Cicadas, Clasificacion, Conceptos y Recomendaciones

00prelims.
qxd 28/11/03 3:40 pm Page i
CYCAD CLASSIFICATION
CONCEPTS AND RECOMMENDATIONS
00prelims.qxd 28/11/03 3:40 pm Page ii
00prelims.qxd 28/11/03 3:40 pm Page iii
Cycad Classification
Concepts and Recommendations
Edited by
Terrence Walters
and
Roy Osborne
CABI Publishing
00prelims.qxd 28/11/03 3:40 pm Page iv
CABI Publishing is a division of CAB International
CABI Publishing CABI Publishing

CAB International 875 Massachusetts Avenue
Wallingford 7th Floor
Oxfordshire OX10 8DE Cambridge, MA 02139
UK USA
Tel: +44 (0)1491 832111 Tel: +1 617 395 4056

Fax: +44 (0)1491 833508 Fax: +1 617 354 6875
E-mail: cabi@cabi.org E-mail: cabi-nao@cabi.org
Web site: www.cabi-publishing.org
CAB International 2004. All rights reserved. No part of this

publication may be reproduced in any form or by any means,
electronically, mechanically, by photocopying, recording or
otherwise, without the prior permission of the copyright owners.
A catalogue record for this book is available from the British

Library, London, UK.
Library of Congress Cataloging-in-Publication Data

Cycad classification : concepts and recommendations / edited by Terrence
Walters and Roy Osborne.
p. cm
Includes bibliographical references (p. ).
ISBN 0-85199-741-4 (alk. paper)
1. Cycads--Classification. I. Walters, Terrence, 1955. II. Osborne,
Roy. III. title
QK494.C93 2004
585.9012--dc21
2003010044
ISBN 0 85199 741 4
Typeset by MRM Graphics Ltd, Winslow, Bucks

Printed and bound in the UK by Cromwell Press, Trowbridge
00prelims.qxd 28/11/03 3:40 pm Page v
Contents
Contributors vii
About the Editors ix
Preface xi
Acknowledgements xv
1. We Hold these Truths 1

Terrence Walters, Roy Osborne and Don Decker
2. Saving Ghosts? The Implications of Taxonomic Uncertainty

and Shifting Infrageneric Concepts in the Cycadales for Red
Listing and Conservation Planning 13
John Donaldson
3. Character Evolution, Species Recognition and Classification

Concepts in the Cycadaceae 23
Ken D. Hill
4. Morphological Characters Useful in Determining Species

Boundaries in Cycas (Cycadaceae) 45
Anders Lindstrm
5. Comments on Cycas, Dyerocycas and Epicycas (Cycadaceae) 57

Chia-Jui Chen, Ken D. Hill and Dennis Wm. Stevenson
6. Classification Concepts in Encephalartos (Zamiaceae) 69

Piet Vorster
v
00prelims.qxd 28/11/03 3:40 pm Page vi
vi Contents
7. Classification Concepts in Macrozamia (Zamiaceae) from

Eastern Australia 85
Paul I. Forster
8. Classification Concepts in Ceratozamia (Zamiaceae) 95

Loran M. Whitelock
9. Relationships and Phytogeography in Ceratozamia

(Zamiaceae) 109
Andrew P. Vovides, Miguel A. Prez-Farrera, Dolores Gonzlez and Sergio Avendao
10. A Morphometric Analysis of the Ceratozamia norstogii

Complex (Zamiaceae) 127
Miguel A. Prez-Farrera, Andrew P. Vovides, Luis Hernndez-Sandoval, Dolores
Gonzlez and Mahinda Martnez
11. Hypotheses on the Relationship between Biogeography and

Speciation in Dioon (Zamiaceae) 137
Timothy J. Gregory and Jeffrey Chemnick
12. Molecular Phylogeny of Zamia (Zamiaceae) 149

Paolo Caputo, Salvatore Cozzolino, Paolo De Luca, Aldo Moretti and Dennis
Wm. Stevenson
13. Systematics of Meso-American Zamia (Zamiaceae) 159

Bart Schutzman
14. Zamiaceae of Bolivia, Ecuador and Peru 173

Dennis Wm. Stevenson
15. In Search of the True Tree: Guidelines for Classification 195

Roy Osborne and Terrence Walters
Appendix 1: The World List of Cycads 219

Ken D. Hill, Dennis Wm. Stevenson and Roy Osborne
Appendix 2: Glossary of Terms Encountered in Cycad

Systematics 237
Roy Osborne and Terrence Walters
Index 259
00prelims.qxd 28/11/03 3:40 pm Page vii
Contributors
Avendao, S., Instituto de Ecologia A.C., Apartado Postal 63, Xalapa, Veracruz 91000,
Mexico.
Caputo, P., Dipartimento di Biologia Vegetale, Universit degli Studi di Napoli Federico II,
Via Foria 223, 80139 Napoli, Italy.
Chemnick, J., Ganna Walska Lotusland, 695 Ashley Road, Santa Barbara, California
93108, USA.
Chen, C.-J., Institute of Botany, Chinese Academy of Sciences, 20 Nanxincum, Xiangshan,
Beijing 100093, China.
Cozzolino, S., Dipartimento di Biologia Vegetale, Universit degli Studi di Napoli Federico
II, Via Foria 223, 80139 Napoli, Italy.
Decker, D., Decker & Associates, Inc., PO Box 222153, Carmel, California 93923, USA.
Donaldson, J., Kirstenbosch Research Center, National Botanical Institute, Private Bag X7,
Claremont 7735, South Africa.
Forster, P.I., Queensland Herbarium, Environmental Protection Agency, Brisbane Botanic
Gardens, Mt Coot-tha Road, Toowong, Queensland 4066, Australia.
Gonzlez, D., Instituto de Ecologia A.C., Apartado Postal 63, Xalapa, Veracruz 91000,
Mexico.
Gregory, T.J., Montgomery Botanical Center, 11901 Old Cutler Road, Miami, Florida
33156-4242, USA.
Hernndez-Sandoval, L., Facultad de Biologia, Universidad Autnoma de Quertaro,
Centro Universitario, Cerro Las Campanas S/N, Quertaro 76010, Mexico.
Hill, K.D., Royal Botanic Gardens, Mrs Macquaries Road, Sydney 2000, Australia.
Lindstrm, A., Nong Nooch Tropical Botanical Garden, 34/1 Sukhumvit Highway,
Najomtien, Sattahip, Chonburi 20250, Thailand.
De Luca, P., Dipartimento di Biologia Vegetale, Universit degli Studi di Napoli Federico II,
vii
00prelims.qxd 28/11/03 3:40 pm Page viii
viii Contributors
Martnez, M., Facultad de Biologa, Universidad Autnoma de Quertaro, Centro

Universitario, Cerro Las Campanas S/N, Quertaro 76010, Mexico.
Moretti, A., Dipartimento di Biologia Vegetale, Universit degli Studi di Napoli Federico II,
Osborne, R., PO Box 244, Burpengary, Queensland 4505, Australia.
Prez-Farrera, M.A., Escuela de Biologa, Universidad de Ciencias y Artes de Chiapas
(UNICACH), Calzada Samuel Len Brindis 151, Tuxtla Gutirrez, Chiapas 29000,
Mexico.
Schutzman, B., Environmental Horticulture Department, University of Florida, 1525
Fifield Hall, Gainesville, Florida 32611-0670, USA.
Stevenson, D.W., Institute of Systematic Botany, New York Botanical Garden, Bronx, New
York 10458, USA.
Vorster, P., Botany Department, University of Stellenbosch, Private Bag X1, Matieland
7602, South Africa.
Vovides, A.P., Instituto de Ecologa A.C., Apartado Postal 63, Xalapa, Veracruz 91000,
Mexico.
Walters, T., Montgomery Botanical Center, 11901 Old Cutler Road, Miami, Florida
33156-4242, USA.
Whitelock, L.M., 4524 Toland Way, Los Angeles, California 90041, USA.
00prelims.qxd 28/11/03 3:40 pm Page ix
About the Editors
Terrence Walters is the executive director of Montgomery Botanical Center,

a botanical garden in Miami, Florida, that concentrates on scientifically docu-
mented population-based collections of cycads and palms. Since 1990, he has
conducted numerous field expeditions to parts of Asia, Africa and the Americas
to investigate and document the cycad flora of the world. He is on the board of
directors of the Cycad Society, a member of IUCNs Cycad Specialist Group
and on the research faculty of Florida International University. (Photography by
Mary Andrews.)
ix
00prelims.qxd 28/11/03 3:40 pm Page x
x About the Editors
Roy Osborne has been studying and growing cycads in Africa and Australia for
more than 20 years. He is a member of the IUCNs Cycad Specialist Group, and
founder and first President of the Cycad Society of South Africa. Now living in
Brisbane, Australia, he has published more than 100 scientific papers, books and
book chapters and has participated in major international conferences on cycad
biology. (Photography by Mary Andrews.)
00prelims.qxd 28/11/03 3:40 pm Page xi
Preface
On 7 April 2002, the Cycad Classification Concepts (CCC) Workshop was con-
vened at Montgomery Botanical Center in Miami, Florida, USA. Seventeen of the
worlds leading authorities on cycad systematics were invited to participate in the
workshop and to submit manuscripts for this volume. Fifteen of these systematists
submitted manuscripts and 14 were able to attend the 3-day CCC Workshop.
The purpose of the CCC Workshop was to develop a suite of classification
guidelines in support of the advancement of an internationally accepted and
stable evolutionary classification system for taxa in the Cycadales. Increased
research activity in the field of cycad systematics has led in some cases to increased
confusion. As researchers across the globe pursue the many new lines of inquiry
provided by technological advances of the past two decades (e.g. DNA sequenc-
ing, random amplified polymorphic DNA analysis, etc.), focus on consensus for
how the approximately 300 species of cycads should be classified has become
clouded. There is an urgent need for guidelines that all cycad systematists can
follow in the designation of species, species boundaries and species groupings. The
CCC Workshop provided the venue for the development of these guidelines.
Although workshops with a similar purpose have been held to examine crit-
ically the systematics of other plant groups, the CCC Workshop was uniquely
designed using progressive business methodologies. Five arenas were identified as
necessary for the planning and management of this event. The Personnel Arena
dealt with the subject of who would be involved as CCC Participants, who would
be on the CCC Support Team and who would be in leadership roles during the
Workshop process. The Site Arena dealt with everything concerning the facilities
required for the Workshop such as rooms for the various events and work ses-
sions, transportation, housing, furniture, catering and audio-visual equipment.
The Operations Arena dealt with identifying and taking those actions required to
xi
00prelims.qxd 28/11/03 3:40 pm Page xii
xii Preface
produce the major product of the Workshop this volume. The Planning Arena
dealt with determining all of the tasks required, their flow, their content and their
sequence from the overall purpose and concept of the Workshop to the minute
details associated with the organization and objectives of the Workshop sessions
themselves. Finally, the Management Arena dealt with how all of the above
would be led and managed.
The first step was to bring in a management consultant, Don Decker, to
support the Management Arena objectives and to oversee development of the
other four arenas. The next steps were to articulate the purpose, or reason, for
having the CCC Workshop, and to determine the products, or results, required
to meet the purpose successfully. The overall process of actions that would be
required to obtain the products was outlined and then the functioning capabili-
ties, or resources, required for the process were identified. These processes and
the development of the above five arenas provided the overall planning and exe-
cution structure for the CCC Workshop.
Bringing together a group of world-renowned cycad systematists represent-
ing several countries, cultures and languages for consensus building can be diffi-
cult. That this event was successful is a tribute to the considerable work that took
place prior to, during and after the Workshop by the CCC Support Team and
the CCC Participants.
The CCC Participants were 14 of the worlds leading and most respected
cycad systematists. Paolo Caputo from the Universit degli Studi di Napoli
Federico II in Italy represented one of the largest concentrations of cycad sys-
tematists at any one institution in the world. The Naples cycad group has worked
extensively on New World taxa.
Participants representing Asia included Chia-Jui Chen from the Institute of
Botany in Beijing, China, an expert on the cycads of China, and Anders
Lindstrm, the cycad curator at Nong Nooch Tropical Gardens in Thailand.
Lindstrm is one of the leading experts on the cycads of Thailand. Cycas lindstromii
was named in his honour.
John Donaldson, from the National Botanical Institute, and Piet Vorster,
from the University of Stellenbosch, were the workshops representatives from
South Africa. Donaldson is Chairman of the IUCN (World Conservation Union)
Cycad Specialist Group. Vorster is currently the President of the Cycad Society
of South Africa and is an authority on the African genus, Encephalartos.
Due to the large number of active cycad systematists in Australia, this country
was well represented at the workshop. Attendees included Paul Forster of the
Queensland Herbarium, Australias expert on Macrozamia. Ken Hill, from the Royal
Botanic Gardens in Sydney, is the worlds expert on the taxonomically difficult genus
Cycas. Roy Osborne, who currently resides in Queensland and formerly lived in
South Africa, began the development many years ago of the world list of cycads.
Hill and Osborne recently published the authoritative work Cycads of Australia.
Andrew Vovides directs the National Cycad Collection of Mexico, and has
developed the concept of local conservation of native cycads by initiating proj-
ects in which local villagers create nurseries to grow native cycads from sustain-
00prelims.qxd 28/11/03 3:40 pm Page xiii
Preface xiii
able seed harvests. Americans Tim Gregory and Jeff Chemnick continue to
undertake extensive systematic fieldwork in Mexico. Their commitment to
walking up every canyon in search of each and every population of a species is
to be admired.
Other participants from the United States included Bart Schutzman of the
University of Florida, the editor of The Cycad Society Newsletter and expert on
Meso-American Zamia. Dennis Stevenson of the New York Botanical Garden is
the leading authority on Central and South American taxa and has published
extensively on evolutionary concepts in the Cycadales. Loran Whitelock from
California, after a decade of fieldwork, research and writing, has recently com-
pleted what will become the major reference work on the cycad flora of the world
The Cycads. Ceratozamia whitelockiana and Encephalartos whitelockii are named in
recognition of Whitelocks extensive research on the worlds cycad flora.
The first session of the CCC Workshop, held on 7 April, created the oppor-
tunity for each CCC Participant to give a 20-minute oral presentation of their
professional views on cycad classification concepts, systematics and taxonomy.
This 1-day work session was organized as a symposium (CCC Symposium) that
included invited guests. The second work session, conducted on day 2, focused on
elucidating the beliefs and philosophies that the participants held to be true con-
cerning cycad systematics. Also on day 2, during work session three, Katherine
Kron of Wake Forest University presented a discussion on a relatively new and
somewhat controversial approach to plant nomenclature called phylocode. On
the third day of the Workshop, the fourth and fifth work sessions required that
the CCC Participants come to alignment on a suite of classification concepts or
guidelines that they, as a group, would support and encourage the use of present-
ly and in the future.
In this volume, Chapter 1 presents why the CCC Workshop was convened
and the beliefs, or working hypotheses and assumptions, that the CCC
Participants hold to be true for cycad classification. This chapter resulted from
work sessions two and three. The final chapter, Chapter 15, based on the products
obtained from work sessions four and five, summarizes the classification guidelines
that the CCC Participants have agreed to follow, support and encourage the use
of to produce a universally accepted stable classification system for the Cycadales.
Prior to the Workshop, each CCC Participant submitted a manuscript to the
editors. These manuscripts were detailed discussions of the oral presentations
presented by the participants during the CCC Symposium (work session one).
These manuscripts constitute Chapters 214 of this volume.
In Chapter 2, John Donaldson discusses the practical need for a durable clas-
sification system in the Cycadales when dealing with cycad conservation issues
and planning. In Chapters 3 and 4, Ken Hill and Anders Lindstrm critically
examine the usefulness of various characters for defining species and species con-
cepts within Cycas. Three of the CCC Participants, Chia-Jui Chen, Ken Hill and
Dennis Stevenson, report on a study of a recently described genus in the
Cycadaceae in Chapter 5. They present a methodology for how cycad systema-
tists should critically evaluate proposed new taxa. Cycad experts Piet Vorster,
00prelims.qxd 28/11/03 3:40 pm Page xiv
xiv Preface
Paul Forster and Loran Whitelock present their individual thoughts on infra-
generic classification concepts for African Encephalartos taxa, Australian
Macrozamia taxa and New World Ceratozamia taxa in Chapters 6, 7 and 8, respec-
tively. Chapters 9 and 10 were submitted by the participants from Mexico,
Andrew Vovides and Miguel A. Prez-Farrera. Unfortunately, Miguel was not
able to attend the Workshop. These two researchers evaluate the usefulness of
characters for defining species and species complexes within Ceratozamia. Tim
Gregory and Jeff Chemnick develop in Chapter 11 an exciting hypothesis that
extant species of Dioon are the result of rapid evolution in a dynamic group of
plants. Two of the participants, Paolo Caputo and Dennis Stevenson, along with
their colleagues, report on a molecular study in Chapter 12 that examines the
usefulness of molecular and morphological data sets when trying to develop a
phylogenetic tree for species of Zamia. Results from Bart Schutzmans extensive
and detailed morphological studies on Meso-American species of Zamia are given
in Chapter 13. Dennis Stevenson, in Chapter 14, presents a monograph on the
Zamiaceae from Bolivia, Ecuador and Peru. His chapter illustrates many of the
guidelines the participants discussed during the last day of the workshop con-
cerning content, style and format for the type of publication resulting from floris-
tic cycad research.
Two appendices are included in this volume. Firstly, the ongoing discovery of
new species and the continuous refinements to the taxonomy of the already-
known taxa mean that the list of officially recognized taxa needs to be timeous-
ly revised. Appendix 1 gives details of the World List of Cycads at the time the
manuscript for this volume was submitted to the publisher and is based on The
Cycad Pages website (http://plantnet.rbgsyd.nsw.gov.au/PlantNet/cycad).
Secondly, the interdisciplinary nature of work on cycad systematics has led to a
large and complex vocabulary of terms, the precise meanings of which are some-
times obscure and occasionally misused. Appendix 2 provides a glossary of these
terms, drawn up after extensive consultations with specialists, and amplified
where possible with cycad-specific examples.
For consistency with author citations for taxa, we have followed the
International Plant Names Index (IPNI Website: http://www.ipni.org/
index.html) for the chapters and appendices. In Chapters 115, authors names
are unabbreviated. They are cited when the taxon first appears within a chapter
and are also cited when appropriate in figure captions and tables. For Appendix
1, authors names are abbreviated.
Taxa known to be distinct by the authors of each chapter, but as yet not offi-
cially published, are indicated by double quotes in Chapters 115.
The work presented in this volume is not only a report on the current state
of affairs in cycad classification, but also highlights areas of difficulty and leads
to guidelines for meaningful future advances. We hope it will become a widely
used reference for the benefit of all cycad researchers, enthusiasts, conservation-
related public and private agencies and students of plant systematics.
Terrence Walters and Roy Osborne

00prelims.qxd 28/11/03 3:40 pm Page xv
Acknowledgements
The directors, members, staff and volunteers of Montgomery Botanical Center

(MBC) deserve thanks and gratitude for their commitment to planning, hosting
and sponsoring the Cycad Classification Concepts (CCC) Workshop. The
Workshop was the first of its kind held at MBC, and therefore demanded an
extremely fast learning curve by the MBC Team. We hope they will consider
organizing and hosting CCC Workshop II in the future.
The Workshop was 11 months in the planning stage. Numerous meetings
were held during this period to define clearly the purpose, expected products and
agenda for the Workshop. Tim Gregory and William Tang provided much
needed guidance during the early planning stages. Don and Sonja Decker are
thanked for hosting a pre-Workshop planning session at their home in December
2001. At this 1-day session, Don Decker, Tim Gregory, Deena Walters and
Terrence Walters clearly identified the purpose and the products expected from
the Workshop, as well as plotting the 3-day agenda. Jeff Chemnick, Tim Gregory
and Loran Whitelock provided encouragement and support throughout the
entire planning period for the Workshop. Their input was much appreciated.
The CCC Participants commitment to the success of the Workshop was
greatly valued, given their own responsibilities and time constraints. They were
truly an amazing group of individuals with whom to collaborate.
The CCC Support Team oversaw the planning and logistics of the
Workshop beginning in April, 2001. Jean Stark of Stark Connections took care
of all the details associated with the participants travel and housing needs before,
during and after the Workshop. Don Decker of Decker & Associates was the
management consultant and leader for the Workshop. Katherine Kron of Wake
Forest University was kind enough to lecture and field numerous questions on
phylocode during one of the work sessions. Evelyn Young planned and coordi-
xv
00prelims.qxd 28/11/03 3:40 pm Page xvi
xvi Acknowledgements
nated all of the on-site meals and events at MBC during the 3 days. Larry
Noblick coordinated the audio-visual equipment and supplies required for the
Workshop. Lee Anderson oversaw facility preparations for all of the events each
day during the Workshop. Deena Walters and Mary Andrews documented the
Workshop photographically. Mayna Hutchinson created magnificent cycad
arrangements for all of the facilities used at MBC during the Workshop. Barbara
Judd, Sue Katz and Eric Shroyer were scribes for the break-out groups. All of the
above individuals, as well as their own support teams, ensured that every aspect
of the Workshop ran smoothly and according to the agenda.
The CCC Workshop was supported through grants from the Bressler
Foundation, Ajax Foundation and General Mills Foundation. Libby and By Besse,
Judith and Richard Bressler, Tim Gregory, Eileen and Loyd Kelly and Linda and
Mark Smith were significant contributors to the Workshop. The commitment of
these foundations and individuals to the Workshop was very much appreciated.
The officers, directors and members of the Central Florida Palm & Cycad Society
are acknowledged for providing support towards the publication of this volume.
A number of individuals were involved in the preparation and publication of
this volume. The Montgomery Botanical Center directors, members and staff
gave the editors the time and resources to prepare the manuscript. The first
editor gratefully acknowledges Deena Walters for graphic design and illustration
production assistance and, most of all, for her support during the preparation of
the manuscript. The second editor similarly acknowledges the support of the
Osborne family and friends.
Finally, we wish to express our thanks to Tim Hardwick and the many indi-
viduals at CAB International who have so professionally managed the publica-
tion aspects of this volume.
Cycad Classification Concepts Workshop support team seated from left to right:
Terrence Walters; Jean Stark; Don Decker and Katherine Kron. (Reprinted by per-
mission of Montgomery Botanical Center, Miami, Florida, USA.)
00prelims.qxd
28/11/03
3:40 pm
Page xvii
Acknowledgements
Cycad Classification Concepts Workshop participants seated from left to right: Andrew P. Vovides; Ken D. Hill; Chia-Jui Chen; Roy
Osborne; Paolo Caputo; John Donaldson; Loran Whitelock; Paul I. Forster; Dennis Wm. Stevenson; Jeffrey Chemnick and Piet Vorster. On
the floor in front: Anders Lindstrm; Bart Schutzman and Timothy J. Gregory. (Reprinted by permission of Montgomery Botanical Center,
xvii
Miami, Florida, USA.)
00prelims.qxd 28/11/03 3:40 pm Page xviii
ch01.qxd 28/11/03 3:42 pm Page 1
We Hold these Truths 1

