"W, ‘Boa Sentença’, 26 Jan. 2001, Harley & Giulietti (K, photos);
Mun. Guanambi, 1964, Ritter 1229 (U); Mun. Palmas de Monte Alto, 1964, Ritter 1229a (U), 21 July
1989, Taylor & Zappi (obs.).
FACHEIROA
A rather variable species in terms of stem size, rib number, flower size and flower-tube
indumentum colour/abundance etc., the variation being only partly correlated on a regional
basis. The characters utilized by Braun & Esteves Pereira to maintain the southern-ranging
variant, F. chaetacantha, as a separate species do not seem to be consistent on the basis of the
materials studied here. The distribution of F. squamosa sens. lat. is markedly disjunct, being
interrupted in central-northern Bahia (drainages of Rio Salitre and Rio Jacaré) due to the
presence of limestone derived substrates, which it appears to avoid, and in the north-
western part of the Chapada Diamantina, where it is replaced by F. ulei.
27. ESPOSTOOPSIS F. Buxbaum
in Krainz, Die Kakteen, Lfg. 38–39 (July 1968). Type and only species: E. dybowskii
(Roland-Gosselin) Buxb. (Cereus dybowskii Gosselin).
A monotypic genus endemic to the caatinga of Bahia and possibly related to Espostoa
Britton & Rose (sens. str.), from the central Andes of southern Ecuador and northern
Peru. It differs from Espostoa in its naked pericarpel and tube, in which it strongly
resembles some cephalium-bearing members of tribe Cereeae (cf. Micranthocereus Backeb.,
sens. lat.), but in other respects it is extremely similar to the Andean genus and could easily
be mistaken for it when in sterile condition. Gerocephalus F. Ritter (see synonymy below)
is an illegitimate generic name based on Cereus dybowskii, but published one month later
than Espostoopsis Buxb. Plate 68.1–68.2.
1. Espostoopsis dybowskii (Gosselin) Buxb., l.c. (1968). Type: Brazil, N Bahia,

Itumirim, Dybowski (living material), Casabianca (fls & frs), apparently not extant (cf.
Kiesling 1986). Neotype (designated here): Brazil, Bahia, Mun. Jaguarari, 38 km N of
Senhor do Bonfim on road BR 407, 15 July 1989, Zappi 125 (SPF; HRCB, ZSS
isoneotypi; K, photos).
Cereus dybowskii Gosselin in Bull. Soc. Bot. France 55: 695 (1908). Cephalocereus dybowskii (Gosselin)
Britton & Rose, Cact. 2: 30 (1920). Espostoa dybowskii (Gosselin) Frič in Kreuzinger, Verzeichnis: 6
(1935). Austrocephalocereus dybowskii (Gosselin) Backeb. in Cact. Succ. J. (US) 23: 149 (1951).
Gerocephalus dybowskii (Gosselin) F. Ritter in Kakt. and. Sukk. 19: 156 (Aug. 1968). Coleocephalocereus
dybowskii (Gosselin) F. Brandt in Kakt. Orchid. Rundschau 6: 124 (1981). Micranthocereus dybowskii
(Gosselin) D. Herbel in Kakt. and. Sukk. 49: 275 (1998), nom. inval. (Art. 33).
vernacular names. Cabeça-de-velho, Homem-velho, Cabeça-branca, Mandacaru-cabeça-branca,

Mandacaru-de-penacho [‘perracho’].
Columnar, erect, shrubby to almost treelike, to 5 m, trunk to 80 cm diam. Stems erect, 4–8 cm diam.,
often braking in the region of the cephalium; vascular cylinder woody, solid, very hard, with scattered
perforations, cortex scarcely mucilaginous; ribs 24–33, 2 × 5–6 mm, strongly tuberculate, sinuses
straight; epidermis light green with numerous pale stomata; areoles 2–4 mm, to 6 mm apart, with
abundant, long, white to grey hairs covering the stem. Spines flexible to brittle, golden or reddish
ESPOSTOOPSIS
brown, central spines 1–2, to 2(–2.5) cm, radials 10–15, 0.3–1.5 cm. Flower-bearing region a strongly
differentiated, lateral cephalium sunken into the stem and markedly protruding from its surface, with
abundant, very compact white wool and bristles to 2.5 cm, the flowers developing in an irregular
manner (not in order of age of cephalium areoles). Flowers 4–4.2 × 3.5–3.8 cm, all parts turning orange
at the slightest touch, then blackening; pericarpel 7 × 10 mm, pale green with minute and scarcely
visible bract-scales; tube to 30 × 22 mm, cylindric, white to pale pink, naked, with broad bract-scales
only at apex; perianth-segments 6–7 × 3 mm, lanceolate, acuminate, white; stamens included in
relation to the perianth-segments; nectar-chamber 12 × 6–8 mm; style 28–35 mm, stigma-lobes c.
15–16, exserted; ovary locule 6 × 3–4 mm, hemiglobose in longitudinal section. Fruit depressed-
globose, 20 × 22 mm, slightly compressed laterally, flower remnants deeply sunken into apex and
carrying with them hairs from the cephalium; pericarpel wrinkled, olive-green, pale pink or reddish,
becoming exposed at cephalium surface or ± buried. Seeds c. 1.5 mm, cochleariform, black, shiny
(Barthlott & Hunt 2000: pl. 50.1).
Eastern caatinga element: dominant or co-dominant on gneiss/granite inselbergs or quartzitic rock

outcrops and in the surrounding caatinga, 300–750 m, central-northern and eastern Bahia (markedly
disjunct). Endemic to Bahia. Map 48.
bahia: cent.-N Bahia, Mun. (?), ‘Juazeiro’, 20 Oct. 1967, A.P. Duarte 10589 (RB); Mun. Jaguarari, 38
km N of Senhor do Bonfim on road BR 407, 15 July 1989, Zappi 125 (SPF, HRCB, ZSS; K, photos);
l.c., 86 km S of Juazeiro, between Flamengo and Juacema, 1 km W of road BR 407, 8 Jan. 1991, Taylor
et al. 1384 (K, HRCB, ZSS, CEPEC); l.c., Itumirim/Flamengo, 8 Feb. 1970, A.L. Costa s.n. (ALCB, in
spirit); ibid., Barrinha, 7–8 June 1915, Rose & Russell 19785 (US, NY); E Bahia, Mun. Maracás, 6–7 km
E of Porto Alegre towards Jequié, 4 Feb. 1991, Taylor et al. 1551 (K, HRCB, ZSS, CEPEC); Mun.
Jequié, E of Porto Alegre on dirt road connecting with road BR 330, gneissic inselberg, 4 Feb. 1991,
Taylor et al. (obs.); Mun. Tanhaçu, nr Sussuarana, Serra Escura, Sep. 2001, Avaldo de Oliveira Soares Filho
(photos), 19 April 2003, Taylor et al. (K, photos).
conservation status. Endangered [EN B2ab(iii)] (3); area of occupancy < 100 km2; PD=4, EI=2,
GD=2. Short-list score (3×8) = 24. All populations of limited extent, the northern one affected by
clearance of the caatinga in some places, the central of the three southern ones recently burnt. Creation
of reserves to accommodate this and associated species at one or more of its southern locations in the Rio
de Contas valley (between Sussuarana and Jequié) is recommended.
The distribution of this isolated species is markedly disjunct between northern and eastern
Bahia (Rio de Contas valley), the southern form showing some differences (notably
smaller stems and more exserted olive-brown fruits) that could justify subspecific status.
The better-known northern population, whence the type came, covers an extensive
area within the município of Jaguarari and in some parts of this region the plant
dominates the vegetation, forming impenetrable groves around and upon quartzitic
outcrops. It may range beyond this area into the neighbouring municípios of Sento Sé,
Campo Formoso, Senhor do Bomfim, Itiúba and further east to Jeremoabo, as implied
by Andrade-Lima (1989: 6) and Luetzelburg (1925–26, 3: 69), but, despite being so
conspicuous, there are no collections or other reports to confirm its presence in these
other localities and in northern Bahia it has not been seen outside Mun. Jaguarari by the
present authors.
The southern populations are found on either gneissic inselbergs or quartzite, but are
associated with a quite different suite of caatinga Cactaceae. Its disjunct distribution and
the differences between the plants from the two areas implies that the species may be a
relict in decline.
ARTHROCEREUS
28. ARTHROCEREUS A. Berger
Kakteen: 337 (1929), nom. cons.; F. Knuth, Den Nye Kaktusbog: 111 (1930). Type:
Monatsschr. Kakt.-Kunde 28: illus. opp. p. 62 (1918), as ‘Cereus damazioi’, typ. cons. (=
A. glaziovii (K. Schum.) N. P. Taylor & Zappi).
Literature: Taylor (1991d, 1992a); Brummitt (1994).

Shrubby or solitary, 1–2(–5) m tall, basally branched, sometimes with a thickened, corky rootstock
(adaptation to withstand fire?). Ribs 9 or more, low. Areoles close together, with felt and long hairs.
Spines golden or ferrugineous, translucent, weakly differentiated into centrals and radials, centrals
ascending or porrect, pungent, radials numerous, adpressed, bristle-like. Flower-bearing part of stem
not differentiated. Flowers salverform, nocturnal, white, or pink and remaining open in the early
morning in A. rondonianus (secondary bird-pollination syndrome?); pericarpel globose, it and the tube
bearing few to many bract-scales with hair-spines in their axils; tube elongate, cylindric to
infundibuliform, sometimes ± S-shaped; perianth-segments mostly narrow; inner segments white to
lilac-pink; stamens usually inserted in 2 series, when distinguishable the outer series in a ring at the
mouth of the tube; ovary locule hemicircular or rounded in longitudinal section. Fruit indehiscent,
globose to obpyriform, with bract-scales and areoles bearing long hair-spines or bristles, funicular pulp
white. Seeds ovoid, 1.2–2.0 mm.
An interesting genus endemic to Brazil, with 3 species in Eastern Brazil (campos rupestres
of central to SW Minas Gerais) and a fourth, A. spinosissimus (Buining & Brederoo) F.
Ritter, which has flowers with striking purple stamens rather similar to No. 2 below,
geographically isolated in Mato Grosso (Chapada dos Guimarães — Map 13). This
disjunct taxon has been separated into subgenus Chapadocereus P. J. Braun & E. Esteves
Pereira (1995: 82), but its similarities to A. melanurus and A. rondonianus would seem to
indicate that subdivision of the genus on geographical lines is untenable. The genus is
assumed to be related to Echinopsis Zucc. (sens. lat.), but differs in its pollen (Leuenberger
1976), indehiscent fruits and unusual habit form.
In the past the genus Arthrocereus has been used in a broader sense than now, including
species presently referred to Echinopsis, ie. E. mirabilis Speg. (Argentina), and Pygmaeocereus
H. Johnson & Backeb. (= Echinopsis), ie. P. bylesianus Backeb. (Peru).
The name of the genus was approved for conservation with a new type, as above, in
1993 (XV Int. Bot. Congress, Tokyo), following the discovery that the former type, Cereus
microsphaericus K. Schum. (and its illegitimate, homotypic synonym, C. damazioi Weingart),
was misapplied by Berger and is actually identifiable with a species of Schlumbergera Lem.
(Rhipsalideae). However, more recently, Eggli & Nyffeler (1996) have cast doubt on
whether the generic name can be accepted as validly published by Berger (1929), although
this has been refuted by Doweld & Greuter (2001). Plates 68.3–70.3.
1. Inner perianth-segments pink, stamens purple (Serra do Cabral & Conselheiro Mata)
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2. A. rondonianus
1. Inner perianth-segments and stamens white or cream (Serra do Espinhaço and Serra da
Mantiqueira) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
2. Flower-tube ± naked, with only few bract-scales and hairs; stems usually decumbent to
ascending, often segmented, sometimes very short to nearly spherical (usually on canga
formation) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3. A. glaziovii
ARTHROCEREUS
2. Flower-tube with conspicuous bract-scales and hairs; stems ± erect, not segmented (on
substrates other than canga) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1. A. melanurus
1. Arthrocereus melanurus (K. Schum.) L. Diers, P. J. Braun & E. Esteves Pereira in

Kakt. and. Sukk. 38: 312–315 (1987). Type (syntypes): ‘Brasilia’, (?) Minas Gerais, Sello
1000 (B†; an extant isosyntype at MO resembles A. glaziovii); Serra de São João del Rey
[Serra do Lenheiro], Glaziou s.n. (living plant, assumed not to have been preserved, but
illustrated by Schumann, l.c. infra, tab. 39, 1890). If an extant duplicate of the Glaziou
collection cannot be found, this name will require conservation, or its application may
have to change in line with the Sello isosyntype at MO (cf. ICBN Art. 9.10).
Cereus melanurus K. Schum. in Martius, Fl. bras. 4(2): 200, tab. 39 (1890). Leocereus melanurus (K. Schum.)
Britton & Rose, Cact. 2: 109 (1920).
A. mello-barretoi Backeb. & Voll. in Arch. Jard. Bot. Rio de Janeiro 9: 157 (1949, publ. 1950). A. melanurus
subsp. mello-barretoi (Backeb. & Voll) P. J. Braun & E. Esteves Pereira in Succulenta 74: 82 (1995).
Holotype: Brazil, Minas Gerais, Serra do Lenheiro [São João del Rei], Mello-Barreto s.n. (RB 65044)
— this is the type locality of A. melanurus!
A. melanurus var. estevesii L. Diers & P. J. Braun in Kakt. and. Sukk. 39(5): 100–105 (1988). A. melanurus
subsp. estevesii (L. Diers & P. J. Braun) P. J. Braun & E. Esteves Pereira in Succulenta 74: 82 (1995).
Holotype: Brazil, Minas Gerais, São João del Rei, Horst et al. 1 (KOELN, n.v.; K, iso.; ZSS iso.,
numbered as Braun 458) — this is the type locality of A. melanurus!
vernacular name. Sabugo-do-capeta.
Shrubby or single-stemmed, 0.4–3.0(–5.0) m tall, generally with a corky, thickened rootstock. Branches
erect, not segmented; epidermis bright green; ribs 9–19. Areoles to 3 mm diam., 4–6 mm apart, felt
greyish, basal areoles with indeterminate growth. Flowers nocturnal, sweetly scented; pericarpel and tube
± covered in triangular bract-scales with brownish woolly hairs; perianth-segments linear, acuminate,
fleshy, pale green with brownish shades, strongly reflexed, touching the outside of the flower tube at full
anthesis, inner segments linear-lanceolate, slender, delicate, cream or whitish, erect; stamens in 2 series;
stigma-lobes exserted. Seed: Barthlott & Hunt (2000: pl. 38.1–2).
conservation status. Vulnerable [VU B2ab(iii)]; area of occupancy estimated to be < 2000 km2,
number of known locations < 10 and some of these subject to anthropogenic alterations (see below).
The following subspecies are recognized:
1. Flowers 9–11.5 cm; inner perianth-segments 23–40 mm (SW Minas Gerais) . . 1a. subsp. melanurus
1. Flowers 10–16.5 cm; inner perianth-segments 40–55 mm (Cent. & S Minas Gerais) . . . . . . . . . . 2
2. Plants with few or no basal branches, reaching 2 m or more, stout rootstock not visibly
developed; ribs 12–19; perianth-segments spathulate, rounded-apiculate (Serra do Ibitipoca)
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1b. subsp. magnus
2. Plants shrubby, branching freely at base, usually not > 1 m, stems slender; rootstock well-
developed; ribs 9–12; perianth-segments tapering more evenly to apex (N of Diamantina
& Serra do Cipó) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1c. subsp. odorus
1a. subsp. melanurus

Shrubby, to 80 cm tall, with a corky, thickened rootstock. Branches to 4.5 cm diam., tapering to 1.5 cm,
erect; ribs 10–17, 3 × 5 mm. Spines yellow to reddish brown, central spines 5–7, stout, porrect, to 3.2
cm, radials numerous, thin to bristle-like and flexible, 4–5 mm. Flowers 9.2–11.5 × 6.5–8.0 cm;
ARTHROCEREUS
pericarpel and tube completely covered in triangular bract-scales with densely hairy areoles and yellow,
brown or ferrugineous stiff hairs to 1.5 cm; tube 5–6 × 1 cm at its narrowest point; outer perianth-
segments to 23–40 × 4–6 mm, inner segments to 32 × 3.5–5.0 mm; the innermost stamens inserted at 2.5
cm above pericarpel, with long filaments, the outermost series at the apex of tube, filaments 10 mm,
anthers to 1.9 mm; ovary locule to 7 mm diam.; style 6–7.7 cm, stigma-lobes 12–13, exserted, 8–10 ×
0.4 mm, slender. Fruit rounded, c. 3.0 cm diam., pericarp green, densely covered in bract-scales and
trichomes. Seeds c. 1.1 mm, testa-cells convex, with cuticular folds.
South-eastern campo rupestre element: in sandy, gravelly and rocky places in campo rupestre, 600–1000 m,
drainage of the Rio Grande, south-western Minas Gerais. Endemic to Minas Gerais. Map 49.
minas gerais: Mun. Alpinópolis, Furnas, 1994, N.M. Dinez (K, photo); Mun. São João del Rei, Serra
do Lenheiro, 1936, Mello-Barreto s.n. (RB 65044); Mun. Tiradentes/São João del Rei, Serra de São José,
24 Feb. 1987, Zappi et al. in CFCR 10338 (SPF), 11 Nov. 1988, Taylor & Zappi in Harley 25524 (SPF,
K), s.d., Horst et al. 1 & Horst 594 (K), Braun et al. 458 (ZSS), 3 Nov. 1990, E. Tameirão Neto 594 (K,
BHCB); Mun. Itutinga, between Lavras and São João del Rei, 56 km from the latter, 13 Oct. 1992, V.C.
Souza & C. Sakuragui 2105 (ESA, SPF); Mun. São Tomé das Letras, 29 Sep. 1968, L. Monteiro 116 &
M.C. Vianna 335 (GUA), 19 Dec. 1971, L. Monteiro s.n. (GUA).
GD=1. Short-list score (2×5) = 10. The type locality is very close to an expanding town and other sites
may be subject to mining activities.
It would not be surprising if this taxon were to be found in NE São Paulo.
1b. subsp. magnus N. P. Taylor & Zappi in Cact. Consensus Initiatives 3: 7 (1997).
Holotype: Brazil, Minas Gerais, Mun. Lima Duarte, Parque Estadual do Ibitipoca, 27 July
1991, Zappi et al. 262 (SPF; CESJ, HRCB, K isos.).
Erect, single-stemmed or somewhat branched at base or above, to 3(–5) m. Stems stout, 3.5–5.0(–8.0)
cm diam.; ribs 12–19, 3 × 7 mm. Spines gold to golden brown, central spines to 2–3.5 cm, radials 8–10
mm. Flowers c. 16.5 × 12.0 cm, pericarpel c. 1.5 × 2.0 cm; nectar-chamber and flower-tube together 10
cm, greenish, expanding from 1.5 cm diam. at base to 2.5 cm diam. at apex, bearing greenish bract-scales
to 2 × 1 mm, with 15 mm brownish hair-spines in their axils, the interior of the tube (throat) with faint
reddish brown markings; perianth-segments spathulate, the outermost 30–45 × 5–10 mm, green with
brownish midrib, inner segments to 55 × 20 mm, pure white, apex rounded-apiculate; stamens in 2 well-
marked series, the outermost (throat circle) to c. 35 mm, the inner to 120 mm, bearing cream-coloured
anthers to 3 × 1 mm; style c. 13.5 cm × 1.5 mm, greenish white; stigma-lobes c. 13, to 10 × 1 mm, cream.
Fruit to 45 × 50 mm; pericarp green with orange tinge, areoles hairy. Seeds 1.9–2.0 mm.
South-eastern campo rupestre element: amongst sand and rocks in campo rupestre, 1200–1700 m, Serra do
Ibitipoca, southern Minas Gerais. Endemic to the core area within Minas Gerais. Map 49.
minas gerais: Mun. Lima Duarte, Parque Estadual do Ibitipoca, 27 July 1991, Zappi et al. 262 (SPF,
CESJ, HRCB, K), ibid., cult. R.B.G. Kew, 6 Aug. 2001 (K, flower in spirit, photos), l.c., H.C. Souza
s.n., (K, photo), 1 Jan. 1982, A. Vilaça 193 (GUA), s.d., anon. in F. Sene s.n. (K, photos), 3–5 Aug. 2003,
Taylor et al. (K, photos); l.c., 17 km from Lima Duarte on road to the Parque Estadual, 21 Sep. 1994,
Splett 616 (K); l.c. (?), ‘Juiz de Fora’, Oct. 1969, Krieger 7379, 7525 & 7527 (CESJ).
PD=3, EI=1, GD=1. Short-list score (1×5) = 5. Population occurs within a protected area (Parque
Estadual do Ibitipoca), the management quality of which will determine whether this plant becomes
threatened in future. At present the park is well managed and the plant appears safe.
Distinct from other races of this species for its much greater size and broader, markedly
ARTHROCEREUS
spathulate-apiculate perianth-segments. The first illustration of this giant subspecies in the

above-cited habitat is that in Leme & Marigo (1993: 68). It develops the densest colonies
and largest specimens in sandy flat areas, where there is less competition.
The plant recently illustrated as from Ibitipoca by Braun & Esteves (2003: 23,F) may
be incorrectly captioned — it appears to be the pale-spined form of A. rondonianus from
Diamantina.
1c. subsp. odorus (F. Ritter) N. P. Taylor & Zappi in Cact. Consensus Initiatives 3: 7
(1997). Holotype: Brazil, Minas Gerais, Serra do Cipó, 1964, Ritter 1354 (U).
? Cereus parvisetus Otto ex Pfeiff., Enum. cact.: 79 (1837); K. Schum., Gesamtb. Kakt.: 67 (1897). Type:
Brazil [Minas Gerais, Serra da Lapa, Riedel, fide Schumann, l.c.], no longer extant.
Arthrocereus odorus F. Ritter, Kakt. Südamer. 1: 225 (1979).
Shrubby, to 1 m tall, branching at base, with a corky, thickened rootstock. Branches to 2–3.3 cm diam.,
erect or hanging from cliffs; ribs 9–12, 2–3 × 5 mm. Spines yellow to reddish, central spines 6–7, stout,
the largest porrect, to 4(–6) cm, radials numerous, thin to bristle-like and flexible, 4–5 mm. Flowers
10–16 × 8–10 cm; pericarpel and tube partially covered in triangular bract-scales with pale to pinkish
brown hairs to 15 mm in their axils; tube 4–8 cm × 8–9 mm at its narrowest point, ridged; outer perianth-
segments 40 × 5–7 mm; inner segments 40–50 × 6–9 mm; the innermost stamens inserted at 4 cm above
pericarpel, with long filaments, the outermost series at the apex of tube, filaments 15–18 mm, anthers to
2.2 mm; ovary locule 8 mm diam.; style 10–13 cm, stigma-lobes 12–13, 10 mm, slender. Fruit rounded
to ovate, 3–4 cm diam., pericarp green, covered in bract-scales and pale brown trichomes. Seeds c.
1.2–1.8 mm, testa-cells convex, with cuticular folds.
South-eastern campo rupestre element: amongst rocks in campo rupestre, 600–900 m, central-southern Minas
minas gerais: Mun. Diamantina, Mendanha, Uebelmann HU 1555, cult. G. Charles (photo), l.c.,
illustrated in Herm et al. (2001: 91); Mun. Santana do Riacho, path to the Serra das Bandeirinhas, near
Cachoeira da Farofa, 13 Sep. 1987, Zappi et al. in CFSC 10419; Mun (?) Santana do Riacho, [W foot of
the] Serra do Cipó, 1964, Ritter 1354 (U); l.c., Mun. Jaboticatubas, along the Rio Cipó, 28 Oct. 1988,
Taylor & Zappi in Harley 25430 (K, SPF), 20 Nov. 1989, Zappi et al. 193 (HRCB, SPF), 1997, Ribeiro de
Andrade & Pimentel Mendonça (K, photo).
conservation status. Vulnerable [VU D2] (2); area of occupancy estimated to be < 20 km2; PD=3,
EI=1, GD=1. Short-list score (2×5) = 10. Certainly known from only 3 localities in two distant and very
limited areas within the Serra do Espinhaço.
2. Arthrocereus rondonianus Backeb. & Voll in Cact. Succ. J. (US) 23: 120 (1951).
Type: Brazil, Minas Gerais, Diamantina, 700 m (presumed not to have been preserved).
Lectotype (designated here): Blätt. Kakteenforsch. 1935(4): [unpaged] (1935), illus.
‘Arthrocereus rondonianus Bckbg. et Voll n.sp.’; the same illustration also in Arch. Jard. Bot.
Rio Janeiro 9: 158, fig. 1 (1949, publ. 1950) and Backeberg, Die Cact. 4: 2110, Abb.
1990 (1960).
Shrubby or single-stemmed, to 1 m tall or more, with a corky, thickened rootstock; stems 1.5–3.0(–4.0)
cm diam., erect, not segmented unless damaged; epidermis bright green; ribs 13–18, 3 × 4 mm. Areoles
to 3 mm diam., 5–7(–10) mm apart, felt greyish. Spines yellow to pale brown (or whitish in cultivation),
central spines 6–7, stout, the largest porrect, to 7 cm, radials numerous, thin to bristle-like and flexible,
5–10 mm. Flowers nocturnal, remaining open until early morning, 10–13 × 9 cm, not scented; pericarpel
ARTHROCEREUS
and tube with scattered triangular bract-scales and pale to pinkish brown hairs to 1.5 cm; tube 6–7 × 0.8
cm at its narrowest point, often S-shaped and ridged, the limb being directed horizontally (erect in other
species); outer perianth-segments to 50 × 4 mm, linear-lanceolate, fleshy, pale green with pinkish shades
or pale pink, strongly reflexed, inner segments to 40 × 6–7 mm, lanceolate, delicate, lilac-pink, erect to
spreading; stamens in two series, the innermost inserted at 4 cm above pericarpel, with long filaments,
the outermost series at the apex of tube, filaments 15–18 mm, anthers purple; style 8–12 cm, stigma-lobes
10–11, exserted, whitish. Fruit rounded to ovoid, 4 × 3 cm; pericarp purplish to brownish, with bract-
scales, bristles and hairs. Seeds c. 2 mm, testa-cells convex, with cuticular folds.
South-eastern campo rupestre element: between rocks and in bushy places, campo rupestre, c. 700–1200 m,
Serra do Cabral and west slope of Serra do Espinhaço, central Minas Gerais. Endemic to the core area
within Minas Gerais. Map 49.
minas gerais: Mun. Joaquim Felício, Ritter 1355 (SGO 124966, fide Eggli et al. 1995: 533); Mun.
Buenópolis, Serra do Cabral, 13 Oct. 1988, Taylor & Zappi in Harley 24989 & 24994 (K, SPF), Horst &
Uebelmann 145 (ZSS); Mun. Diamantina, Conselheiro Mata, s.d., A. Hofacker AH261, cult. M. Winberg
(colour photo on Winberg website, June 2001); l.c. (?), ‘80 km W of Diamantina’, 1983, Braun 439 (ZSS);
l.c. (?), ‘Diamantina’, without collector, cult. Univ. Calif. Bot. Gard., Berkeley, 5 Apr. 1961 (US 2830736);
Mun. Monjolos, 8 June 2002, G. Charles (photos); l.c., nr Rodeador, 8 June 2002, G. Charles (photos).
Short-list score (2×5) = 10. Known to have a very limited range, there being only c. 60 km separating
the two small areas recorded above (Serra do Cabral and W slope of Serra do Espinhaço), which represent
about 5 localities or perhaps just two partially fragmented populations, separated by the lowlands of the
Rio Curimataí and Rio Pardo.
The lilac-pink flowers of this species are spectacular, but the colour is of variable intensity.
3. Arthrocereus glaziovii (K. Schum.) N. P. Taylor & Zappi in Bradleya 9: 84–85