Terrence Walters,1 Roy Osborne2 and Don
Decker3
1Montgomery Botanical Center, Miami, Florida, USA; 2PO

Box 244, Burpengary, Queensland, Australia; 3Decker &
Associates, Inc., Carmel, California, USA
Abstract
In order to develop classification guidelines for the Cycadales, a workshop was held in
April 2002, at Montgomery Botanical Center in Miami, Florida, USA. Fourteen interna-
tionally-renowned cycad systematists spent 3 days identifying and developing guidelines
that would provide a stable, practical and informative classification scheme for cycads. The
participants agreed that convening such a workshop was vital, timely and necessary to
produce a universally accepted evolutionary classification for the Cycadales in the near
future. Before developing the guidelines, the participants first needed to identify the
assumptions, or beliefs, that they hold to be true about cycad classification. These beliefs
are presented under three categories: (i) beliefs about biological relationships; (ii) beliefs
about what systematists can and should do in order to understand biological relationships;
and (iii) beliefs about what cycad systematists can and should do in order to understand
relationships in the Cycadales.
Cycad Classification Concepts Workshop
The field of cycad systematics, which focuses on all members of the plant order
Cycadales, has seen a flurry of activity during the past 20 years. New species are
being discovered and described on an annual basis. Existing species circumscrip-
tions are being critically tested for their scientific soundness. Familial and generic
circumscriptions and relationships are being re-evaluated by a number of labo-
ratories worldwide. Certain key developments in recent years (e.g. advances in
systematic technologies and tools; ease of international travel, including access to
countries previously unavailable to systematists; horticultural demand for rare
CAB International 2004. Cycad Classification: Concepts and Recommendations
(eds T. Walters and R. Osborne) 1
ch01.qxd 28/11/03 3:42 pm Page 2
2 T. Walters et al.
cycads; recognition of the rare and endangered status of cycads; and an urgency
for cycad conservation in many countries) have collectively stimulated cycad sys-
tematists to try to better understand and manage the taxonomy of the worlds
cycad flora.
Today, field and laboratory equipment can quickly generate massive
amounts of systematic data, which often far surpass the immediate needs of sys-
tematists. This is particularly true for molecular data, which are being analysed
in a multitude of ways with a plethora of user-friendly software programs.
Consequently, systematists sometimes find it difficult to decide which analyses are
appropriate for their work and how to interpret the hundreds of statistical sum-
maries produced from these computer programs. Also, with these wonderful new
opportunities and ever-increasing knowledge of cycads, it is easy to lose sight of
the ultimate mission, which is to provide a universally accepted, consistent and
informative evolutionary classification scheme.
Although scientists do not believe that any truth can be exactly known, it is
the purpose of science to approach truths as closely as possible. Scientists are
forced to perform this work in an unsteady grounding of assumptions. These
assumptions are not self-evident, but arise from the observations and experimen-
tations of previous researchers. So, for any particular scientific field, there is a col-
lection of assumptions, or beliefs, based on previous work that forms the
framework for further discovery. As the scientific method proceeds, these assump-
tions are subject to change, usually in the form of minor modifications but some-
times in the form of radical reassessment. But, whatever insights future inquiry
may bring, current hypotheses and guidelines for future research must be rooted
in presently held assumptions.
The major objectives of this volume are to enumerate the currently held
assumptions, or beliefs, in the field of cycad systematics, and to present guidelines
for future systematic work within the Cycadales. These concepts were fleshed out
during a Cycad Classification Concepts (CCC) Workshop held in April 2002, at
Montgomery Botanical Center in Miami, Florida. The CCC Workshop provid-
ed a forum for cycad systematists to regroup and clarify as a team what they
believe to be true (the best working assumptions) and important in the realm of
cycad systematics. The participants then went a step further, agreeing on a suite
of guidelines that they would follow in support of actualizing the teams beliefs
when engaging in future research (see Osborne and Walters, Chapter 15 this
volume). The participants agreed not only to follow these guidelines in their own
systematic studies, but also to encourage the global use of these guidelines by all
cycad systematists and students.
This chapter attempts to record the beliefs raised by the participants
during the CCC Workshop, whereas the final chapter enumerates the pro-
posed guidelines for developing a useful, evolutionary-based classification
system for cycads. Before presenting the beliefs, it is necessary to provide
some background on cycads and systematics and to explain some terms that the
reader will encounter either in the list of beliefs or in other chapters of this
volume.
ch01.qxd 28/11/03 3:42 pm Page 3
What is a Cycad?
Cycads are an ancient group of seed plants that evolved in the Carboniferous or
early Permian, some 280 million years ago (Norstog and Nicholls, 1997). They
reached their zenith of abundance and diversity in the Mesozoic era. Cycads are
one of four groups (cycads, ginkgos, conifers and gnetophytes) that are collec-
tively and commonly referred to as gymnosperms.
The Cycadales (the order containing all cycad families) is considered to be
monophyletic. A monophyletic group is composed of an ancestor and all of its
descendants based on a suite of shared derived characters, called synapomor-
phies. Some synapomorphies within the Cycadales include girdling leaf traces, a
specialized pattern of vascular bundles in the petiole, distinctive meristems,
buffer cells surrounding the archegonium, and the presence of mucilage canals,
methylazoxymethanol glycosides and the non-protein amino acid BMAA
(-n-methylamino-L-alanine). Coralloid roots (specialized roots that host
cyanobacteria) are found in all cycad taxa. Cycads also bear cataphylls, which are
scale-like leaves that serve to protect the apical meristem.
Cycad reproductive structures typically occur in cones, with each strobilus
consisting of an axis and a series of spirally arranged megasporophylls (leaves
bearing ovules) or microsporophylls (leaves bearing pollen sacs). All cycads are
dioecious, with male and female reproductive structures on separate plants.
Insects appear to be the primary vectors for pollination, although wind may
be a factor for some genera (see discussion by Grobbelaar, 2002). Although not
fully substantiated yet, evidence is accumulating to suggest coevolutionary
processes between cycads and their pollinators. Once these processes are uncov-
ered, resulting data will probably have a significant impact on how cycad taxa are
classified.
All genera except Cycas Linnaeus form a determinate female cone. In Cycas,
the female cones are indeterminate. Ovules are borne on loosely arranged
whorls of megasporophylls (for an interpretive discussion on female cones in
Cycas, see Norstog and Nicholls, 1997). Cycad seeds usually have a brightly
coloured, fleshy outer layer called the sarcotesta that encourages dispersal by
animals. Birds, rodents and probably many other animals disperse cycad seeds by
digesting the sarcotesta and dropping the stony layer and its contents away from
the mother plant (Hill and Osborne, 2001). Seeds of some species of Cycas have
a thick layer of spongy tissue, instead of the usual fleshy layer. This spongy layer
allows these seeds to remain buoyant and viable for long periods of time in salt
water. This may explain the wide distribution of this genus compared with the
narrower ranges of other cycad genera.
With the exception of Cycas, all cycad genera are restricted to single land-
masses (Jones, 2002). Encephalartos Lehmann and Stangeria T. Moore occur only on
the continent of Africa. Bowenia Hooker ex Hooker filius, Lepidozamia Regel and
Macrozamia Miquel are endemic to Australia. Microcycas (Miquel) A. de Candolle
is restricted to the island of Cuba. Ceratozamia Brongniart, Chigua D.W. Stevenson,
Dioon Lindley and Zamia Linnaeus are endemic New World genera. Cycas is found
ch01.qxd 28/11/03 3:42 pm Page 4
4 T. Walters et al.
in subtropical and tropical countries of the Old World that have Pacific or Indian
Ocean coastlines and in neighbouring countries.
Although cycad taxa are widely distributed in subtropical and tropical
regions worldwide, extant populations are often widely disjunct. A cycad popula-
tion is frequently found as an isolated pocket of individuals quite far removed
from other such pockets. A major dilemma that faces todays cycad systematists
is understanding the evolutionary histories and futures of these populations. Part
of the problem is determining whether these populations have been artificially
separated because of human fracturing of the habitat, or are naturally occurring
entities that are either gradually going extinct, are restricted to a very specialized
niche, or are continuing to evolve as separate entities.
Cycad Taxonomy and Systematics
Taxonomy is the process of circumscribing and assigning scientific names to the

diversity of taxa, and then ordering this diversity into an appropriate classifica-
tion system. In the realm of biology, a taxon (plural taxa) is a group of indi-
viduals given a proper name or a group that could be given a proper name. For
example, the taxon Dioon includes all named and as yet unnamed groups of indi-
viduals within this genus. An important aim of the cycad systematist is to describe
and name only natural taxa and to place these in a classification system that rep-
resents the order of nature. A natural taxon is a taxon that exists in nature inde-
pendent of human ability to perceive it. It can be discovered, but not invented
(Wiley, 1981). The same assumption can be applied to the order of nature, i.e.
that it can be discovered for what and how it is. The basic assumption for bio-
logical order is that it is based on reproductive ties (genealogy) as they are affect-
ed by the process(es) of evolution.
Classification is the process of organizing knowledge so that it facilitates
communication and comprehension. The objectives of classification are: (i) to
define and distinguish among kinds; and (ii) to position these kinds in a system
that reflects their natural relationships and imparts information about these
kinds. A classification system is a human construct that attempts to make natural
order comprehensible to the human mind.
The classification of biological organisms has its own language and rules of
language use. For assigning a taxonomic name and having the name recognized
by the botanical community, cycad systematists must follow recommendations
outlined in the most recent edition of the International Code of Botanical
Nomenclature (ICBN; Greuter et al., 2000). These recommendations are built on
a hierarchical system of classification wherein each level of the hierarchy is
referred to as a distinct rank. Typical ranks of use in the field of cycad systemat-
ics start with the all-inclusive order (Cycadales) and move down to increasingly
less inclusive ranks such as family (e.g. Zamiaceae), genus (e.g. Macrozamia),
section (e.g. Parazamia) and species (e.g. Macrozamia lucida). The basic rank of
species holds a special place in terms of the usefulness and importance of bio-
ch01.qxd 28/11/03 3:42 pm Page 5
logical entities to humanity; therefore, most cycad systematists undertake studies

at the species level.
Cycad systematics is the study of the cycad diversity that exists on earth
today and the evolutionary history of this diversity. One of the main objectives is
to convey knowledge about the genealogical relationships among cycad taxa in a
hierarchical system of classification. The field of cycad systematics often requires
that systematists have knowledge from many other scientific disciplines, such as
taxonomy, morphology, ecology, molecular biology, pollination biology, anatomy,
embryology, genetics, physiology, phytochemistry and palaeontology, so we are
better able to uncover the true genealogical relationships among taxa.
The number of described cycad species has almost doubled since 1985.
Today, over 300 species are known (see Hill et al., Appendix 1 this volume) and
many researchers believe the number may reach as many as 400 species when all
potential cycad habitats have been investigated and taxonomic studies have been
completed. Exactly what constitutes a cycad species remains unclear. Defining
what makes a species is not a problem limited to cycad taxonomists, but is a basic
source of consternation throughout the biological world. Generally, delimitation
of a population or suite of populations as a new species is based on the training,
background, knowledge and the basic scientific philosophy of the describer. No
unified concept is in place to guide cycad systematists in defining and circum-
scribing new species.
A variety of species concepts are used throughout the biological world. One
of these, the biological species concept, does not work particularly well with
cycads. The major premise of the biological species concept is that individuals
within a species, when tested, are interfertile, while interspecific individuals are
not. However, clearly defined and widely accepted species within a number of
cycad genera can produce viable offspring with one or more other species in the
same genus (Norstog and Nicholls, 1997). Consequently, cycad systematists gen-
erally agree that interspecific fertility, when tested, is just another character for
systematic studies, and that the character of interspecific sterility should not be
unduly weighted in the determination of cycad species. Moreover, the determi-
nation of the production of fertile offspring from putative hybrids is not practi-
cal for those species, like cycads, with long life cycles.
Another out-of-favour species concept for cycad taxonomists is the phenetic
species concept. This concept defines a species based on the overall similarity of
its individuals combined with a significant gap in variation when these individu-
als are compared with individuals of another species. In practice, this qualitative
approach does not always define natural taxa (Judd et al., 1999).
The CCC Workshop participants agreed that the most common (unstated,
but de facto) species concept in use by cycad systematists is what they termed a
morphogeographic species concept. This concept recognizes the importance of
both morphological characters and geographical isolation in circumscribing a
species. The large geographical disjunctions among cycad populations have
greatly influenced the cycad systematists species concept. These disjunct popu-
lations are viewed as maintaining separate identities and having their own evolu-
ch01.qxd 28/11/03 3:42 pm Page 6
6 T. Walters et al.
tionary tendencies and fates. In this respect, the cycad systematists view is a form
of the evolutionary species concept.
By the end of the CCC Workshop, participants were still not able to agree
on exactly what constitutes a cycad species. This was not surprising, since cycad
lineages have a variety of unique and long histories. Species differ to varying
degrees and, therefore, a single species concept does not work for all cycads.
Today, data from a wide variety of sources, including molecular analyses, ecology,
geography, pollination biology and life history strategies, are providing inde-
pendent measures of the evolutionary reality of existing and proposed species in
the Cycadales.
In contrast with species, circumscriptions of cycad genera are clearly defined
and stable, with the possible exception of Chigua, which may be congeneric within
Zamia (Whitelock, 2002; and see Caputo et al., Chapter 12 this volume). Cycad
genera can usually be identified using gross vegetative features and can always be
identified with gross features of the female reproductive structures.
Family circumscriptions within the Cycadales are still somewhat unclear,
being confounded by the age of the group and the inability of cycad system-
atists to decide on the amount of character differentiation required for
family recognition within the order. Three to four families are typically recog-
nized, with the only uncertainties revolving around the placement of two genera,
Stangeria and Bowenia. Given recent advances in molecular systematics and the
number of laboratories actively studying generic and familial relationships, it is
predicted that a stable familial classification will be available in the very near
future.
Historically, the characters chosen, the importance of specific characters for
differentiating genera and species, and the analyses used for describing new cycad
species have been left to the discretion of each investigator. Vegetative characters,
especially those associated with the leaf, along with characters related to various
aspects of the female reproductive structure, are commonly used for distinguish-
ing taxa. Male cone characters are usually not used for differentiating taxa.
Cycad systematists are well aware of the plasticity of various morphological fea-
tures among plants within a taxon, especially when plants are brought under cul-
tivation. However, the degree of plasticity and the taxonomic importance of this
plasticity continue to remain unclear.
Another ongoing problem for cycad systematists is the lack of a consistent
terminology for describing morphological features that are unique within the
Cycadales. This lack of standardized morphological terminology creates prob-
lems when trying to compare characters in one taxonomic description with those
in another description, or when trying to identify an individual plant based on
specific characters. For this reason, a glossary of terms commonly encountered in
cycad systematics is included in this volume (see Osborne and Walters, Appendix
2 this volume).
ch01.qxd 28/11/03 3:42 pm Page 7
Cycad Phylogenetics Uncovering Genealogical

Relationships
Although never really attainable, systematists work toward producing a natural
classification system that arranges taxa in a way that reflects the natural evolu-
tionary order of the taxa. Since the first basic assumption is that the natural order
is created by the process(es) of evolution, systematists typically strive to produce
an evolutionary classification scheme. More specifically, phylogeneticists aim to
recover the broad genealogical lineages within a group of taxa and to produce a
classification system that reflects these genealogical or phylogenetic relationships.
The starting point in phylogenetic analysis is usually the divergence of a pre-
viously occurring lineage into two or more progeny lineages. The next step is to
reconstruct the separation of these lineages by identifying changes, or modifica-
tions, in characters. A character is a feature having one or more states that can
be described, figured, measured, weighed, counted, scored or otherwise commu-
nicated from one systematist to another. Certain characters are biologically con-
nected to the concept of genealogy and these characters can provide cycad
systematists with justification for group membership in a phylogenetic tree. These
types of characters are called apomorphies. An apomorphic character (some-
times referred to as a specialized character or a derived character) has evolved
directly from its pre-existing homologue (Wiley, 1981).
The task of phylogeneticists is to attempt to discover those characters that
reflect the phylogeny of natural taxa. Because species are considered to be natu-
rally occurring entities, by inference, phylogenetic characters are inherent to
species. A phylogenetic character is one in which its occurrence in two or more
taxa is believed to be the product of descent from a shared ancestor. A phyloge-
netic character shared by two organisms implies a phylogenetic relationship. Of
particular importance is the synapomorphy, which is a genealogically shared,
derived character state that arose in an ancestor of a lineage and is present in all
of that lineages descendants (Hennig, 1966). Synapomorphies are the strongest
evidence for shared ancestry. They are distinguished from symplesiomorphies,
which are earlier character states that are shared by members of a lineage and by
a more ancient ancestor to the lineage. In practice, symplesiomorphic versus
synapomorphic character states for a lineage are determined by comparison with
an outgroup (i.e. a related taxon that is not part of the monophyletic lineage
being examined). The outgroup of choice is the sister group to the lineage,
which is genealogically the closest non-ancestral relative of the lineage. In other
words, two or more taxa are sister groups if they share an ancestor not shared by
any other taxon.
Phylogenetic studies depict results by a graphical representation of the
genealogy of one or more descendants from a common ancestor. Phylogenetic
trees are branching diagrams that portray the hypothesized genealogical rela-
tionships and sequence of historical events linking taxa. A clade within a phylo-
genetic tree incorporates the common ancestor of a group and all of its
descendants.
ch01.qxd 28/11/03 3:42 pm Page 8
8 T. Walters et al.
Cycad systematists use phylogenetic trees to try to produce a phylogenetic

classification that reflects the best estimate of the evolutionary history of cycads.
Construction of a classification based on a phylogenetic tree essentially involves
two steps: (i) the delimitation and naming of groups that are monophyletic in the
tree; and (ii) the ranking of these monophyletic groups and placement of them
into a hierarchical classification system (Wiley, 1981). Phylogenetic studies do not
always lead to a new classification. These studies can provide support to an exist-
ing classification. Also, naming every monophyletic group would become cum-
bersome and in some cases not provide any additional information to the
end-user.
Cycad systematists continue to put forth hypotheses about the genealogical
relationships among taxa in the Cycadales. These hypotheses are tested with evi-
dence derived from a wide variety of sources. Hypotheses and test results are
published, usually peer-reviewed, and evaluated, and some phylogenetic trees are
provisionally chosen over others. In other words, the evolutionary tree for the
Cycadales continues to be tested as additional and new types of data become
available.
An important conclusion made by the CCC Workshop participants was that
the knowledge of genealogical relationships among taxa should be placed in an
unambiguous and stable natural classification system that is useful for a multitude
of end-users and purposes. It is believed that such a system can orient human
understanding of life and the world around us.
CCC Workshop Beliefs
The CCC Workshop participants enumerated their beliefs concerning cycad

classification during the second day of the Workshop. Clarity and consensus with
regard to these beliefs were needed so that the participants could go on to
produce a final set of guidelines for future research aimed at establishing a suit-
able classification scheme for cycads (see Osborne and Walters, Chapter 15 this
volume).
For purposes associated with the production of this chapter, the suite of
beliefs generated during the work sessions at the CCC Workshop has been reword-
ed and organized to provide consistency in wording, style and format. The authors
have organized the beliefs under the following three categories: (i) beliefs about
biological relationships; (ii) beliefs about what systematists can and should do in
order to understand biological relationships; and (iii) beliefs about what cycad sys-
tematists can and should do in order to understand relationships in the Cycadales.
Beliefs about biological relationships
We believe there is value in the biological world.

We believe there is a natural order to the biological world.
ch01.qxd 28/11/03 3:42 pm Page 9
We believe that the natural order is based on genealogical relationships.