(1991). Holotype: Brazil, Minas Gerais, Pico d’Itabira do Campo, 20 Dec. 1888, Glaziou
s.n. (B [K, SPF, photos ex B]).
Cereus glaziovii K. Schum. in Martius, Fl. bras. 4(2): 200, tab. 39 (1890). Leocereus glaziovii (K. Schum.)
Britton & Rose, Cact. 2: 109 (1920). Trichocereus glaziovii (K. Schum.) Werderm., Bras. Säulenkakt.:
94 (1933).
Trichocereus campos-portoi Werderm., ibid. 94–95 (1933). Arthrocereus campos-portoi (Werderm.) Backeb. in
Backeb. & Knuth, Kaktus-ABC: 211 (1935, publ. 1936). Holotype: Brazil, Minas Gerais, Serra do
Curral [Mun. Belo Horizonte], June 1932, Werdermann 3994 (B, K [photo ex B]).
A. itabiriticola L. Diers, P. J. Braun & E. Esteves Pereira in Kakt. and. Sukk. 38: 312–315 (1987). Holotype:
Brazil, Minas Gerais, Mun. Itabirito, Pico do Itabirito [formerly Pico do Itabira do Campo], Horst &
Uebelmann HU 330 (KOELN, n.v.; ZSS, iso.) — this is the type locality of A. glaziovii!
? A. damazioi P.V. Heath in Calyx 2(2): 65 (1992), nom. inval. (Art. 32). Based on an invalid name of
Schumann (1903).
[A. microsphaericus sensu (K. Schum.) A. Berger, Kakteen: 337 (1929) et auctt., non Cereus microsphaericus
K. Schum. (= Schlumbergera microsphaerica (K. Schum.) Hövel).]
[Cereus damazioi sensu Quehl in Monatsschr. Kakt.-Kunde 28: illustr. opp. p. 62 (1918) et auctt., non
Weingart, nom. illeg.]
[Trichocereus damazioi sensu (Weingart) Werderm., Bras. Säulenkakt.: 94 (1933), non Cereus damazioi
Weingart, nom. illeg.]
Procumbent or erect, shrubby, 20–30 cm tall, stems rarely to 60 cm or more. Branches 3–4(–4.5) cm
diam., erect or decumbent, generally segmented, sometimes losing the growing point and the segments
becoming very short or almost spherical; epidermis dark green; ribs 10–16, 2–3 × 4–6 mm. Areoles to 2.5
DISCOCACTUS
mm diam., 4 mm apart, felt greyish. Spines yellow to dark reddish brown, central spines 4–8, stout, the
largest porrect, to 4 cm, radials numerous, thin to bristle-like and flexible, to 8 mm. Flowers nocturnal,
8–15 × 5–8.5 cm, sweetly scented; pericarpel and tube bearing few triangular bract-scales with hairy areoles
and few, pale hairs; tube 60–75 × 6–8 mm at its narrowest point, ridged; outer perianth-segments to 30 ×
6 mm, linear, acuminate, fleshy, pale brownish pink or greenish, strongly reflexed; inner segments to 18 ×
5 mm, linear, acute, slender, delicate, cream or whitish, erect; stamens in two poorly distinct series, the
innermost inserted at 2.5–3.5 cm above pericarpel, with long filaments, the outermost series at the apex
of tube, filaments 5–10 mm, anthers to 1.5 mm; ovary locule 8–9 mm diam.; style 5–9 cm, stigma-lobes
7–12, exserted, 6 mm, slender. Fruit depressed-globose to obpyriform, 2–3 cm diam.; pericarp reddish
brown, naked or nearly so. Seeds c. 1.5–1.8(–2.2) mm, testa-cells convex, with cuticular folds.
South-eastern campo rupestre element: on iron-rich rock (canga) in campo rupestre, to 1300–1750 m, east and
south of Belo Horizonte, central-southern Minas Gerais. Endemic to the core area within Minas Gerais.
Map 49.
minas gerais: Mun. Belo Horizonte, Serra do Curral, 22 Dec. 1996, M.F. de Vasconcelos in BHCB
37421 (BHCB, K); l.c., border with Mun. Nova Lima, 13 Jan. 1998, I. Ribeiro de Andrade (K, photo);
Mun. Caeté, Serra da Piedade, Mello-Barreto 2213 (BHMG), 19 Jan. 1971, Irwin et al. 28758 (NY), M.
Braga s.n. (BHCB 7512), 3 Nov. 1988, Taylor & Zappi in Harley 25500 (SPF, K), 12 Dec. 1990, Taylor
& Zappi 750 (HRCB, K, ZSS, BHCB); Mun. Nova Lima, Serra da Moeda, 1989, L. Martens 104 (SPF,
HRCB); Mun. Itabirito, Pico do Itabirito [Itabira, Itabira do Campo], Dec. 1888, Glaziou s.n. (K, SPF,
photos ex B), s.d., Horst & Uebelmann HU 330 (ZSS), 23 July 1966, L. Emygdio de Mello F. 2209 & A.
Andrade 2113 (R), 11 Sep. 1993, W. Antunes Teixeira s.n. (K, BHCB); Mun. Ouro Preto, reported by
Schumann (1903: 38), but awaiting confirmation; without locality or date, Sello 1000 (MO) [isosyntype
for A. melanurus, q.v.].
conservation status. Endangered [EN B1ab(i, ii, iii, iv, v) + 2ab(i, ii, iii, iv, v)] (3); extent of
(historical) occurrence = 672 km2; PD=3, EI=1, GD=2. Short-list score (3×6) = 18. Many of its former
habitats have been eliminated through the mining of iron ore and its isolated populations are sometimes
very small (< 100 individuals). Less than 5 surviving localities are known.
Further field studies are needed to evaluate the status of the erect and prostrate forms of
this species.
29. DISCOCACTUS Pfeiff.
in Allg. Gartenzeitung 5: 241 (1837). Type: D. insignis Pfeiff. (= D. placentiformis (Lehm.)

K. Schum.).
vernacular names. Roseta-do-diabo, Coroa-do-diabo, Coroa-de-frade, Cabeça-de-frade.
Stems depressed-globose, often half buried in the ground, non-mucilaginous, lacking woody tissues,
very small or to c. 10 × 25 cm, remaining solitary or sprouting offsets; ribs 9–26, well-defined or broken
up into tubercles; areoles small to large but soon becoming glabrous, indeterminate growth absent.
Spines relatively few, usually 3–17 per areole, sometimes very stout or minute. At maturity stem apex
transformed into a white-woolly and often bristly cephalium, bearing flowers and fruit from its deeply
sunken apex. Flowers salverform, nocturnal, sweetly scented (hawkmoth syndrome), 1–6 or more
rapidly developing at the same time; pericarpel ovoid, usually quite naked, tube (including the nectar-
chamber) elongate, bearing leafy bract-scales glabrous in their axils; perianth-segments narrowly
oblanceolate, the outermost often greenish, brownish or even deep pink, inner series white, rarely pale
pinkish; stamens numerous, inserted along the inside of the tube, and at its apex, on short filaments,
sometimes the lowermost inserted above a whorl of hairs which protect the nectar-chamber; style long
or short, stigma-lobes c. 5–6, slender, whitish. Fruit clavate, usually naked, dehiscing by longitudinal
DISCOCACTUS
splits in the pericarp, red, orange-yellow, greenish or white, pulp scanty, but very attractive to ants.
Seeds 0.8–2.0 × 0.8–2.2 mm, hat-shaped, black, very shiny, testa tuberculate, the tubercles sometimes
elongated and pointed.
A genus of 7 or more, very closely related species, the majority in Eastern Brazil,
including Maranhão (fide Braun & Esteves Pereira 2002: 79), Ceará, Piauí, Pernambuco
(fide Uebelmann 1996), Bahia and Minas Gerais. Of these, 5 are endemic and mostly
either patchy in distribution or extremely local, rare and often in danger of extinction (the
whole genus has accordingly been placed in Appendix I of C.I.T.E.S. since 1992).
However, the complex species, D. heptacanthus (with a subspecies ranging into NW
Minas Gerais, W & S-cent. Bahia & SW Piauí), ranges as far as north-eastern Paraguay
and eastern Bolivia, through Mato Grosso do Sul, Goiás and Mato Grosso. It is curious
that, except for a single population of D. heptacanthus subsp. catingicola from Mun.
Paramirim, the genus has not been recorded from central and southern Bahia (east of the
Rio São Francisco), although suitable habitats exist, eg. in the Chapada Diamantina.
Instead, such habitats are characterized by similarly adapted Melocactus species (eg. M.
paucispinus, M. conoideus, M. concinnus, M. violaceus).
As the generic name implies, the plants are disc-shaped, and over much of its range
Discocactus is found in habitats through which fire passes regularly (notably the cerrado and
associated campo rupestre). This depressed, ground-hugging habit protects the plant against
the worst effects of burning, since the region a few centimetres above the ground is usually
much cooler from air being drawn in as the fire passes. Three of the rarer, northern species,
however, occur in the caatinga and apparently are not normally subjected to fire.
Discocactus is clearly very closely related to the much larger genus, Gymnocalycium
Pfeiff. ex Mittler (1844), which replaces it to the south and south-west, in Southern
Brazil, Paraguay, Argentina and eastern Bolivia. Gymnocalycium differs mainly in lacking a
cephalium, in having broader, diurnal, brightly coloured flowers and in displaying a
greater diversity in seed-morphology. It may prove to be paraphyletic in respect of
Discocactus and the latter name has priority, potentially raising the fraught question of
which name should be preferred or conserved, although they do not necessarily need to
be combined and are certainly not confusable. Plates 71–73.2.
1. Radial spines 3–8 per areole, often > 1.5 mm thick; stem solitary or offsetting when damaged
by fire, > 11 cm diam. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
1. Radial spines > 8 per areole, to 1.5 mm thick; stems freely offsetting, or solitary and < 11 cm
diam. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
2. Exposed part of vegetative stem with only 3 areoles visible per rib; ribs not broken up by
tubercles (N Bahia to Piauí and Ceará) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2. bahiensis
2. Exposed part of vegetative stem with > 3 areoles per rib and/or ribs tuberculate with deep
sinuses between successive areoles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
3. Seed 1.5–2.0 mm; fruit white or greenish, rarely pinkish red at apex; ribs 9–26, ± tuberculate;
spines often flattened, > 1.5 mm thick . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
3. Seed 1–1.4 mm; fruit yellow-orange at apex; ribs 12–13, scarcely tuberculate; spines ± terete,
to 1.5 mm thick (N Minas Gerais: near Grão Mogol, sand) . . . . . . . . . . . . . . 5. pseudoinsignis
4. Ribs 9–26, acute-edged, with shallow sinuses between areoles on the same rib (Minas Gerais:
E of Rio São Francisco) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4. placentiformis
DISCOCACTUS
4. Ribs 10–12, composed of rounded tubercles with deep, acute sinuses between areoles on the
same rib (SW Piauí, W & cent.-S Bahia & NW Minas Gerais)
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3. heptacanthus subsp. catingicola
5. Ribs not tuberculate; stem solitary, dark purplish green to brownish; spines minute, adpressed,
claw-like (N Minas Gerais: near Grão Mogol, in pure quartz gravel) . . . . . . . . . . . . . . . 6. horstii
5. Ribs ± tuberculate; stem usually offsetting, light green or spines not as above (N Bahia) . . . . . . . 6
6. Tubercles strongly developed, spiralled, obscuring the ribs . . . . . . . . . . . . . . . . . . . . 1. zehntneri
6. Tubercles weakly developed, arranged in clearly defined ± vertical ribs . . . . . . . . . . . 2. bahiensis
1. Discocactus zehntneri Britton & Rose, Cact. 3: 218 (1922); N. P. Taylor in Cact.
Succ. J. Gr. Brit. 43: 40 (1981). Type: Bahia, Mun. Sento Sé, Zehntner in Rose & Russell
19779 (US, lecto. designated here).
Echinocactus zehntneri (Britton & Rose) Luetzelb., Estud. Bot. Nordéste 3: 69 (1926).
D. albispinus Buining & Brederoo in Cact. Succ. J. (US) 46: 252–257 (1974). D. zehntneri f. albispinus
(Buining & Brederoo) Riha in Kaktusy 19: 26 (1983). D. zehntneri var. albispinus (Buining &
Brederoo) P. J. Braun in Succulenta 69: 215 (1990). D. zehntneri subsp. albispinus (Buining &
Brederoo) P. J. Braun & E. Esteves Pereira in Kakt. and. Sukk. 46: 64 (1995). Holotype: Bahia, Mun.
Sento Sé / Campo Formoso, Horst 390 (U).
Solitary or strongly caespitose; stem 3–7 × 2–12 cm, light to dark green; ribs 12–21, broken up into ±
spiralled tubercles; areoles to c. 15 mm apart on the ribs. Spines creamy white to grey, interwoven and
± hiding the stem, centrals 0–2, radials 9–18. Cephalium to 3.5 × 4.5 cm, white-woolly, with 7–60 mm,
erect, dark bristles at its margin. Flowers 33–77 × 35–55 mm, greenish in bud; outer perianth-segments
19–28 × 3.5–6.0 mm, green to yellowish, inner segments 17–25 × 4–6 mm, white. Fruit commonly
bright red, or white to greenish, 18–46 × 6–9 mm. Seed 0.8–2.0 × 0.8–2.2 mm, testa densely covered by
elongate tubercles.
conservation status. Vulnerable [VU B1ab(iii, v)]; extent of occurrence estimated to be < 20000
km2, known from < 10 locations and continuing decline projected (see below).
1a. subsp. zehntneri

Solitary or sparingly offsetting, stems globose to elongate; ribs c. 12–13, tubercles 10 × 15–20 × 15–20
mm; areoles c. 9–10 × 5–6 mm. Central spines 0–2, to 70 × 2 mm, radials c. 11, to 25–75 × 1.5 mm.
Fruit red.
Northern Rio São Francisco caatinga element: on exposed rocks and gravelly soil in caatinga, c. 530 m,
north of the Chapada Diamantina, Bahia. Endemic to Bahia. Map 50.
bahia: Mun. Sento Sé, Zehntner in Rose & Russell 19779 (US), 29 km SE of the [old] town, nr ‘Brejo
Grande’, July 1974, Horst 441 (U, ZSS); Mun. Sento Sé/Campo Formoso, Serra São Francisco, S of
Piçarrão, Horst 390 (U, ZSS).
conservation status. Data Deficient [DD]. Part of its former habitat is believed to have been
submerged beneath the Represa (Lago) de Sobradinho leaving, perhaps, only a single locality intact (that
of the type of the synonymous D. albispinus, which was observed by G. Charles in July 2000). Detailed
knowledge of its extent of occurrence and area of occupancy is currently lacking. However, it is likely
to be Vulnerable (VU).
DISCOCACTUS
1b. subsp. boomianus (Buining & Brederoo) N. P. Taylor & Zappi in Bradleya 9: 86
(1991). Holotype: Bahia, Mun. Morro do Chapéu, Horst 222 (U).
Discocactus boomianus Buining & Brederoo in Succulenta 50: 26 (1971). D. zehntneri var. boomianus
(Buining & Brederoo) P. J. Braun in ibid. 69: 218 (1990).
D. araneispinus Buining et al. in Succulenta 56: 258 (1977); Buining, Gen. Discocactus: 39 (1980). D.
zehntneri var. araneispinus (Buining et al.) P. J. Braun in Succulenta 69: 215 (1990). D. zehntneri subsp.
araneispinus (Buining et al.) P. J. Braun & E. Esteves Pereira in Kakt. and. Sukk. 46: 64 (1995).
Holotype: Bahia, Mun. Sento Sé, Horst 440 (U).
? D. buenekeri W.R. Abraham in Kakt. and. Sukk. 38: 284, with illus. (1987); P. J. Braun in Succulenta
69: 219–221 (1990). D. zehntneri subsp. buenekeri (W.R. Abraham) P. J. Braun & E. Esteves Pereira
in Kakt. and. Sukk. 44: 64 (1993). Holotype: N Bahia, sine loc., CWRA 27 (KOELN
[Succulentarium?], n.v.).
D. zehntneri var. horstiorum P. J. Braun in ibid. 69: 218 (1990). D. zehntneri subsp. horstiorum (P. J. Braun)
P. J. Braun & E. Esteves Pereira in ibid. 46: 64 (1995). Holotype: Bahia, K.I. Horst 667 (ZSS; K, iso.).
Offsetting freely, stems depressed, sometimes ± buried when growing in sand; ribs 13–21, tubercles 3–11
× 2–13 × 2–13 mm; areoles 3–7 × 1.5–3 mm. Central spine 0–1, to 35 × 1 mm, but to 60 mm beneath
the cephalium, radials 4–30 × < 1 mm. Fruit red or white (in the same population). Seed: Barthlott &
Hunt (2000: pl. 44.2).
Northern campo rupestre (Chapada Diamantina) element: on exposed arenitic rocks often with an
accumulation of gravel or in pure quartz sand, caatinga/campo rupestre, c. 700–1000 m, northern Bahia.
Endemic. Map 50.
bahia: Mun. Sento Sé, E part of Serra do Mimoso, N of Limoeiro, Horst 440 (U, ZSS); l.c., serra near
Minas do Mimoso [Serra do Curral Feio, bordering on Mun. Umburanas], 1988, K.I. Horst 667 (K, ZSS);
Mun. Morro do Chapéu, 21–22 km W of town on road BA 052, 11°28'S, 41°20'W, 25 Dec. 1988, Taylor
& Zappi in Harley 27395 (K, SPF, CEPEC), 23 July 1988, Eggli 1274 (ZSS), 3 Aug. 2002, Taylor (K,
photos); l.c. (?), cult. ZSS, Horst & Uebelmann HU 222 (K).
conservation status. Vulnerable [VU D2] (2); PD=1, EI=1, GD=2. Short-list score (2×4) = 8.
Habitats fragmented and subject to collection. There are at least 4 locations for this plant at the southern
end of its range to the west of Morro do Chapéu. Of these, 3 are accessible by road. According to Marlon
Machado, the locations vary in size from 3.5 to 150 ha with variable densities of individuals, numbers
ranging from 50,000 to at least 150,000.
Abraham’s D. buenekeri remains inadequately known in view of the characteristic secrecy

surrounding the location of its wild habitat, but it may be sufficiently distinct to merit
treatment as a third subspecies of D. zehntneri.
2. Discocactus bahiensis Britton & Rose, Cact. 3: 220 (1922); Buining, Gen.
Discocactus: 123–129 (1980); Hofacker in Kakt. and. Sukk. 53: 200–205 (2002).
Holotype: Bahia, Zehntner in Rose 19783 (US).
Echinocactus bahiensis (Britton & Rose) Luetzelb., Estud. Bot. Nordéste 3: 69 (1926).
Discocactus subviridigriseus Buining et al. in Succulenta 56: 262 (1977); Buining, Gen. Discocactus: 115
(1980). D. bahiensis subsp. subviridigriseus (Buining et al.) P. J. Braun & E. Esteves Pereira in Kakt. and.
Sukk. 44: 63 (1993). Holotype: Bahia, Mun. Juazeiro, Horst 438 (U).
D. bahiensis subsp. gracilis P. J. Braun & E. Esteves Pereira in ibid. 52: 286–290, illus. (2001). Holotype:
Bahia, Vereda do Romão Gramacho (Rio Jacaré), São Rafael, s.d., Braun & Esteves Pereira s.n. (UFG
22432, n.v.).
[D. placentiformis sensu Britton & Rose, Cact. 3: 220, fig. 233 (1922) non (Lehm.) K. Schum.]
DISCOCACTUS
vernacular name. Frade-de-cavalo.
Solitary or freely offsetting in age, 4–7 × 5.5–18.0 cm, grey- to light or yellow-green; ribs 10–15, to 30
× 20 mm, sometimes composed of tubercles to c. 7–10 × 15–30 × 15 mm; areoles oval, c. 6–10 × 3–8
mm, 10–23 mm apart, sometimes only 3 visible per rib above ground. Spines 5–11, all radial, strongly
recurved, sometimes hooked at apex, in some forms interwoven with those of adjacent areoles, pale
yellow to pinkish brown at first, later greyish or dark brown to blackish, to 35(–45) × 0.5–2.5(–3.0) mm,
the uppermost 1–3 in each areole very small when present. Cephalium low, to 47 mm diam., with soft
whitish wool, bristles when present hairlike, to 30 mm, yellow or brownish. Flowers 40–72 × 30–54 mm;
outer perianth-segments green, yellow or pinkish brown, inner segments 8–20 × 5–7 mm, white. Fruit
greenish white or red, 25–50 × 8–10 mm. Seed 1.2–1.8 × 1.5–1.9 mm, testa with sparse to dense patches
of pointed tubercles.
Northern caatinga element: on exposed, gravelly river terraces amongst limestone or iron-stained
quartzite, and seasonally inundated river flood plain, under and between jurema-preta (Mimosa
tenuiflora) and carnaúba (Copernicia prunifera) within the caatinga, 380–700 m, Rio São Francisco
drainage of northern Bahia (probably in adjacent Pernambuco), Ceará and north-western Piauí (Rio
Canindé). Map 50.
ceará: unlocalized, photograph communicated by Prof. Mattos, Univ. Fed. Ceará, Fortaleza (K).
piauí: Mun. Francisco Ayres, near the Rio Canindé, Horst & Uebelmann 943 (ZSS).
pernambuco: vaguely reported from near Santa Maria da Boa Vista, Uebelmann (mss).
bahia: N Bahia, Mun. Sento Sé, reported from S of the original town by Buining (1980: 121, photo);
Mun. Juazeiro, near the town, 2–6 June 1915, Rose & Russell 19742, 19783 (US); l.c., SSW of town, 0.5
km S of Rodeadouro, 10 Jan. 1991, Taylor et al. 1387 (K, ZSS, HRCB, CEPEC), Horst 438 (U, ZSS),
9°28'S, 40°33'W, 7 Apr. 2000, E.A. Rocha et al. (K, photos); l.c., E of town, 2–6 June 1915, Rose &
Russell 19764 (US); l.c., [11 km W of] Juremal towards Curral Velho, Horst & Uebelmann 633 (ZSS
T02127 & TP58-114/163/164); Mun. Campo Formoso, c. 60 km W of Maçaroca (Massaroca), W of
Abreus, W bank of the Rio Salitre, Horst 437 (U, ZSS); ibid., reported from Vargem Grande, Laje,
Panelas, Delfino and [Faz.] Vargem do Sal [‘Sol’] by Braun & Esteves Pereira, l.c. (2001); Mun. Itaguaçu
da Bahia, Serra Azul, reported by Braun & Esteves Pereira, l.c. (2001); Mun. Ourolândia (Ouro Branco),
reported by Braun & Esteves Pereira, l.c. (2001); Mun. Jacobina, Caatinga do Moura, reported by Braun
& Esteves Pereira, l.c. (2001); Mun. Morro do Chapéu, E of Rio Jacaré, São Rafael, 5 Aug. 2002, Taylor
(K, photos).
conservation status. Endangered [EN B2ab(i, ii, iii, iv, v)] (3); area of occupancy < 500 km2; PD=2,
EI=1, GD=2. Short-list score (3×5) = 15. Part of its range was eliminated by permanent inundation from
the Represa de Sobradinho (BA/PE) in the 1970s and the remainder has been heavily impacted by
agriculture and road/house construction during the past 10 years. Some of its sites are accessible by road
and have been visited by hobbyist collectors.
The distribution of this taxon seems rather disjunct towards its northern limits, but
remains inadequately known at present. It is related to the preceding species, occasionally
bearing red fruits (cf. Rose & Russell 19742) and both are assumed to have been derived
from within the D. heptacanthus complex, sharing similarities with the geographically close
D. heptacanthus subsp. catingicola (q.v.), especially via the dwarf, southern form called D.
bahiensis subsp. gracilis by Braun & Esteves Pereira, l.c. (2001) — see last locality cited
above. Hofacker (2002) has recently illustrated the diversity of populations in Bahia.
The form found in the vicinity of Juazeiro, which was described as D. subviridigriseus
Buining et al., is connected to typical D. bahiensis by a series of populations from
northern Bahia, comprising Horst 437 (D. bahiensis sensu Buining), the actual type of
the species (Rose & Russell 19783), Horst & Uebelmann 633 and others from the
DISCOCACTUS
extensive area covered by Mun. Campo Formoso. The ‘D. subviridigriseus’ form extends
into Piauí and Ceará and may be expected to occur in Pernambuco. It seems to be an
ecotype of river flood plains in the caatinga, where in northern Bahia, at least, it is
associated with semi-barren habitats characterized by bushes of jurema-preta (Mimosa
tenuiflora (Willd.) Poir.) and carnaúba palms, and is evidently, or was formerly, at times
subjected to temporary inundation.
3. Discocactus heptacanthus (Barb.Rodr.) Britton & Rose, Cact. 3: 218 (1922);

Taylor in Cact. Succ. J. Gr. Brit. 43: 38 (1981). Type: Brazil, Mato Grosso, nr Cuiabá,
Rodrigues (†, see Britton & Rose, l.c.). Lectotype (designated here): Rodrigues, Pl. Mato
Grosso: tab. 11 (1898).
Malacocarpus heptacanthus Barb. Rodr., Pl. Mato Grosso: 29, tab. 11 (1898).
conservation status. Data Deficient [DD]. The status of this species, comprising many
populations found outside the area treated here, remains to be investigated, but it cannot be assumed
to be Least Concern in view of the extensive alterations that are known to have already occurred to
the cerrado biome.
Only the following subspecies is found in Eastern Brazil:
3a. subsp. catingicola (Buining & Brederoo) N. P. Taylor & Zappi in Cact. Consensus
Initiatives 3: 7 (1997). Holotype: W Bahia, Mun. São Desidério, Horst 392 (U).
Discocactus catingicola Buining & Brederoo in Kakt. and. Sukk. 25: 265–267 (1974).
D. griseus Buining & Brederoo in Succulenta 54: 185–190 (1975). D. catingicola var. griseus (Buining &
Brederoo) P. J. Braun & E. Esteves Pereira in Kakt. and. Sukk. 44: 112 (1993). D. catingicola subsp.
griseus (Buining & Brederoo) P. J. Braun & E. Esteves Pereira in ibid. 46: 64 (1995). Holotype: Minas
Gerais, [Mun. João Pinheiro], Serra dos Alegres, 850 m, s.d., Horst & Uebelmann HU 343 (U; ZSS, iso.).
D. rapirhizus Buining & Brederoo in Ashingtonia 2: 44–47 (1975). D. catingicola subsp. rapirhizus
(Buining& Brederoo) P. J. Braun & E. Esteves Pereira in Kakt. and. Sukk. 44: 112 (1993). Holotype:
Goiás, surroundings of Posse, 19 June 1974, Buining & Horst HU 200 (U).
D. spinosior Buining et al. in Succulenta 56: 261 (1977); Buining, Gen. Discocactus: 146 (1980).
Holotype: W Bahia, Mun. Barreiras, Horst 205A (U? [labelled with number only]).
D. nigrisaetosus Buining et al. in Succulenta 56: 260 (1977); Buining, tom. cit.: 129 (1980). D. catingicola
var. nigrisaetosus (Buining et al.) P. J. Braun & E. Esteves Pereira in Kakt. and. Sukk. 44: 112 (1993).
Holotype: W Bahia, Mun. Santana, Horst HU 448 (U? [labelled with number only]).
D. piauiensis P. J. Braun & E. Esteves Pereira in Kakt. and. Sukk. 46: 57–62, with figs (1995). Holotype:
S Piauí, 600 m, E. Esteves Pereira 114 (UFG 14761, n.v.; B?, iso., n.v.), l.c., 1993, Braun 1652, 1655
(B, paratypi, n.v.).
Solitary, 4–9 × 11–20 cm, plain green; ribs 10–12, completely broken into tubercles, which are 15–20
× 22–40 × 15–30 mm; areoles positioned slightly below and sunken into tubercle apex, oval, 6–8 ×
4–5 mm, 15–30 mm apart, 4–5 visible above ground on each rib. Central spine 0(–1), to 20 mm,
radials 5–8, weakly recurved, pale yellow then grey to brownish in age, to 36 × 1.25–2.50 mm.
Cephalium 3–7.5 cm diam., with white wool and 35–45 mm, yellow to dark brownish bristles.
Flowers 50–60 × 40–60 mm, light brown to olive-green in bud; outer perianth-segments 22–24 × 7–8,
white to greenish without, inner segments 18–20 × 5–7, white. Fruit white or faintly pinkish, 40–45
× 8–13 mm. Seed 1.5–2.0 × 1.5–2.0 mm, with irregularly spaced, elongate, pointed tubercles (Barthlott
& Hunt 2000: pl. 44.3–4).
DISCOCACTUS
Western cerrado element: on exposed gravel or sand, cerrado and cerrado-caatinga ecotone, 450–700 m,
south-western Piauí, western and central-southern Bahia and north-western Minas Gerais; to Goiás &
Tocantins. Map 50.
piauí: SW Piauí, Mun. Bom Jesus, 30 km S of town, 1982, Horst & Uebelmann 523 (ZSS).
bahia: W Bahia, Mun. Barreiras, near old airport, Horst 205A (U, ZSS, K); Mun. São Desidério, west
of Sítio Grande at waterfall, Horst 392 (U); Mun. Santana, near Porto Novo, Horst 448 (U, ZSS); l.c. (?),
Braun 341 (ZSS); Mun. Santa Maria da Vitória, airport, Horst & Uebelmann 760 (ZSS, K); Mun. Coribe,
8 km south of Coribe, Horst & Uebelmann 760A (ZSS); cent.-S Bahia, Mun. Paramirim, N bank of rio
Paramirim, 28 Nov. 1988, Taylor in Harley 25558 (K, SPF, CEPEC).
minas gerais: Mun. Januária, Pandeiros, Horst & Uebelmann 720 (ZSS).
PD=1, EI=1, GD=2. Short-list score (2×4) = 8. Vulnerable due to the fragmented nature of its
distribution and small population size. In need of regular monitoring, since its cerrado and caatinga habitats
are undergoing much destructive change.
Despite Buining’s use of the epithet catingicola, this subspecies is more typical of the cerrado
and probably occurs in the caatinga only at its eastern limits. The collections cited from
south-western Piauí and central-southern Bahia (Harley et al. 25558, which is not precisely
localized for conservation reasons) are somewhat intermediate with D. bahiensis.
4. Discocactus placentiformis (Lehm.) K. Schum. in Engler & Prantl, Pflanzenfam.

3(6a): 190 (1894). Type: ‘Brasilia meridionali’, apparently not preserved. Neotype
(designated here): Lehmann in Nov. Act. Nat. Cur. 16(1): tab. 16 (1832) (iconotype).
Cactus placentiformis Lehm., Sem. hort. bot. Hamburg: 17 (1826) and in Nov. Act. Nat. Cur. 16(1):
318–319 (1832). Echinocactus placentiformis (Lehm.) K. Schum. in Martius, Fl. Bras. 4(2): 246 (1890).
Melocactus besleri Link & Otto in Verh. Ver. Beförd. Gartenb. 3: 420 (1827), nom. illeg. (Art. 52.1).
Discocactus lehmannii Pfeiff. in Nov. Act. Nat. Cur. 19(1): 120 (1839), nom. illeg. (Art. 52.1).
D. insignis Pfeiff. in Allg. Gartenz. 5: 241 (1837). Type: a living plant in the collection of Schelhase, not
preserved at the time of publication. Lectotype (designated here): Pfeiffer in Nov. Act. Nat. Cur.
19(1): tab. 15 (1839) — assumed to be an illustration prepared prior to the publication of the name
in 1837.
D. linkii Pfeiff., l.c. (1839). Based on Melocactus besleri Link & Otto, l.c. (1827), typ. excl. Lectotype
(designated here): Link & Otto, l.c., tab. 21 (1827). D. besleri F.A.C. Weber in Bois, Dict. hort.: 450
(1896). Type: as above.
? D. alteolens Lem. ex A. Dietr. in Allg. Gartenzeitung 14: 202 (1846). D. placentiformis var. alteolens (Lem.
ex A. Dietr.) P. J. Braun & E. Esteves Pereira in Kakt. and. Sukk. 44: 105 (1993). D. placentiformis subsp.
alteolens (Lem. ex A. Dietr.) P. J. Braun & E. Esteves Pereira in ibid. 46: 64 (1995). Echinocactus alteolens
(Lem. ex A. Dietr.) K. Schum. in Martius, Fl. bras. 4(2): 246 (1890). Cactus alteolens (Lem. ex A. Dietr.)
O. Kuntze in Deutsch. Bot. Monatsschr. 21: 173 (1903). Type: not known to have been preserved.
D. tricornis Monv. ex Pfeiff., Abbild. Beschr. Cact. 2: tab. 28 (1850). Type: not known to have been
preserved. Lectotype (designated here): Pfeiffer, l.c., tab. 28 (1850).
D. pugionacanthus Buining et al. in Succulenta 56: 260–261 (1977); Buining, Gen. Discocactus: 61 (1980).
D. placentiformis var. pugionacanthus (Buining et al.) P. J. Braun & E. Esteves Pereira in Kakt. and.
Sukk. 44: 105 (1993). D. placentiformis subsp. pugionacanthus (Buining et al.) P. J. Braun & E. Esteves
Pereira in ibid. 46: 64 (1995). Holotype: Minas Gerais, Mun. Grão Mogol (?), Horst 462 [‘275’] (U).
D. pulvinicapitatus Buining & Brederoo in Buining, tom. cit.: 100 (1980). D. latispinus subsp.
pulvinicapitatus (Buining & Brederoo) P. J. Braun & E. Esteves Pereira in Kakt. and. Sukk. 44: 105
(1993). Holotype: Minas Gerais, Mun. Bocaiúva, Horst 425 (U? [fl. labelled with number only]).
D. latispinus Buining et al. in Succulenta 56: 259 (1977); Buining, tom. cit.: 95 (1980). Holotype: Minas
Gerais, Serra do Cabral, Horst 146 (U).
DISCOCACTUS
D. multicolorispinus P. J. Braun & Brederoo in Kakt. and. Sukk. 32: 54–59, with illus. (1981). D.
placentiformis subsp. multicolorispinus (P. J. Braun & Brederoo) P. J. Braun & E. Esteves Pereira in Kakt.
and. Sukk. 44: 105 (1993). Holotype: Minas Gerais, Serra de Minas [collector?] (U, n.v.; ZSS, iso.).
The type plant is illustrated in Cact. Succ. J. (US) 51: 17, fig. 27 (1979).
D. crystallophilus L. Diers & E. Esteves Pereira in ibid. 32: 258–262, with illus. (1981). Holotype: Minas
Gerais, Mun. Augusto de Lima, E. Esteves Pereira E-84 (KOELN [‘Succulentarium’], n.v.; ZSS, iso.).
D. pseudolatispinus L. Diers & E. Esteves Pereira in ibid. 38: 242–247 (1987). D. latispinus subsp.
pseudolatispinus (L. Diers & E. Esteves Pereira) P. J. Braun & E. Esteves Pereira in Kakt. and. Sukk.
44: 105 (1993). Holotype: Minas Gerais, Mun. Claro dos Poções, E. Esteves Pereira 111 (KOELN,
n.v.; UFG, iso.).
Very variable; stem solitary unless damaged by fire, 3.5–10.0 × 12–25 cm, light to dark green; ribs 9–26,
well-defined, but weakly to strongly tuberculate, with shallow sinuses between successive tubercles, to
30 mm high and 40 mm wide near stem base, the tubercles 8–15 × 20–30 × 12–25 mm; areoles circular
to oval, 4–9(–15) × 3–8(–12) mm, somewhat sunken, 15–25 mm apart on the ribs, 3–7 visible per rib
above ground level. Spines 3–8(–10), whitish, dull yellow, brownish, pinkish or reddish, later grey,
central spine 0(–1), to 20 mm, lower 3–5 radials to 40(–45) × 4 mm, mostly flattened and slightly
recurved, sometimes splitting longitudinally, other spines when present much smaller. Cephalium 1–7 ×
4–11 cm, white-woolly, with or without 30–45 mm, brownish bristles. Flowers 45–85 × 40–80 mm, pale
green, brown or deep pink in bud; outer perianth-segments 24–35 × 3–7 mm, inner segments 15–30 ×
2–5 mm, white or pale pinkish. Fruit 30–50 × 5–15 mm, white to pink or reddish at apex. Seed 1.5–1.9
× 1.4–2.0 mm, testa with irregularly spaced, short to elongate, pointed tubercles.
Widespread South-eastern campo rupestre/cerrado element: on arenitic rocks, quartz sand and gravel,
cerrado/campo rupestre, rarely within the southern limits of the caatinga, 550–1275 m, east of the Rio São
Francisco, central and northern Minas Gerais. Endemic to the core area within Minas Gerais. Map 50.
minas gerais: Mun. Manga (?), ‘Mocambinho’, 15°6'S, 43°59'W, [caatinga region!], Mar. 1998, I.
Pimenta (K, photo); Mun. Grão Mogol, 48.5 km west of bridge over Rio Vacaria on road BR 251, Serra
da Bocaina, pass at telecommunications tower, 31 Jan. 1991, Taylor et al. 1512 (K, ZSS, HRCB, BHCB);
Mun. Claro dos Poções, 67 km from Montes Claros on road BR 365, June 1979, E. Esteves Pereira 111
(UFG), 1986, P.J. Braun 707 (ZSS); Mun. Jequitaí, 91.5 km NE of Pirapora on road BR 365, 27 Jan.
1991, Taylor et al. 1450 (K, ZSS, HRCB, BHCB); Mun. Francisco Dumont, c. 2 km from the town
towards Jequitaí, 7 Aug. 1988, Eggli s.n. (ZSS, photo); Mun. Bociaúva, W slope of Serra do Espinhaço,
near Sítio, Faz. Laginha, Horst & Uebelmann 548 (ZSS, K); l.c., Faz. Olho d’Agua, 1978, Horst &
Uebelmann 461 (ZSS); near Rio Jequitaí, Faz. Timboré, Horst & Uebelmann 425 (U, ZSS, K); Mun.
Buenópolis, Serra do Cabral, 6 & 7 km from the town on the road to Barreira da Lapa, 17°53'S, 44°15'W,
13 Oct. 1988, Taylor & Zappi in Harley 24988, 24993 (K, SPF); Mun. Augusto de Lima, above Santa
Barbara industrial site, 31 Aug. 1985, Horst & Uebelmann 356 (K, ZSS); ibid., near road BR 135, 4 km
from the border of Mun. Corinto, 6 Aug. 1988, Eggli 1096 (ZSS, photo); Mun. Diamantina, near
Senador Mourão, 1985, Horst & Uebelmann 707 (ZSS T00227, K); ibid., c. 15 km NE of Diamantina on
road to Mendanha, 30 Jan. 1969, Irwin et al. 22884 (K); ibid., between Diamantina and Biribiri, 18°11'S,
43°36'W, 8 Mar. 1995, Splett 808 (UB); ibid., estrada para o vilarejo de Três Barras, 15 Apr. 1987, Zappi
in CFCR 10540 (SPF); ibid., estrada para Conselheiro Mata, KM 165, 18 July 1987, Zappi et al. in
CFCR 11297 (SPF); 1 km from Conselheiro Mata, Horst & Uebelmann 542 (ZSS); Mun. Couto de
Magalhães de Minas, near the town, 3 Aug. 1988, Eggli 1056 (ZSS), Horst & Uebelmann 232 (ZSS TP58-
46, K); Mun. Datas, 5.5 km N of Datas, 17 Feb. 1988, Supthut 8835 (ZSS); Mun. Gouveia, just east of
Ponte de Paraúna (E of Pres. Juscelino), reported by Buining (1980: 74); Mun. Sete Lagoas, Lagoa
Grande, 16 Oct. 1959, Heringer 7228 (UB); Mun. Ouro Branco, Serra do Ouro Branco, illustrated in
Martius, Flora brasiliensis 1 (1, Tabulae physiognomicae), tab. XLVI (1855).
conservation status. Vulnerable (2) [VU B2ab(iii, v)]; extent of occurrence > 50,000 km2, but area
of occupancy almost certainly < 2000 km2; PD=2, EI=1, GD=2. Short-list score (2×5) = 10. Vulnerable,
since most populations are small and isolated from one another. Especially affected by unsustainable
exploitation in the production of cactus candy, when located close to towns and villages.
DISCOCACTUS
Typical D. placentiformis (Lehm.) K. Schum. is assumed to be the plant found originally

about the towns of Diamantina and Ouro Branco, Minas Gerais. These were visited
before 1827 by botanists such as Riedel (see Urban 1906: 90), who was cited as the
collector of the type of the synonymous D. linkii Pfeiff. (Melocactus besleri Link & Otto pro
parte), based on material contemporary with that of Cactus placentiformis Lehm.
The large-stemmed, many-ribbed, thick-spined form of this species (syn. D.
pulvinicapitatus, D. latispinus, D. pseudolatispinus), from the western slopes of the Serra do
Espinhaço, and from the Serra do Cabral and northwards (municípios Claro dos Poções,
Jequitaí, Bocaiúva, Francisco Dumont & Buenópolis), is distinctive and may be worthy
of recognition as a subspecies. It is connected to typical forms of the species by
populations found near the western edge of Mun. Diamantina (syn. D. multicolorispinus).
The form from the north-eastern population (Mun. Grão Mogol, syn. D. pugionacanthus)
is also distinctive for its ± strongly tuberculate stem and could be mistaken for D.
heptacanthus. Some plants from the region north of Diamantina, eg. Horst & Uebelmann
232 cited above, superficially resemble D. pseudoinsignis (see below) in their spination.
5. Discocactus pseudoinsignis N. P. Taylor & Zappi in Bradleya 9: 86 (1991).

Holotype: Minas Gerais, Mun. Cristália, Zappi et al. in CFCR 12045 (SPF).
[D. insignis sensu Buining, Gen. Discocactus: 81–82 (1980) non Pfeiff. (1837).]
Solitary, 7–9 × 12–21 cm, light to dark green; ribs 12–13, almost straight and even, not tuberculate,
15–25 mm high, 20–40 mm broad; areoles 6 × 4–5 mm, 10–30 mm apart and 5–6 visible per rib above
ground. Spines 5–8(–9), ± terete or at least isodiametric, grey to blackish, central 0–1, 10–30 × 0.5–1.5
mm, radials 5–7(–8), lower 3 c. 25–42 × 1–1.5 mm, straight or variously curved, others much smaller.
Cephalium to 5 × 10 cm, white-woolly, usually with exserted, to 40 mm, dark brownish bristles. Flowers
c. 75 × 60 mm, pale brownish olive-green in bud; outer perianth-segments c. 30 × 4 mm, inner segments
c. 22 × 2 mm, white. Fruit 32–45 × 5–9 mm, orange-yellow to reddish at apex. Seed 1–1.4 × 1.0 mm,
regularly tuberculate.
South-eastern campo rupestre (Grão Mogol) element: in pure quartz sand or sand between arenitic rocks,
campo rupestre, 700–1000 m, Mun. Cristália and Mun. Grão Mogol, northern Minas Gerais. Endemic to
the core area within Minas Gerais. Map 50.
minas gerais: Mun. Grão Mogol, 1972/1974, Horst 347 (U, K, ZSS); l.c., 10 km NW of the town, 22
Oct. 1978, Hatschbach 41364 (MBM); l.c., Vale do Ribeirão das Mortes, 4 Sep. 1986, I. Cordeiro & R.
Mello-Silva in CFCR 10088 (SPF); l.c., Córrego Escurona, 13 May 1990, Zappi et al. in CFCR 12901
(SPF, HRCB); l.c., Várzeas Escuras, 6 Nov. 1997, F. Fernande (K, photo); Mun. Cristália, ‘margem direita
do Rio Itacambiruçu, encosta da Serra das Cabras’, 28 May 1988, Zappi et al. in CFCR 12045 (SPF, holo.).
conservation status. Endangered [EN B1ab(iii) + 2ab(iii)] (3); extent of occurrence = 89 km2
(possibly under-recorded); PD=2, EI=1, GD=1. Short-list score (3×4) = 12. Populations may be affected
by habitat modification, including projected inundation by an hydro-electric dam-lake, but are offered
some protection within the Parque Estadual da Serra do Barão, Grão Mogol.
Following the discovery of this species in the early 1970s, Buining (1980) misidentified it
as D. insignis Pfeiff., a name previously and correctly referred to the synonymy of D.
placentiformis (Lehm.) K. Schum. by earlier authors (see Taylor 1981: 40). Pfeiffer’s 19th
Century description calls for a plant with only 10 ribs (consistently 12–13 in D.
DISCOCACTUS
pseudoinsignis) and, together with his illustration of the type, clearly indicates that the
bract-scales on the flower-tube and outer perianth-segments of the flower were deep
pink, which is a feature of some forms of D. placentiformis, but not of D. pseudoinsignis
(outer segments pale brownish olive-green).
D. pseudoinsignis is similar to the variable D. placentiformis and falls within the
geographical range of the latter (which is recorded from the northern part of Mun. Grão
Mogol, Serra da Bocaina), but can be distinguished by its non-tuberculate ribs, slender
spines, distinctively coloured outer perianth-segments and fruit apex, and smaller seeds.
Its closest relative is probably the following, with which it is partly sympatric.
6. Discocactus horstii Buining & Brederoo in Krainz, Die Kakteen, Lfg 52 (1973);
Buining, Gen. Discocactus: 106–110 (1980). Holotype: Minas Gerais, Mun. Grão Mogol,
Horst 360 (U).
Solitary, to 2 × 6 cm, dark brown- to purplish green; ribs 15–22, straight and narrow, 6–8 mm high, c.
4 mm wide at the rounded edge, and to 10 mm apart, not tuberculate; areoles oval, c. 1.5 × 1.0 mm,
very slightly sunken into the ribs, to 5 mm apart, 4–6 visible per rib above ground level. Spines 9–11, all
radial, pectinate, tightly adpressed to stem, claw-like, brown with a grey coating, to 3.5 × 0.75 mm.
Cephalium to 1.5 × 2 cm, white-woolly, with erect, dark brown, 20 mm bristles. Flowers solitary or
paired, 60–75 × 60 mm, pale yellow-brown in bud; tube only 4–5 mm diam. at narrowest point;
perianth-segments white, outer 30–35 × 3.5–8.0 mm, inner 20–24 × 6–9 mm. Fruit 30 × 4 mm, white.
Seed 1–1.1 × 0.9–1.0 mm, tuberculate (Barthlott & Hunt 2000: pl. 44.1).
South-eastern campo rupestre (Grão Mogol) element: in quartz gravel and sand beneath shrubs in campo
rupestre, c. 1000 m, Serra do Barão, northern Minas Gerais. Endemic to the core area within Minas Gerais.
Map 50.
minas gerais: Mun. Grão Mogol, Horst 360 (U), 1987, P.J. Braun 851 (ZSS).
conservation status. Endangered [EN B1ab(iii) + 2ab(iii)] (3); extent of occurrence < 100 km2;
PD=2, EI=1, GD=1. Short-list score (3×4) = 12. Now known from two adjacent populations, one
heavily impacted by collectors in the recent past and by ongoing habitat modification through quartz
extraction. A protected area has recently been established (Parque Estadual da Serra do Barão) and should
result in increased security, the authorities being well aware of the interest in the plant (M. Machado, in
litt., 20.05.2000).
A dwarf neotenic ally of D. pseudoinsignis, and perhaps the most remarkable of all Brazilian
cacti. Heavily collected in the early 1970s for the European horticultural market (Buining
1974a: 70) and formerly regarded as ‘Critically Endangered’. D. woutersianus Brederoo &
Broek in Succulenta 59: 203 (1980) was said to have been based on material of Horst 360,
but according to Riha in Kaktusy 26: 59 (1990) it is identifiable as the hybrid D. horstii
× D. pseudoinsignis [‘D. insignis’] originating in cultivation and now also reported from the
wild (Uebelmann 1996: HU 1497).
The name D. subnudus Britton & Rose (1922: 217) was based on a photograph of a badly
damaged plant said to emanate from the coast of Bahia. No Discocactus has subsequently
been reported from coastal Brazil and the provenance of this plant and its identity remain
doubtful. Luetzelburg (1925–26, 3: 69, 111) combined this epithet as Echinocactus subnudus.
UEBELMANNIA
30. UEBELMANNIA Buining
in Succulenta 46: 159–163 (1967). Type: U. gummifera (Backeb. & Voll) Buining (Parodia
gummifera Backeb. & Voll).
Globose to elongate, usually single-stemmed, 5–100 cm; vascular cylinder not woody, outer cortex
generally with gummiferous cells (disintegrating and fusing vertically into ducts in U. gummifera) and
positioned within or close to the ribs. Seedlings globose, sometimes developing beneath the surface of
the substrate. Stems erect to decumbent, 5–17 cm diam., epidermis bright to grey-green, whitish, reddish
or dark brown, distinctly roughened; hypodermis with dense accumulations of pectin; ribs c. 13–40,
broken or not into tubercles. Areoles with felt and white or brownish long hairs; spines hard, golden to
greyish, central spines porrect, radials sometimes absent. Flower-bearing region of stem apical, not
differentiated. Flowers diurnal, yellow or greenish, sometimes ageing to pale orange or pinkish, 0.8–2.5
× 0.6–3.0 cm; pericarpel and tube with bract-scales, hairs and eventually bristles; tube rotate; perianth-
segments triangular to lanceolate, spreading to reflexed. Fruit narrowly turbinate, few-seeded, flower
remnant drying whitish, erect; pericarpel with bract-scales and some glabrescent hairs, red; funicular pulp
translucent. Seeds 1.1–2.4 mm, cochleariform, brown to black, with the hilum-micropylar region
depressed; testa-cells flat or convex; cuticular folds absent.
Literature: Nyffeler (1997, 1998); Schulz & Machado (2000).
A remarkable, taxonomically isolated genus of 3 species, endemic to a relatively small

region of central Minas Gerais (campos rupestres, sensu lato) and since 1992 placed in
Appendix I of C.I.T.E.S. in view of its rarity and the potential risks to its survival from
commercial and private collection in the wild. The distribution map for this genus
(Map 51) is partly based on data kindly supplied as GPS co-ordinates by Marlon
Machado, which are not given here in order to protect these rare plants from
unscrupulous collectors.
It is assumed that the flowers are adapted for visits by hymenoptera, as observed by
Schulz & Machado (2000), but Heek & Strecker (1995) have also noted hummingbirds
visiting the flowers of U. gummifera. Plates 74.1–77.1.
1. Ribs not broken up into tubercles; flowers to 18 × 10 mm; fruit reddish to deep pink,
conspicuous; seeds with flat testa-cells (Subg. Leopoldohorstia) . . . . . . . . . . . . . . . . . 3. pectinifera
1. Areoles borne on pronounced tubercles; flowers > 20 × 15 mm; fruit yellowish or greenish,
inconspicuous; seeds with convex testa-cells (Subg. Uebelmannia) . . . . . . . . . . . . . . . . . . . . . . . . . 2
2. Ribs 15–22; stem cortex with mucilage cells but lacking ducts . . . . . . . . . . . . . . . . . 1. buiningii
2. Ribs 22–42; stem cortex with vertically arranged mucilage ducts . . . . . . . . . . . . . 2. gummifera
Subg. Uebelmannia
See Key, above. Found only on quartz sands and gravels, or on rocks adjacent to these
substrates.
1. Uebelmannia buiningii Donald in Nat. Cact. Succ. J. (UK) 23: 2–3 (1968); P. J.
Braun & E. Esteves Pereira in Kakt. and. Sukk. 39: 2–3 (1988); van Heek & Strecker in
ibid. 45: 234–235, illus. (1994). Holotype: Brazil, Minas Gerais, Mun. Itamarandiba,
Serra Negra, Tromba d’Anta, 1966, ‘c. 1000 m’, Horst HU 141 (U; ZSS, iso.).
UEBELMANNIA
Globose to subcylindric, solitary, to 12 × 8 cm; without gummiferous ducts under the epidermis.
Epidermis greenish red to brownish, rough. Ribs to 16–18(–22), strongly tuberculate; tubercles
conical, deflexed, to 7.5 mm high and 6 mm broad, c. 5 mm apart. Areoles on top of the tubercles, 2
mm diam., with sparse greyish wool when young, later glabrescent. Spines yellow-brown, black
tipped, soon becoming whitish, curved upwards, flattened on upper surface, central spines 4, to 15
mm, radials 0–4, to 5 mm. Flower to 27 × 25 mm; tube c. 9 mm, with bract-scales, white hairs and
brownish bristles; perianth-segments deep yellow; style 14 mm; stigma-lobes 6–7. Fruit 5–6 × 4–5 mm,
yellowish. Seed 1.2–1.4 × 0.8–0.9 mm, cochleariform to reniform, black; testa-cells convex (Barthlott
& Hunt 2000: pl. 44.7–8).
South-eastern campo rupestre (Serra Negra) element: on slabs of quartzitic rock amongst gravel, campo
rupestre, c. 1200 m, Serra Negra, Minas Gerais. Endemic to the core area within Minas Gerais. Map 51.
minas gerais: Mun. Itamarandiba, 24 Aug. 1985, Horst & Uebelmann 141 (ZSS, K); Serra Mata Virgem,
south slope, 4 Aug. 1988, Eggli 1074 (ZSS).
conservation status. Critically Endangered [CR C2a] (4); PD=3, EI=1, GD=1. Short-list score
(4×5) = 20. According to Braun & Esteves Pereira (l.c.) and Schulz & Machado (2000) this species is on
the verge of extinction and is scarce in cultivation, where it has proved difficult. Affected by collection
of plants and seed, fire and trampling by cattle.
Marlon Machado (pers. comm.) believes there are good reasons for continuing to accept
this taxon as a species distinct from the following, but it is currently too rare to permit a
proper survey of its described anatomical differences to be carried out.
2. Uebelmannia gummifera (Backeb. & Voll) Buining in Succulenta 46: 159–160

(1967). Holotype: Brazil, Minas Gerais, Serra do Ambrósio [Mun. Rio Vermelho], 1938,
Mello-Barreto s.n. (RB 64065).
Parodia gummifera Backeb. & Voll. in Arq. Jard. bot. Rio de Janeiro 9: 158 (1949, publ. 1950).
Uebelmannia meninensis Buining in Kakt. and. Sukk. 19(8): 151–152 (1969); W. van Heek et al. in ibid.
46: 9–12 (1995). U. gummifera subsp. meninensis (Buining) P. J. Braun & E. Esteves Pereira in
Succulenta 74: 226 (1995). Holotype: Brazil, Minas Gerais, [Mun. Rio Vermelho], Pedra Menina,
Horst HU 108 (U).
U. meninensis var. rubra Buining in Krainz, Die Kakteen, Lfg. 55–56 (1975). U. gummifera var. rubra
(Buining) P. J. Braun & E. Esteves Pereira in Succulenta 74: 226 (1995). Holotype: Brazil, Minas
Gerais, Serra Negra, 1100–1200 m, Horst HU 406 (U).
vernacular names. Coroa-de-frade, cabeça-de-frade.
Globose to cylindric, elongate, solitary, 5–30(–40) × 6–15 cm; with gummiferous ducts under the
epidermis. Epidermis greenish red to brownish, rough. Ribs to c. 30–40, strongly tuberculate; tubercles
conical, deflexed, to 4.5 mm high and 4–5 mm broad, c. 4 mm apart. Areoles on top of the tubercles, 2
mm diam., when young with sparse greyish wool, later glabrescent. Spines greyish or whitish, porrect or
curved upwards, flattened on upper surface, central spines 2–6, 10–25 mm long, radials 0–4, to 8 mm.
Flower 20–25 × 18–30 mm; tube c. 10 mm, with triangular bract-scales and white hairs; perianth-
segments deep yellow; style 12 mm; stigma-lobes 7–8. Fruit 6–8 × 4–6 mm, yellowish or greenish. Seed
1.1–1.3 × 0.8–0.9 mm, black or dark reddish brown; testa-cells convex.
South-eastern campo rupestre (Serra Negra/Serra do Ambrósio) element: in quartz sand, campo rupestre,
900–1600 m, Serra Negra, Minas Gerais. Endemic to the core area within Minas Gerais. Map 51.
minas gerais: Mun. Itamarandiba, above Penha de França, 24 Aug. 1985, Horst & Uebelmann 282 (ZSS,
K), 3 Aug. 1988, Eggli 1059 (ZSS); 25 km SW of town, 4 Aug. 1988, Eggli 1062 (ZSS); c. 20 km SSW
UEBELMANNIA
of town, Serra d’Anta, 4 Aug. 1988, Eggli 1069 (ZSS); Mun. Rio Vermelho, Pedra Menina, 22 Aug.
1985, Horst & Uebelmann 406 (ZSS, K); l.c., c. 9 km SSE, (?) Rio Mundo Velho, 22 Aug. 1985, Horst &
Uebelmann 108 (U, ZSS, K), 5 Aug. 1988, Eggli 1086, 1088 (ZSS); l.c., Serra do Ambrósio, 1938, Mello-
Barreto s.n. (RB); l.c., Faz. do Sr José Batista, 15 July 1984, A.M. Giulietti et al. in CFCR 4511 (SPF, SP),
6 Mar. 1988, Zappi & Prado in CFCR 11821 (SPF).
conservation status. Vulnerable [VU B1ab(iii, v)] (2); extent of occurrence = 198 km2; PD=3,
EI=1, GD=2. Short-list score (2×6) = 12. Populations with often large or very large numbers of
individuals, but overall range of species very limited and some habitats affected by charcoal production
and collection of plants and seed. Local morphological variation significant.
The plant described as U. meninensis may be worthy of recognition as a subspecies (fide

M. Machado), but has not been studied adequately in habitat by the authors.
Subg. Leopoldohorstia P. J. Braun & E. Esteves Pereira in Succulenta 74: 134 (1995).
See Key, above. Found on or beside crystalline rocks. Type and only species:
3. Uebelmannia pectinifera Buining in Nat. Cact. Succ. J. (UK) 22: 86–87 (1967).
Holotype: Brazil, Minas Gerais, Mun. Couto de Magalhães de Minas, Feb. 1966, Horst
HU 106 (U).
U. pectinifera var. pseudopectinifera Buining in Kakt. and. Sukk. 23(5): 125 (1972). Holotype: Brazil, Minas
Gerais, SE of Datas, towards Serro, Horst HU 280 (U).
U. pectinifera var. multicostata Buining & Brederoo in Krainz, Die Kakteen, Lfg 62 (1975). Holotype:
Brazil, Minas Gerais, 10 km E of Mendanha towards the Rio Jequitinhonha, Horst HU 362 (U).
‘U. ammotrophus’ hort., nom. nud.
Globose to fusiform or cylindric, elongate, solitary or aggregated, 10–50(–100) × 10–17 cm; with
gummiferous ducts under the epidermis. Epidermis green to grey-green, with or without waxy plates,
rough. Ribs 13–29, straight, not tuberculate, to 5 mm high and 5–7 mm broad. Areoles 2 mm diam.,
very condensed, to 3 mm apart, with sparse greyish or brownish wool when young, later glabrescent.
Spines greyish, whitish or yellowish brown, porrect or spreading, straight, flattened on upper surface,
central spines 2–6, 10–40 mm, radials 0–3, spreading. Flower 8–16 × 6–10 mm; tube to c. 8 mm, with
triangular bract-scales, white hairs and bristles; perianth-segments pale yellow, tinged reddish or greenish;
style 6–7 mm; stigma-lobes 7–8. Fruit 15–25 × 6–8 mm, pericarp with few glabrescent or slightly woolly
bract-scales, reddish to bright pinkish. Seed 1.7–2.4 × 1.3–1.6 mm, cochleariform, keeled, brown, shiny,
smooth; testa cells flat.
conservation status. Vulnerable [VU B1ab(iii, v)]; extent of occurrence < 20000 km2. Although it
has a total of at least 18 distinct locations/populations, the severe fragmentation of these, their generally
very small size, the decline in the quality of habitat and ongoing collection of adult individuals warrants
an assessment of Vulnerable.
A variable species, comprising a complex of numerous local forms, the following

circumscription into subspecies possibly representing a convenient over-simplification of
the situation in nature (cf. Schulz & Machado 2000):
UEBELMANNIA
1. Plants with grey-green, white-scaly epidermis; ribs 13–20(–26) . . . . . . . . . 3a. subsp. pectinifera
1. Plants with green epidermis, white scaly plates absent; ribs (16–)18–29 . . . . . . . . . . . . . . . . . . . . 2
2. Spines yellow, ascending, organized in rows following the edge of the rib; plants to 50 cm
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3b. subsp. flavispina
2. Spines greyish, spreading, not organized in rows; plants to 100 cm . . . . . . . . . 3c. subsp. horrida
3a. subsp. pectinifera

Plants 20–85 cm tall, with grey-green, white-scaly epidermis; ribs 13–20(–26). Spines greyish or whitish,
ascendent, organized in rows following the edge of the ribs, central spines 2–4, 10–25 mm, radials 0–3.
South-eastern campo rupestre (Diamantina) element: crystalline rocks in campo rupestre, c. 650–1250 m,
Serra do Espinhaço, north-east and south of Diamantina, Minas Gerais. Endemic to the core area within
minas gerais: Mun. Couto de Magalhães de Minas, Feb. 1966, Horst HU 106 (U, ZSS, K); Mun.
Presidente Kubitschek, near junction with road BR 259, 25 July 1998, Hatschbach et al. 68223 (K); l.c.,
Trinta Réis, Apr. 1986, Zappi (obs.); SE of Datas, towards Serro, Horst HU 280 (ZSS, K); Mun.
Diamantina, 29 Aug. 1985, Horst & Uebelmann 362 (ZSS, K); l.c., 10 km E of Mendanha, Horst &
Uebelmann 362 (U); l.c., Inhaí, reported by van Heek & Strecker (1994: 135).
EI=1, GD=2. Short-list score (2×6) = 12. Endangered from collection of plants and seed throughout its
limited range, where its 9 known populations are mostly small (area of occupancy < 500 km2).
The form reported from Inhaí seems somewhat morphologically and geographically
intermediate between all three subspecies recognized here.
3b. subsp. flavispina (Buining & Brederoo) P. J. Braun & E. Esteves Pereira in Succulenta
74: 135 (1995). Holotype: Brazil, Minas Gerais, Barão do Guaçuí, 1280 m, Horst &
Uebelmann 361 (U).
Uebelmannia flavispina Buining & Brederoo in Succulenta 52: 9–10 (1973).