We believe that the pattern of genealogical relationships naturally produces a
hierarchical structure of lineages.
We believe that each species (as a natural group) is a monophyletic lineage that
evolves independently of other such lineages.
Beliefs about what plant systematists can and should do in order to

understand biological relationships
We believe we can construct hypotheses that are testable, and that the process
of testing and refining hypotheses leads to a better understanding of the
natural world.
We believe that genealogical relationships can be recovered through hypothe-
sis testing.
We believe that as technology, resources and data increase and change, we will
be better able to construct a classification scheme that approximates true
genealogical relationships.
We believe that we should construct hierarchical classification schemes that
best reflect actual genealogical relationships. Such schemes have greater pre-
dictive power, have greater heuristic value, and improve our ability to under-
stand and communicate about the biological world as it existed, as it exists, and
as it may exist in the future.
We believe that the process of refining classification schemes brings us closer
to approximating true genealogical relationships and therefore converges
towards stability of the classification.
We believe that the most important evolutionary entity to define and circum-
scribe is the species.
We believe that species are not evolutionarily static (i.e. they change through
time).
We believe that species can be difficult to recognize, and, therefore, the defini-
tions and circumscriptions that we apply to particular species are hypotheses to
be tested.
We believe that the exploration of species and species concepts will provide the
common language for understanding speciation, species interactions and plant
systematics.
We recognize the existence of the International Code of Botanical
Nomenclature (ICBN) and support the beliefs, philosophies, and principles of
the Code to provide one correct name for each taxonomic group within a
stable classification system.
We believe that systematists should share information through the publication
of data and analyses.
ch01.qxd 28/11/03 3:42 pm Page 10
10 T. Walters et al.
Beliefs about what cycad systematists can and should do in order to

understand relationships in the Cycadales
We believe that the Cycadales forms a distinct monophyletic lineage, and

genealogical relationships within this lineage can be inferred through the col-
lection and analysis of data.
We believe that higher ranks in the Cycadales (e.g. genera and above) are easily
recognizable and definable.
We believe that the greatest challenge in cycad systematics is recognizing
appropriate units to call species.
We believe that there are differing opinions concerning cycad species defini-
tions and circumscriptions.
Given the uncertainty of species definitions and the lack of infraspecific data
on cycads, we believe that it is not yet appropriate to try to define and identify
relationships of taxa below the species level.
We believe that to better understand cycad species, we must concentrate our
resources on variation and relationships at the population level.
We believe that there is a wealth of available data on cycads that still must be
captured and analysed.
We believe that a classification system should be valuable for a variety of
known and unknown end-users and purposes.
We believe that the extinction of cycad species is accelerating and that access
to native populations is decreasing rapidly. Actions must be undertaken imme-
diately to describe, classify, conserve and preserve species for continuing scien-
tific studies.
We believe that the process of understanding cycad systematics should be a
collaborative endeavour.
Acknowledgements
The Cycad Classification Concepts Participants spent many hours developing
and discussing the suite of beliefs presented in this chapter. Their commitment to
and support of undertaking this long and arduous task is greatly appreciated.
The authors are deeply indebted to Deena Walters for her critical comments on
early drafts of this manuscript.
References
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. and Hawksworth,
D.L. (2000) International Code of Botanical Nomenclature (Saint Louis Code). Koeltz
Scientific Books, Kningstein, Germany, 474 pp.
ch01.qxd 28/11/03 3:42 pm Page 11
Grobbelaar, N. (2002) Cycads with Special Reference to the Southern African Species. Published
by the author, Pretoria, South Africa, 331 pp.
Hennig, W. (1966) Phylogenetic Systematics. University of Illinois Press, Champaign-Urbana,
Illinois, 263 pp.
Hill, K. and Osborne, R. (2001) Cycads of Australia. Kangaroo Press, New South Wales,
Australia, 116 pp.
Jones, D.L. (2002) Cycads of the World Ancient Plants in Todays Landscape, 2nd edn. Reed
New Holland, Sydney, Australia, 456 pp.
Judd, W.S., Campbell, C.S., Kellogg, E.A. and Stevens, P.F. (1999) Plant Systematics: a
Phylogenetic Approach. Sinauer Associates, Inc., Sunderland, Massachusetts, 464 pp.
Norstog, K.J. and Nicholls, T.J. (1997) The Biology of the Cycads. Cornell University Press,
Ithaca, New York, 363 pp.
Whitelock, L.M. (2002) The Cycads. Timber Press, Portland, Oregon, 374 pp.
Wiley, E.O. (1981) Phylogenetics: the Theory and Practice of Phylogenetic Systematics. John Wiley &
Sons, New York, 439 pp.
ch01.qxd 28/11/03 3:42 pm Page 12
ch02.qxd 28/11/03 3:42 pm Page 13
Saving Ghosts? The 2

Implications of Taxonomic
Uncertainty and Shifting
Infrageneric Concepts in the
Cycadales for Red Listing and
Conservation Planning
John Donaldson
Kirstenbosch Research Centre, National Botanical Institute,

Claremont, South Africa
Abstract
A comparison of the cycad Red Lists from 1978 to 2002 shows that taxonomy has had a
profound influence on the outcomes of the Red List process. The descriptions of new
species, as well as taxonomic revisions, have increased the number of recognized cycad
taxa from 136 in 1978, to over 300 at the time of publication. During this time, the pro-
portion of threatened taxa has fluctuated from a low of 46% of all cycad taxa in 1978, to
a high of 82% in 1997, and is currently estimated to be 52%. At least one-third of the
changes in the Red List are due to taxonomic changes, which reflects an increase in taxo-
nomic activity between 1978 and 2002. In addition to new species, there were 48 changes
in the Red List between 1978 and 2002 as a result of uncertainty about the infrageneric
status of cycad taxa. Frequent changes in threatened status are not helpful for conserva-
tion planning and could even undermine the Red List process. The IUCN has introduced
robust criteria for Red Listing to deal with inadequate population data and ecological
uncertainty. The results of this analysis show that cycad taxonomists need to develop con-
sistent and widely accepted concepts for infrageneric taxa to reduce the influence of tax-
onomic uncertainty on the Red Listing process.
Introduction
Ongoing assessment of the threatened status of the worlds cycad flora (Red
Listing) is an essential process for conservation planning and actions (Osborne,
ch02.qxd 28/11/03 3:42 pm Page 14
14 J. Donaldson
1995; Donaldson, 2003). Having a clear picture of which cycads are most threat-
ened with extinction, and where they are located, enables conservation agencies
and funding bodies to allocate resources to the most threatened taxa or to the
most effective conservation actions.
Red Listing and conservation planning are dynamic processes. Typically,
Red Listing assessments are based initially on inadequate information and
they are modified as more information becomes available. As a result,
changes in threatened status are unavoidable, and they can come about for
several reasons:
Actual changes in plant distribution or abundance due to ongoing threats or to
the success of conservation actions.
The discovery of new populations or additional demographic data that
provide more accurate estimates of population numbers and trends.
Changes in the criteria used for assigning threatened status.
The discovery of previously unknown species.
Taxonomic revisions that result in the splitting of existing species into several
taxa or the lumping of several described species into one species.
Frequent changes in Red List status are not helpful for conservation planning and
they may even undermine the credibility of Red Listing as a conservation tool.
The aim of Red Listing should be to create a reasonably stable and scientifically
based system that can portray real changes in the distribution and abundance of
threatened species as well as the relative status of different species. To achieve
such a system, the IUCN introduced revised categories of threat and more rig-
orous criteria for assessing threatened status (IUCN, 1994; IUCN/SSC Criteria
Review Working Group, 1999; Mace, 2000). The revised system provides a
framework for dealing with the uncertainty that arises from using inadequate
ecological data to assess threatened status. However, the IUCN system
does not deal specifically with an important area of uncertainty in threat assess-
ments, i.e. the taxonomic status of threatened species. Clearly, a credible Red List
relies on a consistent and widely accepted system of classification. Within the
Cycadales, there is still considerable uncertainty about what a taxon is, especial-
ly at the infrageneric level, and this problem has important implications for cycad
Red Listing.
This chapter examines changes in the threatened status of cycad taxa from
1978 to 2002 to show how taxonomy influences the outcome of Red Listing
assessments and conservation planning. During this time there have been five
Red Listing accounts of the world cycad flora, starting with the list drawn up by
the IUCN Threatened Plant Unit and based on the criteria outlined in The IUCN
Plant Red Data Book (Lucas and Synge, 1978). The initial list was followed by
Gilbert (1984), Osborne (1995), the 1997 IUCN Red List of Threatened Plants
(Walter and Gillett, 1998) and the Cycad Action Plan (Donaldson, 2003). These
accounts provide a basis for evaluating changes in threatened status over a 24-
year period and for determining the relative contributions of taxonomic uncer-
tainty and ecological uncertainty to these changes.
ch02.qxd 28/11/03 3:42 pm Page 15
Red Listing and Conservation Planning 15
Taxonomy and the Threatened Status of Cycads
In the first cycad Red List, based on The IUCN Plant Red Data Book (Lucas and
Synge, 1978), there were 136 recognized cycad taxa (species and subspecies),
comprising one Extinct species (Encephalartos woodii Sander), 63 Threatened taxa
(46%), 45 taxa of Unknown threatened status (33%) and only 15 taxa (11%) that
were Not Threatened (Fig. 2.1). In 1984, the number of cycad taxa had risen to
168, comprising one Extinct species, 65 Threatened taxa (38.6%), 86 taxa of
Unknown status (51%) and 16 taxa (9.5%) that were classified as Not Threatened
(Gilbert, 1984). Eleven years later, using the same criteria, Osborne (1995) rec-
ognized 197 taxa, comprising one Extinct species, 124 Threatened taxa (63%),
42 taxa that could not be assessed (21%) and 30 taxa that were Not Threatened
(15%) (Fig. 2.1). In the last assessment using these criteria, Walter and Gillett
(1997) recognized only 180 taxa, comprising three Extinct species, 147
Threatened taxa (82%), three taxa of Unknown status (1.6%) and 27 taxa that
were Not Threatened (15%) (Fig. 2.1). Finally, the Cycad Action Plan
Not Threatened
Fig. 2.1. The number of cycad taxa (species, subspecies and undescribed species)
classified as Extinct, Threatened, Not Threatened or Unknown, according to the
IUCN Threatened Plant Unit database (1978), Gilbert (1984), Osborne (1995), the
1997 IUCN Red List of Threatened Plants (Walter and Gillett, 1998) and the 2002
assessments for the Cycad Action Plan (Donaldson, 2003). The total height of each
bar represents the total number of recognized taxa. Taxa of unknown conservation
status include those classified as Indeterminate, Unknown and Data Deficient.
ch02.qxd 28/11/03 3:42 pm Page 16
16 J. Donaldson
(Donaldson, 2003), based on the latest revised IUCN criteria (Mace, 2000),
included 297 taxa, comprising two species that are Extinct in the Wild (no
Extinct species), 154 Threatened taxa (52%), 18 taxa that are Data Deficient
(6%) and a further 123 taxa (41%) that are not classified as threatened (Fig. 2.1).
Clearly there are substantial differences between these assessments and it is nec-
essary to examine these differences in greater detail to determine how taxonom-
ic uncertainty has influenced the outcome of the Red Listing process.
The substantial increase in the number of recognized cycad taxa (at the
species and subspecies level) between 1977 and 2002 (Fig. 2.1) has obviously
influenced the overall threatened status of the worlds cycads. Even at this level,
there is a degree of confusion about what constitutes a valid taxon for Red List
evaluation. Osborne (1995) recognized 197 taxa whereas Walter and Gillett
(1998) recognized only 180 taxa. The taxa omitted by Walter and Gillett (1998)
comprised six species of Cycas Linnaeus, both species of Bowenia Hooker ex
Hooker filius, and nine taxa in the Zamiaceae, including one undescribed species
of Encephalartos Lehmann. Walter and Gillett (1998) also included Cycas celebica
Miquel, which is now considered synonymous with C. rumphii Miquel (Hill et al.,
2003).
Species classified as Extinct (pre-1994 categories) or Extinct in the Wild
(IUCN, 1994) provide a useful starting point for exploring the influence of taxo-
nomic uncertainty on Red Listing. Until 1995, only Encephalartos woodii was clas-
sified as an Extinct species and this status has been consistent in later assessments.
Walter and Gillett (1998) also included Cycas szechuanensis C.Y. Cheng, W.C.
Cheng & L.K. Fu and Zamia monticola Chamberlain as Extinct taxa. Cycas szechua-
nensis was first classified as Extinct because it was known only in cultivation. It was
later thought to be conspecific with the more widespread C. guizhouensis K.M. Lan
& R.F. Zou (Hill et al., 2003), which would warrant a downlisting to Near
Threatened. However, it is now regarded as a valid species (Chen, 2000), but it
no longer warrants a status of Extinct in the Wild because two wild populations
have been discovered. In this case, changes in both taxonomic interpretation and
new discoveries of wild populations have influenced the Red List status. In con-
trast, the taxonomic status of Zamia monticola has not been in doubt and the
species was downgraded to Critically Endangered (Stevenson et al., 2003) after
the discovery of new populations.
The Cycad Action Plan (Donaldson, 2003) also lists Encephalartos relictus P.J.H.
Hurter as Extinct in the Wild. This species was described from a cultivated spec-
imen. Two undescribed taxa that were classified as Extinct (herbarium specimen)
and Extinct in the Wild (cultivated specimen) in an assessment of African cycads
by J. Golding and P.J.H. Hurter (unpublished results) were not included in the
Cycad Action Plan, due to their uncertain taxonomic status. A clear and consis-
tent approach to infrageneric delimitations would help to resolve the Red List
status of these taxa.
Some of the same problems recur with other categories of threat (Fig. 2.2).
The overall change in the number of threatened taxa shows a gradual increase
in threatened taxa from 1978 to 1997 using the pre-1994 criteria. Based on the
ch02.qxd 28/11/03 3:42 pm Page 17
Pre-1994 Rare Post-1994 Vulnerable

Pre-1994 Vulnerable Post-1994 Endangered
Pre-1994 Endangered Post-1994 Critical
180
160
140
120
Number of taxa
100
80
60
40
20
0
1978 1984 1995 1997 1995 2002
Year of assessment
Fig. 2.2. The number of cycad taxa (species, subspecies and undescribed species)
assigned to IUCN threatened categories according to the IUCN Threatened Plant
Unit database (1978), Gilbert (1984), Osborne (1995), the 1997 IUCN Red List of
Threatened Plants (Walter and Gillett, 1998) and the 2002 assessments for the
Cycad Action Plan (Donaldson, 2003). Assessments for 1978 and 1997 used the
old IUCN (pre-1994) criteria whereas the 2002 assessment used the criteria
introduced after 1994. Osborne (1995) used both sets of criteria to evaluate cycad
species.
post-1994 criteria, there is also an increase in the number of threatened taxa

from Osborne (1995) to the Cycad Action Plan in 2002 (Donaldson, 2003) (Fig.
2.2). The low number of taxa of unknown status in the later assessments (Fig. 2.1)
indicates that there is increasing ecological information on threatened taxa so
that changes in the number of taxa assigned to different threatened categories
may be due to better ecological information. However, an analysis of changes
due to better ecological information (distribution and abundance data) compared
with those due to taxonomic changes (new species, change in infrageneric status)
(Fig. 2.3) shows that taxonomy accounts for a large number of changes in the
number of threatened cycads.
Clearly, many changes in the number of threatened species could arise from
ch02.qxd 28/11/03 3:42 pm Page 18
18 J. Donaldson
as
ia
a
nia
s
on
ia
a
ca
r to
eri
igu
mi
mi
am
am
cyc
Dio
we
Cy
ng
za
Za
ala
Ch
oz
oz
cro
Bo
cro
Sta
ph
pid
rat
Mi
Ma
ce
Ce
Le
En
ia
a
nia
ia
s
eri
a
a
a
on
r to
cyc
ca
am
mi
igu
mi
am
ng
Dio
we
ala
Cy
Za
za
oz
Ch
cro
oz
Sta
Bo
cro
ph
rat
pid
Mi
ce
Ma
Ce
Le
En
Fig. 2.3. The number of cycad taxa where a change in threatened status was
recorded between assessments. (A) Changes between the TPU Red List based on
Lucas and Synge (1978) and Osborne (1995) using the pre-1994 IUCN criteria. (B)
Changes between the 1997 IUCN Red List of Threatened Plants (Walter and
Gillett, 1998) and the 2002 assessments for the Cycad Action Plan (Donaldson,
2003). Changes that resulted from taxonomic revisions (new species, split taxa
and combined taxa) are represented by black bars, and those that resulted from
more complete ecological data are represented by white bars.
ch02.qxd 28/11/03 3:42 pm Page 19
the description of new species as a result of increased field work in areas such as
Australia, Central Africa, Mexico and South-East Asia between 1978 and 2002.
It is therefore important to distinguish between the influence of new taxa on Red
List assessments, compared with the influence of other changes that reflect
uncertainty about the infrageneric status of cycad taxa. The number of taxa
included in one Red List assessment that were no longer considered to be valid
taxa in the following assessment (Fig. 2.4A) can be compared with the number of
new taxa that were included in each Red List assessment (Fig. 2.4B). The results
show that a considerable number of changes could be attributed to the descrip-
tion of new species (Fig. 2.4B), especially in Cycas, but also in Encephalartos,
Macrozamia Miquel and Zamia Linnaeus. However, a large number of changes
also occurred because of changes in the infrageneric status of described species,
especially in Zamia, Cycas and Macrozamia. The data indicate that there were fewer
changes due to uncertain taxonomic status between 1997 and 2002 than between
earlier assessments (Fig. 2.4A), suggesting that the species limits of cycad taxa are
better understood now than they were in 1978.
The data summarized in Figs 2.12.4 do not reveal the many subtle influ-
ences that taxonomic changes have on cycad Red Listing. The description of new
species and changes in species concepts can have different outcomes on Red
Listing depending on the genus under review. For example, the delimitation of
new species and subspecies of Cycas resulted in an increase in recognized taxa
from eight in 1978 to 98 in 2002, while the number of threatened taxa increased
from four in 1978 (50%) to 38 in 2002 (39%). As a result, there has been a sub-
stantial decrease in the proportion of Cycas species that are listed as threatened
with extinction. In Africa, the delimitation of new species and subspecies of
Encephalartos over the same period resulted in an increase from 48 taxa in 1978 to
67 taxa in 2002, with an increase in threatened taxa from 32 (66%) in 1978 to 46
(68%) in 2002. In the case of Cycas, many new species were discovered and
described in areas where cycads are still abundant. In contrast, several new
species of Encephalartos were described as a result of the revision of already threat-
ened taxa, such as the split of E. eugene-maraisii I. Verdoorn into several taxa, all
classified as threatened.
The influence of higher-level classification on conservation planning
The focus of this chapter is primarily on what happens at the infrageneric level,
but it is worth noting that taxonomic changes at higher levels also affect conser-
vation planning. Vane-Wright et al. (1991) argued that systematics should be used
as a criterion for prioritizing conservation actions. Their reasoning was that,
given the high number of species that are threatened with extinction, priority
should be given to those species that represent more threatened lineages (repre-
sented by higher taxonomic groups). For cycads, the genus Chigua D.W. Stevenson
comprises two Critically Endangered species (Stevenson et al., 2003) and repre-
sents the second most threatened genus within the Cycadales [after Microcycas
ch02.qxd 28/11/03 3:42 pm Page 20
20 J. Donaldson
s
ia
on
a
s
a
ca
r to
mi
mi
am
Dio
Cy
za
Za
ala
oz
cro
ph
rat
Ma
ce
Ce
En
a
s
on
a
ia
r to
mi
ca
mi
am
Dio
Za
Cy
ala
za
oz
cro
ph
rat
ce
Ma
Ce
En
Fig. 2.4. Changes in recognized cycad taxa between different Red List
assessments, from Gilbert (1984) to Osborne (1995), from 1995 to the 1997 IUCN
Red List of Threatened Plants (Walter and Gillett, 1998), and from 1997 to 2002
(Donaldson, 2003). (A) The number of taxa listed on the first date but not
recognized as valid taxa when the next list was published. (B) The number of taxa
listed in the later assessment that were not recognized as distinct taxa in the
earlier assessment. Only the major genera are included, as there were no
differences in the genera Bowenia, Lepidozamia, Microcycas and Stangeria.
ch02.qxd 28/11/03 3:42 pm Page 21
(Miquel) A. de Candolle]. Based on Vane-Wright et al. (1991), the two species of

Chigua should receive higher conservation priority than equally threatened
species in larger genera (Walters, 2003). However, if Chigua is viewed as part of
the genus Zamia (sensu Schutzman and Dehgan, 1993), then the two species cur-
rently included in Chigua become part of a much larger genus and therefore have
lower conservation priority.
Conclusions
A comparison of the cycad Red Lists from 1978 to 2002 shows that taxonomy
has had a profound influence on the outcomes of the Red List process. By far the
greatest contribution has been from the description of new taxa, due to a resur-
gence of interest in cycad taxonomy within the past 25 years. Greater taxonom-
ic activity has also resulted in revised species concepts and the sorting out of
nomenclatural problems so that many early names are no longer recognized as
valid species. In both cases, the changes reflect developments within the science
of cycad taxonomy that are a positive contribution to our knowledge of the
Cycadales. These changes do, however, influence the process of Red Listing and
it is essential to ensure that taxonomic changes are both necessary and consistent.
To achieve this, taxonomists need to agree on concepts used to delimit infra-
generic taxa and then apply these concepts consistently.
References
Chen, C.J. (2000) Cycadaceae. In: Fu, L.K. et al. (eds) Higher Plants of China, Vol. 3.
Qingdao Press, Qingdao, China, pp. 111.
Donaldson, J.S. (ed.) (2003) Cycads. Status Survey and Conservation Action Plan. IUCN/SSC
Cycad Specialist Group, IUCN, Gland, Switzerland and Cambridge, UK,
ix + 86 pp.
Gilbert, S.G. (1984) Cycads: Status, Trade, Exploitation and Protection 19771982. TRAFFIC,
Washington, DC, 74 pp.
Hill, K.D., Chen, C.J. and Loc, P.K. (2003) Regional overview: Asia. In: Donaldson, J.S.
(ed.) Cycads. Status Survey and Conservation Action Plan. SSC Cycad Specialist Group,
IUCN, Gland, Switzerland and Cambridge, UK, 2530 pp.
IUCN (1994) IUCN Red List Categories. Prepared by the IUCN Species Survival
Commission, IUCN, Gland, Switzerland, 22 pp.
IUCN/SSC Criteria Review Working Group (1999) IUCN Red List Criteria review pro-
visional report: draft of the proposed changes and recommendations. Species 3132,
4357.
Lucas, G. and Synge, H. (1978) The IUCN Plant Red Data Book. IUCN Threatened Plants
Committee, Kew, UK, 540 pp.
Mace, G. (2000) Summary of the results of the review of IUCN Red List categories and
criteria 19962000. In: Hilton-Taylor, C. (compiler) 2000 Red List of Threatened Species.
IUCN, Gland, Switzerland and Cambridge, UK, pp. 5761.
ch02.qxd 28/11/03 3:42 pm Page 22
22 J. Donaldson
Osborne, R. (1995) The world cycad census and a proposed revision of the threatened
species status for cycad taxa. Biological Conservation 71, 112.
Schutzman, B. and Dehgan, B. (1993) Computer assisted systematics in the Cycadales. In:
Stevenson, D.W. and Norstog, K.J. (eds) The Biology, Structure, and Systematics of the
Cycadales. Proceedings of the Second International Conference on Cycad Biology. Palm & Cycad
Societies of Australia Ltd, Milton, Queensland, Australia, pp. 281289.
Stevenson, D.W., Vovides, A. and Chemnick, J. (2003) Regional overview: New World. In:
Donaldson, J.S. (ed.) Cycads. Status Survey and Conservation Action Plan. SSC Cycad
Specialist Group, IUCN, Gland, Switzerland and Cambridge, UK, 3138 pp.
Vane-Wright, R.I., Humphries, C.J. and Williams, P.H. (1991) What to protect?
Systematics and the agony of choice. Biological Conservation 55, 235254.
Walter, K.S. and Gillett, H.J. (eds) (1998) 1997 IUCN Red List of Threatened Plants. Compiled
by the World Conservation Monitoring Centre, IUCN, Gland, Switzerland and
Cambridge, UK, 862 pp.
Walters, T. (2003) Off-site collections. In: Donaldson, J.S. (ed.) Cycads. Status Survey and
Conservation Action Plan. SSC Cycad Specialist Group, IUCN, Gland, Switzerland and
Cambridge, UK, 4853 pp.
ch03.qxd 28/11/03 3:43 pm Page 23
Character Evolution, Species 3

Recognition and Classification
Concepts in the Cycadaceae
Ken D. Hill
Royal Botanic Gardens, Sydney, Australia
Abstract
A number of systems of infrageneric classification of the family Cycadaceae have been

presented by different authors. No two systems have been the same in structure or species
recognition, and some have been strikingly discordant. Phylogenetic analysis of combined
morphological and molecular datasets has yielded a cladogram with good resolution and
support on many deeper branches. Plotting characters used by different authors in devel-
oping infrageneric systems of classification allows an independent assessment of the value
of these characters and of the systems of classification derived from them. No classifica-
tion system presented to date is wholly concordant with the results of the phylogenetic
studies, and a number of characters previously heavily relied upon in defining groups are
shown to be highly plastic. In particular, the recently described segregate genus Epicycas is
shown to be polyphyletic, as are the four subgenera of Cycas erected by the same authors.
The analysis supports the recognition of a single genus with five sections, although a
number of taxa are insufficiently known to be clearly placed. Species recognition and
species concepts are discussed.
Introduction
The genus Cycas Linnaeus has long been accepted as the single constituent genus
of the family Cycadaceae, itself the basal divergence within the extant Cycadales
or Cycadophyta (Johnson, 1959; Stevenson, 1992). The genus (and family) has a
present-day distribution concentrated in a zone between northern Australia and
southern China, and extending westwards to Madagascar, the Comoros and the
adjacent African mainland, and eastwards to Tonga. Many regional taxonomic
ch03.qxd 28/11/03 3:43 pm Page 24
24 K.D. Hill
treatments have been published over the years, but all suffer to some extent from
the lack of an up-to-date comprehensive monographic treatment (e.g. Backer and
Bakhuizen van den Brink, 1963; Smitinand, 1971; Fu et al., 1978; Hip and
Vidal, 1996; Chen and Stevenson, 1999). The comprehensive treatment by
Schuster (1932) is generally acknowledged to have created more problems than it
solved (Johnson, 1959), and the most recent overall treatment by de Laubenfels
and Adema (1998), wherein a second genus is erected, has proved controversial
and has not been widely accepted (see Chen et al., Chapter 5 this volume). The
different regional and comprehensive treatments also differ markedly in species
recognition and circumscription.
Analysis of combined morphological and molecular data yields a resolved
cladogram with good support on many deeper branches (Fig. 3.1). Plotting char-
acters used by different authors in developing infrageneric systems of classifica-
tion allows an independent assessment of the value of these characters and of the
systems of classification derived from them.
Historical background
A number of authors have attempted to divide the genus Cycas internally and
present systems of infrageneric classification. These are most notable in their dis-
cordancy. This is partly an indication that the known species are closely related
and form a coherent group, but also a reflection of the poor understanding of
specific limits and relationships within the group. The first attempt to subdivide
the genus was that of Miquel, who firstly separated C. revoluta Thunberg from all
other species on the basis of the revolute leaflet margins (Miquel, 1843), and later
recognized two informal groups separated by tomentose vs. glabrous ovules
(Miquel, 1861). He was followed by de Candolle (1868). Other early authors rec-
ognized two informal groups on the basis of the degree of division of the lamina
of the megasporophyll (Carruthers, 1893). Warburg (1900) divided the genus,
again informally, on the basis of the number of ovules per megasporophyll.
Warburg was followed by Pilger (1926). Schuster (1932) attempted the first formal
infrageneric classification of Cycas (discussed below), recognizing three major
groups as sections. A second formal system was presented by Smitinand (1971),
with a very different major division into two sections. Dehgan and Dehgan (1988)
published an informal infrageneric classification based on seed structure and
pollen morphology that was completely different from either of the above, rec-
ognizing two subgenera.
Hill (1994, 1996) recognized four sections, each with two or three subsec-
tions, in part merging the systems of Schuster and Smitinand. Wang (1996)
erected two subgenera and within these a number of sections, subsections and
series, many corresponding in circumscription, but not rank or placement, to
taxa recognized by Hill.
De Laubenfels and Adema (1998) presented an entirely different system, sep-
arating the new genus Epicycas de Laubenfels and dividing the remainder of Cycas
ch03.qxd 28/11/03 3:43 pm Page 25
Classification Concepts in the Cycadaceae 25
C. wadei
C. curranii
C. revoluta
C. panzihuaensis
C. changjiangensis
C. guizhouensis
C. sexseminifera
C. bifida
C. diannanensis
C. pectinata
C. clivicola
C. lindstromii
C. pranburiensis
C. condaoensis
C. siamensis
C. inermis
C. scratchleyana
C. beddomei
C. circinalis
C. spherica
C. riuminiana
C. thouarsii
C. rumphii
C. micronesica
C. bougainvilleana
C. seemannii
C. apoa
C. tuckeri
C. yorkiana
C. cairnsiana
C. ophiolitica
C. maconochiei
C. calcicola
C. furfuracea
C. silvestris
C. media
C. papuana
C. armstrongii
Fig. 3.1. Combined molecular and morphological analysis for species of Cycas
(strict consensus tree). Bootstrap support values shown above branches. From Hill
(1999 and in preparation).
into four subgenera. The classification systems of Schuster and subsequent

authors are examined in the light of recent cladistic studies below.
Species concepts
Species definition in Cycadaceae is complicated by the variability of some of the
characters that have been traditionally used to separate taxa and by the inade-
ch03.qxd 28/11/03 3:43 pm Page 26
26 K.D. Hill
quacy of specimens and recorded observations in representing the differences

and similarities occurring within and between populations and taxa. For
example, development of spines on petioles is often variable within populations,
and changes with age in most taxa. Microsporophylls vary considerably in size
and shape from base to apex of the cone; in particular, the apical spine is reduced
or absent on the lowermost sporophylls, gradually increasing in size towards the
apex or towards the centre of the cone (Amoroso, 1986). Megasporophylls also
vary, often widely, in size, shape of lamina and number of ovules. This can
depend on their position within a growth flush, the first and last produced often
being markedly smaller and less elaborate than those in the centre.
Because herbarium collections are frequently fragmentary, incomplete and
often sterile, many characters are not represented or recorded. Comprehensive
field study is thus essential to understanding the ranges of variation of characters
that may distinguish taxa. An example is in the recent segregation of seven taxa
in Cape York Peninsula, Australia, from what had long been accepted as a single
species (Hill, 1995). Many of the characters that discriminate these taxa were
unrepresented in specimens or notes, and it was only by comprehensive and sys-
tematic field observation that the totality of morphological variation could be
observed. This is achieved in a rigorous and reproducible manner by following a
standard pro forma (Fig. 3.2) and making a standard set and number of obser-
vations on each population. Data recorded can then be statistically analysed if
required, and groups can be defined. This synthesis can then lead to satisfactory
species recognition, following which identification keys and classifications can be
developed (Table 3.1). In the Cape York example, the cataphyll characters criti-
cal in discriminating taxa had not been previously recorded. It is also of note that
these taxa occur in discrete, geographically separated populations.
The genus Cycas in general shows a similar geographical replacement
pattern throughout its range, often with many closely related entities. Although
these taxa are often similar in many respects, the homoplasy evident in the defin-
ing characters does not always allow unequivocal aggregation into groups that
could be treated as species with subordinate infraspecific taxa. In order to satis-
factorily separate and recognize groups of populations that show real, albeit
sometimes small differences, it is strongly recommended that these groups be
treated as distinct species. This rather narrow view of species is considered
preferable to the arbitrary submerging of these recognizably distinct and true-
breeding groups of populations into broader and less meaningful species and
the consequent loss of information on the real diversity of these plants.
Phylogenetic studies
Phylogenetic analyses of the family Cycadaceae sensu stricto have been conducted
using sets of data obtained from DNA molecular sequences and from morpho-
logical and anatomical characters. Sequence data of the internal transcribed
spacer region from the nuclear genome for 40 terminal taxa were analysed (Hill,
ch03.qxd 28/11/03 3:43 pm Page 27
Fig. 3.2. Pro forma used for rigorous systematic observations in populations of
Cycas.
ch03.qxd 28/11/03 3:43 pm Page 28
28 K.D. Hill
Table 3.1. Key to the species of Cycas in Cape York Peninsula, Australia.
Characters observed in field studies are shown in bold.
1 Cataphylls hard, strongly spinescent 2

1 Cataphylls soft, not spinescent 4
2 Hypodermis absent; leaf margins flat; leaflet midrib
equally raised above and below C. xipholepis K.D. Hill
2 Hypodermis present; leaf margins recurved; leaflet midrib almost flat above,
strongly raised below 3
3 Pinnae > 9 mm wide; cataphylls deciduous C. silvestris K.D. Hill
3 Pinnae < 9 mm wide; cataphylls persistent C. media R. Brown
4 Crown and cataphylls densely orange-woolly
or floccose 5
4 Crown and cataphylls very shortly grey to
white-sericeus 6
5 Megasporophyll apex 60100 mm long, with 2432
lateral spines 36 mm long, apical spine
216 mm long C. yorkiana K.D. Hill
5 Megasporophyll apex 4055 mm long, with 1624
lateral spines 14 mm long, apical spine
2125 mm long C. badensis K.D. Hill
6 Megasporophylls short (1013 cm); apex dilated
(3550 mm wide, 50 65 mm long) C. tuckeri K.D. Hill
6 Megasporophylls longer (1522 cm); apex not
dilated (1835 mm wide, 6075 mm long) C. semota K.D. Hill
in preparation), and morphological and anatomical data were taken for the same
40 taxa from a data set used in morphological studies (Hill, 1999 and in prepa-
ration). The combined data yielded a result that was not completely consistent
with any published morphological studies or recent taxonomic classifications (Fig.
3.1). The resultant cladogram from these studies will be taken as a basis for char-
acter analysis below.
Character Analysis
Previously published classifications have been examined and characters used to
define groups recorded (Table 3.2). These were then plotted on to the cladogram
previously obtained from analysis of combined morphological and molecular
data (Fig. 3.1). Consistency indices (CI) are assigned to each character as a
measure of the degree of homoplasy shown by that character, and also as a
measure of the reliability and efficacy of that character in defining and recog-
nizing natural groups.
Most of the characters that have been widely used in infrageneric classifica-
tion in the past are shown to be useful, although character polarities were not
always as previously assumed. Each character state is discussed individually below.
ch03.qxd 28/11/03 3:43 pm Page 29
Table 3.2. Characters used by different workers in defining groups in Cycas.
Outgroup
Character States conditions Reference
Ovules Glabrous/tomentose Glabrous Miquel, 1861;

Schuster, 1932
Sclerotesta Smooth/striate/ribbed/ Smooth/striate Hill, 1996;
verrucose/crested Wang, 1996
Spongy Absent/present Absent Deghan and Yuen,
megagametophytic 1983; Hill, 1994
tissue
Fibrous sarcotesta Absent/present Absent Hill, 1994; Wang,
1996
Ovule number 114 2 Warburg, 1900
Seed cone Closed/open Not comparable Wang, 1996
Megasporophyll Pectinate/dentate/ Entire Schuster, 1932
apex entire
Megasporophyll Long/short Not comparable Schuster, 1932
length
Pollen cones Soft/rigid Soft/rigid Smitinand, 1971;
Hill, 1996, 1999
Pollen cones Cylindrical/ovoid Cylindrical/ovoid de Laubenfels and
Adema, 1998
Microsporophyll Attenuate/truncate Not comparable de Laubenfels and
apex Adema, 1998
Base of plant Swollen/not swollen Swollen/not de Laubenfels and
swollen Adema, 1998
Habit Caulescent/ Caulescent/ Smitinand, 1971;
acaulescent acaulescent de Laubenfels and
Adema, 1998
Petiole Long/short Long/short de Laubenfels and
Adema, 1998
Leaflet Entire/divided Entire/divided Wang, 1996
Leaflet margins Flat/revolute Flat de Laubenfels and
Adema, 1998
Leaflet apex Soft/spinescent Soft/spinescent Schuster, 1932
Leaflet midrib Raised/flat Raised/flat de Laubenfels and
Adema, 1998
Ovule tomentum
Two species, Cycas revoluta and C. taitungensis C.F. Shen, K.D. Hill, C.H. Tsou &
C.J. Chen, possess tomentose ovules. This is a synapomorphy uniting these taxa
(CI 100%, Fig. 3.3A) but of no value in further grouping or taxonomically clas-
sifying this small group. Miquel (1861), Schuster (1932) and Hill (1996, 1999) had
used this character state as a key defining character for section Asiorientales. This
section proves to be a natural group, supported by several other synapomorphies,
ch03.qxd 28/11/03 3:43 pm Page 30
30 K.D. Hill
and the character of tomentose ovules is clearly a useful character in identifica-

tion of this group.
Ribbed sclerotesta
Two species, Cycas wadei Merrill and C. curranii (J. Schuster) K.D. Hill, possess a
strongly ribbed sclerotesta. This is a synapomorphy uniting these taxa (CI 100%,
Fig. 3.3A) but of no value in further grouping or taxonomically classifying this
small group. This section proves to be a natural group, supported by several other
synapomorphies, and the character of ribbed sclerotesta is clearly a useful char-
acter in identification of this group. This clade was recognized as subsection
Wadeanae by Hill (1996) and Wang (1996) on the basis of this character.
Verrucose sclerotesta
A verrucose sclerotesta is shown to be a well-corroborated synapomorphy (CI

100%, Fig. 3.3A) supporting quite a large natural group of mainland Asian species
occurring in southern China and northern Indochina. This group was formally
named by Smitinand (1971) as section Stangerioides, although at that time with a
single constituent species and not on the basis of sclerotesta characters. Section
Stangerioides was adopted by Hill (1996) and Wang (1996), although with wider cir-
cumscriptions that differed from each other and that in both cases rendered the
natural group non-monophyletic. The verrucose sclerotesta was recognized as a
defining character for smaller groups within the broader groups by these authors.
Crested sclerotesta
A small group of taxa from the Western Pacific and eastern Malesian regions
possesses a crested sclerotesta. This is a synapomorphy uniting these taxa (CI
100% in the best-case scenario, Fig. 3.3B) but of no value in further grouping or
taxonomically classifying this small group. While the crested sclerotesta is a useful
character in recognition of the group, the relationships of this group with other
more distantly allied species are unclear, and no formal nomenclatural recogni-
tion of this group has yet been attempted.
Spongy megagametophytic tissue
The presence of spongy megagametophytic tissue is shown to be a synapomor-

phy (CI 50% in the strict consensus tree, 100% in the best-case scenario, Fig.
3.3A) supporting a small and widely distributed group of species that occurs in
near-coastal situations from Tonga westwards to East Africa. When present, the
ch03.qxd 28/11/03 3:43 pm Page 31
Fig. 3.3. Character state distribution on the consensus cladogram. Species of

Cycas indicated by their specific epithet. (A) Seed characters. (B) Seed characters.
(C) Megasporophyll characters. (D) Female cone characters.
ch03.qxd 28/11/03 3:43 pm Page 32
32 K.D. Hill
truncate microsporophyll
Fig. 3.3. (continued) Character state distribution on the consensus cladogram.

Species of Cycas indicated by their specific epithet. (E) Megasporophyll length
(cm). (F) Pollen cone. (G) Microsporophyll characters. (H) Stem characters.
ch03.qxd 28/11/03 3:43 pm Page 33
Fig. 3.3. (continued) Character state distribution on the consensus cladogram.