‘U. pectinifera var. crebrispina’ Strecker in Kakt. and. Sukk. 45: 34–35, illus. (1994), nom. nud. (Horst &
Uebelmann 642).
‘U. warasii’ F. Ritter, nom. nud.
Plants to 50 cm tall, with green epidermis; ribs (16–)21–29. Spines yellow, ascendent, organized in rows
following the edge of the rib, central spines 2–4, 10–25 mm.
South-eastern campo rupestre (Diamantina) element: crystalline rocks in campo rupestre, c. 1200–1350 m,
Serra do Espinhaço, west of Diamantina, Minas Gerais. Endemic to the core area within Minas Gerais.
Map 51.
minas gerais: Mun. Diamantina, 8 km on the road to Conselheiro Mata, 18°15'S, 43°44'W, 30 Oct.
1988, Taylor & Zappi in Harley 25454 (K, SPF), 17 Apr. 1987, Zappi et al. in CFCR 10593 (SPF), 29 July
1988, Horst & Uebelmann 854 (ZSS, K); [N of] Barão do Guaçuí, Horst & Uebelmann 361 (U, ZSS, K),
20 Aug. 1985, Horst & Uebelmann 642 (ZSS, K).
EI=1, GD=2. Short-list score (2×6) = 12. Subjected to the collection of plants and seed throughout its
very restricted range, comprising c. 6 populations of variable size, but mostly rather small.
UEBELMANNIA
3c. subsp. horrida (P. J. Braun) P. J. Braun & E. Esteves Pereira in Succulenta 74: 135
(1995). Holotype: Brazil, Minas Gerais, Mun. Bocaiúva, near Sítio, 1982, Horst &
Uebelmann 550 (ZSS; K, iso.).
Uebelmannia pectinifera var. horrida P. J. Braun in Kakt. and. Sukk. 35: 264–266 (1984).
Plants to 100 cm tall, with green epidermis; ribs 23–27. Spines greyish, spreading, not organized in rows;
central spines 3–6, 20–40 mm.
South-eastern campo rupestre (Rio São Francisco drainage) element: on sandstone outcrops on the western
slopes of the Serra do Espinhaço (‘Serra Mineira’), c. 700–850 m, Mun. Bocaiúva, Minas Gerais. Endemic
to the core area within Minas Gerais. Map 51.
minas gerais: Mun. Bocaiúva, NE of Engenheiro Dolabela, near Sítio, 1982, Horst & Uebelmann 550
(ZSS, K), 8 Aug. 1988, Eggli 1124 (ZSS).
conservation status. Vulnerable [VU D2] (2); PD=3, EI=1, GD=1. Short-list score (2×5) = 10.
Known from only a single small group of localities, close to agricultural activities and highly sought-after
by collectors, but currently not experiencing marked decline (Schulz & Machado 2000).
Apparently disjunct from the remainder of the genus, but the region in between is very
poorly known.
The untypifiable name Uebelmannia centeteria (Lehm. ex Pfeiff.) Schnabel (Echinocactus

centeterius Lehm. ex Pfeiff., Enum. Cact.: 65. 1837) is doubtfully referred to the Chilean
species, Eriosyce curvispina (Colla) Kattermann, by Hunt et al. (1994: 146). It was stated by
Pfeiffer, l.c., to have originated from Minas Gerais, but this is assumed to be an error.
BIBLIOGRAPHY AND ACKNOWLEDGEMENTS
6. BIBLIOGRAPHY AND
ACKNOWLEDGEMENTS
6.1 LITERATURE CITED OR CONSULTED

Ab’Sáber, A. N. (1974). O domínio morfoclimático semi-árido das caatingas brasileiras.
Geomorphologia (São Paulo) 43: 1–39.
Agra, M. de F. (1996). Plantas da medicina popular dos Cariris Velhos, Paraíba, Brasil. Editora
União, João Pessoa. Pp. 125.
Anderson, E. F. (1999). The IOS Cactaceae Working Party: A Major Contribution of the
IOS. I.O.S. Bull. 7: 18.
Andrade-Lima, D. de (1950). Catálogo do Herbário da Escola Superior de Agricultura da
Tapera. Bol. Secr. Agric. (Recife) 17: 279–281.
—— (1966). Cactaceae de Pernambuco (Cactaceous plants of Pernambuco). Anais do IX
Congresso Int. Pastagens (São Paulo) 2: 1453–1458.
—— (1975). A vegetação da bacia do Rio Grande, Bahia. Nota preliminar. Revta Brasil.
Biol. 35: 223–232.
—— (1977). A flora de áreas erodidas de calcário Bambuí, em Bom Jesus da Lapa, Bahia.
Ibid. 37: 179–194.
—— (1981). The caatingas dominium [sic]. Revta Brasil. Bot. 4: 149–163.
—— (1989). Plantas das Caatingas. Academia Brasileira de Ciências, Rio de Janeiro. Pp.
243.
Assis, J.G.A., Oliveira, A.L.P.C., Resende, S.V., Senra, J.F.V. & Machado, M. (2003).
Chromosome numbers in Brazilian Melocactus (Cactaceae). Bradleya 21: 1–6.
Azevedo, A. de, ed. (1972). Brasil, a terra e o homem, ed. 2, 1. Companhia Editora
Nacional, São Paulo.
Backeberg, C. (1938). Blätter für Kakteenforschung 1938(6): [unpaged].
—— (1942). Cactaceae Lindley. Jahrb. Deutsch. Kakteen Ges. 1941(2): 1–80 (publ. 1942).
—— (1950). Nova genera et subgenera. Cact. Succ. J. (US) 22: 153–154.
—— (1958–62). Die Cactaceae. Handbuch der Kakteenkunde, 1–6. VEB Gustav Fischer
Verlag, Jena.
—— (1966). Das Kakteenlexikon. Enumeratio diagnostica Cactacearum. Ibid.
—— & Knuth, F. M. (1936). Kaktus-ABC. Gyldendalske Boghandel-Nordisk Forlag,
Copenhagen. Pp. 432 (1935, publ. 1936).
Bandeira, R. L. (1995). Chapada Diamantina, história, riquezas e encantos. Onavlis Editora,
Salvador, Bahia.
BIBLIOGRAPHY
Barthlott, W. (1976). In: Löve, A., IOPB chromosome reports, LIV. Taxon 25: 644–645.
—— (1977). Kakteen. Belser, Stuttgart.
—— (1979). Cacti. Stanley Thornes, Cheltenham, U.K.
—— (1987). New names in Rhipsalidinae (Cactaceae). Bradleya 5: 97–100.
—— (1988). Über die systematischen Gliederung der Cactaceae. Beitr. Biol. Pflanz. 63:
17–40.
—— & Hunt, D. R. (1993). Cactaceae. In: Kubitsky, K., The Families and Genera of
Vascular Plants, 2: 161–197. Springer Verlag, Berlin etc.
—— & —— (2000). Seed diversity in the Cactaceae subfamily Cactoideae. Succulent Pl.
Res. 5: 1–173.
——, Poremski, S., Kluge, M., Hopke, J. & Schmidt, L. (1997). Selenicereus wittii
(Cactaceae): an epiphyte adapted to Amazonian Igapó inundation forests. Pl. Syst.
Evol. 206: 175–185.
—— & Rauh, W. (1975). Notes on the morphology, palynology and evolution of the
genus Schlumbergera. Cact. Succ. J. (US), Yearbook 1975: 5–21.
—— & Taylor, N. P. (1995). Notes towards a monograph of Rhipsalideae (Cactaceae).
Bradleya 13: 43–79.
—— —— (1996). Hatiora — die Osterkakteen und ihre Verwandten. Kakt. and. Sukk.
47: 73–77.
—— & Voit, G. (1979). Mikromorphologie der Samenschalen und Taxonomie der
Cactaceae. Pl. Syst. Evol. 132: 205–229.
Bauer, R. (2003). A synopsis of the tribe Hylocereeae F. Buxb. Cact. Systematics Initiatives
17: 3–64.
Benson, L. (1969). The cacti of the United States and Canada — new names and
nomenclatural combinations. Cact. Succ. J. (US) 41: 124–128.
—— (1982). The Cacti of the United States and Canada. Stanford Univ. Press, California.
Pp. 1044.
Berger, A. (1905). A systematic revision of the genus Cereus. Ann. Rep. Missouri Bot. Gard.
16: 57–86.
—— (1920). Einiges über Rhipsalis. Monatsschr. Kakt.-Kunde 30: 1–4.
—— (1926). Die Entwicklungslinien der Kakteen. Gustav Fischer, Jena.
—— (1929). Kakteen. Ulmer Verlag, Stuttgart.
Bohle, B. (2000). In der Chapada Diamantina, Brasilien. Kakt. and. Sukk. 51: 43–48.
Boyle, T. H. & Idnurm, A. (2001). Physiology and genetics of self-incompatibility in
Echinopsis chamaecereus (Cactaceae). Sex Plant Reprod. 13: 323–327.
Bradley, R. (1716–1727). The History of Succulent Plants. W. Mears, London.
Brandão, M. (2000). Caatinga. In: Mendonça, M. P. & Lins, L. V., Lista Vermelha das
Espécies Ameaçadas de Extinção da Flora de Minas Gerais. Fundação Biodiversitas,
Fundação Zoo-Botânica de Belo Horizonte, Minas Gerais, Brazil. Pp. 160.
Brasil (1998). Primeiro Relatório Nacional para a Convenção sobre Diversidade Biológica. Brasil.
Ministério do Meio Ambiente, dos Recursos Hídricos e da Amazônia Legal, Brasília,
D.F., Brazil. Pp. 283.
BIBLIOGRAPHY
—— (2000). Política Nacional de Biodiversidade. Roteiro de consulta para elaboraçâo de uma

proposta. Ibid., ibid. Pp. 48.
Braun, P. J. (1986). Melocactus lanssensianus P. J. Braun species nova — een nieuwe soort
uit Pernambuco, Brazilië. Succulenta 65: 29.
—— (1988a). Eindrücke aus dem nordostbrasilienischen Sertão. Kakt. and. Sukk. 39:
169–171.
—— (1988b). On the taxonomy of Brazilian Cereeae (Cactaceae). Bradleya 6: 85–99.
—— (1990). Arquipélago de Fernando de Noronha (Brasilien) — Eindrücke von einer
Reise zu einem der letzten Inselparadiese und zum Standort von Cereus insularis
Hemsley. Kakt. and. Sukk. 41: 254–258.
—— & Esteves Pereira, E. (1986–89). Revision der Gattung Facheiroa Britton et Rose.
Ibid. 37: 56, 74–79; 38: 26–33, 82–85, 184–187; 39: 64–68, 126–131; 40: 198–203,
298–301.
—— —— (1988). Uebelmannia buiningii Donald — eine bald ausgerottete Art. Ibid. 39: 2–3.
—— —— (1989). The Opuntia inamoena complex in Brazil, I. Cact. Succ. J. (US) 61:
268–273.
—— —— (1990). Siccobaccatus Braun et Esteves — Een nieuw cactus geslacht uit
Brasilië. Succulenta 69: 3–8.
—— —— (1991a). Micranthocereus Backeberg subgenus Austrocephalocereus (Backeberg)
Braun & Esteves. Een nieuw ondergeslacht (Cactaceae) uit Brazilië. Ibid. 70: 62–67.
—— —— (1991b). Arrojadoa dinae Buining & Brederoo var. nana Braun & Esteves. Eine
neue Arrojadoa-Sippe aus Minas Gerais / Brasilien. Kakt. and. Sukk. 42: 190–195.
—— —— (1992). Pilosocereus rizzoianus Braun & Esteves Pereira, a new species of
Cactaceae from Goiás, Brazil. Cact. Succ. J. (US) 64: 148–155.
—— —— (1995a). The Opuntia inamoena complex in Brazil, VI. Ibid. 67: 111.
—— —— (1995b). Nieuwe combinaties en namen voor Cactussen uit Brazilië, Bolivia
en Paraquay. Succulenta 74: 81–85, 130–135, 226.
—— —— (1999a). Pilosocereus chrysostele (Vaupel) Byles & Rowley subsp. cearensis P. J.
Braun & Esteves — A new subspecies of Pilosocereus from the state of Ceará, Brazil.
Brit. Cact. Succ. J. 17: 21–27.
—— —— (1999b). Eine neue Kakteensippe aus Bahia, Brasilien. Pilosocereus densiareolatus
subsp. brunneolanatus P. J. Braun & Esteves. Kakt. and. Sukk. 50: 283–289.
—— —— (2002). Kakteen und andere Sukkulenten in Brasilien. Schumannia 3: 1–235.
—— —— (2003). Brazil and its Columnar Cacti — 70 years after Werdermann.
Kaktusy (CZ) 39 (Special 2002/1): 1–47.
Britton, N. L. & Rose, J. N. (1919–1923). The Cactaceae, 1–4. Carnegie Institution,
Washington, D.C. (Vol. 1, 1919, vol. 2, 1920, vol. 3, 1922, vol. 4, 1923.)
Brummitt, R. K. (1994). Report of the Committee for Spermatophyta: 41. Taxon 43:
271–277.
—— (1996). Ibid.: 44. Ibid. 45: 677–678.
—— & Taylor, N. P. (1990). To correct or not to correct? Ibid. 39: 298–306.
Buining, A. F. H. (1967). Een nieuw geslacht der Cactaceae uit Minas Gerais. Succulenta
46: 159–163.
BIBLIOGRAPHY
—— (1974a). Discocactus in Brazil. I.O.S. Bull. 3(3): 67–72.

—— (1974b). Melocactus ernestii Vaupel. In: Krainz, H., Die Kakteen, Lfg 58.
—— (1975a). Melocactus florschützianus [sic] Buin. et Bred. spec. nova. Ashingtonia 2:
25–27.
—— (1975b). Austrocephalocereus purpureus (Gürke) non Backeberg. Ibid. 28–29.
—— (1980). The genus Discocactus Pfeiffer. Buining-fonds, Succulenta, The Netherlands.
—— & Brederoo (1974a). Pseudopilocereus superfloccosus Buin. & Brederoo sp. nova. Cact.
Succ. J. (US) 46: 60–63, 96.
—— & —— (1974b). A new Discocactus species from Brazil. Ibid. 252–257.
Burnham, R. J. (1995). A new species of winged fruit from the Miocene of Ecuador:
Tipuana ecuatoriana (Leguminosae). Amer. J. Bot. 82: 1599–1608.
—— & Graham, A. (1999). The history of neotropical vegetation: new developments
and status. Ann. Missouri Bot. Gard. 86: 546–589.
Buxbaum, F. (1950–54). Morphology of cacti. Parts I–III. Abbey Garden Press, Pasadena.
—— (1958). The phylogenetic division of the subfamily Cereoideae [sic], Cactaceae.
Madroño 14: 177–206.
—— (1959). Die behaartblütigen Cephalienträger Südamerikas. Oesterr. Bot. Zeitschr.
106: 138–158.
—— (1962). Das phylogenetische System der Cactaceae. In: Krainz, H., Die Kakteen,
Lfg. 21.
—— (1968a). Die Entwicklungslinien der Tribus Cereeae Britt. et Rose emend. F.
Buxb. (Cactaceae, Cactoideae), 1. Beitr. Biol. Pflanzen 44: 215–276.
—— (1968b). Ibid., 2. Versuch einer phylogenetischen Gliederung innerhalb der Tribus.
Ibid. 389–433.
—— (1971). Buiningia. In: Krainz, H., Die Kakteen, Lfg. 46–47.
—— (1972). Coleocephalocereus luetzelburgii. Ibid., Lfg. 48–49.
—— (1975). Provisorium einer Neugliederung der Tribus Notocacteae (C VI). Ibid.,
Lfg. 60.
Byles, R. S. & Rowley, G. D. (1957). Pilosocereus Byl. & Rowl. nom. gen. nov.
(Cactaceae). Cact. Succ. J. Gr. Brit. 19: 66–67, 69.
Cavalcante Bernardes, L. M. (1951). Notas sobre o clima da bacia do São Francisco. Revta
Bras. Geogr. 13: 473–479.
Costa, C. M. R., Hermann, G., Martins, C. S., Lins, L. V. & Lamas, I. R. (1998).
Biodiversidade em Minas Gerais: um atlas para sua conservação. Biodiversitas, Belo
Horizonte. Pp. 94.
Das, A. B., Mohanty, S. & Das, P. (1998a). Variation in Karyotype and 4C DNA content
in six species of Melocactus of the family Cactaceae. Cytologia 63: 9–16.
—— —— —— (1998b). Interspecific variation in Nuclear DNA content and
chromosome analysis in Melocactus. Ibid. 63: 239–247.
Davis, S. D., Heywood, V. H., Herrera-MacBryde, O., Villa-Lobos, J., & Hamilton, A.
C. (1997). Centres of Plant Diversity. A guide and Strategy for their Conservation, 3. The
Americas. The World Wide Fund For Nature (WWF) and IUCN – The World
Conservation Union, IUCN Publications Unit, Cambridge, U.K.
BIBLIOGRAPHY
Dickie, J. B., Ellis, R. H., Kraak, H. L., Ryder, K. & Tompsett, P. B. (1990).
Temperature and seed storage longevity. Ann. Bot. 65: 197–204.
Dobat, K. & Piekert-Holle, T. (1985). Blüten und Fledermäuse: 238–243. Senckenbergischen
Naturforschenden Gesellschaft, Frankfurt.
Domingues, A. J. P. (1973). Relevo. Pp. 11–27 in Instituto Brasileiro de Geografia,
“Novo” Paisagens do Brasil, ed. 2 (Biblioteca Geográfica Brasileira, Serie D, publ. no.
2). Ministério do Planejamento e Coordenação Geral, Rio de Janeiro.
Doweld, A. & Greuter, W. (2001). Nomenclatural notes on Notocactus and on Alwyn
Berger’s “Kakteen”. Taxon 50: 879–885.
Eerkens, G. (1983). Braziliaanse Melocactussen met een HU nummer. Succulenta 62:
275–278.
—— (2000). Melocactus species HU 534. Ibid. 79: 274–275.
Eggli, U., Muñoz Schick, M. & Leuenberger, B. E. (1995). Cactaceae of South America:
The Ritter Collections. Englera 16: 1–646.
—— & Taylor, N., eds (1991). List of Cactaceae names from Repertorium Plantarum
Succulentarum (1950–1990). Royal Botanic Gardens, Kew, Richmond, U.K. Pp. 222.
Egler, W. (1951). Contribuição ao estudo da caatinga Pernambucana. Revta Bras. Geogr.
13: 577–590.
Eiten, G. (1983). Classificação da vegetação do Brasil. Conselho Nacional de
Desenvolvimento Científico e Tecnológico (CNPq), Brasília.
Endler, J. & Buxbaum, F. (1974). Die Pflanzenfamilie der Kakteen, ed. 3. A. Philler, Minden.
Esteves Pereira, E. (1997). Pierrebraunia Esteves, a new genus of Cactaceae from Brazil.
Cact. Succ. J. (US) 69: 296–302.
Farjon, A. & Page, C. N., comps (1999). Conifers — Status Survey and Conservation Action
Plan. IUCN/SSC Conifer Specialist Group, IUCN, Gland, Switzerland and
Cambridge, U.K.
Figueira, J. E. C., Vasconcellos-Neto, J., Garcia, M. A. & Teixeira de Souza, A. L. (1993).
O cactus e o lagarto. Ciência Hoje (BR) 15(89): 12–13.
—— —— —— —— (1994). Saurocory in Melocactus violaceus (Cactaceae). Biotropica
26: 295–301.
Funk, V. A. & Brooks, D. R. (1990). Phylogenetic Systematics as the Basis of
Comparative Biology. Smithsonian contributions to botany, 73. Pp. 45.
Gibson, A. C. & Horak, K. E. (1978). Systematic anatomy and phylogeny of Mexican
columnar cacti. Ann. Missouri Bot. Gard. 65: 999–1057 (1977, publ. 1978).
Giulietti, A. M. & Pirani, J. R. (1988). Patterns of geographic distribution of some plant
species from the Espinhaço range, Minas Gerais and Bahia, Brazil. In: Heyer, W. R.
& Vanzolini, P. E., eds, Proceedings of a Workshop on Neotropical Distribution Patterns:
39–69. Academia Brasileira de Ciências, Rio de Janeiro.
Glaziou, A. F. M. (1909). Plantae Brasiliae centralis a Glaziou lectae. Mem. Soc. Bot. France
1(3): 325–327.
Golding, J. S. & Smith, P. P. (2001). A 13-point strategy to meet conservation challenges.
Taxon 50: 475–477.
BIBLIOGRAPHY
Greuter, W., McNeill, J., Barrie, F. R., Burdet, H. M., Demoulin, V., Filgueiras, T. S.,
Nicolson, D. H., Silva, P. C., Skog, J. E., Trehane, P., Turland, N. J. & Hawksworth,
D. L., eds (2000). International Code of Botanical Nomenclature (Saint Louis Code).
Regnum Vegetabile, 138.
Gürke, M. (1908). Neue Kakteen-Arten aus Brasilien. Monatsschr. Kakt.-Kunde 18: 52–57.
Gusmão, M. (1999). O sertão virou pó. Veja (BR) 32(35): 122–125 (publ. 1 Sep. 1999).
Gutman, F., Bar-Zvi, D., Nerd, A. & Mizrahi, Y. (2001). Molecular typing of Cereus
peruvianus clones and their genetic relationships with other Cereus species evaluated by
RAPD analysis. J. Hort. Sci. Biotech. 76: 709–713.
Harley, R. M. (1988). Evolution and distribution of Eriope (Labiatae), and its relatives, in
Brazil. In: Heyer, W. R. & Vanzolini, P. E., eds, Proceedings of a Workshop on Neotropical
Distribution Patterns: 71–120. Academia Brasileira de Ciências, Rio de Janeiro.
—— & Mayo, S. J. (1980). Towards a Checklist of the flora of Bahia. Royal Botanic
Gardens, Kew, Richmond, U.K. Pp. 250.
—— & Simmons, N. A. (1986). Florula of Mucugê. Royal Botanic Gardens, Kew,
Richmond, U.K. Pp. 228.
Heath, P. V. (1992). The type of Monvillea Britton & Rose (Cactaceae). Taxon 41: 85–87.
—— (1994). New combinations in Pseudopilocereus Buxbaum. Calyx 4: 140–141.
Heek, W. van & Strecker, W. (1994). An den Standorten von Uebelmannia, 4.
Uebelmannia pectinifera Buining var. multicostata Buining & Brederoo (HU 362). Kakt.
and. Sukk. 45: 134–136.
—— —— (1995). Uebelmannia gummifera (Backeberg & Voll) Buining HU 282 und HU
859. Ibid. 46: 121–123.
—— —— (2002). Am Fundort von Micranthocereus streckeri. Ibid. 53: 113–116.
Herm, K., Hofacker, A., Charles, G., Heek, W. van, Bohle, B., Strecker, W. & Heimen,
G. (2001). Kakteen in Brasilien / Cacti in Brazil. Published by the authors (ISBN 3-00-
007573-9). Pp. 176.
Herschkovitz, M. A. & Zimmer, E. A. (1997). On the evolutionary origin of the cacti.
Taxon 46: 217–232.
Hoehne, F. C. (1915). Annexo 5, História Natural: Botânica 6: 55–56. In: Commissão de
Linhas Telegráphicas Estratégicas de Matto Grosso ao Amazonas. Rio de Janeiro.
Hofacker, A. (2001). Pilosocereus bohlei Hofacker spec. nov. – eine neue Art aus Bahia.
Kakt. and. Sukk. 52: 253–257, illus.
—— (2002). Der Formenkreis um Discocactus bahiensis Britton & Rose in Bahia,
Brasilien. Ibid. 53: 200–205, illus.
—— & Braun, P. J. (1998). Staubfäden in zwei Reihen: Die Gattung Brasilicereus
Backeberg. Ibid. 49: 265–268, illus.
Hooker, J. D. (1899). Cereus paxtonianus. Curtis’s Bot. Mag. 125: tab. 7648.
Hooker, W. J. (1838). Melocactus depressus. Ibid. 65: tab. 3691.
Howard, R. (1989). Flora of the Lesser Antilles, Leeward and Windward Islands, 5: 398–422.
Arnold Arboretum, Harvard Univ., Jamaica Plain, Massachusetts.
—— & Touw, M. (1982). Opuntia species in the Lesser Antilles. Cact. Succ. J. (US) 54:
170–179.
BIBLIOGRAPHY
Hunt, D. R. (1967). Cactaceae. In: Hutchinson, J., Families and Genera of Flowering Plants,
ed. 2. Oxford.
—— (1969). A synopsis of Schlumbergera Lem. (Cactaceae). Kew Bull. 23: 255–263.
—— (1984). The Cactaceae of Plumier’s Botanicum Americanum. Bradleya 2: 39–64.
—— (1988). New and unfamiliar names for use in the European Garden Flora: Addenda
and Corrigenda. Ibid. 6: 100.
—— (1989). Notes on Selenicereus (A. Berger) Britton & Rose and Aporocactus Lemaire
(Cactaceae-Hylocereinae). Ibid. 7: 89–96.
—— (1991). Stabilization of Names in Succulent Plants. Pp. 151–155, in: Hawksworth,
D. L., Improving the Stability of Names: needs and options. Regnum Vegetabile, 123.
Koeltz Scientific Books, Königstein.
—— (1992a). Selenicereus. A further note on the ‘Pitaya’. In: Hunt, D. R. & Taylor, N.
P., eds, Notes on miscellaneous genera of Cactaceae (2). Bradleya 10: 32.
—— (1992b). CITES Cactaceae Checklist. Royal Botanic Gardens, Kew, Richmond, U.K.
—— (1999a). Ibid., ed. 2. Ibid.
—— (1999b). The opuntioids: all sections or all genera? Cact. Consensus Initiatives 8: 3–6.
—— (2000). Classification of subfam. Cactoideae. Cact. Systematics Initiatives 9: 3.
—— & Taylor, N. P., eds (1986). The genera of the Cactaceae: towards a new consensus.
—— —— (1990). The genera of Cactaceae: progress towards consensus. Ibid. 8: 85–107.
—— ——, eds (1991). Notes on miscellaneous genera of Cactaceae. Ibid. 9: 81–92.
—— ——, eds (1992). Notes on miscellaneous genera of Cactaceae (2). Ibid. 10: 17–32.
––––– ––––– & Zappi, D. (1994). Appendix IV: Documentation of accepted names and
synonyms. In: Kattermann, F., Eriosyce (Cactaceae), The genus revised and amplified.
Succulent Pl. Res. 1: 131–169.
Ibisch, P. L., Boegner, A., Nieder, J. & Barthlott, W. (1996). How diverse are neotropical
epiphytes? An analysis based on the “Catalogue of the Flowering Plants and
Gymnosperms of Peru”. Ecotropica 2: 13–28.
Innes, C. & Glass, C. (1991). Illustrated Encyclopaedia of Cacti. Quarto Publishing, London.
Pp. 320.
IUCN (2001). IUCN Red List Categories: Version 3.1. IUCN, Gland, Switzerland and
Cambridge, UK. Pp. 23.
Judd, W. S., Campbell, C. S., Kellogg, E. A. & Stevens, P. F. (1999). Plant Systematics. A
phylogenetic approach. Sinauer Associates, Sunderland, Massachusetts. Pp. 464.
Kiesling, R. (1986). Tipos de Cactáceas Sudamericanas en herbarios extranjeros. Bol. Soc.
Argent. Bot. 24: 381–386.
—— (1990). Cactus de la Patagonia. Reprinted with English summary from Flora
Patagónica 5: 218–243 (1988).
—— (1994). Monvillea kroenleinii, a new species from Paraguay. Cact. Succ. J. (US) 66:
157–165.
—— (1999). Origin, Evolution and Distribution of Opuntia ficus-indica. XVI International
Botanical Congress Abstracts: 36.
King, L. C. (1956). A Geomorfologia do Brasil Oriental. Revta Bras. Geogr. 18: 147–265.
BIBLIOGRAPHY
Krook, R. & Mottram, R. (2001). Opuntia Index. Part 7: Nomenclatural note and P–Q.
Labouret, J. (1853). Monographie des Cactées. Dusacq, Librarie Agricole de la Maison
Rustique, Paris.
Lemaire, C. C. (1838). Cactearum aliquot novarum ac insuetarum in Horto Monvilliano cultarum
accurata descriptio. Levrault, Paris. Pp. 40.
—— (1862). Histoire et Révision du Gènre Pilocereus. Rev. Hort. (Paris) 34: 426–430.
Leme, E. M. C. & Marigo, L. C. (1993). Bromeliads in the Brazilian wilderness. Marigo
Comunicação Visual, Rio de Janeiro.
Leuenberger, B. E. (1976). Die Pollenmorphologie der Cactaceae und ihre Bedeutung
für die Systematik. Diss. Bot. 31: 1–321.
—— (1978). Type specimens of Cactaceae in the Berlin-Dahlem Herbarium. Cact. Succ.
J. Gr. Brit. 40: 101–104.
—— (1986). Pereskia (Cactaceae). Mem. New York Bot. Gard. 41: 1–140.
—— (1991). Interpretation of Cactus ficus-indica L. and Opuntia ficus-indica (L.) Miller
(Cactaceae). Taxon 40: 621–627.
—— (1993). Interpretation of Cactus opuntia L., Opuntia vulgaris Mill., and O. humifusa
(Rafin.) Rafin. (Cactaceae). Ibid. 42: 419–429.
—— (1997). Cactaceae. In: Görts van Rijn, A. R. A. & Jansen-Jacobs, M. J., eds, Flora
of the Guianas, 31(18). Royal Botanic Gardens, Kew, Richmond, U.K.
—— (2002). The South American Opuntia ser. Armatae (= O. ser. Elatae) (Cactaceae).
Bot. Jahrb. Syst. 123: 413–439.
Lewis, G. P. (1987). Legumes of Bahia. Royal Botanic Gardens, Kew, Richmond, U.K.
Pp. 369.
—— (1998). Caesalpinia. A revision of the Poincianella-Erythrostemon Group. Ibid. Pp. 233.
Linder, H. P. (1995). Setting Conservation Priorities: The Importance of Endemism and
Phylogeny in the Southern African Orchid Genus Herschelia. Conserv. Biol. 9: 585–595.
Linke, A. (1858). Nachtrag zum Verzeichniss der Cacteen-Sammlung bei August Linke. Berlin.
Locatelli, E. & Machado, I. C. (1999a). Comparative Study of the Floral Biology in Two
Ornithophilous Species of Cactaceae: Melocactus zehntneri and Opuntia palmadora.
—— & —— (1999b). Floral Biology of Cereus fernambucensis: a sphingophilous cactus of
restinga. Ibid. 17: 86–94.
—— —— & Medeiros, P. (1997). Floral Biology and Bat Pollination in Pilosocereus
catingicola (Cactaceae) in Northeastern Brazil. Ibid. 15: 28–34.
Luetzelburg, P. von (1925–26). Estudo Botânico do Nordéste, 1–3. Inspetoria Federal de
Obras Contra as Secas, Rio de Janeiro (dated 1922–23, but vol. 1 publ. 1925, vols 2–3
publ. 1926).
Madsen, J. E. (1989). 45. Cactaceae. In: Harling, G. & Andersson, L., Flora of Ecuador, 35.
Pp. 79.
Maffei, M., Meregalli, M. & Scannerini, S. (1997). Chemotaxonomic significance of
surface wax n-alkanes in the Cactaceae. Biochem. Syst. Ecol. 25: 241–253.
BIBLIOGRAPHY
Martius, C. P. F. von (1832). Beschreibung einiger neuen Nopaleen. Nova Acta Leop. 16:
321–362.
—— (1841). Flora brasiliensis, 1(1). Oldenbourg, Munich.
Mauseth, J. D. (1996). Comparative Anatomy of Tribes Cereeae and Browningieae
(Cactaceae). Bradleya 14: 66–81.
McMillan, A. J. S. & Horobin, J. (1995). Christmas Cacti: The genus Schlumbergera and
its hybrids. Succulent Pl. Res., 4.
Menezes, A. I. (1949). Flora da Bahia. Companhia Ed. Nacional, São Paulo.
Mittler, L. (1844). Taschenbuch für Cactusliebhaber, 2. Schreck, Leipzig.
Moser, G. (1985). Kakteen. Adolfo Maria Friedrich und sein schönes Paraguay. G. Moser,
Kufstein, Austria.
Myers, N., Mittermeier, R. A., Mittermeier, C. G., Fonseca, G. A. B. da & Kent, J.
(2000). Biodiversity hotspots for conservation priorities. Nature 403: 853–858.
Nassar, J. M., Hamricks, J. L. & Fleming, T. H. (2002). Genetic variation and population
structure of the mixed-mating cactus, Melocactus curvispinus (Cactaceae). Heredity 87:
69–79.
Nimer, E. (1973). Clima. Pp. 33–49 in Instituto Brasileiro de Geográfia, “Novo” Paisagens
do Brasil, ed. 2 (Biblioteca Geográfica Brasileira, Serie D, publ. no. 2). Ministério do
Planejamento e Coordenação Geral, Rio de Janeiro.
Nixon, K. C. & Wheeler, Q. D. (1990). An amplification of the phylogenetic species
concept. Cladistics 6: 211–223.
Nyffeler, R. (1997). Stem anatomy of Uebelmannia (Cactaceae) — with special reference
to Uebelmannia gummifera. Bot. Acta 110: 489–495.
—— (1998). The genus Uebelmannia Buining (Cactaceae: Cactoideae). Bot. Jahrb. Syst.
120: 145–163.
—— (1999). The Evolution of Stem Characters in the Cacti. I.O.S. Bull. 7: 27.
—— (2000a). A Sketch of the Cactus Tribe[s] — What do the Molecules Tell Us? I.O.S.
Bull. 8: 25.
—— (2000b). Molecular Systematics and the Polyphyly of the Tribe Notocacteae
(Cactaceae). Ibid. 8: 25–26.
—— (2000c). Should Pfeiffera be resurrected? Cact. Systematics Initiatives 10: 10–11.
—— (2001). What about the tribe Notocacteae? Ibid. 12: 25–27.
—— (2002). Phylogenetic relationships in the cactus family (Cactaceae) based on
evidence from trnK/matK and trnL-trnF sequences. Amer. J. Bot. 89: 312–326.
—— & Eggli, U. (1996). Berger’s “Kakteen” — the end of a nomenclatural nightmare
in sight? Taxon 45: 301–304.
Oldfield, S., comp. (1997). Cactus and Succulent Plants — Status Survey and Conservation
Action Plan. IUCN/SSC Cactus and Succulent Specialist Group. IUCN, Gland,
Switzerland and Cambridge, U.K. Pp. 10 + 212.
O’Leary, M. C. & Boyle, T. H. (1998). Segregation distortion at isozyme locus Lap-1 in
Schlumbergera (Cactaceae) is caused by linkage with the gametophytic self-
incompatibility (S) locus. J. Heredity 89(3): 206–210.
BIBLIOGRAPHY
Oliveira-Filho, A. T. de (1993). Gradient analysis of an area of coastal vegetation in the