Species of Cycas indicated by their specific epithet. (I) Petiole length (cm). (J)
Leaflet characters.
spongy tissue may cause the seeds to float and facilitate aquatic dispersal on
oceanic currents (Dehgan and Yuen, 1983), although not all seeds with spongy
tissue float. Although the taxonomic implications of this character were recog-
nized by these authors, misidentifications of study materials and a general lack of
understanding of specific limits hindered application of these observations. The
group of species, defined by presence of spongy megagametophytic tissue, was
recognized as subsection Rumphiae by Hill (1994). The 50% CI arises from the
uncertain placement of Cycas riuminiana Porte ex Regel, an inland forest species
from the Philippines that is not aquatically dispersed and lacks other features of
subsection Rumphiae. Placement of this species requires further study.
Fibrous sarcotesta
A fibrous or corky layer within the sarcotesta that apparently grows outwards
from the outer surface of the sclerotesta is present in a range of taxa, and has
been recognized as a grouping character by Hill (1996) and Wang (1996). This
character state appears to have been gained near the base of the tree, below Cycas
pectinata Buchanan-Hamilton (Fig. 3.3B), and lost below C. riuminiana, although
deltran optimization suggests two independent acquisitions on the clades below
ch03.qxd 28/11/03 3:43 pm Page 34
34 K.D. Hill
C. pectinata and below C. inermis Loureiro. The moderate level of homoplasy

shown (CI 33%) highlights the uncertain placement and need for further study of
C. apoa K.D. Hill, an inland forest species from New Guinea. Nevertheless, this is
shown to be a useful character for defining and recognizing several groups.
Ovule number
The number of ovules per megasporophyll has been used both in defining species
(Mueller, 1874; Wei, 1994; Chang et al., 1998) and in subdividing the genus
(Warburg, 1900; Pilger, 1926). This character is highly variable in most if not all
species, with only a slight tendency for mainland Asian species to carry fewer
ovules (CI 9%, Fig. 3.3C). The degree of overlap, however, renders this charac-
ter ineffective both in recognizing species and in defining groups.
Megasporophyll pectinate
The pectinate state is shown to be ambiguous (Fig. 3.3D). In one scenario, the
transition to the dentate or entire condition is a synapomorphy defining a large
clade including Cycas inermis and all species above (corresponding to section
Cycas), with the exception of two apparent reversals in C. scratchleyana F. Mueller
and C. riuminiana (CI 33%). This section was first recognized informally by
Carruthers (1893) and formally by Schuster (1932). This group has been recog-
nized by most workers at some level, ranging from subsection (Smitinand, 1971)
through section (above and Hill, 1996) to subgenus (Wang, 1996). In fact, only the
systems presented by Deghan and Deghan (1988) and de Laubenfels and Adema
(1998) have failed to discriminate this group.
Seed cone closed or open
The closed seed cone is shown to be plesiomorphic (Fig. 3.3D). The transition to
the open condition is a synapomorphy (CI 50%) defining a large clade including
Cycas inermis and all species above (section Cycas) as well as a clade of two species
C. lindstromii S.L. Yang, K.D. Hill & Hip and C. pranburiensis S.L. Yang, K.D.
Hill, W. Tang & Vatcharakorn. This character is correlated with the pectinate
megasporophyll, but with less ambiguity.
Megasporophyll long or short
A long megasporophyll was regarded by Schuster (1932) as a diagnostic charac-

ter for subsection Pandemicae. The analysis indicates that this condition has arisen
independently in at least four lineages (CI 17%, Fig. 3.3E). This character is con-
ch03.qxd 28/11/03 3:43 pm Page 35
sequently of no value in recognizing major groupings, although it is useful in dis-

criminating related species and is a potential synapomorphy for Cycas thouarsii R.
Brown ex Gaudichaud and C. micronesica K.D. Hill.
Pollen cone cylindrical or ovoid
A cylindrical pollen cone was regarded by de Laubenfels and Adema (1998) as a

diagnostic character for the new genus Epicycas, although the cylindrical state was
acknowledged also to occur in Cycas. The analysis indicates that this condition is
present in three separate lineages (CI 14% for cone shape, including the states
ovoid and narrow-ovoid, Fig. 3.3F) and that the cylindrical condition may in fact
be plesiomorphic for Cycas. Cone shape is also shown to be homoplastic within
clades, and is consequently of little value in recognizing natural groups.
Microsporophyll soft, waxy or hard
Microsporophyll texture shows no clear evolutionary progression, with the pres-

ence of all three states near the base of the tree and an ambiguous ancestral con-
dition (Fig. 3.3G). The different states are, however, highly consistent (CI 100%).
The soft condition defines the group discussed above as section Stangerioides; the
waxy condition characterizes two small clades (the Cycas revoluta clade and the C.
wadei clade) and C. panzhihuaensis L. Zhou & S.Y. Yang; and the hard condition pre-
scribes a large clade that includes C. pectinata and all species above. The soft pollen
cone character was used by Smitinand (1971) in defining section Stangerioides.
Microsporophyll truncate
Microsporophylls with a shortened apex (truncate) were regarded by de

Laubenfels and Adema (1998) as the diagnostic character for subgenus Truncata.
The analysis indicates that such shortening has arisen independently in three sep-
arate clades (CI 22%, Fig. 3.3H) and is consequently of little value in recogniz-
ing major groupings, although it does suggest the possibility that Cycas rumphii
Miquel, C. micronesica and C. seemannii A. Braun might form a clade.
Caudex with a bulbous base
A caudex with a bulbous base was regarded by de Laubenfels and Adema (1998)
as the key diagnostic character for the new genus Epicycas. The analysis indicates
that this condition has arisen independently in at least two and possibly six sepa-
rate lineages (CI 11%, Fig. 3.3H) and is consequently of little value in recogniz-
ing major groupings.
ch03.qxd 28/11/03 3:43 pm Page 36
36 K.D. Hill
Caudex wholly subterranean
Wholly hypogeous growth habits are shown by a few species. This character and the
bulbous base (above) were concatenated by de Laubenfels and Adema (1998) in
defining the genus Epicycas. The truly hypogeous habit is much less frequent than the
bulbous base, and is shown to have arisen in two groups independently (CI 25%, Fig.
3.3H). This character was used by Smitinand (1971) in defining section Stangerioides.
Petiole long or short
A short petiole (less than 30 cm long) was regarded by de Laubenfels and Adema
(1998) as a key diagnostic character for Cycas subgenus Revoluta, and a long petiole
for subgenus Pectinata. The analysis indicates that the long condition has arisen inde-
pendently in at least seven different lineages (CI 20%, Fig. 3.3I), possibly with sub-
sequent reversals, and is consequently of little value in recognizing major groups.
Leaflet dichotomously divided
Dichotomously divided leaflets are a striking character present in four species

occurring in southern China and Vietnam. This section was recognized as a
series Multipinnata by Wang (1996), who regarded the divided state as primitive.
This analysis shows the divided state to be advanced (Fig. 3.3J), but is not suffi-
ciently detailed to examine the value of this character as a grouping character.
Other morphological characters do not separate these four taxa and the divided
leaflets are a useful grouping character on this basis.
Leaflet margins revolute
Revolute leaflet margins were first used to subdivide the genus by Miquel (1843).
In contrast, flat, often undulate leaflet margins were considered characteristic of
the genus Epicycas by de Laubenfels and Adema (1998). Flat leaflets are here
shown to be ancestral. The change to revolute margins is shown by the analysis
to occur in several lineages (CI 22%, Fig. 3.3K), and is consequently ineffective
in defining groups, although it is at times a useful character in discriminating
species within groups.
Leaflet apex spinescent
A spinescent leaflet apex was regarded by Schuster (1932) as a diagnostic char-

acter for subsection Endemicae. The analysis indicates that this condition has
arisen independently in all major lineages (CI 10%, Fig. 3.3K), possibly more
ch03.qxd 28/11/03 3:43 pm Page 37
than once in some of these clades, and is consequently of little value in recog-
nizing major groups.
Leaflet midrib raised or flat
A flat midrib was regarded by de Laubenfels and Adema (1998) as a diagnostic

character for Cycas subgenus Revoluta. The analysis indicates that this condition
has arisen independently in three separate lineages (CI 12%, Fig. 3.3K), possibly
more than once in some clades, and is consequently of little value in recognizing
major groups. Alternatively, flat midribs can be almost as parsimoniously
explained by only two origins and multiple subsequent losses.
Assessment of Previous Classifications
Schuster (1932) attempted the first formal subgeneric classification of Cycas (Fig.
3.4A). He recognized three major groups (sections), corresponding basically to
the pectinate megasporophyll (section Indosinenses) and non-pectinate megasporo-
phyll (section Lemuricae) groups, with the further separation of C. revoluta (in which
he included C. taiwaniana Carruthers) as a monotypic group defined by narrow
revolute leaflets and tomentose ovules (section Asiorientales). He further subdivid-
ed section Lemuricae into two subsections, Pandemicae and Endemicae, on differences
in tips of pinnae and length of megasporophylls. Although the major groups
erected by Schuster take nomenclatural priority, two of his three sections are
paraphyletic as he circumscribed them. He also allocated species and infraspecif-
ic taxa within these groups erratically and often apparently on the basis of guess-
work, placing many previously recognized species as infraspecific taxa within
complex hierarchies under C. circinalis Linnaeus and C. rumphii. Many of these are
clearly distinct and distantly related species, rendering his concepts of C. circinalis
and C. rumphii polyphyletic. In addition, Schusters work does not comply with
the requirement for subgroups including the type of a genus to carry automati-
cally the generic name (the autonym rule International Code of Botanical
Nomenclature (ICBN), Greuter et al., 2000). Hence section Lemuricae should cor-
rectly be section Cycas and subsection Pandemicae should be subsection Cycas. In
summary, Schuster recognized five infrageneric taxa, three of which are shown
to be polyphyletic (Fig. 3.4A).
Only four infrageneric groups are marked on Fig. 3.4A. One of the marked
groups (Endemicae) is monophyletic, while the other is monotypic (and thus mono-
phyletic by definition). If the unmarked lineages collectively form another of his
taxa (section Lemuricae?) then it might be said that it is poly- and not paraphylet-
ic. Similarly, Pandemicae would be most appropriately described as polyphyletic.
Smitinand (1971) proposed a very different major division into two groups
(Fig. 3.4B), separating Cycas micholitzii Dyer in the monotypic section Stangerioides
on the basis of the soft, shortly apiculate microsporophylls in very small, slender
ch03.qxd 28/11/03 3:43 pm Page 38
38 K.D. Hill
Fig. 3.4. Previous systems of classification. Incorrect names under the ICBN are
indicated by*. Monophyletic clades supported in this analysis shown in bold.
Species of Cycas indicated by their specific epithet. Classifications by: (A)
Schuster (1932); (B) Smitinand (1971); (C) Dehgan (1987); (D) Hill (1996, 1999).
ch03.qxd 28/11/03 3:43 pm Page 39
Fig. 3.4. (continued) Previous systems of classification. Incorrect names under the
ICBN are indicated by*. Monophyletic clades supported in this analysis shown
in bold. Species of Cycas indicated by their specific epithet. Classifications by:
(E) Wang (1996); (F) de Laubenfels (1998).
cones, and the dwarf, mainly subterranean habit with very few leaves
(Stangerioid habit). The remainder of the genus he placed in section Cycas,
which he divided into two subsections on the basis of pectinate megasporophylls
(subsection Pectinatae, also given as Pinnatidae in the key) and non-pectinate megas-
porophylls (subsection Circinnalidae). The former subsection included C. revoluta.
Again, the requirement for autonyms was not strictly followed, and subsection
Circinnalidae is correctly (and automatically) subsection Cycas. Subsection Pectinatae
(or Pinnatidae) is also illegitimately described, no type species being designated. Of
the four infrageneric taxa recognized by Smitinand, three are shown to be poly-
phyletic (Fig. 3.4B).
Dehgan and Dehgan (1988) alluded to an infrageneric classification and
published a series of names with a major division based on presence or absence
of spongy tissue in seeds and incorporating differences in pollen structure (Fig.
3.4C). This classification was not formally published, and neither the require-
ment for autonyms nor the rule of priority was followed. Although the character
basis for this subdivision was sound, incorrect identifications and a lack of under-
standing of specific limits made the arrangement untenable. The proposed clas-
sification recognized two subgenera with three sections and three subsections. No
assignment of species to the lower groups was made, however, apart from the
ch03.qxd 28/11/03 3:43 pm Page 40
40 K.D. Hill
listing of a single typical species. On the basis of these inclusions, the two sub-
genera are shown to be polyphyletic, as is the one section that has more than one
included species (Fig. 3.4C).
Hill (1996, 1999) presented an arrangement incorporating four sections cor-
responding in name to the three erected by Schuster plus section Stangerioides as
described above (Fig. 3.4D), in all cases with different circumscriptions, on the
basis of morphological cladistic analyses. Nine subordinate groups correspon-
ding to subsections were also proposed, most of them informal. Of the four sec-
tions, two are shown to be polyphyletic, and one of the nine subsections is shown
to be polyphyletic, with two others possibly paraphyletic (Fig. 3.4D).
Wang (1996) presented another formal infrageneric classification, recogniz-
ing two subgenera on the basis of pectinate vs. non-pectinate megasporophylls.
Within these a number of sections, subsections and series were erected (Fig.
3.4E). These are not altogether internally consistent and several of the names are
in contravention of the ICBN (Greuter et al., 2000). Of the two subgenera, one
(subgenus Panzhihuaensis) is shown to be polyphyletic, and one of the sections is
paraphyletic (Fig. 3.4E). Full enumeration of species included in a number group
is not attempted, making assessment uncertain.
De Laubenfels and Adema (1998) presented an entirely different system, rec-
ognizing a new genus Epicycas for taxa with a largely subterranean habit and a
bulbous underground base. They divided the remainder of Cycas into four sub-
genera based on a combination of leaf, microsporophyll and megasporophyll
morphology. Every generic and infrageneric group as circumscribed by these
authors is shown to be polyphyletic (Fig. 3.4F).
Conclusions
No previously published system of classification is wholly in accord with the

results of molecular and morphological phylogenetic analyses. Many of the key
characters on which previously published classifications were based are shown to
be plesiomorphic, homoplastic or autapomorphic and of little value in defining
infrageneric relationships and developing classifications based on these. In par-
ticular, separation of the genus Epicycas is clearly unwarranted (see Chen et al.,
Chapter 5 this volume).
However, certain of the characters discussed above are shown to be synapo-
morphic characters that are useful in defining natural groups. Seed characters in
particular are shown to be highly consistent, as are some aspects of sporophyll
morphology. These characters are useful in developing a workable key to groups
(Table 3.3) and to species within the genus Cycas. Data are still lacking on many
recognizable species in the genus, and consequently placement of a number of
species on the basis of the characters discussed above is uncertain. It is at present
premature to develop a comprehensive system of infrageneric classification on
this basis, but an interim arrangement recognizing demonstrably monophyletic
groups and the nearest possible placements of species or groups incertae sedis can
ch03.qxd 28/11/03 3:43 pm Page 41
Section Asiorientales
Section Wadeanae
Section Stangerioides
Section Indosinenses
Species incertae sedis
Section Cycas
Subsection Cycas
Subsection Rumphiae
Subsection Endemicae
Fig. 3.5. Comprehensive enumeration of species in Cycas. Consensus morphologi-

cal and molecular tree shown in bold, other species interpolated from morpho-
logical cladistic analysis using available data.
ch03.qxd 28/11/03 3:43 pm Page 42
42 K.D. Hill
Table 3.3. Key to major recognizable subdivisions of the genus Cycas.
1 Ovules tomentose Section Asiorientales

1 Ovules glabrous 2
2 Megasporophyll lamina pectinate 3
2 Megasporophyll lamina not pectinate Section Cycas
3 Male cones soft, microsporophyll apices not 4
deflexed
3 Male cones rigid, waxy, microsporophyll apices 5
deflexed
4 Sclerotesta not ribbed Cycas panzhihuaensis
L. Zhou & S.Y. Yang
4 Sclerotesta ribbed Section Wadeanae
5 Microsporophylls flexible, rounded; sarcotesta not
fibrous; sclerotesta verrucose Section Stangerioides
5 Microsporophylls rigid, acuminate; sarcotesta
fibrous; sclerotesta smooth Section Indosinenses
be informally presented here (Table 3.4). The primary division is made into sec-
tions rather than subgenera for the reason that basal branches are short and addi-
tional data may well collapse these or elucidate different relationships. The
sections recognized are distinct and well-supported clades.
A comprehensive enumeration of taxa in the genus Cycas that are recogniz-
able as species or comparable terminal taxa is presented in Fig. 3.5. The tree is
based on the strict consensus of a combined morphological and molecular analy-
sis (shown in bold). Species for which molecular data are unavailable are inter-
polated on the basis of morphological cladistic analysis using available data (Hill,
1999 and in preparation). Not all species are adequately known, and data are
incomplete for a number of taxa. This again represents an informal interim
arrangement that shows known groups and relationships and highlights areas
requiring further study.
Acknowledgements
The Hermon Slade Foundation is warmly thanked for the financial support that
allowed this and related studies to proceed. The Vietnamese Institute of Ecology
and Biological Resources and the Chinese Academy of Science are thanked for
assistance with laboratory and field studies in Vietnam and China. Kampon
Tansacha and the Nong Nooch Tropical Garden are gratefully acknowledged for
hospitality and logistical assistance. Anders Lindstrm assisted in the field and in
valuable discussions of the taxonomic and distributional limits of the cycads of
Asia. The keepers of the herbaria at A, B, BKF, BM, G, K, L, NY and P are
acknowledged for access to their collections. Peter Weston is thanked for con-
structive comment on earlier drafts of the manuscript.
ch03.qxd 28/11/03 3:43 pm Page 43
Table 3.4. A provisional classification of Cycas.
Section Asiorientales
Synapomorphies: encrypted stomata
3 species. No subgroups
Section Wadeanae
Synapomorphies: ribbed seeds
2 species. No subgroups
Section Stangerioides
Synapomorphies: verrucose seeds, soft pollen cones
About 25 species. No clear subgroups.
Section Indosinenses
Synapomorphies: fibrous sclerotesta, hard pollen cones
About 15 species. No clear subgroups.
Section Cycas
Synapomorphies: open seed cones, non-pectinate megasporophylls
About 53 species. Three distinct monophyletic subgroups and a number of
unplaced species
Subsection Cycas: about 4 species
Synapomorphies: fibrous sclerotesta
Subsection Rumphiae: about 10 species
Synapomorphies: spongy megagametophyte
Subsection Endemicae: about 32 species
Synapomorphies: palisade tissue in lower mesophyll
species incertae sedis; about 8 species
References
Amoroso, V.B. (1986) Morphological study of the sporophylls of Philippine Cycas.
Philippine Journal of Science 115(3), 177198.
Backer, C.A. and Bakhuizen van den Brink, R.C. (1963) Cycadaceae. In: Flora of Java, Vol.
1. Noordhoff, Groningen, The Netherlands, p. 87.
Carruthers, W. (1893) On Cycas taiwaniana sp. nov. and C. seemannii A.Br. Journal of Botany
31, 13; t. 330331.
Chang, H.T., Huang, Y.Y. and Zheng, H.X. (1998) Acta Sci. Nat. Univ. Sunyatseni 37, 8.
[Cycas septemsperma.]
Chen, C.J. and Stevenson, D.W. (1999) Cycadaceae. In: Wu, Z.Y. and Raven, P.H. (eds)
Flora of China, Vol. 4, Cycadaceae through Fagaceae. Science Press, Beijing and Missouri
Botanical Garden Press, St Louis, Missouri, pp. 17.
De Candolle, A.P. (1868) Cycadeae. In: Prodromus Systema Natura and Regnum Vegetabile 16(2).
Victor Massen, Paris, pp. 361521.
De Laubenfels, D.J. and Adema, F. (1998) A taxonomic revision of the genera Cycas and
Epicycas gen. nov. (Cycadaceae). Blumea 43, 351400.
Dehgan, B. and Dehgan, N.B. (1988) Comparative pollen morphology and taxonomic
affinities in Cycadales. American Journal of Botany 75, 15011516.
ch03.qxd 28/11/03 3:43 pm Page 44
44 K.D. Hill
Dehgan, B. and Yuen, C.K.K.H. (1983) Seed morphology in relation to dispersal, evolu-
tion and propagation of Cycas L. Botanical Gazette 144, 412418.
Fu, S.H., Cheng, W.C., Fu, L.K. and Chen, C.J. (1978) Cycadaceae. In: Cheng, W.C. and
Fu, L.K. (eds) Flora Reipublicae Popularis Sinicae 7. Science Press, Beijing, China,
pp. 417.
Hip, N.T. and Vidal, J.E. (1996) Cycadaceae. In: Morat, Ph. (ed.) Flore du Cambodge, du
Laos et du Vitnam, Vol. 28, Gymnospermae. pp. 623.
Hill, K.D. (1994) The Cycas rumphii complex (Cycadaceae) in New Guinea and the Western
Pacific. Australian Systematic Botany 7, 543567.
Hill, K.D. (1995) Infrageneric relationships, phylogeny and biogeography of the genus
Cycas (Cycadaceae). In: Vorster, P. (ed.) Proceedings of the Third International Conference on
Cycad Biology. Cycad Society of South Africa, Stellenbosch, South Africa, pp.
139162.
Hill, K.D. (1996) A taxonomic revision of the genus Cycas (Cycadaceae) in Australia.
Telopea 7, 164.
Hill, K.D. (1999) Cycas an evolutionary perspective. In: Chen, C.J. (ed.) Biology and
Conservation of Cycads, Proceedings of the Fourth International Conference on Cycad Biology.
International Academic Publishers, Beijing, China, pp. 98115.
Johnson, L.A.S. (1959) The families of cycads and the Zamiaceae of Australia. Proceedings
of the Linnaean Society of New South Wales 84, 64117.
Miquel, F.A.W. (1843) Genera et species Cycadearum viventium. Linnaea 17, 675744.
Miquel, F.A.W. (1861) Prodromus Systematis Cycadearum. Van der Post, Utrecht, Holland, 35
pp.
Mueller, F.A.W. (1874) Fragmenta Phytographie Australiae, Vol. 8. Government Printer,
Melbourne, Australia, 304 pp.
Pilger, R. (1926) Cycadaceae. In: Engler, A. (ed.) Die Naturlichen Pflanzenfamilien, 2nd edn 2,
13, pp. 4482.
Schuster, J. (1932) Cycadaceae. In: Engler, A. (ed.) Das Pflanzenreich, Fascicle 99, Vol. 4,
Part 1, pp. 1168.
Smitinand, T. (1971) The genus Cycas Linn. (Cycadaceae) in Thailand. Natural History
Bulletin of the Siam Society 24, 163175.
Stevenson, D.W. (1992) A formal classification of the extant cycads. Brittonia 44, 220223.
Wang, D.Y. (1996) Systematic classification and brief introduction to Cycadales (Chapter
2) and Taxonomy of Cycas in China (Chapter 3). In: Wang, F.X. and Liang, H.B. (eds)
Cycads in China. Guangdong Science and Technology Press, Guangdong, China, pp.
9142.
Warburg, O. (1900) Cycadaceae. In: Warburg, O. (ed.) Monsunia. Engelmann, Leipzig,
Germany, pp. 178181.
Wei, F.N. (1994) A new cycad from Guangxi. Guihaia 14, 300. [Cycas ferruginea.]
ch04.qxd 28/11/03 3:44 pm Page 45
Morphological Characters 4
Useful in Determining Species
Boundaries in Cycas
(Cycadaceae)
Anders Lindstrm
Nong Nooch Tropical Botanical Garden, Najomtien, Sattahip,

Chonburi, Thailand
Abstract
Morphological characters within the genus Cycas, based on previously published descrip-
tions and on recent field research, are evaluated for their usefulness in distinguishing
among distinct, yet morphologically similar, pairs of taxa within the genus. A standard-
ized suite of taxonomically useful morphological characters in support of future designa-
tion of species within Cycas is recommended.
Introduction
Taxonomy has traditionally used certain morphological characters or suites of
morphological characters to define taxa. These characters are often chosen in a
way that permits convenient measurement from herbarium specimens. For
cycads, this usually restricts the choice to those characters derived from limited
aspects of leaf and sometimes cone morphology, since these are usually the only
parts represented in herbarium accessions. Annotations of herbarium vouchers
often lack data significant to taxonomic studies or even identification. Leaf
length, petiole length, stem height, stem profile and branching pattern are exam-
ples of these characters that cycad researchers require in their evaluation of
taxa.
Cycads, especially representatives of the genus Cycas Linnaeus, typically have
long leaves that are difficult to process as herbarium specimens. Therefore, the
ch04.qxd 28/11/03 3:44 pm Page 46
46 A. Lindstrm
majority of Cycas vouchers in herbaria are represented only by portions of leaves

usually a group of about four to six pinnae. Specimens often lack the petiole, a
feature of the plant that can be of particular taxonomic significance. Labels
accompanying Cycas specimens most often lack the necessary data to describe the
whole leaf and seldom indicate from which portion of the leaf the pinnae were
obtained. Within a leaf, terminal, central and basal pinnae each have distinctive
features.
Herbarium accessions only infrequently include reproductive material from
Cycas plants and representation of both male and female carpological material
for any given taxon is rare.
A major problem associated with past herbarium-based Cycas studies has
been that different suites of characters had to be used in the analysis of different
specimens. The taxonomist is often faced with having to study specimens that
have different portions of a leaf represented, or a portion of a leaf on one spec-
imen and a complete leaf on another, different reproductive material present or
reproductive material absent, and few descriptive or quantitative data in the
annotations.
A further limitation arises in that herbarium specimens often do not neces-
sarily represent the average plant for any given population. It is known that
plants within a population of cycads express a wide variation for numerous mor-
phological characters. Identifying a plant that represents a population can be
time consuming, difficult and is often not a priority when herbarium specimens
are collected. Very few specimen labels include details of the morphological vari-
ation within the population.
The assignment of sectional and subsectional ranks in Cycas has been based
primarily on the sharing of characters associated with the seed, and secondarily
with other reproductive and pinna features (see Hill, Chapter 3 this volume). For
any given section or subsection, seed structure is in several aspects linked to
animate and inanimate dispersal processes and thus implicit in part for infra-
generic genetic exchange. However, these seed characters are usually not avail-
able from herbarium specimens or accompanying data, making it difficult to
assign the sectional or subsectional ranks. A suite of useful characters, that can
be conveniently documented in the field and later included in the specimen anno-
tation, is needed to assist future taxonomic studies at the section and subsection
ranks.
To facilitate future taxonomic studies at the sectional, subsectional and spe-
cific ranks within Cycas, it is critical to identify a suite of morphological charac-
ters that can be accessed, whether through herbarium specimens with their
associated annotations and/or through publications. Hill (see Chapter 3 this
volume) proposes a suite of such characters.
Understanding the significance of a specific suite of characters is essential in
the classification of new taxa into the appropriate section or subsection within the
genus. The aim of this chapter is to test a suite of characters that may reflect
measurable and taxonomically useful features in a consistently reliable classifica-
tion within Cycas.
ch04.qxd 28/11/03 3:44 pm Page 47
Determining Species Boundaries in Cycas 47
Materials and Methods

The author has visited wild populations of almost every non-Australian species
in the genus Cycas and herbarium specimens have been made from each popula-
tion. Prior to collecting specimens, measurements were taken from a number of
mature plants within each population using a pre-defined suite of characters.
The characters chosen for this study were based on their use in past taxo-
nomic studies within the genus, as well as those that showed potential for sup-
porting future taxonomic studies. The 30 quantitative and qualitative characters
(Table 4.1) are associated with seven aspects of a plant: plant habit, leaf, pinna,
cataphyll, megastrobilus, microstrobilus and seed. Most of these characters and
their measurement aspects are presented in Fig. 4.1.
To understand the potential value of the chosen morphological characters
for future studies within Cycas, three pairs of distinct yet morphologically similar
species were identified, and measurements were obtained from selected wild pop-
ulations. One population was sampled for four of the six species (C. edentata de
Laubenfels, C. elephantipes A. Lindstrm & K.D. Hill, C. micronesica K.D. Hill and
C. siamensis Miquel) and two populations were sampled for the remaining species
(C. seemannii A. Braun and C. rumphii Miquel).
The first pair of taxa studied was Cycas siamensis and C. elephantipes. Both
species are arborescent and native to Thailand. The second pair, C. edentata and
C. rumphii, are tall, erect, arborescent cycads from islands in the Pacific Ocean.
Cycas edentata and C. rumphii are believed to hybridize naturally where their distri-
butions overlap. The third pair examined was C. seemannii and C. micronesica. Cycas
seemannii is a tall, erect, arborescent and usually unbranched species found in Fiji,
Vanuatu, the Tonga Islands and New Caledonia. Cycas micronesica (often misiden-
tified in collections as either C. circinalis Linnaeus or C. rumphii) is an erect,
arborescent species up to 8 m tall from the Mariana Islands. Populations of these
two species are widely separated geographically and no spontaneous hybridiza-
tion is likely to occur. Nevertheless, molecular evidence indicates that, in addition
to being morphologically similar, they are evolutionarily closely related (Ken Hill,
Australia, 2002, personal communication).
Measurements were recorded for multiple plants within each population for
the six species. In some cases, not all characters were measured male cones and
seeds were often not available for all species during the time of the fieldwork. For
each population sampled, the number of plants measured, the mean for each
character and the standard deviation for each character are listed in Tables
4.24.4. Data derived from each species pair were compared using a t-test with
an alpha level of 0.05. The t-test assesses whether the means of two groups are
statistically different from each other and yields the same results as a one-way
analysis of variance (ANOVA). Results from the t-tests are included in Tables
4.24.4.
ch04.qxd 28/11/03 3:44 pm Page 48
48 A. Lindstrm
Table 4.1. Vegetative and reproductive characters for taxonomic studies in

Cycas. Characters of questionable use for taxonomic studies are indicated by *.
Characters determined to be unstable and variable among species at the onset
of the investigation are indicated by **. Unstable and variable characters were
removed from the analyses.
Plant part Character/character state
Plant habit Stem height (cm)

Stem minimum diameter (cm)
Leaf Leaf number
Length (cm)*
Petiole length (cm)
Petiole thickness (mm)
Number of spines on petiole**
Petiole length percentage spines
Terminus (spine, single leaflet, leaflet pair)**
Pinna Number
Length (cm)
Width (mm)
Basal width (mm)
Arrangement of pinnae (opposing, alternate)**
Spacing between pinnae (mm)
Pinna length (cm)
Cataphyll Length (cm)
Megastrobilus Megasporophyll length (cm)
Megasporophyll lamina length (mm)
Megasporophyll lamina width (mm)*
Megasporophyll apical spine length (mm)*
Megasporophyll lateral spine number
Megasporophyll number of ovules
Microstrobilus Microstrobilus length, excluding peduncle (cm)
Microstrobilus diameter (cm)
Microsporophyll length (mm)
Microsporophyll width (mm)
Microsporophyll apical spine length (mm)
Seed Length (mm)
Width (mm)
Results
Characters of questionable use
Three characters preliminarily included in the study as potentially useful taxo-

nomic characters were proven to be unstable and variable within and among
ch04.qxd 28/11/03 3:44 pm Page 49
Fig. 4.1. Vegetative and reproductive characters measured from Cycas specimens:
(1) leaf length; (2) petiole length; (3) pinna number; (4, 8) pinna length; (5) pinna
width; (6) pinna basal width; (7) spacing between pinnae; (9) megasporophyll
length; (10) megasporophyll lamina length; (11) megasporophyll lamina width;
(12) megasporophyll apical spine length; (13) megasporophyll lateral spine
number; (14) megasporophyll number of ovules; (15) microstrobilus length
(excluding peduncle); (16) microstrobilus diameter; (17) microsporophyll length;
(18) microsporophyll width; (19) microsporophyll apical spine length. Characters
not shown (see Table 4.1), but included in the study, are those associated with the
stem, cataphyll, seed, petiole thickness and petiole length percentage spines.
ch04.qxd
28/11/03
Table 4.2. Sample sizes (n), means, standard deviations (SD) and results from t-tests (P = probability; * = P < 0.05) for morphological
50
characters of Cycas siamensis Miquel (SIA) and C. elephantipes A. Lindstrm & K.D. Hill (ELE). Material was not available to allow
analyses of microstrobilus and seed characters for these taxa.
SIAa ELEb
3:44 pm
Character n Mean SD n Mean SD P
Plant habit
Stem height (cm) 7 8.43 4.72 10 87.6 48.20 < 0.001*
Page 50
Stem minimum diameter (cm) 7 11.43 1.72 10 22.7 3.23 < 0.001*
Number of leaves 7 18.86 3.24 10 37.4 13.60 < 0.001*
Leaf
Length (cm) 12 75.58 10.35 10 167.3 12.08 < 0.001*
Petiole length (cm) 12 13.79 3.22 10 25.7 6.52 < 0.001*
Petiole thickness (mm) 12 7.4 1.3 10 15.0 2.7 < 0.001*
A. Lindstrm
Petiole length percentage spines 12 69.75 26.37 10 54.0 19.6 = 0.124
Pinna
Number 12 73.54 8.67 10 117.3 20.0 < 0.001*
Length (cm) 12 10.04 1.6 10 18.76 2.8 < 0.001*
Width (mm) 12 5.83 0.58 10 9.1 0.9 < 0.001*
Basal width (mm) 12 3.25 0.45 10 3.9 0.7 = 0.028*
Spacing between pinnae (mm) 12 4.33 0.78 10 8.2 1.3 < 0.001*
Pinna length (cm) 12 2.82 1.05 9 39.3 23.4 < 0.001*
Cataphyll
Length (cm) 6 2.68 0.84 9 8.14 2.0 < 0.001*
Megastrobilus
Megasporophyll length (cm) 2 9.25 1.06 5 15.5 1.9 = 0.006*
Megasporophyll lamina length (mm) 2 24 8.48 5 41.6 3.8 = 0.187
Megasporophyll lamina width (mm) 2 20 8.48 5 48.8 9.9 = 0.051
Megasporophyll apical spine length (mm) 2 25.5 14.85 5 44.0 10.0 = 0.297
Megasporophyll lateral spine number 2 11 4.24 5 25.2 0 = 0.144
aCycas siamensis plants from Tak Province, Thailand.
bCycas elephantipes plants from Chaiyaphum Province, Thailand.
ch04.qxd
28/11/03
Table 4.3. Sample sizes (n), means, standard deviations (SD) and results from t-tests (P = probability; * = P < 0.05; calculated for species com-
parisons only) for morphological characters of Cycas seemannii A. Braun (SEE) and C. micronesica K.D. Hill (MIC). Material was not available
to allow analyses of cataphyll and microstrobilus characters for these taxa.
3:44 pm
SEEa SEEb MICc
Character n Mean SD n Mean SD n Mean SD P
Plant habit
Stem height (cm) 8 6.5 2.39 14 2.68 2.00 11 3.64 1.23 = 0.544
Page 51
Stem minimum diameter (cm) 8 20.13 6.49 14 20.29 11.29 11 15.19 3.22 = 0.034*
Determining Species Boundaries in Cycas

Leaf
Length (cm) 8 230.5 27.21 14 186.1 26.70 15 185.8 72.62 = 0.423
Petiole length (cm) 8 41.63 10.86 14 31.57 6.27 15 60.47 27.98 = 0.003*
Petiole thickness (mm) 8 2.63 0.23 14 2.32 0.42
Petiole length percentage spines 8 34.5 38.92 14 20.71 22.86 15 1.4 2.53 < 0.001*
Pinna
Number 8 101.7 18.13 14 96.43 7.08 15 71.93 16.17 < 0.001*
Length (cm) 8 28.25 1.83 14 24.43 3.37 15 16.44 13.47 = 0.018*
Width (mm) 8 14.75 4.00 14 10.86 1.10 15 8.2 6.69 = 0.041*
Basal width (mm) 8 7.25 1.49 14 5.43 0.65 15 2.47 1.73 < 0.001*
Spacing between pinnae (mm) 8 8.5 2.78 14 8.5 2.35 15 8.4 6.96 = 0.958
Pinna length (cm) 7 16.43 4.16 12 13.34 3.73 15 10.23 8.58 = 0.936
Megastrobilus
Megasporophyll length (cm) 6 32.5 4.68 12 32.42 5.42 3 36 7.81 = 0.516
Megasporophyll lamina length (mm) 6 50.83 2.04 12 53.33 4.44 5 35.6 2.61 < 0.001*
Megasporophyll lamina width (mm) 6 41.83 5.95 12 36.17 4.90 5 31.4 6.70 = 0.091
Megasporophyll apical spine length (mm) 6 22.5 6.90 12 13.42 4.66 5 25.6 18.67 = 0.338
Megasporophyll lateral spine number 6 11.5 2.26 2 11.5 0.71 5 6 1.22 < 0.001*
Seed
Length (mm) 1 55 2 57.5 3.54 5 52 2.74 = 0.085
Width (mm) 1 42 2 49 1.41 5 49 6.52 = 0.560
aCycas seemannii plants from Tanna Island, Vanuatu.
bCycas seemannii plants from Bourail, New Caledonia.
cCycas
51
micronesica plants from islands of Micronesia (Rota, Guam, Yap and Palau Island).
ch04.qxd 28/11/03 3:44 pm Page 52
52 A. Lindstrm
populations of non-Australian Cycas species (Table 4.1). Therefore, the following

three characters were excluded early in the study:
1. Terminus of the leaf. Three separate character states have been found:
1. (a) terminating in a spine;
1. (b) terminating in a single leaflet;
1. (c) terminating in leaflet pair.
Although sometimes quite uniform in several of the taxa studied, the level of
variation seen in the majority of Cycas species indicates that this is not a consis-
tent reliable character for taxonomic studies. This character was removed from
the analyses.
2. Number of spines on the petiole. The number of spines is variable and
depends on a plants age, size and condition. Number of spines on the petiole was
therefore removed from the analyses. A more consistent and useful character
associated with petiole spines is the percentage of the petiole length that bears
spines.
3. The arrangement of the pinnae. The arrangement of pinnae, opposite
or alternate to each other on either side of the leaf rachis, was highly variable
from individual to individual within a population and was removed from the
analyses.
During the course of the study, leaf length was often found to be a variable char-
acter within a population and within a species. Leaf length can be dependent on
light conditions and overall condition of the plant. However, this character was
retained in the analyses because it was found to be useful in distinguishing Cycas
siamensis and C. elephantipes (Table 4.2).
For all three pair comparisons (Tables 4.2 and 4.3), megasporophyll lamina
width and megasporophyll apical spine length were the only characters not sig-
nificant for distinguishing between any of the three pairs of species. These two
characters were retained in the analyses because they show promise for delimit-
ing other species in the genus.
Species-specific characters
A number of characters examined were found to be taxonomically useful when

comparing morphologically similar but distinct species pairs. Table 4.2 presents
the results of t-tests for 19 characters for Cycas siamensis and C. elephantipes. Based
on the results of the t-tests, 14 of the 19 characters associated with the stem, leaf,
pinna, cataphyll and megastrobilus were valuable for distinguishing these two
taxa. Although not presented here, similar types of analysis indicated that C. sia-
mensis is morphologically more similar to C. nongnoochiae K.D. Hill and C. elephan-
tipes is morphologically more similar to C. pachypoda K.D. Hill. Fewer than 14
characters differentiated C. siamensis from C. nongnoochiae and C. elephantipes from
C. pachypoda.
Table 4.3 presents the results for the analyses of data obtained from
ch04.qxd
28/11/03
Table 4.4. Sample sizes (n), means, standard deviations (SD) and results from t-tests (P = probability; * = P < 0.05; calculated for
species comparisons only) for morphological characters of Cycas edentata de Laubenfels (EDE) and C. rumphii Miquel (RUM).
Material was not available to allow analysis of stem and microstrobilus characters for these taxa.
3:44 pm
EDEa RUMb RUMc
Character n Mean SD n Mean SD n Mean SD P
Leaf
Page 53
Length (cm) 7 163.7 34.77 4 213.7 43.04 2 211 45.25 = 0.100
Determining Species Boundaries in Cycas

Petiole length (cm) 7 44.71 11.98 4 59.25 15.22 2 50.5 12.02 = 0.159
Petiole length percentage spines 7 88.57 17.73 4 82.5 5 2 40 14.14 = 0.422
Pinna
Number 7 140.3 20.70 4 80.5 20.07 2 96.5 26.16 = 0.002*
Length (cm) 7 30.29 3.99 4 26 6.48 2 25 5.66 = 0.292
Width (mm) 7 13.71 1.704 4 16.25 2.5 2 12.5 2.12 = 0.135
Basal width (mm) 7 6.14 0.90 4 5.25 0.5 2 4.5 0.71 = 0.063
Spacing between pinnae (mm) 7 19.71 4.46 4 13.75 3.5 2 11 1.41 = 0.040*
Pinna length (cm) 6 22.83 5.85 4 1.25 0.5 1 18 0 < 0.001*
Cataphyll
Length (cm) 2 7 0 2 1.9 0.85 = 0.074
Megastrobilus
Megasporophyll length (cm) 4 30.23 6.263 3 29.67 4.51 1 26 = 0.896
Megasporophyll lamina length (mm) 4 63.75 13.77 3 34.33 10.07 1 45 = 0.022*
Megasporophyll lamina width (mm) 4 35.5 5.45 3 27.67 6.81 1 25 = 0.180
Megasporophyll apical spine length (mm) 4 27.5 8.67 3 26 10.39 1 33 = 0.849
Megasporophyll lateral spine number 3 8.67 4.167 3 2.337 0.58 1 2 = 0.116
Seed
Length (mm) 4 6 0 3 4 0 1 8 > 0.860
aCycas edentata plants from Culion Island, Philippines.
bCycas rumphii plants from Halmahera Island, Indonesia.
53
cCycas rumphii plants from Sulawesi, Indonesia.
ch04.qxd 28/11/03 3:44 pm Page 54
54 A. Lindstrm
measuring plants of the morphologically similar Cycas seemannii and

C. micronesica. Nine of the 20 characters examined were found to be valuable for
discriminating between the two species. At least one of the characters
associated with the stem, leaf, pinna and megastrobilus was significant in
distinguishing between the species.
Table 4.4 presents the summary for the analyses of Cycas edentata and C.
rumphii. These two species can be distinguished based on four of the 16 charac-
ters examined: pinna number, spacing between pinnae, proximal pinna length
and megasporophyll lamina length. Although not presented here, the presence
of stout spines terminating the microsporophylls can also be used to distinguish
these two species.
Most importantly, it was the number of pinnae that was the only character
in all three analyses that was consistently significant in distinguishing between
morphologically similar species (Tables 4.2 to 4.4).
Discussion
A suite of morphological characters has been selected that can be used success-
fully to distinguish and compare different species in Cycas, even for species that
are morphologically similar to one another. The analysis of characters tested in
this project indicates that if measurements were obtained using the prescribed
suite of characters (Table 4.1), taxonomists would be able to document more rig-
orously the degree of similarities and differences among taxa. This type of data
would strongly augment herbarium specimen studies, as well as provide for a
more reliable classification system within Cycas.
In support of future taxonomic studies that are dependent on herbarium
specimens, specimens from plants within a population should, whenever possible,
represent both male and female material. Specimen collectors should obtain
specimens that best represent the norm within the range of morphological vari-
ation in any population. When possible, a whole leaf for each specimen should
be preserved. If this is not feasible, the fieldworker should thoroughly document
all of the relevant leaf characters (see Table 4.1) for inclusion in the specimen
annotation. Herbarium specimens that contain cataphylls, seeds, megasporo-
phylls and male cones should, if possible, represent each population sample.
Dried specimens can have loosely distinguishing features such as shape, size and
colour. Therefore, these features must be noted in the field and then transcribed
to the specimen label. A photographic image of the plant should accompany the
herbarium specimen.
It is recommended that a form similar to Hill (see Fig. 3.2, Chapter 3
this volume) is used when sampling a population prior to collecting herbar-
ium specimens for any Cycas population. Data derived from the completion
of such a form will: (i) document the variation within the population; (ii)
ensure that appropriate and taxonomically useful characters are assessed in the
field; and (iii) provide the necessary information for incorporation in the herbar-
ch04.qxd 28/11/03 3:44 pm Page 55
ium specimen annotations, thus serving the needs of future generations of Cycas
scientists.
Acknowledgements
I thank Ken Hill of the Royal Botanic Gardens (Sydney, Australia) who initiated
the use of a character sheet for the genus Cycas (see Hill, Chapter 3 this volume).
Kampon Tansacha, Director of Nong Nooch Tropical Botanical Garden
(Thailand) is most gratefully acknowledged for his committed support, both
financially and inspirationally. Kyle Williams helped in preparing this text.
ch04.qxd 28/11/03 3:44 pm Page 56
ch05.qxd 28/11/03 3:45 pm Page 57
Comments on Cycas, 5
Dyerocycas and Epicycas
(Cycadaceae)
Chia-Jui Chen,1 Ken D. Hill2 and Dennis Wm.