state of Paraíba, northeastern Brazil. Edinb. J. Bot. 50: 217–236.
—— & Ratter, J. A. (1995). A study of the origin of central Brazilian forests by the
analysis of plant species distribution patterns. Ibid. 52: 141–194.
Pennington, R. T., Prado, D. E. & Pendry, C. A. (2000). Neotropical seasonally dry
forests and Quaternary vegetation changes. J. Biogeography 27: 261–273.
Petit, S. (1999). The effectiveness of two bat species as pollinators of two species of
columnar cacti on Curaçao. Haseltonia 6: 22–31 (1998, publ. 1999).
Pfeiffer, L. (1837). Enumeratio diagnostica cactearum hucusque cognitarum: 43–46. L.
Oehmigke, Berlin.
Pickel, B. (1938). Catálogo do Herbário da Escola Superior de Agricultura em Tapera
(Pernambuco). Bol. Mus. Nac. (Rio de Janeiro) 13: 63–132.
Pinkava, D. J. (2002). On the evolution of the continental North American
Opuntioideae (Cactaceae). Succulent Pl. Res. 6: 59–98.
Piso, W. (1648): Piso, W. & Marcgrave, G. (1648). Historia naturalis Brasiliae ..... Frans
Hack, Leiden & Ludovic Elzevier, Amsterdam.
Porembski, S., Martinelli, G., Ohlemüller, R. & Barthlott, W. (1998). Diversity and
ecology of saxicolous vegetation mats on inselbergs in the Brazilian Atlantic rainforest.
Diversity and Distributions 4: 107–119.
Prado, D. E. (1991). A critical Evaluation of the Floristic Links between Chaco and Caatingas
Vegetation in South America. PhD thesis, University of St Andrews, Scotland.
[Unpublished.]
—— & Gibbs, P. E. (1993). Patterns of species distributions in the dry seasonal forests of
South America. Ann. Missouri Bot. Gard. 80: 902–927.
Prance, G. T. & Mori, S. A. (1982). In Memoriam, Dárdano de Andrade-Lima
(1919–1981). Brittonia 34: 351–354.
Raw, A. (1996). Territories of the ruby-topaz hummingbird Chrysolampis mosquitus at
flowers of the “turk’s-cap” cactus, Melocactus salvadorensis in the dry caatinga of north-
eastern Brazil. Revta Bras. Biol. 56: 581–584.
Reid, W. V. (1998). Biodiversity hotspots. Trends Ecol. Evol. 13: 275–280.
Renvoize, S. A. (1984). Grasses of Bahia. Royal Botanic Gardens, Kew, Richmond, U.K.
Pp. 301.
Resende, S. V. (2002). Estudos Citogenéticos em Melocactus (Cactaceae). Depto de Biologia
Geral, Instituto de Biologia, Universidade Federal da Bahia, Salvador. Pp. 39.
[Unpublished.]
Riccobono, V. (1909). Studii sulle Cattee del R. Orto Botanico di Palermo. Boll. Ort.
Bot. Palermo 8: 215–266.
Ritter, F. (1968). Die Cephalien tragenden Kakteen Brasiliens. Kakt. and. Sukk. 19:
87–96, 119–123, 140, 156–162.
—— (1979). Kakteen in Südamerika, 1. F. Ritter Selbstverlag, Spangenberg.
—— (1980). Ibid., 2. Ibid.
Rizzini, C. T. (1979). Tratado de Fitogeografia do Brasil. HUCITEC, Universidade de São
Paulo, São Paulo.
BIBLIOGRAPHY
—— (1982). Melocactus no Brasil. Instituto Brasileiro de Defesa Florestal (IBDF), Rio de

Janeiro.
—— (1986). Sobre a Cactácea dendróide do calcário de Minas Gerais. Rev. Brasil. Biol.
46: 781–784.
—— & Mattos-Filho, A. (1992). Contribuição ao conhecimento das floras do nordeste
de Minas Gerais e da Bahia Mediterrânea. Jardim Botânico do Rio de Janeiro. Estudos
e Contribuições, 9. Pp. 95.
Rocha, E. A. & Agra, M. F. (2002). Flora do Pico do Jabre, Paraíba, Brasil: Cactaceae
Juss. Acta bot. bras. 16: 15–21.
Rodal, M. J. N., Sales, M. F. de, Mayo, S. J. (1998). Florestas Serranas de Pernambuco.
Localização e Conservação dos Remanescentes dos Brejos de Altitude. Universidade Federal
Rural de Pernambuco, Recife, Brazil. Pp. 30.
Ross, R. (1981). Chromosome counts, cytology and reproduction in the Cactaceae.
Amer. J. Bot. 68: 463–470.
Rowley, G. D. (1992). Didiereaceae. ‘Cacti of the Old World’. British Cactus & Succulent
Society, Richmond, U.K.
Ruiz, A. M. S., Soriano, P. J., Cavelier, J. & Cadena, A. (1997). Relaciones mutualísticas
entre el murciélago Glossophaga longirostris y las cactáceas columnares en la zona árida
de La Tatacoa, Colombia. Biotropica 29: 469–479.
Sales, M. F. de, Mayo, S. J. & Rodal, M. J. N. (1998). Plantas Vasculares das Florestas
Serranas de Pernambuco. Um Checklist da Flora Ameaçada dos Brejos de Altitude,
Pernambuco, Brasil. Universidade Federal Rural de Pernambuco, Recife, Brazil. Pp.
130.
Santos, F. A. R., Watanabe, H. M. & Alves, J. L. de H. (1997). Pollen morphology of
some Cactaceae of North-Eastern Brazil. Bradleya 15: 84–97.
Savolainen, V., Fay, M. F., Bank, M. van der, Powell, M., Albach, D. C., Weston, P.,
Backlund, A., Johnson, S. A. & Chase, M. W. (2000). Phylogeny of the eudicots: a
nearly complete familial analysis based on rbcL gene sequences. Kew Bull. 55: 257–309.
Scheinvar, L. (1985). Cactáceas. In: Reitz, R., Flora Ilustrada Catarinense, I. Itajaí, Santa
Catarina, Brazil.
—— (1988). Cactus triangularis Linne, Cereus ocamponis Salm-Dyck, Cactus triangularis
Vellozo, Cereus undatus Haworth és Cereus trigonus Haworth. Kaktusz Világ (Hungary)
1988(4): 63–69.
Schlindwein, C. & Wittmann, D. (1997). Stamen movements in flowers of Opuntia
(Cactaceae) favour oligolectic pollinators. Pl. Syst. Evol. 204: 179–193.
Schulz, R. & Machado, M. (2000). Uebelmannia and their environment. Schulz
Publishing, Teesdale, Australia. Pp. 160.
Schumann, K. M. (1890). Cactaceae. In: Martius, C. P. F. von, Flora brasiliensis, 4(2).
—— (1894). Cactaceae. In: Engler, A. & Prantl, K., Das Pflanzenfamilien, 3(6a): 156–205.
—— (1897–98). Gesamtbeschreibung der Kakteen. J. Neumann, Neudamm.
—— (1903). Ibid., Nachträge 1898 bis 1902. Ibid.
Stafleu, F. & Cowan, R. S. (1983). Taxonomic Literature, ed. 2, 4. Regnum Vegetabile, 110.
BIBLIOGRAPHY
Stannard, B. E., ed. (1995). Flora of the Pico das Almas, Chapada Diamantina, Bahia. Royal
Botanic Gardens, Kew, Richmond, U.K.
Stuppy, W. (2002). Seed characters and generic classification of the Opuntioideae.
Succulent Pl. Res. 6: 25–58.
—— & Huber, H. (1991). Samenmerkmale und Gattungsgliederung der Opuntioideae.
Kakt. and. Sukk. 42: 122–126, 144–147, 165–167, 208–211.
Supthut, D. (1999). A short Chronicle of 50 Years of IOS. I.O.S. Bull. 7: 6–16.
Taylor, N. P. (1980). Notes on the genus Melocactus (1): E Brazil. Cact. Succ. J. Gr. Brit.
42: 63–70.
—— (1981). Reconsolidation of Discocactus Pfeiff. Ibid. 43: 37–40.
—— (1982). Additional notes on E Brazilian Melocactus. Ibid. 44: 7–8.
—— (1991a). The genus Melocactus in Central and South America. Bradleya 9: 1–80.
—— (1991b). Cereus, Micranthocereus. In: Hunt, D. R. & Taylor, N. P., eds, Notes on
miscellaneous genera of Cactaceae. Ibid. 9: 85–86, 89.
—— (1991c). [Submission to the Republic of Brazil’s CITES Management Authority
(IBAMA), via the CITES Secretariat, Lausanne, Switzerland, proposing the inclusion
of Discocactus, Uebelmannia and 4 species of Melocactus in Appendix I of the
Convention on International Trade in Endangered Species of fauna & flora.]
—— (1991d). Proposal to conserve 5401a Arthrocereus A. Berger (Cactaceae) with a new
type. Taxon 40: 660–662.
—— (1992a). Arthrocereus, Cereus. In: Hunt, D. R. & Taylor, N. P., eds, Notes on
miscellaneous genera of Cactaceae (2). Bradleya 10: 17, 20–25.
—— (1992b). Plants in peril, 17. Melocactus conoideus. Kew Magazine 9: 138–142.
—— (1994). Proposal to conserve Cactus cruciformis Vell. (Cactaceae). Taxon 43: 469–470.
—— (1995a). Hylocereus. In: Rowley, G. D., ed., The Sessé & Mociño cactus plates – A
postscript. Bradleya 13: 119–120.
—— (1995b). Validation of Monvillea kroenleinii Kiesling as Cereus kroenleinii and a note
on extension of its range. Kew Bull. 50: 819–820.
—— (1999). Rhipsalis puniceodiscus. Cactaceae. Curtis’s Bot. Mag. 16: 29–33.
—— (2000). Taxonomy and Phytogeography of the Cactaceae of Eastern Brazil. PhD thesis.
Life Sciences, The Open University, UK. Pp. 398+16. [Unpublished.]
—— (mss. [1996]). Cladistic analysis of 25 taxa representative of major elements of
Rhipsalideae (Cactaceae), based on 29 morphological characters and utilising PAUP3
with Adams consensus tree.
—— & Eggli, U. (1991). Stephanocereus. In: Hunt, D. R. & Taylor, N. P., eds, Notes on
miscellaneous genera of Cactaceae. Bradleya 9: 91.
—— Stuppy, W. & Barthlott, W. (2002). Realignment and revision of the Opuntioideae
of Eastern Brazil. Succulent Pl. Res. 6: 99–132.
—— & Zappi, D. C. (1989). An alternative view of generic delimitation and
relationships in tribe Cereeae (Cactaceae). Bradleya 7: 13–40.
—— —— (1990). Notes on Leocereus Britton & Rose. Ibid. 8: 108.
—— —— (1991). Arthrocereus, Discocactus. In: Hunt, D. R. & Taylor, N. P., eds, Notes
on miscellaneous genera of Cactaceae. Ibid. 9: 84–86.
BIBLIOGRAPHY
—— —— (1992a). Cactaceae do Vale do Rio Jequitinhonha, Minas Gerais. Acta Bot.

Bras. 51: 63–69 (1991, publ. 1992).
—— —— (1992b). Leocereus. In: Hunt, D. R. & Taylor, N. P., eds, Notes on
miscellaneous genera of Cactaceae (2). Bradleya 10: 25–26.
—— —— (1992c). Proposal to conserve Cereus jamacaru DC. (Cactaceae) with a new
type. Taxon 41: 590–591.
—— —— (1995). Appendix II: Herbarium records. In: McMillan, A. J. S. & Horobin,
J. F., Christmas Cacti. The genus Schlumbergera and its hybrids. Succulent Pl. Res. 4:
80–81.
—— —— (1996). Two species of Arrojadoa from Eastern Brazil. Curtis’s Bot. Mag. 13:
70–78.
—— —— (1997). Nomenclatural adjustments and novelties in Brazilian Cactaceae.
Cact. Consensus Initiatives 3: 7–8.
—— —— & Eggli, U. (1992). Pseudoacanthocereus. In: Hunt, D. R. & Taylor, N. P., eds,
Notes on miscellaneous genera of Cactaceae (2). Bradleya 10: 28–31, pl. V.
Tricart, J. (1985). Evidence of Upper Pleistocene dry climates in northern South
America. Chapter 10, pp. 197–217, in: Douglas, I. & Spencer, T., Environmental change
and tropical geomorphology. Allen & Unwin, London.
Uebelmann, W. J. (1996). Horst & Uebelmann Feldnummernliste. W. J. Uebelmann,
Zufikon, Switzerland. Pp. 204.
—— & Braun, P. J. (1984). 25 Jahre HU Horst–Uebelmann. Beschriebene Arten und
Feldnummern. Flora-Buchhandel, Titisee-Neustadt, Germany. Pp. 64.
Ule, E. (1908). Kautschukgewinnung und Kautschukhandel in Bahia. Notizbl. Bot. Gart.
Berlin 5(41a): Tab. IV [map].
Urban, I. (1906). Vitae itineraque collectorum botanicorum. In: Martius, C. P. F. von,
Flora brasiliensis 1(1): 1–154.
Velloso, A. L., Sampiao, E. V. S. B. & Pareyn, F. G. C., eds (2002). Ecorregiões Propostas
para o Bioma Caatinga. Associação Plantas do Nordeste & Instituto de Conservação
Ambiental, The Nature Conservancy do Brasil, Recife. Pp. 75.
Wallace, R. S. (1995). Molecular systematic study of the Cactaceae: Using chloroplast
DNA variation to elucidate Cactus phylogeny. Bradleya 13: 1–12.
—— (1997). The phylogenetic position of Mediocactus hahnianus. Cact. Consensus Initiatives
4: 11–12.
—— & Cota, J. H. (1996). An intron loss in the chloroplast gene rpoC1 supports a
monophyletic origin for the subfamily Cactoideae of the Cactaceae. Curr. Genet. 29:
275–281.
—— & Dickie, S. L. (2002). Taxonomic implications of chloroplast DNA sequence
variation in subfam. Opuntioideae (Cactaceae). Succulent Pl. Res. 6: 9–24.
—— & Gibson, A. C. (2002). Evolution and Systematics. Pp. 1–21. In: Nobel, P. S.,
Cacti. Biology and Uses. Univ. California Press, Berkeley etc.
Walters, S. M., Alexander, J. C. M., Brady, A., Brickell, C. D., Cullen, J., Green, P. S.,
Heywood, V. H., Matthews, V. A., Robson, N. K. B., Yeo, P. F. & Knees, S. G.
(1989). The European Garden Flora, 3. Cambridge Univ. Press, Cambridge, U.K. etc.
BIBLIOGRAPHY
Warming, J. E. B. (1908). Lagoa Santa. Trans. A. Löfgren. In: Warming, J. E. B. & Ferri,
M. G. (1973), q.v.
—— & Ferri, M. G. (1973). Lagoa Santa & A Vegetação dos Cerrados Brasileiros. Editora
Univ. São Paulo and Itatiaia Editora, Belo Horizonte.
WCMC (1992). Protected Areas of the World: A review of national systems, 4. Neartic and
Neotropical. IUCN, Gland, Switzerland and Cambridge, U.K.
Werdermann, E. (1932). Cephalocereus Lehmannianus Werd. etc. In: Fedde, F., Rep. Sp.
Nov., Sonder-Beih. C, Lfg. 11.
—— (1933). Brasilien und seine Säulenkakteen. J. Neumann, Neudamm.
—— (1934). Cactaceae novae. Notizbl. Bot. Gart. Berlin 12: 228–229.
—— (1942). Brazil and its Columnar Cacti. Abbey Garden Press, Pasadena. [Werdermann
(1933), trans. R. W. Kelly; pp. 84–120 trans. & ed. L. Croizat.]
Whitehead, P. J. P. & Boeseman, M. (1989). A portrait of Dutch 17th century Brazil.
Animals, plants and people by the artists of Johan Maurits of Nassau. North-Holland
Publishing Company. Amsterdam / Oxford / New York.
Wurdack, J. (1970). Erroneous data in Glaziou collections of Melastomataceae. Taxon 19:
911–913.
Yang, X.-Y. (1999). Storage, germination and characterisation of Cactaceae seed. PhD thesis,
Jilin Agricultural Univ., P. R. China. [Unpublished.]
Zappi, D. C. (1989). A família Cactaceae nos Campos Rupestres da Cadeia do Espinhaço, Minas
Gerais. MSc thesis, Univ. São Paulo. [Unpublished.]
—— (1991). Flora da Serra do Cipó, Minas Gerais: Cactaceae. Bol. Bot. Univ. São Paulo
12: 43–59 (1990, publ. 1991).
—— (1992). Revisão taxonómica do gênero Pilosocereus (Cactaceae) no Brasil. PhD thesis,
Univ. São Paulo. [Unpublished; subsequently updated and published as the
following:]
—— (1994). Pilosocereus (Cactaceae). The genus in Brazil. Succulent Pl. Res. 3: 1–160.
—— & Taylor, N. P. (1991). Cipocereus. In: Hunt, D. R. & Taylor, N. P., eds, Notes on
miscellaneous genera of Cactaceae. Bradleya 9: 86.
—— —— (1994). Cactaceae del Brasile orientale / Eastern Brazilian Cacti. Piante Grasse
13(4), suppl.: 65–78 (1993, publ. 1994).
ACKNOWLEDGEMENTS
6.2 ACKNOWLEDGEMENTS & PHOTO CREDIT CODES