Stevenson3
1Institute of Botany, Chinese Academy of Sciences, Xiangshan,

Beijing, China; 2Royal Botanic Gardens, Sydney, Australia;
3Institute of Systematic Botany, New York Botanical Garden,
Bronx, New York, USA
Abstract
The recently described genus Epicycas and the obscure genus Dyerocycas (both separated
from Cycas) are evaluated, and generic concepts in Cycadaceae are discussed. Epicycas is a
superfluous and illegitimate name, based on the same type as Dyerocycas (i.e. Cycas
micholitzii). Dyerocycas is a valid name for a group distinguished by differences in leaflet mor-
phology, but the differences with Cycas are slight, and inconstant in other correlative char-
acters. A taxonomic treatment for the genus Cycas is presented, subsuming Dyerocycas as a
generic synonym. Relative diagnostic characters are tabulated and different stem types are
illustrated.
Introduction
Recently D.J. de Laubenfels described a new genus, Epicycas de Laubenfels, based

on a subterranean bulbous stem base (de Laubenfels and Adema, 1998), with
eight species from south-eastern Asia. He combined Schusters (1932) section
Indosinensis and Smitinands (1971) section Stangerioides in the new genus with the
type Epicycas micholitzii (Dyer) de Laubenfels. The core character for the genus is
plants with a mostly underground bulbous base, one or more leafy apices on the
surface of the bulb, each eventually developing a short, usually slender trunk or

ch05.qxd 28/11/03 3:45 pm Page 58
58 C.-J. Chen et al.
underground bulb branched and generally reaching the fertile stage before any
trunk forms.
This is unlike the case in cycad genera in other families, where above-ground
trunks are an unmodified continuation of the underground part, or where the
bulbous bases are not underground at all (Bowenia Hooker ex Hooker filius is an
exception, with a bulbous underground base and multiple slender underground
stems). Other characters of Epicycas are: (i) pinnae with flat, often undulate
margins; (ii) pinnae in some species dichotomously forked; (iii) pollen cones taper-
ingly cylindrical, while those of Cycas are mostly ovoid; and (iv) terminal lobes in
the sterile blade of megasporophylls often dichotomous, even in species with
simply pinnate leaves.
Fifty-five years before Epicycas was published, Dyerocycas was described by
Nakai (1943), named for botanist W.T. Thistleton-Dyer and based principally on
the presence of dichotomous pinnae. Other distinguishing characters cited were
a subterranean trunk, cylindrical or narrowly oblong pollen cones, and
microsporophylls with evenly scattered sporangia beneath unlike those in Cycas
sensu stricto with clusters of sporangia (sori) beneath. Dyerocycas micholitzii (Dyer)
Nakai (basionym Cycas micholitzii Dyer) was designated as the type of the genus.
Epicycas thus becomes a superfluous and illegitimate name based on the same
type as Dyerocycas (St Louis Code Articles 7.4, 11.5, and 52.1; see Greuter et al.,
2000).
Dyerocycas has a simple description but is a valid name. Although Epicycas is
an illegitimate name, it is described in detail and includes almost half the species
of Cycas in south-eastern Asia, especially in China. As the authors (de Laubenfels
and Adema, 1998) of Epicycas said:
Our knowledge of the species of the here newly described genus Epicycas is quite
incomplete, much of what we know has only recently been uncovered and
several specialists, particularly from China, are actively studying this material.
Therefore the treatment of Epicycas will have to be necessarily less complete than
that of Cycas.
In our recent fieldwork and laboratory studies of Chinese cycads, in conjunction

with herbarium studies both in China and abroad, we have attempted to examine
the new described genus in detail and establish its justification. The more deeply
our studies proceed, the more firmly we hold to Cycas as the only genus in
Cycadaceae.
Chen and Stevenson (1999) placed Epicycas as a synonym of Cycas. Hill
et al. (K.D. Hill, N.T. Hip and P.K. Loc, 2003, unpublished results) followed
this treatment, and made a short comment on it. To date, no other authors
have recognized either Dyerocycas or Epicycas, and no detailed assessment
of these names has been published. We comment below on the newly
described genera and discuss the genus concept in Cycadaceae sensu
stricto.
ch05.qxd 28/11/03 3:45 pm Page 59
Comments on Cycas, Dyerocycas and Epicycas 59
Characters Among Different Genera

In order to make comments on the genera Dyerocycas and Epicycas, and to compare
them with Cycas, it is necessary to summarize their diagnostic characters and bio-
geography (Table 5.1). The species selected in the table include all members of
the segregated genera and selected species of Cycas, especially from south-eastern
Asia, that show some morphological similarities. Species names are as published
in the segregate genera, without implying any acceptance of the segregates. The
observations and measurements in Table 5.1 originate from recent literature (Fu
et al., 1978; Jones, 1993; Chen and Yang, 1994a,b, 1996; Hill and Chen, 1994;
Walters and Yang, 1994; Wei, 1994; Chen et al., 1995; Wang, 1995; Wang and
Deng, 1995; Guan and Zhou, 1996; Hill, 1996; Wang et al., 1996; Chen and
Zhong, 1997; Yang et al., 1997; Chen and Saren, 1998; de Laubenfels and
Adema, 1998; Liu, 1998; Chen, 1999; Chen and Stevenson, 1999; Hill and Yang,
1999; Hill and Osborne, 2001; Hill et al. (K.D. Hill, N.T. Hip and P.K. Loc,
2003, unpublished results)); additional data were obtained from PE herbarium
and from our own fieldwork. The stem forms in Cycas are illustrated diagram-
matically in Fig. 5.1AF.
Discussion
Although the name is superfluous and illegitimate, the generic circumscription of
Epicycas becomes that of Dyerocycas under the rules of priority. Since formal trans-
fers of all species included in Epicycas have not been made to Dyerocycas, the group
will be discussed below under the name Epicycas. The circumscription of Epicycas,
including Smitinands (1971) section Stangerioides and Schusters (1932) section
Indosinenses, delimits a far wider range than the original Dyerocycas. In de
Laubenfels genus concept, Epicycas far exceeds the eight species listed by him in
1998 probably to at least 17 species in south-eastern Asia, about half the species
of Cycadaceae in this region.
Diagnostic characters inconsistent
As shown in Table 5.1, the diagnostic characters of Epicycas are almost all unsta-
ble and inconsistent. A mainly underground base as in E. micholitzii and E. multi-
pinnata (C. J. Chen & S.Y. Yang) de Laubenfels (Fig. 5.1A) is a principal defining
character for the new genus, but this character is neither wholly consistent within
the genus nor restricted to the species placed in the genus (Hill et al., 2002). For
example, the trunk of E. elongata (D. Yue Wang ex Leandri as Cycas elonga) de
Laubenfels sometimes gradually thickens toward the base without any subter-
ranean bulbous structure, and this species has an arborescent, often forked, stem
to 8 m in height (Fig. 5.1F). In E. siamensis (Miquel) de Laubenfels, the stem is
acaulescent or arborescent and to 2 m tall, but always abruptly swollen at the
ch05.qxd
28/11/03
Table 5.1. Character states among Dyerocycas, Epicycas and Cycas.
60
Stems Pinnae Pollen cones
Bulbous Pinna Width Flat
3:45 pm
Taxon Distribution base Branching Height (m) dissection (mm) margin Shape Size (cm) MLB
Dyerocycas micholitzii SE Asia + Unbranched 0.6 0.3 2 1528 + fc 3555 68 +

(Dyer) Nakai
[Epicycas micholitzii (Dyer)
Page 60
de Laubenfels]
Epicycas multipinnata ChinaVietnam + Unbranched 0.6 0.25 3 1024 + fc 1530 46 +
(C.J. Chen & S.Y. Yang)
de Laubenfels
C.-J. Chen et al.

E. tonkinensis (Linden & SE Asia + Unbranched 0.8 .3 1 1520 + c 1525 47 +
Rodigas) de Laubenfels
E. sp. novum #1 Vietnam + Dichotomous 0.54 0.10.3 1 1013 Unknown +
E. miquelii (Warburg) ChinaVietnam + Dichotomous 1 0.2 1 611 of 2030 68 +
de Laubenfels
E. elongata (D. Yue Wang SE Asia Dichotomous 48 0.50.8 1 511 + o 2025 89 +
ex Leandri as C. elonga)
de Laubenfels
E. siamensis (Miquel) SE Asia + Unbranched 2 0.3 1 58 + o 22 8
de Laubenfels
E. lindstromii (S.L. Yang, Vietnam + Dichotomous 0.1 0.1 1 711 + f 1220 45 +
K.D. Hill & Hip)
de Laubenfels
Cycas angulata R. Brown Australia + Unbranched 12 0.4 1 45 o 1525 1215
C. beddomei Dyer India + Congested 0.4 0.35 1 23.5 o 1520 57.5
C. changjiangensis N. Liu China + Congested 0.5 0.15 1 49 c 1525 46 +
ch05.qxd
28/11/03
C. circinalis Linnaeus S Asia Congested 5 0.45 1 912 + oc 3080 1011
C. debaoensis Y.C. China + Unbranched 0.7 0.4 3 815 + fc 1325 49 +
Zhong & C.J. Chen
C. ferruginea F.N. Wei ChinaVietnam + Dichotomous 0.8 0.18 1 69 f 2035 610 +
3:45 pm
C panzhihuaensis L. Zhou China Unbranched 13 0.250.3 1 67 c 2545 68 +
& S.Y. Yang
C. pectinata Asia Dichotomous 117 0.150.9 1 68 o 3055 1522
BuchananHamilton
Comments on Cycas, Dyerocycas and Epicycas

0.84 0.20.4 1520 811
Page 61
C. platyphylla K.D. Hill Australia + Unbranched 1 46 o
C. rumphii Miquel Asia, Pacific Adventitious 1.512 0.20.5 1 1218 oc 3545 1220 +
C. segmentifida D.Yue China + Unbranched 0.150.8 0.10.4 1 1218 + c 3060 512 +
Wang & C.Y. Deng
C. szechuanensis C.Y. China + Dichotomous 0.6 3 0.30.5 1 1015 + c 3560 5.510
Cheng, W.C. Cheng &
L.K. Fu
C. taiwaniana Carruthers China Dichotomous 28 0.20.8 1 816 + c 3560 711 +
C. tanqingii D.Yue Wang ChinaVietnam + Unbranched 0.82 0.150.35 1 1522 + c 3040 58 +
C. wadei Merrill Philippines + Unbranched 5 0.4 1 58 c 4070 910 +
+, present; , variable; , absent. MLB, megasporophyll lamina lobes branched. Pinna dissection: 1, none; 2, twice; 3, thrice. Pollen cone shape: c, cylindrical;
f, fusiform; o, ovoid.
61
ch05.qxd 28/11/03 3:45 pm Page 62
Fig. 5.1. Stem forms of Cycas and Epicycas. (A) Stem subterranean, mostly under-
ground [e.g. C. debaoensis Y.C. Zhong & C.L Cheng, E. micholitzii (Dyer) de
Laubenfels, E. multipinnata (C.J. Chen & S.Y Yang) de Laubenfels and E. tonkinen-
sis (Linden & Rodigas) de Laubenfels (= C. balansae Warburg)]. (B) Bulbous under-
ground base producing a slender aerial stem or branched short stem [e.g. C.
ferruginea F.N. Wei and E. miquelii (Warburg) de Laubenfels (= C. sexseminifera
F.N. Wei)]. (C) Stem almost all subterranean and frequently branched [e.g. C.
chevalieri Leandri and E. lindstromii (S.L. Yang, K.D. Hill & Hip) de Laubenfels].
(D) Trunk abruptly swollen at base [e.g. C. changjiangensis N. Liu and E. siamensis
(Miquel) de Laubenfels]. (E) Trunk simple and erect as in most species [e.g. C.
media R. Brown and C. panzhihuaensis L. Zhou & S.Y Yang]. (F) Trunk often
dichotomously branched [e.g. C. pectinata Buchanan-Hamilton and E. elongata
(D. Yue Wang ex Leandri as "C. elonga") de Laubenfels].
base (Fig. 5.1D). These stems are unlike the subterranean bulbous stem consid-
ered typical of Epicycas (Fig. 5.1AC). In Cycas, however, there are some species
possessing a more or less thickened or swollen stem at the base like those in
ch05.qxd 28/11/03 3:45 pm Page 63
Epicycas, e.g. Cycas changjiangensis N. Liu in Hainan, China (Fig. 5.1D), which has a
stem similar to Epicycas siamensis. Similar stems are seen sometimes in Cycas bed-
domei Dyer in south-east India, C. wadei Merrill in the Philippines, and even in C.
circinalis Linnaeus (the type of Cycas) in southern India.
Pinnae with flat and often undulate margins were also regarded as distinctive
of Epicycas. However, this character is not always consistent in Epicycas (Table 5.1),
especially in E. miquelii (Warburg) de Laubenfels and E. sp. novum #1, and it is also
common in Cycas in both Asia and Australasia. Even in C. circinalis, the type of
the genus, the pinnae show a flat margin and are often undulate. Dichotomous
pinnae appear in only a few species, e.g. Epicycas micholitzii, E. multipinnata and
Cycas debaoensis Y.C. Zhong & C.J. Chen. Dichotomous lobing of the sterile
lamina of the megasporophyll is common in many species of Cycas, especially in
Asia, but typical only in C. segmentifida D. Yue Wang & C.Y. Deng. Pollen cones in
Epicycas are not always tapering and cylindrical in shape; for example, E. elongata
and E. siamensis have ovoid pollen cones. Most species of Cycas in Asia also have
cylindrical pollen cones.
Subterranean and thickened stems shown to be specialized forms
Cycas, in general, like Dioon Lindley, Lepidozamia Regel and Encephalartos Lehmann,
is characteristically arborescent, except for a few species. The trunks retain a close
covering of narrow cataphylls and large overlapping leaf bases, and sometimes
show a prominent megasporophyll base ring for many years. These persistent
processes have a supportive and protective function (Norstog and Nicholls, 1997).
In Cycas there are two main stem types and six subtypes (Fig. 5.1):
I. Subterranean
1. Stem subterranean, at least half of stem underground, but with nearly
equal thickness in both ends (Fig. 5.1A), e.g. Cycas debaoensis, Epicycas
micholitzii, E. multipinnata and E. tonkinensis (Linden & Rodigas) de
Laubenfels (= C. balansae Warburg).
2. Bulbous underground base producing a slender aerial stem or many
branched small and short stems (Fig. 5.1B), e.g. Cycas ferruginea F.N. Wei
and Epicycas miquelii (= Cycas sexseminifera F.N. Wei).
3. Branched subterranean stem almost without aerial stems (Fig. 5.1C), e.g.
Cycas chevalieri Leandri and Epicycas lindstromii (S.L. Yang, K.D. Hill &
Hip) de Laubenfels.
II. Arborescent
4. Trunk swollen at base (Fig. 5.1D), e.g. Cycas changjiangensis and Epicycas sia-
mensis.
5. Trunk simple and erect (Fig. 5.1E), e.g. Cycas media R. Brown and Cycas
panzhihuaensis L. Zhou & S.Y. Yang.
6. Trunk erect and dichotomously branched (Fig. 5.1F), e.g. Cycas pectinata
Buchanan-Hamilton and Epicycas elongata.
ch05.qxd 28/11/03 3:45 pm Page 64
Epicycas, as circumscribed by de Laubenfels, includes five of the above stem types

(Fig. 5.1AD, F), i.e. E. micholitzii, E. multipinnata and E. tonkinensis with stem type
I-1, E. miquelii and E. sp. novum #1 with type I-2, E. lindstromii and partly E. miquelii
with type I-3, E. siamensis with type II-4 and E. elongata with type II-6. All stem
types also exist in Cycas (sensu stricto in de Laubenfels definition). This stem char-
acter is therefore not diagnostic for either genus.
The growth habitats of many of these species are often dry or seasonally dry
and sunny. The swollen or thickened stem in Epicycas and in some Cycas species
might therefore be a result of convergence in relatively similar environments.
The subterranean and thickened stems could be considered as specialized forms
adopted to survive particular harsh environmental conditions, such as drought
and repeated burning, minimizing the possibility of damage (Norstog and
Nicholls, 1997). These species, however, have been placed in different taxonom-
ic groups (Hill, 1995, 1999; Wang et al., 1996) and both morphology and micro-
morphology suggests that relationships of some of these species are not close (Xi
and Wang, 1989; Wang et al., 1996).
Epicycas miquelii (= Cycas sexseminifera) and C. ferruginea growing in crevices in
bare outcrops in rugged karst limestone hills in west Guangxi, China, and in
north Vietnam, are dwarf in habit with short, above-ground bulbous stems, often
with multiple branches when older. However, they are not closely related to C.
micholitzii (Wang et al., 1996). In contrast to the short stems adapted to dry envi-
ronments on limestone cliffs, Epicycas micholitzii (= Cycas micholitzii), E. multipinnata
and E. tonkinensis (= Cycas balansae) have mostly subterranean stems, and are
adapted to dense tropical forests. In these conditions, the plants may receive
limited light. Reduction of the emerging aerial stem, leaf number and size of the
megasporophyll cluster may decrease energy consumption, while lengthening
leaves and enlarging and even dividing the leaflets as in E. multipinnata would
allow more light assimilation in dense forests.
Sole character inadvisable
Epicycas is defined almost solely on having an underground bulbous base on

which one or more leafy apices grow, each eventually developing a short, usually
slender trunk, or branched bulb (de Laubenfels and Adema, 1998). In actual fact,
in most species of Epicycas, such as E. micholitzii, E. multipinnata, E. tonkinensis, E.
siamensis and at least partly in E. miquelii, the emergent trunks are an unmodified
continuation of the underground part, like those of Cycas (Fig. 5.1A). Only in E.
lindstromii and E. miquelii (Fig. 5.1C) are subterranean stems often branched to
produce modified, shortly slender aerial stems.
The subterranean habit is maintained by the activity of contractile elements
in roots and stems themselves, a phenomenon known in several groups of plants
and investigated in cycads most recently by Stevenson (1980) and Norstog and
Nicholls (1997). Stevenson found that, in arborescent cycads, the contractile
property is confined to the roots, but all subterranean forms also have contractile
ch05.qxd 28/11/03 3:45 pm Page 65
stems. In these, the stem itself is often observed to have external wrinkles. Stem
contraction, almost certainly a further adaptation to keep the stem underground,
correlates well with the development of a subterranean habit rather than with
evolutionary affinities (Norstog and Nicholls, 1997).
Although the subterranean stem appears to be a specialized, advanced char-
acter, no other consistent characters are correlated with this habit in defining an
Epicycas group. Dichotomous pinnae appear only in two or three species. Pinnae
with flat and sometimes undulate margins can be found in many species in both
genera. Most species in both genera, especially in Asia, commonly possess more
or less dichotomous lobes in the sterile blade of megasporophylls. The cylindri-
cal pollen cones are common in the two genera, especially in Asia. Therefore,
Epicycas is defined based solely on the bulbous underground base.
Sister species referred to the different genera
Some pairs of clearly sister species are placed in the different genera in de
Laubenfels and Adema (1998). A good example is that Cycas elongata (Leandri) D.
Yue Wang as C. elonga (Fig. 5.1F) and C. siamensis (Fig. 5.1D) are placed in
Epicycas, while Cycas pectinata (Fig. 5.1F) is referred to Cycas. All occur in south-
eastern Asia, and are closely allied to one another in their narrow pinnae with a
midrib flat above and raised below, ovoid pollen cone with a long apical spine on
each microsporophyll, and seed structure with a fibrous layer over a smooth scle-
rotesta.
Dyerocycas a natural genus?
Nakai (1943) separated Dyerocycas from Cycas and designated Dyerocycas micholitzii
as the type of the genus. This genus was defined principally by the subterranean
stems and dichotomous pinnae, correlated with cylindrical or narrowly oblong
pollen cones, microsporophylls with separate sporangia beneath, and megas-
porophylls with a flabellate-pectinate sterile blade. Dyerocycas when erected includ-
ed only D. micholitzii, but would now include two or three recently described
species in Cycas. This small group is readily distinguishable from Cycas (sensu stricto)
by the dichotomous pinnae. In this circumscription, Dyerocycas is more
likely a natural group, more restricted in membership than Epicycas and with
diagnostic characters discontinuous with those of Cycas and correlated with other
characters.
However, the members of Cycas sensu lato present a uniform appearance, the
same chromosome number (2n = 22), and common morphology such as megas-
porophylls foliiform, ovules more than two, seeds zygomorphic (platyspermic),
young pinnae coiled (circinate ptyxis), and pinnae with a midrib and no lateral
veins. The degree of difference between Nakais genus and Cycas sensu stricto is
small, and inconstant in other correlative characters such as these and the sub-
ch05.qxd 28/11/03 3:45 pm Page 66
terranean trunk and pollen cone shape. Therefore, the present authors consider
Cycas a natural genus, and Cycadaceae a monotypic family. The many current
infrageneric categories need further study to determine relationships between
species, but this does not hinder us in continuing to use the old binomial. The
changes of genera should not be undertaken unless there are strong taxonomic
reasons for the change (Davis and Heywood, 1963) and these have not yet been
demonstrated.
Nakai (1943) also described microsporophyllasubtus tota facie sporangiata in
his Dyerocycas, but as subtus sorifera in Cycas. In our examination of materials in
the different groups, no apparent pattern in microsporangium arrangement has
been revealed and in fact, at least in our examined species, microsporangia are
arranged radially in clusters of three to five microsporangia (sori) on the abaxial
surface of microsporophylls. Possibly, the radial symmetry of cycad sori suggests
a relationship to those of pteridosperms, e.g. Potoniea (Norstog and Nicholls, 1997)
or eusporangiate ferns such as the Marattiaceae (Stevenson, 1990).
Conclusions
1. Dyerocycas Nakai is an effectively and validly published name. Epicycas de