Any study in systematics and biogeography of this size inevitably relies on the help and
kindness of many individuals. These acknowledgements cannot hope to be
comprehensive and thus, at the outset, the authors wish to apologize to anyone who feels
under-valued by the omission of their name or organization from the following list.
To many colleagues, both current, former and retired, at the Royal Botanic Gardens,
Kew, we owe more than a mention: Dr DAVID HUNT (Honorary Research Fellow and
Cactaceae specialist), Dr RAY HARLEY (shared field knowledge & photos), Dr ALAN
PATON (NT’s PhD supervisor), Dr GWIL LEWIS (for taking the trouble to photograph
cacti on his travels and for help with literature), Dr EIMEAR NIC LUGHADHA (for varied
advice), Dr NICHOLAS HIND (for cactus collections and photographs), Dr DICK
BRUMMITT (for much nomenclatural advice), Dr WOLFGANG STUPPY (for collaboration
on Opuntioideae and IUCN Red Listing), Dr MIKE MAUNDER (conservation literature),
STEVE RUDDY (for help with software on countless occasions), JUSTIN MOAT (Herbarium
GIS Unit, for help with mapping), PHIL GRIFFITHS (field companion), Dr CHARLIE
BUTTERWORTH and ROBYN COWAN (for molecular phylogenies), Dr ILIA LEITCH (for
obtaining cytogenetics literature), Dr SIMON MAYO (for helpful discussions), MARILYN
WARD (illustrations research, Kew Library), ANDREW MCROBB (Photographic Unit,
Information Services Dept) and Dr BOB ALLKIN (CNIP, for logistical support in Brazil).
The following Brazilian friends are sincerely thanked for facilitating field excursions
and herbarium visits (some generously providing accommodation in their own homes)
and/or for supplying information and photographs (institutions in which they are based
identified by Index Herbariorum codes): Prof. EMERSON ROCHA (JPB, PEUFR and now
Univ. Santa Cruz, Ilhéus, Bahia) & Profa MARIA DE FÁTIMA AGRA (JPB); the late Dr
ANDRÉ DE CARVALHO & JOMAR JARDIM (CEPEC, the latter now at HUEFS); MARLON
MACHADO (HUEFS, for much valuable information, photographs and assistance) and also
his collaborators from Vitória da Conquista, Bahia, namely RAYMUNDO REIS JUNIOR &
JOSÉ CARLOS FALCON (both members of the Associação Conquistense de Orquidófilos),
AVALDO DE OLIVEIRA SOARES FILHO & MARCELLO MOREIRA (Univ. Estadual do
Sudoeste da Bahia) and RITA AUGUSTO (Cruz das Almas, BA, who acted as guide at the
Serra da Jibóia, Mun. Santa Teresinha); HÉLIO BOUDET FERNANDES (MBML); Dr
FRANCISCO DE ASSIS DOS SANTOS (SP, now HUEFS); Profa MARIA LUIZA SANTOS
(Aracaju, Sergipe, for facilitating access to ASE and the Serra da Itabaiana); Dra ANA
MARIA GIULIETTI (SPF, now HUEFS, DZ’s PhD supervisor), Dra NANUZA LUIZA DE
MENEZES (SPF, ‘que maravilha !’), Dr J. R. PIRANI (SPF, DZ’s mentor); JOÃO BATISTA
(field companion) & Dra MANOELA DA SILVA (MG); Dr OBERDAN PEREIRA (Univ. Fed.
Vitória, ES); Dr LUCIANO PAGANUCCI QUEIROZ, Dr EDUARDO BORBA, CÁSSIA TATIANA
ANDRADE, Dra LIGIA FUNCH & postgraduate students (HEUFS); ALOÍSIO CARDOSO
(Administrador, APA de Gruta dos Brejões, BA); Irmão DELMAR (Presidente, Grupo
Ambientalista Morrense, Morro do Chapéu, BA); MÍRIAM PIMENTEL MENDONÇA, INÊS
RIBEIRO DE ANDRADE, MARIA GUADALUPE CARVALHO & JULIANA ORDONES REGO
(Fundação Zoo-Botânica de Belo Horizonte); Padre CÉLIO M. DELL’AMORE (Director,
ACKNOWLEDGEMENTS
RPPN do Caraça, Catas Altas, Minas Gerais); JULIO ANTÔNIO LOMBARDI & JOÃO
RENATO STEHMANN (BHCB); Dra MARIA DO CARMO E. AMARAL, Dr VOLKER
BITTRICH & LIDYANNE AONA (UEC); PATRÍCIA SOFFIATTI & RAFAELA FORZZA (SPF, the
latter now at RB); Dra LÚCIA ROSSI & E. L. CATHARINO (SP); Dr FABIO DE MELO SENE
(Univ. São Paulo, Ribeirão Preto); Dr GERT HATSCHBACH (MBM); LÚCIA HELENA DE
OLIVEIRA (IBAMA, Brasília) and ROSÂNGELA LYRA-LEMOS (MAC). The Brazilian
government’s agencies, CNPq and IBAMA, who approved collecting permits and special
visas in 1988, 1990 & 1999 (EX-14/88, 181/90-DEVIS), with IBAMA also part funding
NT’s participation in a Brazilian ecosystems conservation symposium in 1997, are
gratefully acknowledged for their support.
European and North American friends and acquaintances whose help has been
invaluable include Dr URS EGGLI (ZSS, field companion, photographer and
bibliographer), Prof. Dr WILHELM BARTHLOTT (BONN, collaborator on Rhipsalideae
and seed-morphology, many photographs), Dr STEFAN POREMBSKI (ROST, inselberg
cacti), Dr BEAT LEUENBERGER (B, photographs of types & botanists), Dr MATS HJERTSON
(supplier of bibliography and field companion), Dr RALF BAUER (epiphytes), ANDREAS
HOFACKER and acquaintances (Böblingen, Germany; photographs and discussions), Dr
TOBY PENNINGTON (E, biogeography discussions and references), Dr PETER GIBBS (STA,
NT’s PhD external examiner), Prof. Dr PAUL MAAS (U, for facilitating study visits), Dr
ROBERT WALLACE (AMES, for information on gene sequence phylogenies), Dr DAN
NICOLSON (US), the late Dr JOHN WURDACK (US), MYRON KIMNACH (HNT), Dr W.
WAYT THOMAS (NY) and the directors/curators of other herbaria listed in Chapter 1.
U.K. cactophiles, GRAHAM CHARLES, PAUL KLAASSEN, TERRY HEWITT (Holly Gate
Nursery) and ROY MOTTRAM (Whitestone Gardens) kindly assisted with plant material
and/or information/photographs (especially G. Charles!), while BRITISH AIRWAYS’
‘Assisting Conservation’ programme provided free flights to NT for travel to Brazil and
Europe on two occasions, enabling participation in conservation planning, presentation
of papers and receipt of feedback.
PHOTO CREDIT CODES

B = © courtesy of the Botanic Garden and Botanical Museum Berlin-Dahlem, WR =
Werner Rauh† (courtesy of Wilhelm Barthlott), NT = Nigel Taylor, DZ = Daniela
Zappi, GC = Graham Charles, WB = Wilhelm Barthlott, UE = Urs Eggli, K = ©
Trustees of the Royal Botanic Gardens Kew, AJ = Andrea L. Jones, MM = Marlon
Machado, RB = Ralf Bauer, ER = Emerson Rocha, RH = Ray Harley, AH = Andreas
Hofacker, KH = Konrad Herm, WvH = Werner van Heek.
GLOSSARY AND INDEXES
7. GLOSSARY AND INDEXES
7.1 GLOSSARY OF SPECIALIZED BOTANICAL TERMS

(CACTACEAE)
A few specialized terms used by botanists to describe cactaceous structures are defined
below. Some other botanical terms are explained where they are first used and, therefore,
are not repeated here. For the terminology used to describe vegetation types, see
Chapter 3.2.
areoles are the felted cushions (actually highly telescoped short shoots) found on various
parts of the cactus plant, bearing the spines (= modified leaves), trichomatous hairs etc.
Their meristems, when active, can give rise to new shoots and flower-buds.
aril (funicular envelope) in the sense used here is the bony and sometimes fibrous covering
of the true seed developed in Opuntioideae, being a highly specialized development
of the funicle and hence correctly described as an aril, although it is very different
from so-called structures (or strophioles) in other parts of the Cactus Family (Barthlott
& Hunt 2000: 24).
brachyblast leaves (Pereskioideae) are those produced by the areoles (see above), as opposed
to those subtending areoles on the main extension axes (long shoots).
bract-scales are the minute to conspicuous scalelike appendages seen on the pericarpel and
flower-tube below (ie. anterior to) the perianth-segments and often intergrading with
the latter. They may be naked in their axils (eg. most Cereeae) or subtend areoles
bearing trichomes, spines etc.
cephalium is a ± modified part of the stem, whether apical or lateral, whence the flowers
and fruits are borne (cf. Barthlott & Hunt 1993: 164). It is in effect a kind of
inflorescence structure in which the areoles may be enlarged or reduced relative to
those in the vegetative part, and often compressed together, bearing abundant
trichomes and/or dense spines/bristles, distinguishing the fertile part of the stem from
the purely vegetative. It may be either a chlorophyllous or non-chlorophyllous part
of the stem, whose cross-section in the case of lateral cephalia (plural) may remain
normal (terete) or deformed, as in the so-called ‘sunken’ cephalia.
cereoid cacti are those with usually erect, leafless, ± cylindrical stems, as in Cereus, which
can be contrasted with the globular to depressed stems of, eg., Melocactus or Discocactus.
epilithic is used in this book in contrast to ‘epiphytic’ to describe cacti found growing upon
rocks or rock surfaces when the plants in question are representative of taxa whose
life-form is commonly that of an epiphyte.
GLOSSARY
flower-tube is the hollow or partially hollow structure above the pericarpel which
comprises fused floral and receptacular tissues; the latter on the exterior, often bearing
bract-scales; the former within and subtending the perianth-segments at its apex.
funicular pulp is the term used for the either solid or semi-liquid pulp found within the
pericarp of cactus fruits, surrounding the seeds, which is derived from the ovule
funicles.
glochids are normally short, strongly barbed, easily detached, specialized spines produced
by the areoles of Opuntioideae and distinct from the generally much larger normal
spines, which are also barbed in this subfamily.
hair-spines are specialized true cactus spines (ie. multicellular structures representing
modified leaves) with a soft and often woolly, hairlike quality, often found on flowers
of Trichocereeae. True areolar hairs are normally single-celled trichomes, short and
felt-like, but sometimes longer and woolly (see long hairs).
hypertrophic spines are those developed in an exaggerated manner from areoles of
indeterminate growth at the base of stems in certain rock-dwelling cacti, especially
Coleocephalocereus and Micranthocereus spp.
long hairs are fine, elongate, woolly hairs produced by the areoles of a variety of Cereeae
species and distinguishable from the short felt-like areolar trichomes and thicker hair-
spines (see above)
pericarp is the fruit wall formed by the fusion of stem (receptacular) and floral tissues (see
below). The visible exterior is the stem component and may bear bract-scales, areoles,
spines etc., or be almost or quite naked.
pericarpel is the structure comprising the lower part of the specialized stem or receptacle
into which the ovary of the inverted cactus flower is sunken.
podaria (sing. podarium) are the swellings often subtending areoles that represent the points
of attachment of leaves or bracts that have been lost, or almost lost, in the course of
evolution of the highly succulent habit.
stem-segments refers to stems where the seasonal growth is begun and terminated at a
constriction, or marked by the development of an entirely new growth segment. This
is best seen in Rhipsalideae (Rhipsalis, Hatiora, Schlumbergera), but also occurs, for
example, in the cereoid taxa bearing terminal cephalia, where the subsequent vegetative
phase grows ‘through’ the cephalium (Stephanocereus leucostele, Arrojadoa spp.).
GLOSSÁRIO
GLOSSÁRIO DE TERMOS BOTÂNICOS (CACTACEAE)
Alguns termos especializados são atualmente utilizados por botânicos para descrever
estruturas típicas de Cactaceae. A seguir, são apresentadas as traduções para a língua
portuguesa e respectivas interpretações.
areoles — aréolas são estruturas parecidas com pequenas almofadas felpudas, consistindo em
ramos extremamente encurtados encontrados em várias partes das plantas, dando
origem aos espinhos, tricomas, cerdas, e, no caso de apresentarem meristemas ativos
no seu interior, podem originar novos ramos e botões florais.
aril (funicular envelope) — arilo (envelope funicular) no sentido aplicado aqui trata-se da
cobertura óssea e às vezes fibrosa cobrindo a semente em Opuntioideae, tratando-se
de uma estrutura desenvolvida a partir de uma modificação extrema do funículo e
portanto corretamente descrita como um arilo, apesar de mostrar-se bastante diferente
das estruturas chamadas arilos (ou estrofíolos) em outros membros da família Cactaceae
(Barthlott & Hunt 2000: 4)
brachyblast leaves — folhas do braquiblasto (Pereskioideae) são aquelas produzidas pelas
aréolas (ver acima), contrastando com as folhas que subtendem aréolas nos eixos
principais de extensão (ramos longos ou ‘long shoots’).
bract-scales — brácteas escamiformes são apêndices diminutos até bem visíveis encontrados
no pericarpelo e no tubo-floral abaixo dos segmentos do perianto e muitas vezes
intergradando-se com os mesmos. Essas brácteas podem apresentar axilas nuas
(maioria dos membros da tribo Cereeae) ou subtender aréolas apresentando tricomas,
espinhos, etc.
cephalium — cefálio é uma parte mais ou menos modificada do ramo, podendo ser apical
ou lateral, a partir da qual surgem flores e frutos (Barthlott & Hunt 1993: 164). Trata-
se de uma estrutura do tipo inflorescência, na qual as aréolas podem apresentar-se
maiores ou reduzidas com respeito àquelas da parte vegetativa, e muitas vezes
comprimidas, comportando tricomas ou espinhos e cerdas abundantes, distinguindo a
parte fértil daquela puramente vegetativa. Pode tratar-se de uma parte tanto
fotossintetizante como não fotossintetizante do ramo, sendo que uma secção
transversal de um dito cefálio lateral pode permanecer cilíndrico ou apresentar-se
deformado, no caso de um cefálio definido como ‘aprofundado’.
cereoid — cereóide é o adjetivo utilizado para caracterizar aqueles cactos com ramos
normalmente eretos, sem folhas, mais ou menos cilíndricos, como no caso de Cereus,
contrastando com aqueles que apresentam ramos globosos ou depressos, como por
exemplo Melocactus ou Discocactus.
epilithic — epilítico, utilizado neste livro em contraste com ‘epifítico’ ou ‘epífita’, é relativo
a cactos encontrados crescendo sobre rochas ou superfícies rochosas, quando o tipo de
planta em questão é mais comumente encontrado crescendo como epífita.
GLOSSÁRIO
flower-tube — tubo-floral trata-se da estrutura oca ou partialmente oca acima do

pericarpelo, constituída de tecidos florais e receptaculares adnados; externamente
apresentando brácteas-escamiformes e internamente subtendendo segmentos do
perianto na porção apical.
funicular pulp — polpa funicular trata-se da polpa sólida ou líquida encontrada no interior
do pericarpo de frutos de Cactaceae, e que envolve as sementes, sendo derivada do
funículo dos óvulos.
glochids — gloquídeos são normalmente espinhos curtos, com os lados fortemente serreados
e que se destacam facilmente, presentes nas aréolas das espécies da subfamília
Opuntioideae e distintos dos espinhos normais, que são muito maiores, mas também
serreados na mesma subfamília.
hair-spines — tricomas-capiliformes são espinhos verdadeiros especializados, ou seja,
estruturas multicelulares representando folhas modificadas, com textura macia e muitas
vezes lanosa, assemelhando-se a fios de cabelo, encontrados muitas vezes em flores de
membros da tribo Trichocereeae. Tricomas areolares são normalmente unicelulares,
curtos e com aspecto aveludado, mas podem às vezes apresentar-se mais longos e
lanosos (ver ‘tricomas lanosos’ – ‘long hairs’).
hypertrophic spines — espinhos hipertróficos são aqueles desenvolvidos de maneira exagerada
a partir de aréolas de crescimento indeterminado presentes na base dos ramos em
certos cactos que crescem sobre rochas, especialmente Coleocephalocereus e
Micranthocereus spp.
long hairs — tricomas lanosos são tricomas finos, longos, lanosos, produzidos pelas aréolas de
muitas espécies de Cereeae, distintos dos tricomas curtos das aréolas e dos ‘tricomas
capiliformes’ – ‘hair-spines’ (ver acima).
pericarp — pericarpo é a parede do fruto formada pela fusão do ramo (receptáculo) e dos
tecidos florais (ver ‘pericarpel’ — ‘pericarpelo’ abaixo). A parte visível é o exterior do
ramo e pode apresentar brácteas-escamiformes, aréolas, espinhos, costelas, etc., ou ser
quase ou completamente liso.
pericarpel — pericarpelo trata-se da estrutura composta pela parte inferior do ramo
especializado ou receptáculo no interior do qual o ovário invertido da flor de
Cactaceae encontra-se imerso.
podaria (sing. podarium) — podários trata-se dos espessamentos encontrados subtendendo
aréolas e que representam o ponto de conexão de folhas ou brácteas que sofreram
redução ou perda no curso da evolução de ramos com hábito extremamente suculento
e fotossintetizante.
stem-segment — segmento do ramo refere-se aos ramos onde o crescimento estacional inicia-
se e finaliza em uma constricção, ou é marcado pelo desenvolvimento do crescimento
de um segmento inteiramente novo. Este fenômeno pode ser observado na tribo
Rhipsalideae, mas também ocorre em cactos cereóides dotados de cefálio terminal
(Stephanocereus leucostele, Arrojadoa spp.).
INDEX TO BOTANICAL NAMES & EPITHETS
7.2 INDEX TO BOTANICAL NAMES AND EPITHETS

Accepted names (or their epithets) are indicated in bold type, synonyms, inadequately
known taxa, orthographic variants and invalid names by italics. Synonyms based on types
collected outside the Eastern Brazil area, in adjacent regions of Brazil, are also included.
Other names of taxa from outside the region or those that are only doubtfully
represented, which are mentioned in discussion or cited as the types of generic names, are
not usually included, unless given in plain Roman type. Autonyms representing accepted
names of included taxa are also listed. The accepted names of introduced or widely
cultivated taxa are followed by an asterisk (*). Authority citations are omitted except for
misapplied names and homonyms (the citation, if given in abbreviated form being of the
author who misapplied the name or published the homonym in question). Entries are
referenced by the number of the accepted generic name followed, after a colon, by the
species number and, if applicable, the letter(s) indicating the accepted infraspecific
classification. Binomials referenced with ‘see’ to the generic number only are discussed
under the given genus heading in Chapter 5; similarly, those followed by a genus:species
number. Hybrid taxa are indicated by means of a multiplication sign ( ) connecting the
species numbers, eg. Tacinga quipa (T. palmadora T. inamoena) is indicated as
3:53:6. A few names listed here do not appear as formally cited synonyms in chapter 5,
especially some archaic and ‘nude’ (description-less) names in Lepismium and Uebelmannia.
Acanthocereus albicaulis 16:2 ‘marylanae’ (nom. prov.) 19:4b

brasiliensis 10:1 multiflora 19:2
ARROJADOA 19 subsp. hofackeriana 19:2b
albiflora 19:219:4 penicillata 19:3
aureispina 19:4 var. decumbens 19:3
var. anguinea 19:4 var. spinosior 19:3
var. guanambensis 19:4 polyantha 21:6
bahiensis 19:1 rhodantha 19:4
beateae 19:2a subsp. anguinea 19:4
canudosensis 19:4 subsp. aureispina 19:4
cremnophila 19:1 subsp. canudosensis 19:4
dinae 19:2 subsp. guanambensis 19:4
subsp. dinae 19:2a subsp. reflexa 19:4
subsp. eriocaulis 19:2b var. anguinea 19:4
subsp. nana 19:2a var. guanambensis 19:4
var. nana 19:2a var. occibahiensis 19:4
eriocaulis 19:2b var. theunisseniana 19:4
subsp. albicoronata 19:2b theunisseniana 19:4
var. albicoronata 19:2b ARTHROCEREUS 28
var. rosenbergeriana 19:2b campos-portoi 28:3
heimenii 19:2a damazioi P.V. Heath 28:3
horstiana 19:4 glaziovii 28:3
INDEX
itabiriticola 28:3 cruciformis 11:3

melanurus 28:1 dillenii 6:2
subsp. estevesii 28:1a ficus-indica 6:3
subsp. magnus 28:1b grandifolius 1:2
subsp. melanurus 28:1a heptagonus Vell. non L. 20:5
subsp. mello-barretoi 28:1a heterocladus 4:1
subsp. odorus 28:1c hexagonus Vell. non L. 20:5
var. estevesii 28:1a melocactoides 23:16?
mello-barretoi 28:1a melocactus Vell. non L. 22:2
microsphaericus 14:2 opuntia 6:3
microsphaericus sensu A. Berger 28:3 oreas 23:1
odorus 28:1c pereskia 1:1
rondonianus 28:2 phyllanthus 9:1
spinosissimus see 28 placentiformis 29:4
Austrocephalocereus albicephalus 21:2 teres 12:11
dolichospermaticus 21:8 triangularis Vell. non L. 7:1
dybowskii 27:1 urumbeba 6:1
fluminensis 22:2 zehntneri 23:11
lehmannianus 21:3 Cassyta baccifera 12:12
purpureus 21:3 Cephalocereus arrabidae 20:5
purpureus Backeb. 22:4 bradei 17:3
salvadorensis 20:3b brasiliensis 20:6
BRASILICEREUS 15 catingicola 20:3
breviflorus 15:1 chrysostele 20:19
markgrafii 15:2 cuyabensis 20:14
phaeacanthus 15:1 dybowskii 27:1
subsp. breviflorus 15:1 exerens sensu (K. Schum.) Rose 20:5
BRASILIOPUNTIA 4 fluminensis 22:2
bahiensis 4:1 glaucescens see below 20:20
brasiliensis 4:1 glaucochrous 20:9
subsp. bahiensis 4:1 gounellei 20:2
subsp. subacarpa 4:1 hapalacanthus 20:3b
subacarpa 4:1 lehmannianus 21:3
Buiningia aurea 22:5 leucostele 18:1
brevicylndrica 22:5 luetzelburgii 18:2
var. elongata 22:5 machrisii 20:14
var. longispina 22:5 melocactus 22:2
purpurea 22:6 minensis 17:4
Cactus alteolens 29:4? penicillatus 19:3
arboreus 4:1 pentaedrophorus 20:8
bahiensis 23:3 phaeacanthus 15:1
brasiliensis 4:1 piauhyensis 20:18
cochenillifera 5:1 piauhyensis Britton & Rose 20:12b
polyanthus 21:6 insularis 16:4

purpureus 21:3 jamacaru 16:5
purpureus Britton & Rose 22:4 subsp. calcirupicola 16:5b
rhodanthus 19:4 subsp. goiasensis 16:5a
rupicola 20:3b subsp. jamacaru 16:5a
salvadorensis 20:3b jamacaru sensu Scheinvar 1985 16:6a
sergipensis 20:3b leucostele 18:1
ulei Gürke 26:1 luetzelburgii 18:2
zehntneri 20:2b macrogonus sensu K. Schum. 1890 20:5
Cereastreae pentaedrophorus 20:8 macrogonus sensu Warming 1892 20:10
CEREUS 16 markgrafii 15:2
adscendens 24:1 melanurus 28:1
alacriportanus 16:6? melocactus 22:2
albicaulis 16:2 microsphaericus 14:2
bahiensis 25:1 milesimus 16:6a
brasiliensis 10:1 minensis 17:4
calcirupicola 16:5b mirabella 16:1
subsp. cabralensis 16:5b myosurus 11:3
subsp. cipoensis 16:5b neonesioticus var. interior 16:6a
var. albicans 16:5b neotetragonus 16:3?
var. cabralensis 16:5b obtusangulus 14:2
var. cipoensis 16:5b obtusus 16:3?
var. pluricostatus 16:5b parvisetus 28:1c?
catingicola 20:3 parvulus 14:2?
chrysostele 20:19 penicillatus 19:3
coerulescens sensu Warming 16:5b pentaedrophorus 20:8
crassisepalus 17:2 pernambucensis 16:3
cruciformis 11:3 peruvianus auctt. p.p. 16:6a
damazioi 14:2 phaeacanthus 15:1
damazioi auctt. non Weingart 28:3 var. breviflorus 15:1
dybowskii 27:1 piauhyensis 20:18
extensus see 7:1 pitajaya Hook. non (Jacq.) DC 16:3a
fernambucensis 16:3 rhodanthus 19:4
subsp. fernambucensis 16:3a ridleii 16:4?
subsp. sericifer 16:3b sericifer 16:3b
fluminensis 22:2 setaceus 7:1
glaziovii 28:3 setosus 11:3
goebelianus 22:4 setosus Gürke non Loddiges 20:2a
goiasensis 16:5a squamosus 26:3
hexagonus see 16.5 squamulosus 11:3
hildmannianus 16:6 sublanatus 20:6?
subsp. hildmannianus 16:6a tenuis 11:3
subsp. uruguayanus see 16:6a tenuispinus 11:3
triangularis K. Schum. p.p. 7:1 minensis 17:4

trigonodendron see 16:5c pachystele 22:4
ulei sensu Luetzelb. 20:12a paulensis 22:2a
undatus 7:2 pluricostatus 22:3
undatus sensu Luetzelb. 7:1 subsp. uebelmanniorum 22:3
variabilis auctt. non Pfeiff. 16:3 pleurocarpus 17:4b
vellozoi 22:2 purpureus 22:6
warmingii 20:5 DISCOCACTUS 29
CIPOCEREUS 17 albispinus 29:1a
bradei 17:3 alteolens 29:4?
crassisepalus 17:2 araneispinus 29:1b
laniflorus 17:1 bahiensis 29:2
minensis 17:4 subsp. gracilis 29:2
subsp. leiocarpus 17:4a subsp. subviridigriseus 29:2
subsp. minensis 17:4b besleri 29:4
subsp. pleurocarpus 17:4b boomianus 29:1b
pleurocarpus 17:4b buenekeri 29:1b?
pusilliflorus 17:5 catingicola 29:3a
COLEOCEPHALOCEREUS 22 subsp. griseus 29:3
albicephalus 21:2 subsp. rapirhizus 29:3
aureispinus 20:16 var. griseus 29:3
aureus 22:5 var. nigrisaetosus 29:3a
braunii 22:1a crystallophilus 29:4
brevicylindricus 22:5 griseus 29:3
var. elongatus 22:5 heptacanthus 29:3
var. longispinus 22:5 subsp. catingicola 29:3a
buxbaumianus 22:1 horstii 29:6
subsp. buxbaumianus 22:1a insignis 29:4
subsp. flavisetus 22:1b insignis sensu Buining 29:5
decumbens 22:2b latispinus 29:4
diersianus 22:2a subsp. pseudolatispinus 29:4
dybowskii 27:1 subsp. pulvinicapitatus 29:4
estevesii 22:1b lehmannii 29:4
flavisetus 22:1b linkii 29:4
fluminensis 22:2 multicolorispinus 29:4
subsp. braamhaarii 22:2a nigrisaetosus 29:3a
subsp. decumbens 22:2b piauiensis 29:3a
subsp. fluminensis 22:2a placentiformis 29:4
subsp. paulensis 22:2a subsp. alteolens 29:4?
var. braamhaarii 22:2a subsp. multicolorispinus 29:4
goebelianus 22:4 subsp. pugionacanthus 29:4
lehmannianus 21:3 var. alteolens 29:4?
luetzelburgii 18:2 var. pugionacanthus 29:4
placentiformis sensu Britton & Erythrorhipsalis cereuscula 12:17

Rose 29:2 pilocarpa 12:18
pseudoinsignis 29:5 Espostoa dybowskii 27:1
pseudolatispinus 29:4 ulei 26:1
pugionacanthus 29:4 ESPOSTOOPSIS 27
pulvinicapitatus 29:4 dybowskii 27:1
rapirhizus 29:3 FACHEIROA 26
spinosior 29:3a cephaliomelana 26:2
subnudus see below 29:6 subsp. cephaliomelana 26:2a
subviridigriseus 29:2 subsp. estevesii 26:2b
tricornis 29:4 chaetacantha 26:3
woutersianus see 29:6 var. montealtoi 26:3
zehntneri 29:1 deinacanthus 26:3
f. albispinus 29:1a estevesii 26:2b
subsp. albispinus 29:1a pilosa 26:2a
subsp. araneispinus 29:1b publiflora 26:1
subsp. boomianus 29:1b pubiflora 26:1
subsp. buenekeri 29:1b? squamosa 26:3
subsp. horstiorum 29:1b tenebrosa 26:2a
subsp. zehntneri 29:1a ulei 26:1
var. albispinus 29:1a Floribunda bahiensis 19:1
var. araneispinus 29:1b pusilliflora 17:5
var. boomianus 29:1b Gerocephalus dybowskii 27:1
var. horstiorum 29:1b Gummocactus F. Ritter (nom. nud.) 30
Echinocactus alteolens 29:4? Hariota alternata 12:6
bahiensis 29:2 cereuscula 12:17
centeterius see below 30:3c clavata 12:16
placentiformis 29:4 cribrata see 12:14
subnudus see below 29:6 cruciformis 11:3
zehntneri 29:1 knightii 11:3
Epiphyllanthus candidus 14:2 paradoxa 12:6
microsphaericus 14:2 prismatica 12:11
obovatus 14:3 ramosissima 11:3?
obtusangulus 14:2 saglionis 12:17
opuntioides 14:3 salicornioides 13:1
EPIPHYLLUM 9 squamulosa 11:3
crispatum 12:4 villigera 13:1
obovatum 14:3 HARRISIA 24
obtusangulum 14:2 balansae see 24
opuntioides 14:3 adscendens 24:1
oxypetalum* see below 9:1 HATIORA 13
phyllanthus 9:1 clavata 12:16
Eriocereus adscendens 24:1 cylindrica 13:2
herminiae see below 13:2 vollii 11:3

salicornioides 13:1 warmingianum 11:2
f. cylindrica 13:2 Malacocarpus heptacanthus 29:3
f. villigera 13:1 Mediocactus coccineus auctt. 7:1
HYLOCEREUS 7 setaceus 7:1
extensus see 7:1 MELOCACTUS 23
setaceus 7:1 acispinosus 23:3a(ii)
undatus* 7:2 amethystinus 23:3b
LEOCEREUS 25 ammotrophus 23:3b
bahiensis 25:1 arcuatispinus 23:11
subsp. barreirensis 25:1 axiniphorus 23:14
subsp. exiguospinus 25:1 azulensis 23:2a
subsp. robustispinus 25:1 azureus 23:7
subsp. urandianus 25:1 subsp. ferreophilus 23:8
var. barreirensis 25:1 var. krainzianus 23:7
var. exiguospinus 25:1 bahiensis 23:3
var. robustispinus 25:1 f. acispinosus 23:3a(ii)
var. urandianus 25:1 f. bahiensis 23:3a(i)
estevesii 25:1 f. inconcinnus 23:3a(iii)
melanurus 28:1 subsp. amethystinus 23:3b
paulensis 22:2a? subsp. bahiensis 23:3a
squamosus 26:3 bahiensis sensu Werderm. 23:11
urandianus 25:1 besleri 29:4
LEPISMIUM 11 brederooianus 23:3a(ii)?
anceps 11:3 canescens 23:11
cavernosum 11:3 var. montealtoi 23:11
cereoides 12:8 concinnus 23:14
commune 11:3 subsp. axiniphorus 23:14
cruciforme 11:3 conoideus 23:4
floccosum 12:5 conoideus sensu Rizzini 23:1a
gibberulum 12:5c cremnophilus 23:1b
houlletianum 11:1 curvicornis 23:11
f. regnellii 11:1 deinacanthus 23:5
knightii 11:3 f. mulequensis 23:2b
lineare 11:2 subsp. florschuetzianus 23:2b
myosurus 11:3 subsp. longicarpus 23:2b
pacheco-leonii 12:7 depressus 23:16a
pacheco-leonis 12:7 diersianus 23:6
paradoxum 12:6 f. rubrispinus 23:6
pulvinigerum 12:5c douradaensis 23:11
radicans 11:3 ellemeetii 23:16c
ramosissimum 11:3? ernestii 23:2
tenue 11:3 subsp. ernestii 23:2a
subsp. longicarpus 23:2b subsp. cremnophilus 23:1b

erythracanthus 23:2a subsp. ernestii 23:2a
ferreophilus 23:8 subsp. oreas 23:1a
florschuetzianus 23:2b subsp. rubrisaetosus 23:1a
giganteus 23:11 var. bahiensis 23:3a
glaucescens 23:13 var. bahiensis sensu Rizzini 23:1a
glauxianus 23:3b var. longispinus 23:2a
griseoloviridis 23:3b var. rubrisaetosus 23:1a
grisoleoviridis 23:3b var. submunitis 23:1a
helvolilanatus 23:11 oreas auctt. non Miq. 23:2a
inconcinnus 23:3? pachyacanthus 23:9
var. brederooianus 23:3a(ii)? subsp. pachyacanthus 23:9a
inconcinnus sensu F. Ritter 23:10 subsp. viridis 23:9b
interpositus 23:2a paucispinus 23:15
krainzianus 23:7 pentacentrus 23:16c?
lanssensianus 23:12 pruinosus 23:14?
lensselinkianus 23:3b var. concinnus 23:14
levitestatus 23:6 pruinosus sensu P.J. Braun 23:13
f. securituberculatus 23:6 robustispinus 23:14
longicarpus 23:2b rubrisaetosus 23:1a
longispinus 23:2a rubrispinus 23:6
macrodiscus 23:11 salvadorensis 23:10
var. minor 23:16b salvadorensis sensu Rizzini 23:11
macrodiscus sensu Rizzini 23:10 saxicola 23:11
macrodiscus var. macrodiscus securituberculatus 23:6
sensu F. Ritter 23:14 uebelmannii 23:6
margaritaceus 23:16c violaceus 23:16
var. disciformis 23:16c subsp. margaritaceus 23:16c
var. salvadoranus 23:16c subsp. natalensis 23:16a
melocactoides 23:16? subsp. ritteri 23:16b
f. depressus 23:16a subsp. violaceus 23:16a
var. depressus 23:16a warasii 23:6
var. natalensis 23:16a zehntneri 23:11
var. violaceus 23:16a subsp. arcuatispinus 23:11
montanus 23:2b subsp. canescens 23:11
mulequensis 23:2b subsp. robustispinus 23:14
neomontanus 23:2b var. ananas 23:11
nitidus 23:2a var. curvicornis 23:11
oreas 23:1 var. viridis 23:11
f. azulensis 23:2a MICRANTHOCEREUS 21
f. erythracanthus 23:2a albicephalus 21:2
subsp. bahiensis 23:3a aureispinus 21:2
subsp. bahiensis sensu Rizzini 23:1a auriazureus 21:4
auri-azureus 21:4 monacantha 6:1

densiflorus 21:7 opuntia 6:3
dolichospermaticus 21:8 palmadora 3:4
dybowskii 27:1 subsp. catingicola see below 3:6
estevesii see 21 & 21:8 rubescens sensu K. Schum. 3:2
flaviflorus 21:7 saxatilis 3:5
subsp. densiflorus 21:7 subsp. minutispina 3:5a
var. uilianus 21:7 subsp. occibahiensis 3:5a
haematocarpus 21:3 subsp. pomosa 3:5a
lehmannianus 21:3 var. minutispina 3:5a
monteazulensis 21:2 var. occibahiensis 3:5a
polyanthus 21:6 var. pomosa 3:5a
purpureus 21:3 stricta var. dillenii 6:2
ruficeps 21:3 umbrella 6:1
streckeri 21:5 urumbeba 6:1
uilianus 21:7 vulgaris 6:3
violaciflorus 21:1 vulgaris auctt. p.p. non Mill. 6:1
Mirabella albicaulis 16:2 werneri 3:3
minensis 16:1 Parodia gummifera 30:2
Monvillea albicaulis 16:2 PERESKIA 1
minensis 16:1 aculeata 1:1
NOPALEA* 5 aureiflora 1:5
cochenillifera* 5:1 bahiensis 1:3
OPUNTIA 6 grandifolia 1:2
arborea 4:1 subsp. grandifolia 1:2a
arechavaletae 6:1 subsp. violacea 1:2b
arechavaletai 6:1 var. violacea 1:2b
arechaveletai 6:1 pereskia 1:1
bahiensis 4:1 stenantha 1:4
brasiliensis 4:1 zehntneri 2:1
subsp. bahiensis 4:1 Pierrebraunia bahiensis 19:1
subsp. subacarpa 4:1 brauniorum see below 20:20
catingicola see below 3:6 Pilocereus arenicola 20:3a
var. fulviceps see below 3:6 arrabidae 20:5
cochenillifera 5:1 aurisetus 20:15
dillenii* 6:2 bradei 17:3
var. reitzii 6:2 brasiliensis 20:6
estevesii 3:5b catingicola 20:3
ficus-indica* 6:3 chrysostele 20:19
fusicaulis 6:3? cuyabensis 20:14
inamoena 3:6 exerens 20:5
f. spinigera see below 3:3 floccosus Backeb. & Voll 20:10
var. flaviflora 3:6 glaucescens A. Linke see below 20:20
glaucescens sensu Werderm. 20:12a subsp. catingicola 20:3a

glaucochrous 20:9 subsp. hapalacanthus 20:3b
gounellei 20:2 subsp. robustus 20:3a
var. zehntneri 20:2b subsp. salvadorensis 20:3b
leucostele 18:1 cenepequei 20:12a
luetzelburgii 18:2 chrysostele 20:19
melocactus 22:2 subsp. cearensis 20:18?
minensis 17:4 circinnuspetalus 20:14
pentaedrophorus 20:8 coerulescens see below 20:20
piauhyensis 20:18 coerulescens sensu (Lem.)
piauhyensis sensu Werderm. 20:12b F. Ritter 20:15a
polyedrophorus 20:8 cristalinensis 20:14
phaeacanthus 15:1 cuyabensis 20:14
rupicola 20:3b cyaneus 20:12a
sergipensis 20:3b densiareolatus 20:20
setosus 20:2a subsp. brunneolanatus 20:20
sublanatus 20:6? densivillosus 20:14
tuberculatus 20:1 estevesii 20:14?
virens sensu Ule 20:5 flavipulvinatus 20:7
PILOSOCEREUS 20 subsp. carolinensis 20:7
arenicola 20:3a var. carolinensis 20:7
arrabidae 20:5 floccosus 20:10
atroflavispinus 20:12a subsp. floccosus 20:10a
aurilanatus 20:15b subsp. quadricostatus 20:10b
aurisetus 20:15 fulvilanatus 20:11
subsp. aurilanatus 20:15b subsp. fulvilanatus 20:11a
subsp. aurisetus 20:15a subsp. rosae 20:11b
subsp. densilanatus 20:15a gaturianensis 20:18
subsp. supthutianus 20:15a glaucescens see below 20:20
subsp. werdermannianus 20:15a glaucescens Rizzini & A. Mattos
aureispinus 20:16 1992 20:13
azulensis 20:4 glaucochrous 20:9
azureus 20:12a goianus 20:14
bohlei see below 20:20 gounellei 20:2
bradei 17:3 subsp. gounellei 20:2a
brasiliensis 20:6 subsp. zehntneri 20:2b
subsp. brasiliensis 20:6a var. zehntneri 20:2b
subsp. ruschianus 20:6b hapalacanthus 20:3b
braunii 20:2b jauruensis 20:14
carolinensis 20:7 juaruensis 20:14
var. robustispinus 20:7 laniflorus 17:1
catingicola 20:3 leucocephalus see 20
subsp. arenicola 20:3a lindaianus 20:14
lindanus 20:14 subsimilis 20:10b20:13

luetzelburgii 18:2 superfloccosus 20:2b
machrisii 20:14 superfloccosus auctt. 20:20
subsp. cristalinensis 20:14 var. brunneolanatus 20:20
magnificus 20:13 supthutianus 20:15a
minensis 17:4 tuberculatus 20:1
mucosiflorus 20:18 werdermannianus 20:15a
multicostatus 20:17 var. densilanatus 20:15a
occultiflorus see 20:20 (20:2020:12a?) var. diamantinensis 20:15a
oreus 20:12a zehntneri 20:2b
pachycladus 20:12 Piptanthocereus alacriportanus 16:6?
subsp. pachycladus 20:12a cabralensis 16:5b
subsp. pernambucoensis 20:12b calcirupicola 16:5b
parvus see below 20:20 var. pluricostatus 16:5b
pentaedrophorus 20:8 cipoensis 16:5b
subsp. pentaedrophorus 20:8a crassisepalus 17:2
subsp. robustus 20:8b goiasensis 16:5a
pernambucoensis 20:12b neonesioticus var. interior 16:6
var. caesius 20:12b obtusus 16:3?
var. montealtoi 20:12a sericifer 16:3b
piauhyensis 20:18 Platyopuntia inamoena 3:6
subsp. gaturianensis 20:18 f. spinigera see below 3:3
subsp. mucosiflorus 20:18 saxatilis 3:5
pleurocarpus 17:4b PSEUDOACANTHOCEREUS 10
pusillibaccatus 20:14 boreominarum 10:1
pusilliflorus 17:5 brasiliensis 10:1
quadricostatus 20:10b f. boreominarum 10:1
robustus 20:3a Pseudopilocereus arrabidae 20:5
rosae 20:11b atroflavispinus 20:12a
rupicola 20:3b aurilanatus 20:15b
ruschianus 20:6b aurisetus 20:15
salvadorensis 20:3b azureus 20:12a
saudadensis 20:14 azureus P. V. Heath 20:12a
saxatilis 20:15a bradei 17:3
var. densilanatus 20:15a carolinensis 20:7
schoebelii 20:12a var. robustispinus 20:7
sergipensis 20:3b catingicola 20:3
splendidus 20:12a? chrysostele 20:19
suberbus 20:12a cuyabensis 20:14
var. gacapaensis 20:12a densiareolatus 20:20
var. lanosior 20:12a flavipulvinatus 20:7
var. regius 20:12a floccosus 20:10
sublanatus 20:6? fulvilanatus 20:11
gaturianensis 20:18 alboareolata 12:11

glaucescens see below 20:20 alternata 12:6
glaucochrous 20:9 anceps 11:3
gounellei 20:2 baccifera 12:12
hapalacanthus 20:3b subsp. baccifera 12:12a
jauruensis 20:14 subsp. hileiabaiana 12:12b
juaruensis 20:14 brevibarbis 11:3
luetzelburgii 18:2 burchellii 12:14
machrisii 20:14 capilliformis 12:11a
magnificus 20:13 campos-portoana see 12:14
mucosiflorus 20:18 cavernosa 11:3
multicostatus 20:17 cassutha 12:12a
oreus 20:12a cassuthopsis 12:12a
pachycladus 20:12 cassythoides 12:12a
parvus see below 20:20 cassythoydes 12:12a
pentaedrophorus 20:8 cereoides 12:8
pernambucensis 20:12b cereuscula 12:17
pernambucoensis 20:12b chloroptera 12:2
var. caesius 20:12b clavata 12:16
var. montealtoi 20:12a clavata Luetzelb. see 12
piauhyensis 20:17 clavellina 12:11
polyedrophorus 20:8 cribrata (Lem.) N.E. Brown see 12:14
pusillibaccatus 20:14 cribrata sensu K. Schum. 12:17
quadricostatus 20:10b crispata 12:4
robustus 20:3a crispimarginata 12:3
rupicola 20:3b cruciformis 11:3
ruschianus 20:6b cylindrica Vaupel, illeg. 13:2
salvadorensis 20:3b densiareolata 12:10
sergipensis 20:3b elliptica 12:2
splendidus 20:12a? floccosa 12:5
superbus 20:12a subsp. floccosa 12:5a
var. gacapaensis 20:12a subsp. oreophila 12:5b
var. lanosior 20:12a subsp. pulvinigera 12:5c
var. regius 20:12a gibberula 12:5c
superfloccosus 20:2b gonocarpa 11:2
tuberculatus 20:1 gracilis 12:11
werdermannianus 20:15a heteroclada 12:11
var. densilanatus 20:15a hoelleri 12:19
var. diamantinensis 20:15a houlletiana 11:1
zehntneri 20:2b var. regnellii 11:1
QUIABENTIA 2 houlletii 11:1
zehntneri 2:1 juengeri 12:15
RHIPSALIS 12 knightii 11:3
lindbergiana 12:10 warmingiana 11:2

linearis 11:2 Rhodocactus grandifolius 1:2
macahensis 12:13 SCHLUMBERGERA 14
macropogon 11:3 kautskyi 14:1
maricaensis 12:11 microsphaerica 14:2
mittleri 11:3 f. obtusangula 14:2
monteazulensis 12:5b f. parvula 14:2?
myosurus 11:3 subsp. candida 14:2
neves-armondii see above 12:19 obtusangula 14:2
oblonga 12:3 opuntioides 14:3
pacheco-leoni 12:7 truncata subsp. kautskyi 14:1
pacheco-leonis 12:7 var. kautskyi 14:1
subsp. catenulata 12:7a SELENICEREUS* 8
paradoxa 12:6 anthonyanus* see 8
subsp. septentrionalis 12:6a extensus see 7:1
var. catenulata 12:7a rizzinii 7:1
penduliflora 12:17 setaceus 7:1
penduliflora sensu K. Schum. 12:11 Siccobaccatus dolichospermaticus 21:8
pilocarpa 12:18 STEPHANOCEREUS 18
platycarpa Luetzelb. see 12 leucostele 18:1
prismatica 12:11 luetzelburgii 18:2
pulchra 12:13 TACINGA 3
pulvinigera 12:5c atropurpurea 3:1
radicans 11:3 var. zehntnerioides 3:1
ramosissima 11:3? braunii 3:2
regnellii 11:1 funalis 3:1
rhombea sensu Loefgren 12:4 subsp. atropurpurea 3:1
robusta Luetzelb. see 12 var. atropurpurea 3:1
russellii 12:1 inamoena 3:6
saglionis 12:17 luetzelburgii 3:1
salicornioides 13:1 palmadora 3:4
var. cylindrica 13:2 quipa 3:53:6
var. villigera 13:1 saxatilis 3:5
salicornoides 13:1 subsp. estevesii 3:5b
squamulosa 11:3 subsp. saxatilis 3:5a
sulcata 12:9 werneri 3:3
teres 12:11 zehntneri 3:1
f. capilliformis 12:11a Trichocereus campos-portoi 28:3
f. heteroclada 12:11 damazioi 14:2
f. prismatica 12:11 damazioi sensu Werderm. 28:3
tetragona 12:11 glaziovii 28:3
virgata 12:11 UEBELMANNIA 30
vollii 11:3 ammotrophus (nom. nud.) 30:3a
INDEX TO VERNACULAR NAMES
antonensis (nom. nud.) 30:2 subsp. flavispina 30:3b

buiningii 30:1 subsp. horrida 30:3c
buiningiana (orth. var.) 30:1 subsp. pectinifera 30:3a
centeteria see below 30:3c var. crebrispina (nom. nud.) 30:3b
cinerea (nom. nud.) 30:2 var. elegans (nom. nud.) 30:3
crebrispina (nom. nud.) 30:3b var. heteracantha (nom. nud.) 30:3
flavispina 30:3b var. horrida 30:3c
var. longispina (nom. nud.) 30:3b var. multicostata 30:3a
gummifera 30:2 var. pseudopectinifera 30:3a
subsp. meninensis 30:2 pseudopectinifera (nom. nud.) 30:3a
var. gigantea (nom. nud.) 30:2 warasii 30:3b
var. rubra 30:2 Zehntnerella chaetacantha 26:3
horrida (nom. nud.) 30:3c var. montealtoi 26:3
meninensis 30:2 polygona 26:3
var. antenaensis (nom. nud.) 30:2 squamulosa 26:3
var. rubra 30:2 Zygocactus candidus 14:2
pectinifera 30:3 obtusangulus 14:2
subsp. crebrispina (nom. nud.) 30:3b opuntioides 14:3
7.3 INDEX TO VERNACULAR NAMES

As all users of vernacular plant names should be aware, the application of such
nomenclature is very haphazard, so that some plants have many different names, while
many different species are known by the same. Yet other, common species, eg.
Brasilicereus phaeacanthus, have no recorded vernacular, while sometimes rare, narrow
endemic taxa possess a distinctive local name. Some commonly applied vernaculars are
used more or less indescriminately for a wide variety of cacti, so that, eg., the name
‘Mandacaru’ (commonly Cereus jamacaru) can be used to identify almost any Cactaceae in
Brazil; similarly ‘Facheiro’ and sometimes ‘Xique-xique’ (strictly speaking = Pilosocereus
gounellei) are used for almost all columnar or cylindrical-stemmed cacti in the Nordeste,
while ‘Quipá’, ‘Palma’ and ‘Palmatória’ are applied to anything vaguely opuntioid in
appearance. There are countless spelling variants, requiring the user to look for other
possible ways of spelling a name when consulting the index below. It is also important to
note that the following list is probably far from complete:
Alastrado Pilosocereus gounellei Facheiro-azul Pilosocereus pachycladus

subsp. gounellei Facheiro-da-lapa P. densiareolatus
Ambeba Brasiliopuntia brasiliensis Facheiro-da-praia P. arrabidae, P. catingicola
Arumbeva B. brasiliensis subsp. salvadorensis
Azedinha Pereskia aculeata Facheiro-da-serra Facheiroa cephaliomelana
Cabeça-branca Espostoopsis dybowskii subsp. estevesii
Cabeça-de-frade Melocactus spp., Facheiro-das-pedras Coleocephalocereus
buxbaumianus
Discocactus spp.,
Facheiro-de-serra Pilosocereus chrysostele
Uebelmannia gummifera
Facheiro-fino P. pentaedrophorus
Cabeça-de-frade-do- Melocactus conoideus
Periperi Facheiro-preto Facheiroa spp.
Cabeça-de-velho Espostoopsis dybowskii, Facho Pereskia aureiflora
Pilosocereus aurisetus, Facho-de-renda Brasiliopuntia brasiliensis
P. densiareolatus Flor-de-cera Pereskia bahiensis,
Cai-cai Quiabentia zehntneri P. stenantha,
Calumbi Pilosocereus pachycladus Quiabentia zehntneri
subsp. pernambucoensis Figo-da-Espanha Opuntia ficus-indica
Cardeiro Cereus jamacaru Figo-da-Índia O. ficus-indica
Cardo-ananá C. fernambucensis Figueira-do-inferno Cereus fernambucensis
Cardo-de-praia C. fernambucensis Frade-de-cavalo Discocactus bahiensis
Cardo-vinagre C. fernambucensis Gerumbeba Brasiliopuntia brasiliensis
Catana-de-jacaré Pseudoacanthocereus Gogóia Tacinga inamoena
brasiliensis Guibá T. inamoena
Caxacubri Pilosocereus tuberculatus Guipá T. inamoena
Cheque-cheque P. gounellei subsp. zehntneri Homem-velho Espostoopsis dybowskii
Chique-chique P. gounellei subsp. zehntneri Inhabento Pereskia bahiensis,
Cipó-Santo Pereskia aculeata P. stenantha
Cipó-de-espinho(s) Tacinga braunii, T. funalis Irahuka’arã Rhipsalis baccifera
Conambaia Rhipsalis baccifera Iviro Tacinga inamoena
Coroa-de-frade Discocactus spp., Jamacaru Cereus jamacaru
Melocactus spp., Jamaracá Opuntia ficus-indica
Uebelmannia gummifera Jamaracurú Cereus jamacaru
Coroa-do-diabo Discocactus spp. Jumarucú C. jamacaru
Cumbeba Brasiliopuntia brasiliensis, Jumbeba Pereskia bahiensis,
Cereus jamacaru P. stenantha
Dedinho Rhipsalis floccosa subsp. Jumucurú Cereus jamacaru
pulvinigera
Jurubeba Brasiliopuntia brasiliensis
Entrada-de-baile Pereskia bahiensis,
Jurumbeba Opuntia ficus-indica
P. grandifolia, P. stenantha
Lobolôbô Pereskia aculeata
Enxerto Rhipsalis baccifera,
Mandacaru Cereus jamacaru
R. lindbergiana
Mandacaru-babão Pilosocereus catingicola
Espada-de-jacaré Hylocereus setaceus
Mandacaru-cabeça- Espostoopsis dybowskii
Espinho-de-Santo- Pereskia aculeata, branca
Antônio P. bahiensis, P. stenantha, Mandacaru-da-serra Rhipsalis russellii
Quiabentia zehntneri Mandacaru-de-boi Cereus jamacaru
Espinho-preto Pereskia aculeata Mandacaru-de-facho Pilosocereus catingicola,
Facheiro Pilosocereus spp., P. pachycladus
Facheiroa squamosa Mandacaru-de-faixo Cereus jamacaru
Mandacaru-de-laço Pilosocereus tuberculatus Quiabo-da-lapa Cipocereus bradei,

Mandacaru-de- Espostoopsis dybowskii C. minensis,
penacho Pilosocereus aurisetus,
Mandacaru-de- Coleocephalocereus P. fulvilanatus,
topete buxbaumianus
Uebelmannia pectinifera
Mandacaru-de-três- Hylocereus setaceus
Quiabo-do-inferno Cipocereus minensis,
quinas
Pilosocereus aurisetus
Mandacaru-de-veado Pilosocereus pentaedrophorus
Quipá Tacinga inamoena,
Mandacaru-facheiro Cereus jamacaru
T. palmadora, T. quipa
Monducuru Opuntia monacantha
Quipá-de-espinho T. palmadora
Mumbebo Brasiliopuntia brasiliensis
Quipá-voador T. funalis
Mumbeca B. brasiliensis
Rabo-de-arara Lepismium houlletianum
Ora-pro-nobis Pereskia aculeata,
Rabo-de-gato Tacinga funalis
P. bahiensis, P. grandifolia
Rabo-de-espinhos T. braunii
Ora-pro-nobis- P. aureiflora
da-mata Rabo-de-onça Arrojadoa rhodantha,
Orelha-de-onça Opuntia ficus-indica Tacinga palmadora
Palma Nopalea cochenillifera, Rabo-de-raposa Arrojadoa dinae,
Opuntia ficus-indica, A. penicillata, A. rhodantha,
Tacinga palmadora, Cipocereus minensis,
T. saxatilis Harrisia adscendens,
Palma-de-engorda Nopalea cochenillifera Micranthocereus auriazureus,
Palma-de-espinho(s) Opuntia dillenii, Pilosocereus aurisetus,
Tacinga palmadora P. chrysostele, P. piauhyensis
Palma-de-ovelha T. inamoena Rabo-de-rato Tacinga braunii, T. funalis
Palma-doce Nopalea cochenillifera Rainha-de-noite Hylocereus setaceus
Palmadora Brasiliopuntia brasiliensis Rosa-mole Pereskia grandifolia
Palma-gigante Opuntia ficus-indica Roseta-do-diabo Discocactus spp.
Palmatória Brasiliopuntia brasiliensis, Rumbeba Brasiliopuntia brasiliensis
Nopalea cochenillifera, Sabonete Pereskia grandifolia
Opuntia dillenii, Sabugo-do-capeta Arthrocereus melanurus
O. ficus-indica, Sem-vergonha Pereskia grandifolia
O. monacantha, Surucucú P. aculeata, P. bahiensis,
Tacinga inamoena, P. stenantha
T. palmadora Trança-perna Tacinga funalis
Palmatória-de-espinho T. palmadora Tripa-de-galinha Rhipsalis baccifera
Palmatória-de-quipá T. palmadora Urumbeba Brasiliopuntia brasiliensis,
Palmatória-do-diabo Brasiliopuntia brasiliensis Cereus jamacaru,
Palmatória grande B. brasiliensis Opuntia monacantha
Palmatória-miúda Nopalea cochenillifera, Xique-xique Pilosocereus gounellei subsp.
gounellei
Tacinga inamoena
Xique-xique-amarelo Facheiroa cephaliomelana
Palminha T. palmadora
subsp. estevesii
Passa-prá-lá Harrisia adscendens Xique-xique-das- Pilosocereus gounellei subsp.
Quiabento Pereskia bahiensis, pedras zehntneri
P. grandifolia, P. stenantha, Xiquexique-do-sertão Brasiliopuntia brasiliensis
Quiabentia zehntneri Xique-xique-preto Facheiroa cephaliomelana
APPENDICES
7.4 APPENDICES I & II
APPENDIX 1
Table 7.1
Data on the scientific discovery and description of the native Cactaceae of
Eastern Brazil
TAXA (IN SYSTEMATIC Earliest illustration or herbarium Date of earliest correctly applied
ORDER) Key: † = endemic to record as native in Eastern Brazil; valid Latin name; origin of type;
total area; ‡ = endemic to core area state code; author ref./collector(s) author(s)
Pereskia aculeata 1896; MG; Silveira 1753; ex cult.; Linnaeus

P. grandifolia subsp. grandifolia 1815/1816; ES/RJ; Wied-Neuwied 1819; Rio de Janeiro; Haworth
P. grandifolia subsp. violacea ‡ 1922; MG; Santos 1986; Minas Gerais; Leuenberger
P. bahiensis ‡ 1906; BA; Ule 1908; Bahia; Gürke
P. stenantha ‡ 1912; BA; Zehntner 1979; Bahia; Ritter
P. aureiflora ‡ 1950; BA; Pinto 1979; Minas Gerais; Ritter
Quiabentia zehntneri ‡ 1912; BA; Zehntner 1919; Bahia; Britton & Rose
Tacinga funalis ‡ Before 1915; BA; Zehntner 1919; Bahia; Britton & Rose
T. braunii ‡ Before 1890; MG; anon. in Glaziou 1989; Minas Gerais; Esteves
Pereira
T. (Opuntia) werneri ‡ 1915; BA; Rose & Russell 1992; Bahia; Eggli
T. (Opuntia) palmadora ‡ Before 1645; NE Brazil; see 1919; Bahia; Britton & Rose
Whitehead & Boeseman
(1989: tt. 89b, 99b)
T. (Opuntia) saxatilis subsp. saxatilis 1959; MG; Ritter 1979; Minas Gerais; Ritter
T. (Opuntia) saxatilis subsp. estevesii ‡ 1984; BA; Esteves Pereira 1990; Bahia; Braun
T. (Opuntia) inamoena † Before 1890; MG; anon. in Glaziou 1890; Minas Gerais; Schumann
Brasiliopuntia (Opuntia) brasiliensis Before 1645; NE Brazil; see 1814; NE Brazil; Willdenow
(1989: tt. 3a, 7a, 51b)
Opuntia monacantha 1959; MG; Ritter 1819; ‘Barbados’; Haworth
Hylocereus setaceus Before 1867; MG; Warming (1908) 1828; Brazil; Salm-Dyck
Epiphyllum phyllanthus [data not available] 1753; ex cult.; Linnaeus
Pseudoacanthocereus brasiliensis ‡ 1912; BA; Zehntner 1920; Bahia; Britton & Rose
Lepismium houlletianum 1978; MG; Martinelli 1858; Brazil; Lemaire
L. warmingianum Before 1867; MG; Warming (1908) 1890; Minas Gerais; Schumann
L. cruciforme 1944; MG; Heringer 1827; ex cult.; Haworth
Rhipsalis russellii 1915; BA; Rose & Russell 1923; Bahia; Britton & Rose
R. elliptica 1959; MG; Castellanos 1890; São Paulo; Schumann
R. oblonga 1917; ES; Luetzelburg 1918; Rio de Janeiro; Löfgren
R. crispata 1970; PE; Andrade-Lima 1830; Brazil; Haworth
APPENDICES
R. floccosa subsp. floccosa 1915; BA; Rose & Russell 1837; Brazil; Pfeiffer
R. floccosa subsp. oreophila ‡ 1964; MG; Ritter 1979; Minas Gerais; Ritter
R. floccosa subsp. pulvinigera 1915; MG; Hoehne 1889; Brazil; Lindberg
R. paradoxa subsp. septentrionalis ‡ Before 1966; BA; Martins 1995; Bahia; Taylor & Barthlott
R. pacheco-leonis subsp. catenulata 1986; ES; Rauh & Kautsky 1992; Rio de Janeiro; Kimnach
R. cereoides 1986; ES; Rauh & Kautsky 1936; Rio de Janeiro; Backeberg
& Voll
R. sulcata ‡ 1986; ES; Rauh & Kautsky 1898; ex cult.; Weber
R. lindbergiana Before 1823; MG; Saint-Hilaire 1890; Rio de Janeiro; Schumann
R. teres 1898; MG; Jaguaribe 1829; Rio de Janeiro; Vellozo
R. baccifera subsp. baccifera 1945; CE; Cutler 1770–1777; ex cult.; Mueller
R. baccifera subsp. hileiabaiana ‡ Before 1857; BA; Blanchet 1995; Bahia; Taylor & Barthlott
R. pulchra 1970; MG; Krieger 1915; Rio de Janeiro; Löfgren
R. burchellii 1986; ES; Rauh 1923; São Paulo; Britton & Rose
R. juengeri 1991; MG; Zappi 1995; ex cult.; Barthlott & Taylor
R. clavata 1986; ES; Rauh & Kautskyi 1892; Rio de Janeiro; Weber
R. cereuscula 1893; MG; Silveira 1830; Brazil; Haworth
R. pilocarpa 1986; ES; Rauh & Kautsky 1903; São Paulo; Löfgren
R. hoelleri ‡ 1987; ES; Orssich 1995; Espírito Santo; Barthlott &
Taylor
Hatiora salicornioides 1892; MG; Ule 1819; Rio de Janeiro; Haworth
H. cylindrica 1989; BA; Augusto 1915; Rio de Janeiro; Britton &
Rose
Schlumbergera kautskyi ‡ 1986; ES; Kautsky 1991; Espírito Santo; Horobin &
McMillan
S. microsphaerica 1941; MG; Brade 1890; Rio de Janeiro; Schumann
S. opuntioides 1991; MG; Zappi 1905; Rio de Janeiro; Löfgren &
Dusén
Brasilicereus phaeacanthus ‡ 1906; BA; Ule 1908; Bahia; Gürke
B. markgrafii ‡ 1938; MG; Markgraf et al. 1950; Minas Gerais; Backeberg &
Voll
Cereus mirabella 1964; MG; Ritter 1979; Minas Gerais; Ritter
C. albicaulis ‡ 1915; BA; Rose & Russell 1920; Bahia; Britton & Rose
C. fernambucensis subsp. Before 1645; NE Brazil; see 1839; Pernambuco; Lemaire
fernambucensis Whitehead & Boeseman
(1989: tt. 45, 93, 97)
C. fernambucensis subsp. sericifer 1917; ES; Luetzelburg 1979; Rio de Janeiro; Ritter
C. insularis † Before 1877; F. de Noronha; Moseley 1884; F. de Noronha; Hemsley
C. jamacaru subsp. jamacaru †? Before 1645; NE Brazil; see 1828; NE Brazil; De Candolle
(1989: tt. 70, 78b, 89b, 99b)
C. jamacaru subsp. calcirupicola ‡ Before 1867; MG; Warming (1908) 1979; Minas Gerais; Ritter
C. hildmannianus Before 1890; MG; Glaziou (1909) 1890; SE Brazil; Schumann
Cipocereus laniflorus ‡ 1987; MG; Zappi & Scatena 1997; Minas Gerais; Taylor &
Zappi
APPENDICES
C. crassisepalus ‡ 1959; MG; Ritter 1973; Minas Gerais; Buining &

Brederoo
C. bradei ‡ Before 1935; MG; Brade ? 1942; Minas Gerais; Backeberg &
Voll
C. minensis subsp. leiocarpus ‡ 1934; MG; Brade 2004; Minas Gerais; Taylor &
Zappi
C. minensis subsp. minensis ‡ 1932; MG; Werdermann 1933; Minas Gerais; Werdermann
C. pusilliflorus ‡ 1964; MG; Ritter 1979; Minas Gerais; Ritter
Stephanocereus leucostele ‡ 1906; BA; Ule 1908; Bahia; Gürke
S. luetzelburgii ‡ 1913; BA; Luetzelburg 1923; Bahia; Vaupel
Arrojadoa bahiensis ‡ 1981; BA; Furlan et al. 1993; Bahia; Braun & Esteves
Pereira
A. dinae subsp. dinae ‡ 1964; BA; Ritter 1973; Bahia; Buining & Brederoo
A. dinae subsp. eriocaulis ‡ 1972; MG; Horst 1973; Minas Gerais; Buining &
Brederoo
A. penicillata ‡ 1906; BA; Ule 1908; Bahia; Gürke
A. rhodantha ‡ 1907; PI; Ule 1908; Bahia; Gürke
Pilosocereus tuberculatus ‡ 1932; PE; Werdermann 1933; Pernambuco; Werdermann
P. gounellei subsp. gounellei † Before 1645; NE Brazil; 1897; Pernambuco; Weber
see Whitehead & Boeseman
(1989: tab. 80)
P. gounellei subsp. zehntneri ‡ 1917; BA; Zehntner 1920; Bahia; Britton & Rose
P. catingicola subsp. catingicola ‡ c. 1907; BA; Ule 1908; Bahia; Gürke
P. catingicola subsp. salvadorensis † 1915; BA; Rose & Russell 1933; Bahia; Werdermann
P. azulensis ‡ 1968; BA/MG; Castellanos 1997; Minas Gerais; Taylor &
Zappi
P. arrabidae 1973; ES; Horst & Uebelmann 1862; Rio de Janeiro; Lemaire
P. brasiliensis subsp. brasiliensis 1990; ES; O.J. Pereira 1920; Rio de Janeiro; Britton &
Rose
P. brasiliensis subsp. ruschianus ‡ 1917; ES; Luetzelburg 1980; Espírito Santo; Buining &
Brederoo
P. flavipulvinatus † 1935; CE; Drouet 1979; Piauí; Buining & Brederoo
P. pentaedrophorus subsp. pentaedr. ‡ Before 1853; BA; Morel 1858; Bahia; Cels
P. pentaedrophorus subsp. robustus ‡ 1972; BA; Horst & Uebelmann 1994; Bahia; Zappi
P. glaucochrous ‡ 1932; BA; Werdermann 1933; Bahia; Werdermann
P. floccosus subsp. floccosus ‡ Before 1867; MG; Warming (1908) 1950; Minas Gerais; Backeberg &
Voll
P. floccosus subsp. quadricostatus ‡ 1965; MG; Ritter 1979; Minas Gerais; Ritter
P. fulvilanatus subsp. fulvilanatus ‡ 1968; MG; Horst 1973; Minas Gerais; Buining &
Brederoo
P. fulvilanatus subsp. rosae ‡ 1982; MG; Horst & Uebelmann 1984; Minas Gerais; Braun
P. pachycladus subsp. pachycladus ‡ 1907; BA; Ule (1908) 1975; Bahia; Buining & Brederoo
P. pachycl. subsp. pernambucoensis † 1915; BA; Rose & Russell 1979; Pernambuco; Ritter
P. magnificus ‡ 1964; MG; Ritter 1972; Minas Gerais; Buining &
Brederoo
APPENDICES
P. machrisii Before 1975; BA; Horst & Uebelmann 1957; Goiás; Dawson
P. aurisetus subsp. aurisetus ‡ Before 1862; MG; anon. 1933; Minas Gerais; Werdermann
P. aurisetus subsp. aurilanatus ‡ 1964; MG; Ritter 1979; Minas Gerais; Ritter
P. aureispinus ‡ c. 1972; BA; Horst & Uebelmann 1974; Bahia; Buining & Brederoo
P. multicostatus ‡ 1965; MG; Ritter 1979; Minas Gerais; Ritter
P. piauhyensis † 1907; PI; Ule 1908; Piauí; Gürke
P. chrysostele † 1920; PB/PE; Luetzelburg 1923; Paraíba/Pernambuco;
Vaupel
P. densiareolatus ‡ 1936; MG; Duarte 1979; Minas Gerais; Ritter
Micranthocereus violaciflorus ‡ 1968; MG; Buining 1969; Minas Gerais; Buining
M. albicephalus ‡ 1964; MG; Ritter 1979; Minas Gerais; Buining &
Brederoo
M. purpureus ‡ c. 1906; BA; Ule 1908; Bahia; Gürke
M. auriazureus ‡ 1971; MG; Buining 1973; Minas Gerais; Buining &
Brederoo
M. streckeri ‡ 1985; BA; Van Heek & Van Criekinge 1986; Bahia; Van Heek & Van
Criekinge
M. polyanthus ‡ 1932; BA; Werdermann 1932; Bahia; Werdermann
M. flaviflorus ‡ 1912; BA; Zehntner 1974; Bahia; Buining & Brederoo
M. dolichospermaticus ‡ 1972; BA; Horst & Uebelmann 1974; Bahia; Buining & Brederoo
Coleocephaloc. buxbaum. subsp. b. ‡ 1972; MG; Buining & Horst 1974; Minas Gerais; Buining
C. buxbaumianus subsp. flavisetus ‡ 1965; MG; Ritter 1979; Minas Gerais; Ritter
C. fluminensis subsp. fluminensis 1972; MG; Horst & Uebelmann 1838; Rio de Janeiro; Miquel
C. fluminensis subsp. decumbens ‡ 1965; MG; Ritter 1968; Minas Gerais; Ritter
C. pluricostatus ‡ 1917; MG/ES; Luetzelburg 1971; Minas Gerais; Buining &
Brederoo
C. goebelianus ‡ Before 1920; BA; Zehntner 1923; Bahia; Vaupel
C. aureus ‡ 1964; MG; Ritter 1968; Minas Gerais; Ritter
C. purpureus ‡ 1972; MG; Horst & Uebelmann 1973; Minas Gerais; Buining &
Brederoo
Melocactus oreas subsp. oreas ‡ Before 1840; BA; anon. 1840; Bahia; Miquel
M. oreas subsp. cremnophilus ‡ 1971; BA; Harley 1972; Bahia; Buining & Brederoo
M. ernestii subsp. ernestii † c. 1906; BA; Ule 1920; Bahia; Vaupel
M. ernestii subsp. longicarpus ‡ 1964; BA; Ritter 1974; Minas Gerais; Buining &
Brederoo
M. bahiensis subsp. bahiensis ‡ 1915; BA; Rose & Russell 1922; Bahia; Britton & Rose
M. bahiensis subsp. amethystinus ‡ 1959; MG; Ritter 1972; Bahia; Buining & Brederoo
M. conoideus ‡ 1972; BA; Horst 1973; Bahia; Buining & Brederoo
M. deinacanthus ‡ 1971; BA; Horst 1973; Bahia; Buining & Brederoo
M. levitestatus ‡ 1964; MG; Ritter 1973; Bahia; Buining & Brederoo
M. azureus ‡ 1968; BA; Horst 1971; Bahia; Buining & Brederoo
M. ferreophilus ‡ 1967; BA; Horst 1973; Bahia; Buining & Brederoo
M. pachyacanthus subsp. pachyac. ‡ 1972; BA; Horst 1975; Bahia; Buining & Brederoo
M. pachyacanthus subsp. viridis ‡ 1988; BA; Taylor & Zappi 1991; Bahia; Taylor
APPENDICES
M. salvadorensis ‡ 1932; BA; Werdermann 1934; Bahia; Werdermann

M. zehntneri † 1915; BA; Rose & Russell 1922; Bahia; Britton & Rose
M. lanssensianus ‡ 1977; PE; Horst 1986; Pernambuco; Braun
M. glaucescens ‡ 1967; BA; Horst 1972; Bahia; Buining & Brederoo
M. concinnus ‡ 1964; BA; Ritter 1972; Bahia; Buining & Brederoo
M. paucispinus ‡ 1981; BA; Heimen et al. 1983; Bahia; Heimen & Paul
M. violaceus s.l. Before 1645; NE Brazil; see 1835; Brazil; Pfeiffer
(1989: tab. 48c)
M. violaceus subsp. violaceus 1837; PE; Gardner as above
M. violaceus subsp. ritteri ‡ 1964; BA; Ritter 1979; Bahia; Ritter
M. violaceus subsp. margaritaceus ‡ Before 1858; BA; anon. 1857/58; Bahia; Miquel
Harrisia adscendens ‡ Before 1645?; NE Brazil; see 1908; Bahia; Gürke
(1989: tt. 89b, 99b)?
Leocereus bahiensis ‡ 1912; BA; Zehntner 1920; Bahia; Britton & Rose
Facheiroa ulei ‡ 1907; BA; Ule 1908; Bahia; Gürke
F. cephaliomelana subsp. cephaliom. ‡ 1959; MG; Ritter 1975; Bahia; Buining & Brederoo
F. cephaliomelana subsp. estevesii ‡ 1984; BA; E. Esteves Pereira 1986; Bahia; Braun
F. squamosa ‡ 1907; PI; Ule 1908; Piauí; Gürke
Espostoopsis dybowskii ‡ Before 1908; BA; Dybowski 1908; Bahia; Gosselin
Arthrocereus melanurus subsp. melan. Before 1890; MG; Glaziou 1890; Minas Gerais; Schumann
A. melanurus subsp. magnus ‡ 1969; MG; Krieger 1997; Minas Gerais; Taylor &
Zappi
A. melanurus subsp. odorus ‡ 1964; MG; Ritter 1979; Minas Gerais; Ritter
A. rondonianus ‡ Before 1935; MG; Brade ? 1951; Minas Gerais; Backeberg
& Voll
A. glaziovii ‡ Before 1832; MG; Sello 1890; Minas Gerais; Schumann
Discocactus zehntneri subsp. zehntn. ‡ c. 1915; BA; Zehntner 1922; Bahia; Britton & Rose
D. zehntneri subsp. boomianus ‡ 1967; BA; Horst 1971; Bahia; Buining & Brederoo
D. bahiensis † c. 1915; BA; Zehntner 1922; Bahia; Britton & Rose
D. heptacanthus subsp. catingicola 1972; BA; Horst 1974; Bahia; Buining & Brederoo
D. placentiformis ‡ Before 1826; MG; Riedel ? 1826; Brazil; Lehmann
D. pseudoinsignis ‡ 1972; MG; Horst 1991; Minas Gerais; Taylor &
Zappi
D. horstii ‡ 1972; MG; Horst 1973; Minas Gerais; Buining &
Brederoo
Uebelmannia buiningii ‡ 1966; MG; Horst 1968; Minas Gerais; Donald
U. gummifera ‡ 1938; MG; Mello-Barreto 1950; Minas Gerais; Backeberg
& Voll
U. pectinifera subsp. pectinifera ‡ 1966; MG; Horst 1967; Minas Gerais; Buining
U. pectinifera subsp. flavispina ‡ 1972; MG; Horst & Uebelmann 1973; Minas Gerais; Buining &
Brederoo
U. pectinifera subsp. horrida ‡ 1982; MG; Horst & Uebelmann 1984; Minas Gerais; Braun
APPENDICES
APPENDIX 2
Checklists of Cactaceae from adjacent Brazilian Regions
The states for species of restricted occurrence are indicated by means of 2-letter codes,
explained below; otherwise a listed taxon is known or expected to occur in all states
included in the Region, those with a dagger (†) being endemic to the region or area
indicated in the heading to the list. The most up-to-date published sources on the Cactaceae
native to Brazil as a whole are Hunt (1999a) and Braun & Esteves Pereira (2002, 2003).
Northern Brazil
Comprising the states/territories of Acre (AC), Amazonas (AM), Rondônia (RO), Roraima (RR),
Amapá (AP), Pará (PA) and Tocantins (TO). This region has only 16 species in total:
Cereus hexagonus (syn. C. perlucens, Pilocereus Micranthocereus (Siccobaccatus) estevesii (TO)

perlucens) Pilosocereus flexibilispinus† (TO)
C. sp. nov.† (TO: Mun. Palmas, limestone)* P. oligolepis (RR)
Discocactus heptacanthus subsp. catingicola (TO) P. machrisii (sens. lat.) (PA,TO)
Disocactus amazonicus (Wittia amazonica, Pseudorhipsalis ramulosa (AC)
Wittiocactus amazonicus) (AM) Rhipsalis baccifera subsp. baccifera
Epiphyllum phyllanthus Hylocereus cf. setaceus (PA,RR)
Melocactus estevesii† (RR) Selenicereus wittii (Strophocactus wittii ) (AM)
M. neryi (AM,RR,RO)
M. smithii (syn. M. roraimensis, Echinocactus
amazonicus?) (RR)
Central-western Brazil
Comprising the states/districts of Mato Grosso (MT), Mato Grosso do Sul (MS), Goiás (GO) and Distrito
Federal (DF). This region has a total of c. 36 species:
Arrojadoa sp. nov (?), cf. A. dinae† (GO) D. ferricola† (MS)

Arthrocereus spinosissimus† (MT) Echinopsis calochlora (MS,MT)
Brasiliopuntia brasiliensis (MT,MS) E. rhodotricha (MS)
Cereus adelmarii† (MT) Epiphyllum phyllanthus
C. bicolor Frailea cataphracta (syn. F. matoana) (MS)
C. kroenleinii (MT,MS) Gymnocalycium anisitsii (MS)
C. lanosus (MS)? G. marsoneri (MS)
C. pierre-braunianus (GO) Harrisia balansae (syn. H. guelichii, Eriocereus
C. saddianus† (MT) guelichii, Cereus balansae; ‘Harrisia pomanensis’
C. spegazzinii (syn. Monvillea spegazzinii) (MS) misapplied) (MS)
C. mirabella (GO,DF)? Hylocereus setaceus (MS,MT)
Cleistocactus baumannii subsp. horstii (MT,MS) Lepismium cruciforme (MS)
Discocactus heptacanthus sens. lat. Micranthocereus (Siccobaccatus) estevesii (GO)
* Seen in the nursery of the Jardim Botânico de Brasília, July 2000. Rootstock said not to be tuberous; stems creeping over
rocks, elongate cylindric, 2 cm diam., plain green, not glaucous; ribs 4–6(–8), low; spines brownish with pale tips, centrals 1(–4),
to 12 mm, radials 6, shorter. Flowers and fruit unknown.
APPENDICES
Opuntia anacantha (MS) P. saxicola (Cereus ritteri) (MS)

O. cf. quimilo (MS) Pilosocereus diersianus† (GO)
O. roborensis (MS) P. machrisii, sens. lat.
Pereskia aculeata (GO) P. parvus (GO)
P. sacharosa (MS,MT) P. vilaboensis (syn. P. rizzoanus) † (GO)
Praecereus euchlorus subsp. euchlorus (syn. Monvillea Rhipsalis russellii (GO,MT)
alticostata, ‘M. cavendishii’) (MS)
Southern Brazil and the parts of South-eastern Brazil not included within E Brazil
Comprising the states of Minas Gerais (MG, in part), Rio de Janeiro (RJ, in part), São Paulo (SP), Paraná
(PR), Santa Catarina (SC) and Rio Grande do Sul (RS). This area has a total of c. 103 species:
Cereus aethiops (RS) P. carambeiensis (PR,SC)

C. hildmannianus (sens. lat.) P. concinna, sens. lat. (RS)
C. mirabella (MG) P. crassigibba† (RS)
Discocactus heptacanthus (MG) P. erinacea (RS)
Echinopsis calochlora (RS) P. fusca (RS)
E. eyriesii (RS) P. haselbergii† (RS,SC)
E. rhodotricha (RS) P. herteri (RS)
Epiphyllum phyllanthus P. horstii† (RS)
Frailea castanea (RS) P. langsdorfii (RS)
F. cataphracta (RS) P. leninghausii (RS)
F. curvispina† (RS) P. linkii (RS,SC,PR)
F. gracillima, sens. lat. (RS) P. magnifica† (RS)
F. mammifera (RS) P. mammulosa (RS)
F. perumbilicata† (RS) P. muricata (RS)
F. phaeodisca (RS) P. neoarechavaletae (RS)
F. pumila (RS) P. neohorstii† (RS)
F. pygmaea, sens. lat. (RS) P. nothominuscula† (RS)
Gymnocalycium denudatum (RS) P. nothorauschii (RS)
G. horstii† (RS) P. ottonis (RS,SC,PR)
G. netrelianum (RS) P. oxycostata (RS)
Hatiora cylindrica (MG?,RJ,SP?) P. rudibuenekeri† (RS)
H. epiphylloides† (SP,RJ) P. rutilans (RS)
H. gaertneri† (PR,SC,RS) P. scopa (RS)
H. herminiae† (SP) P. tenuicylindrica† (RS)
H. rosea† (PR,SC,RS) P. warasii (RS)
H. salicornioides (PR,SP,RJ,MG) Pereskia aculeata
Hylocereus setaceus (PR,SP,RJ,MG) P. grandifolia (SP)
Lepismium cruciforme P. nemorosa (RS)
L. houlletianum Pilosocereus albisummus† (MG)
L. lumbricoides P. brasiliensis (RJ)
L. warmingianum P. machrisii (MG,SP)
Micranthocereus (Siccobaccatus) estevesii (MG) Praecereus euchlorus subsp. euchlorus (SP,PR)
Opuntia monacantha (syn. O. arechavaletae) Rhipsalis agudoensis† (RS)
Parodia alacriportana† (RS,SC) R. baccifera subsp. shaferi (SP)
P. arnostiana† (RS) R. burchellii (SP,PR)
P. buiningii (RS) R. campos-portoana
APPENDICES
R. cereoides (RJ) R. pachyptera†

R. cereuscula R. paradoxa (subsp. paradoxa†)
R. clavata (SP,RJ) R. pentaptera† (RJ)
R. crispata (SP,RJ,SC) R. pilocarpa
R. dissimilis† (SP,PR) R. pulchra (MG,SP,RJ)
R. elliptica R. puniceodiscus†
R. ewaldiana† (RJ) R. russellii (PR)
R. floccosa (sspp. floccosa & pulvinigera) R. teres
R. grandiflora† (RJ,SP,PR) R. trigona† (SP,PR,SC)
R. lindbergiana (RJ,SP) Schlumbergera microsphaerica (RJ)
R. mesembryanthemoides† (RJ) S. opuntioides (SP,RJ)
R. neves-armondii S. orssichiana† (SP,RJ)
R. oblonga (RJ,SP) S. russelliana† (RJ)
R. olivifera† (SP,RJ) S. truncata† (RJ)
R. ormindoi† (RJ) Tacinga saxatilis subsp. saxatilis (MG).
R. pacheco-leonis (RJ)

02.taylor & Zappi 2004-Cacti of Eastern Brazil (LIVRO)

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

02.taylor & Zappi 2004-Cacti of Eastern Brazil (LIVRO)

Uploaded by

Copyright:

Available Formats

CACTI

Facheiroa squamosa (Photo: Nigel Taylor)

Nigel Taylor and Daniela Zappi

First published in 2004 by

Production editor: Ruth Linklater

For information or to purchase all Kew titles please visit

PREFACE by Prof. Dr Wilhelm Barthlott ................................................................ xi

2. HISTORY OF DISCOVERY, NAMING AND CLASSIFICATION .............. 13

4. CONSERVATION ..................................................................................... 139

5. TAXONOMIC INVENTORY ................................................................... 159

TEXT TABLES, MAPS, HALFTONE ILLUSTRATIONS

Map 5. Distribution records of all taxa restricted to campo rupestre (excluding

Map 34. Distribution of Pilosocereus subg. Gounellea ........................................... 102

7. Tacinga saxatilis subsp. saxatilis, T. inamoena ................................................. 126

38. Pilosocereus tuberculatus, P. gounellei subsp. gounellei ........................................ 253

62. Melocactus glaucescens..................................................................................... 405

natural inheritance — a richness which is slipping away day-by-day as man’s reckless

1.2 BACKGROUND TO THE CACTI OF EASTERN

1.3 STUDY METHODS AND TAXONOMIC CONCEPTS

Dates Areas visited Main purpose of Collaborator(s)

July 1989 Bahia, N Minas Gerais Collecting material for

Jan.–Feb. Cent. & SW Ceará, As above

Sep. 1990 Cent.-N Minas Gerais Collecting material for

Dec. 1990 NE & SE Minas Gerais, As above (Taylor/Zappi)

Jan.–Feb. Bahia, E Goiás, N Minas As above U. Eggli (ZSS)

April 1992 N & E Bahia Noting/photographing

Feb. 1995 E Bahia, E Alagoas, As above BA: A.M. Carvalho

March– W Pernambuco, N & W Noting/photographing E.A. Rocha, Univ. Estadual

Nov. 2001 E Bahia Cactaceae Course, Univ. J. Jardim et al. (HUEFS)

disarticulating old/diseased stem-segments and expanded flowers with ± colourless

spreading yellow perianth-segments. However, the previously ill-defined ‘dustbin genus’

important if similarities or differences and suites of linked characters that relate to

2. HISTORY OF DISCOVERY, NAMING

2.2 COUNT JOHAN MAURITS IN NORTH-EASTERN

2.3 19 TH CENTURY COLLECTORS

2.4 THE GOLDEN AGE OF CACTUS DISCOVERY:

2.5 MODERN COLLECTORS: POST 1950

catingicola, D. pseudoinsignis, D. horstii, Uebelmannia buiningii, U. pectinifera (3 subspp.). The

2.6 TAXONOMIC HISTORY OF THE CACTACEAE

Stephanocereus, Pilosocereus, Micranthocereus, Austrocephalocereus and Coleocephalocereus. The

2.7 CACTACEAE SYSTEMATICS 1980–2003 AND THE

1976), where the slender-stemmed, small-flowered South American rhipsaloid epiphytes

E Brazilian genus IOS Consensus status Comments/justification

C ACTOIDEAE (HPE clade) — Hylocereeae:

C ACTOIDEAE (HPE clade?) — Echinocereeae(?):

C ACTOIDEAE (BCT clade) — Cereeae:

22. Coleocephalocereus Accepted by a majority of IOS WP The presence of fruits dehiscent by

C ACTOIDEAE (BCT clade) — Trichocereeae:

30. Uebelmannia Accepted less than unanimously by Strongly supported as monophyletic

Subfamily Characters / molecular markers

O PUNTIOIDEAE K. Schum. Stems succulent, at least at first, bearing cylindrical to

C ACTOIDEAE Stems succulent, leaves reduced to minute ephemeral

Tribes/subtribes of Buxbaum, Status based on molecular Comments

H YLOCEREEAE Buxb.: 5 subtribes

PACHYCEREEAE Buxb.: 4 subtribes

T RICHOCEREEAE Buxb.: 4 subtribes

NOTOCACTEAE Buxb.: 5 subtribes

TAXON / STATE MA PI CE RN PB PE AL SE BA MG1 ES RJ2

P. flocc. subsp. quadricostatus ‡ +

M. violaceus subsp. violaceus + + + + + + +

STATE MA PI CE RN PB PE AL SE BA MG1 ES RJ2

Pereskioideae. Pereskia, nowadays the only genus included in this subfamily, is

Opuntioideae. In terms of species, the 4 rather distinct genera of Opuntioideae native