Laubenfels is a later name based on the same type as Dyerocycas Nakai (i.e. Cycas
micholitzii); therefore Epicycas must be a superfluous and illegitimate name.
2. Epicycas is defined principally based on a single character (the subterranean
stem), not on the sum total of characters manifested in Cycas. Even this charac-
ter is inconsistent within the genus Epicycas. There are also examples of sister
species placed in different genera. The Epicycas concept is not a natural group and
can not be accepted. Similarly, the name Dyerocycas based on the Epicycas circum-
scription cannot be accepted.
3. Dyerocycas in the original circumscription of Nakai may be a natural group,
but the relationship with the remaining species of Cycas is unclear and does not
appear to be a simple sister relationship. Therefore, Dyerocycas should be placed in
synonymy with Cycas.
The nomenclatural treatment for the genus Cycas thus becomes:
Cycas Linneaus (1753) Species Plantarum, 1188. LECTOTYPE: C. circinalis

Linnaeus, designated by Stevenson in Jarvis et al. (1993).
Synonym: Dyerocycas Nakai, 1943. Ordines, familiae, tribi, genera, sectiones,
species, varietates, formae et combinationes novae, a Prof. Nakai-Takenosin adhuc ut novis
edita, Appendix quaestiones characterium naturalim plantarum. Universitatis
Imperialis Tokyoensis. p. 208. TYPE: Dyerocycas micholitzii (Dyer) Nakai (=
Cycas micholitzii Dyer).
Synonym: Epicycas de Laubenfels in de Laubenfels and Adema (1998),
Blumea 43, 388. TYPE: Epicycas micholitzii (Dyer) de Laubenfels (= Cycas
micholitzii Dyer).
ch05.qxd 28/11/03 3:45 pm Page 67
References
Chen, C.J. (1999) Taxonomical and biogeographical studies on Cycas L. (Cycadaceae) in
China. In: Chen, C.J. (ed.) Biology and Conservation of Cycads. Proceedings of the Fourth
International Conference on Cycad Biology. International Academic Publishers, Beijing,
China, pp. 116128.
Chen, C.J. and Saren, J. (1998) Cycads in Asia, with notes on the conservation of cycads
in China. In: Chiu, S.T. and Peng, C.I. (eds) Proceedings of the Cross-strait Symposium on
Floristic Diversity and Conservation. National Museum of Natural Science, Taichung,
Taiwan, pp. 4764.
Chen, C.J. and Stevenson, D.W. (1999) Cycadaceae. In: Wu, Z.Y. and Raven, P.H. (eds)
Flora of China, Vol. 4, Cycadaceae through Fagaceae. Science Press, Beijing, China, and
Missouri Botanical Garden Press, St Louis, Missouri, pp. 17.
Chen, C.J. and Yang, S.Y. (1994a) Cycas multipinnata C.J. Chen & S.Y. Yang a remarkable
new cycad from China. Acta Phytotaxonomica Sinica 32, 239.
Chen, C.J. and Yang, S.Y. (1994b) Supplementary description of Cycas multipinnata C.J.
Chen & S.Y. Yang. Acta Phytotaxonomica Sinica 32, 480481.
Chen, C.J. and Yang, S.Y. (1996) Notes on Cycas multipinnata C.J. Chen & S.Y. Yang. Acta
Phytotaxonomica Sinica 34, 563564.
Chen, C.J. and Zhong, Y.C. (1997) Cycas debaoensis Y.C. Zhong & C.J. Chen a new cycad
from China. Acta Phytotaxonomica Sinica 35, 571.
Chen, C.J., Zhou, L., Yang, S.Y. and Han, Z.S. (1995) Cycas in China, with notes on its
conservation status. In: Vorster, P. (ed.) Proceedings of the Third International Conference on
177191.
Davis, P.H. and Heywood, V.H. (1963) Principles of Angiosperm Taxonomy. Oliver and Boyd,
Edinburgh, 556 pp.
De Laubenfels, D.J. and Adema, F. (1998) A taxonomic revision of the genera Cycas and
Epicycas gen. nov. (Cycadaceae). Blumea 43, 351400.
Fu, S.H., Cheng, W.C., Fu, L.K. and Chen, C.J. (1978) Cycadaceae. In: Cheng, W.C. and
Fu, L.K. (eds) Flora Reipublicae Popularis Sinicae 7. Science Press, Beijing, China, pp.
417.
Guan, Z.T. and Zhou, L. (1996) Cycads of China. Sichuan Science and Technology Press,
Chengdu, China, 242 pp.
Hill, K.D. (1995) Infrageneric relationships, phylogeny and biogeography of the genus
Cycas (Cycadaceae). In: Vorster, P. (ed.) Proceedings of the Third International Conference on
139162.
Telopea 7, 164.
Hill, K.D. (1999) Cycas an evolutionary perspective. In: Chen, C.J. (ed.) Biology and
Conservation of Cycads. Proceedings of the Fourth International Conference on Cycad Biology.
International Academic Publishers, Beijing, China, pp. 98115.
Hill, K.D. and Chen, C.J. (1994) On Cycas taiwaniana Carruthers (Cycadaceae) and the
cycads of south-eastern China. Acta Phytotaxonomica Sinica 32, 538548.
ch05.qxd 28/11/03 3:45 pm Page 68
Hill, K.D. and Osborne, R. (2001) Cycads of Australia. Kangaroo Press, Sydney, Australia,
116 pp.
Hill, K.D. and Yang, S.L. (1999) The genus Cycas (Cycadaceae) in Thailand. Brittonia 51,
4873.
Jarvis, C.E., Barrie, F.R., Allan, D.M. and Reveal, J.L. (1993) A list of Linnaean generic
names and their types. Regnum Vegetabile 127, 1100.
Jones, D.L. (1993) Cycads of the World Ancient Plants in Todays Landscape. Reed, Chatswood,
Australia, 312 pp.
Liu, N. (1998) A new species of the genus Cycas from Hainan Island, China. Acta
Phytotaxonomica Sinica 36, 552554. [Cycas changjiangensis.]
Nakai, T. (1943) Ordines, familiae, tribi, genera, sectiones, species, varietates, formae et combinationes
novae, a Prof. Nakai-Takenosin adhuc ut novis edita, Appendix Quaestiones characterium natural-
im plantarum. Universitatis Imperialis Tokyoensis, Tokyo, Japan, 208 pp.
Schuster, J. (1932) Cycadaceae. In: Engler, A. (ed.) Das Pflanzenreich, Fascicle 99, Vol. 4, Part
1, pp. 1168.
Smitinand, T. (1971) The genus Cycas Linn. (Cycadaceae) in Thailand. Natural History
Bulletin of the Siam Society 14 (12), 163175.
Stevenson, D.W. (1980) Observations on root and stem contraction in cycads (Cycadales)
with special reference to Zamia pumila L. Botanical Journal of the Linnaean Society 81,
275281.
Stevenson, D.W. (1990) Morphology and systematics of the Cycadales. In: Stevenson,
D.W. (ed.) The Biology, Structure, and Systematics of the Cycadales. Proceedings of the Symposium
CYCAD 87. Memoirs of the New York Botanical Garden 57, pp. 855.
Walters, T.W. and Yang, S.L. (1994) The cycads of China: findings from the Montgomery
Foundation/Fairchild Tropical Garden 1992 expedition. Journal of the Cycad Society 1,
611.
Wang, D.Y. (1995) A preliminary study of the Cycas micholitzii complex. Encephalartos 44,
3138.
Wang, D.Y. and Deng, C.Y. (1995) A new species of Cycas (Cycadaceae) from China.
Encephalartos 43, 1114. [Cycas segmentifida.]
Wang, F.X., Liang, H.B., Chen, T.Q. and Wang, D.Y. (1996) Cycads in China. Guangdong
Science and Technology Press,. Guangzhou, China, 295 pp.
Wei, F.N. (1994) A new cycad from Guangxi. Guihaia 14, 300. [Cycas ferruginea.]
Xi, Y.Z. and Wang, F.H. (1989) Pollen exine ultrastructure of extant Chinese gym-
nosperms. Cathaya 1, 119142.
Yang, S.L., Hill, K.D. and Hip, N.T. (1997) Cycas lindstromii. Novon 7, 213215.
ch06.qxd 28/11/03 3:45 pm Page 69
Classification Concepts in 6
Encephalartos (Zamiaceae)
Piet Vorster
Botany Department, University of Stellenbosch, Stellenbosch,

South Africa
Abstract
Much progress has been made towards recognizing and circumscribing species of
Encephalartos, and currently 65 species plus two subspecies are recognized. Some of the
characteristics used for circumscription of species could be used to construct a phyloge-
netic tree, but others have merely diagnostic value. When constructing a phylogenetic tree,
it is essential to keep in mind basic biological principles concerning reproductive behav-
iour, isolation and geographical distribution, because the first two especially are key factors
in evolution.
The following evidence is available for phylogenetic reconstruction: vegetative mor-
phology, morphology of reproductive structures, pollen morphology, leaflet anatomy,
chromosome number and morphology, chemical characteristics, isoenzyme profiles, DNA
analyses and geographical information.
Introduction
Taxonomy consists of several phases. The first phase is to recognize and circum-
scribe taxonomic units, be it families, genera, species or subspecies. In practice
this is done on morphological evidence, but it is meaningless unless reproductive
behaviour as a driving force of evolution is recognized. The second phase is
naming of the circumscribed taxonomic units, an activity often denigrated by
taxonomists who do not understand the significance of a correct and practical
nomenclature. The third phase comprises the determination of evolutionary
relationships between the previously circumscribed taxonomic units. This is the
ch06.qxd 28/11/03 3:45 pm Page 70
70 P. Vorster
most difficult phase; it involves assessing all the available evidence and deducing
the relationships that we find most satisfactory. What makes this phase even more
difficult is that practically all the available evidence represents a status at the
present time, whereas evolution is a process that in the case of cycads spans a
period of tens of millions of years.
Much progress has been made towards recognizing and circumscribing
species of Encephalartos Lehmann and currently 65 species plus two subspecies are
recognized (Hill et al., Appendix 1 this volume). Much of the known distribution
range outside South Africa is still imperfectly explored, even in respect of such
relatively large and conspicuous plants as is the case for most species of
Encephalartos, and it seems likely that more species will be discovered. The leading
South African cycad botanist, Dr R.A. Dyer (personal communication), was
against recognizing infraspecific categories in Encephalartos and the present author
has followed that principle. Recognition of infraspecific categories implies a
closer evolutionary relationship between the infraspecific taxa than between
species, and such views cannot be substantiated with currently available evidence.
In this regard, caution should also be exercised when using genetic distance when
analysing molecular data. As long ago as 1930 Du Rietz argued convincingly that
the taxonomic discontinuity between infraspecific taxa belonging to the same
species may be greater than between different species, because reproductive
behaviour is the prime isolating factor in speciation. It follows that an apparent-
ly minor genetic change may cause total reproductive isolation between two oth-
erwise genetically identical taxa which would then behave like separate species;
and conversely that two taxa at opposite ends of a somewhat discontinuous range
of variation may differ genetically to a considerable extent, but still be reproduc-
tively compatible and exhibit many common features.
Some of the characteristics used for circumscription of species could be used
to construct a phylogenetic tree, but others have merely diagnostic value. At this
stage of knowledge of Encephalartos, the evolutionary significance of morphological
characteristics simply is not known and caution about convergence must remain.
When constructing a phylogenetic tree, it is essential to keep in mind basic
biological principles concerning reproductive behaviour, isolation and geograph-
ical distribution, because the first two especially are key factors in evolution.
Available Evidence
Vegetative morphology
The vegetative morphology of Encephalartos has never been studied as such, but
is well known although poorly understood in respect of its evolutionary signifi-
cance. In comparison with other living genera of the Cycadales, Encephalartos is
well endowed with taxonomically useful vegetative characteristics (Vorster, 1993).
From observations on cultivated plants it is evident that some of these character-
istics are plastic.
ch06.qxd 28/11/03 3:45 pm Page 71
Classification Concepts in Encephalartos 71
Firstly, some characteristics change unrecognizably with the age of the plant.
For example, in immature plants, leaves of basal offshoots and the first leaves pro-
duced after defoliation tend to have heavily dentate leaflets, while a few leaf
flushes later the leaflets on the same plant may be completely entire. Thus in
Encephalartos woodii Sander the very characteristic pattern of dentation in such
early leaves vanishes completely in later-produced leaves. Similarly, many species
produce copious amounts of trichomes at the apices of the stems prior to the
emergence of cones or new foliage, and in some cases the emerging leaves them-
selves are hairy; but in some species the trichomes on the stem apices are not per-
sistent, and in most species the hairs on the foliage are caducous. Useful
information may yet be discovered in a study of trichome types at species level
(Stevenson, 1981). For this same reason seedling plants often bear no similarity to
mature plants and are consequently difficult to identify. Apart from illustrations
of seedling leaflets of a few species (Giddy, 1974; Jones and Wynants, 1997), it is
not known what seedlings look like.
Secondly, some characteristics are influenced by environmental factors. For
example, in Encephalartos equatorialis P.J.H. Hurter, leaflets characteristically
overlap (Vorster and Heibloem, 1995), but if a leaf develops under less bright
conditions, leaflets are more widely spaced and do not overlap. Similarly, the
angle between opposing leaflets is used in descriptive taxonomy, but this angle
can also be much larger than normal under more moist or shady conditions.
Morphology of reproductive structures
The morphology of reproductive structures in Encephalartos is also reasonably well

known, but again poorly understood. Until fairly recently, species of Encephalartos
were defined in terms of vegetative characteristics, but Dyer (1956, 1965) distin-
guished E. princeps R.A. Dyer from E. lehmannii Lehmann, E. arenarius R.A. Dyer
from E. latifrons Lehmann, and E. trispinosus (Hooker) R.A. Dyer from E. horridus
(Jacquin) Lehmann on their cone morphologies. Vorster (1993) listed cone char-
acteristics used in taxonomy and described a number of new species in terms of
their cones (Robbertse et al., 1988a,b; Vorster, 1990, 1995, 1996a,b,c). The
underlying philosophy is that the reproductive processes are of prime importance
in affecting and maintaining reproductive isolation as a factor in speciation.
Although it is possible to divide many species into groups on the basis of cone
morphology, the evolutionary significance of these characteristics is unclear.
Pollen morphology
The pollen morphology has been studied in a number of South African cycad
species (Dehgan and Dehgan, 1988; Marshall et al., 1989a). Unlike in some other
plant groups, the pollen grains within Encephalartos seem to be very similar
between species in respect of shape and surface sculpturing, and unlikely to con-
ch06.qxd 28/11/03 3:45 pm Page 72
72 P. Vorster
tribute towards the taxonomy and understanding of the phylogeny of the genus.
Nevertheless, the pollen grains of unstudied species should be studied to com-
plete the database.
Leaflet anatomy
Leaflet anatomy has been studied in some South African cycads (Koeleman et al.,
1981; Spreeth and Vorster, 1995), but should be investigated for all species. It is
also necessary to study several samples of every species, taken throughout their
distribution range, in order to assess variation within species. This neglected type
of evidence is very likely to yield phylogenetically useful information, though
convergence and divergence are ever-present pitfalls. Useful information may be
discovered in a study of the morphology of wax platelets deposited on the leaf
surfaces (Osborne and Stevens, 1996; Hill and Stanberg, 1999).
Chromosome number and chromosome morphology
The chromosome number has been determined for only a few Encephalartos
species (Marchant, 1968; Mogford, 1981; Moretti, 1990), and in all these is 2n =
18. Breeding experiments with cultivated plants suggest reproductive barriers
between some species within the same genus. For instance, the group of species
(Group 1, see later) comprising E. brevifoliolatus Vorster, E. cycadifolius (Jacquin)
Lehmann, E. friderici-guilielmi Lehmann, E. ghellinckii Lemaire, E. humilis
I. Verdoorn, E. laevifolius Stapf & Burtt Davy and E. lanatus Stapf & Burtt Davy
seems to have a cytological reproductive barrier against hybridization outside this
group. Similarly, E. inopinus R.A. Dyer seems to be reproductively incompatible
with any other species, and attempts to hybridize E. ferox Bertolini filius with other
species have met with a very low success rate. Chromosome number for all
species should be determined as a matter of priority, while individual karyotypes,
which may explain reproductive incompatibility between species, should be
documented.
Associated organisms
Associated organisms have not been studied in proportion to their possible

importance to the taxonomy, and indeed survival, of Encephalartos. Apart from the
leaf-eating larvae of Zerenopsis leopardina Felder (Lepidoptera: family
Geometridae), a considerable number of snout beetle species (Coleoptera: family
Curculionidae) appear to be associated exclusively in nature with different species
or species groups of Encephalartos (Oberprieler, 1995) and have been used in the
taxonomy of Encephalartos (Vorster and Oberprieler, 1999). Some of these have
been demonstrated to be pollinators (Donaldson et al., 1995); others, like species
ch06.qxd 28/11/03 3:45 pm Page 73
of Antliarrhinus Schoenherr, are seed predators; but the majority seem to live and
breed on the plants without playing any part in the plants reproductive process-
es. Oberprielers ongoing studies on the taxonomy and phylogeny of these beetles
may well be significant in reconstructing the phylogeny of Encephalartos. Norstog
(1987) and Norstog and Nicholls (1997) speculated that the beetlecycad rela-
tionship is very old, at least in respect of pollinating beetles, and the phylogeny
of the beetles may be surmised to mirror that of Encephalartos, adding another
type of phylogenetic evidence. The same argument applies to the symbiotic
Cyanophyta which inhabit and probably cause development of coralloid roots.
Numerous studies have been published on the physiology of these (e.g. Pate et al.,
1988), but of more importance to taxonomy is the work on the identity and speci-
ficity of the Cyanophyta by Grilli Caiola (1975), Grobbelaar et al. (1987),
Marshall et al. (1989b) and Grobbelaar and Marshall (1993). Grobbelaar et al.
(1987) isolated seven species of Cyanophyta, in two genera, from 31 species of
Encephalartos. It is possible that these Cyanophyta have a wide tolerance of host
species (Grobbelaar et al., 1987) and thus may not contribute much to phyloge-
netic reconstruction in Encephalartos. Lastly, Vovides (1991) demonstrated the asso-
ciation of mycorrhiza with Mexican cycads. Should it be proved that mycorrhiza
are associated with all the Cycadales, it may be profitable to study the taxonomy
and partner-specificity of the fungus involved. Regrettably it seems unlikely that
the associated Cyanophyta and mycorrhiza will be studied in sufficient detail in
the foreseeable future.
Chemical characteristics
The chemical characteristics of Encephalartos are poorly known. All species inves-
tigated have methylazoxymethanol (MAM) glycosides (De Luca et al., 1980;
Moretti et al., 1983), but finer distinctions that may illuminate evolutionary
history have not been recorded, and it seems as if the MAM glycosides do not
show the kind of variability found, for example, in alkaloids in other plant
groups. Most, if not all, species have deposits of alkane waxes on leaf surfaces
(Osborne and Stevens, 1996), the thickness and crystallization pattern of the
deposits being responsible for the range of leaf colours from bright green to
silvery grey. At least the thickness of the deposits seems to have taxonomic impli-
cations, and it is conceivable that thick deposits evolved independently in differ-
ent parts of the geographical range. The chemistry of these waxes was studied
by Osborne et al. (1989) and a study of their chemical evolution may shed light
on the phylogeny of the species. Stephens and Stephen (1988) found that sugars
resulting from acid hydrolysis of gummy exudates varied considerably between
three species examined. Flavonoids should be investigated, because the analysis
is relatively inexpensive and has proved useful in other plant groups.
Perhaps the phylogenetically most important chemical compounds associat-
ed with Encephalartos are the volatile substances released by cones, especially male
cones, some details of which have been reported by Pellmyr et al. (1991) and Tang
ch06.qxd 28/11/03 3:45 pm Page 74
74 P. Vorster
(1993). These have far-reaching taxonomic implications, because they serve to

attract very particular species of pollinating vectors. This is especially important
in areas where several species of Encephalartos occur together and cone simulta-
neously, yet where hybridization seldom occurs. Apart from characterizing the
constituent compounds responsible for a specific odour, the study should take
account of the synthesis of the compounds and mechanisms by which the differ-
ent compounds can evolve. The relationship between volatile substances of dif-
ferent Encephalartos species is likely to mirror that of the plants themselves.
Isoenzymes
Isoenzymes have yielded promising results in other cycad genera (Walters and
Decker-Walters, 1991) and have been compared for a few South African species
(Van der Bank et al., 1998, 2001). This method is less expensive than DNA com-
parisons, and in a pilot study (Van der Bank et al., 2001) yielded results that are
more compatible with morphological evidence than DNA. While this type of evi-
dence has relatively high resolution and is a useful tool in circumscribing and
even identifying species, its value in reconstructing phylogenetic relationships
is less tested. The method is also difficult to use, because freshly collected mate-
rial has to be fixed in liquid nitrogen, which presents logistic difficulties in the
field. The technical constraints of the method largely preclude its application to
wild plants, but it is ideally suited to cultivated material, which fortunately
abounds.
DNA analysis
DNA analysis is the current fashionable evidence of phylogenetic history. At face

value, it certainly seems to be an appropriate approach, because it is the genes
that change in evolutionary processes. However, it should be kept in mind that
only a very small part of the genome can at present be analysed and compared
between species. It is not known whether this portion of the genome plays any
part in the evolution of any species, because many completely different factors
may result in reproductive isolation and consequent evolution.
Van der Bank et al. (2001) used rbcL and internal transcribed spacer 1 and 2
sequences in an exploratory study of some Eastern Cape cycads, which were
assessed together with isoenzyme and morphological evidence. On the whole the
results were acceptable against the background of morphological and geograph-
ical evidence, but should be accepted with reservations because the work implies
that the rare phenomenon of strong and characteristic lobing of the leaflets arose
independently in different Eastern Cape species but nowhere else in the genus. In
view of the perceived antiquity of cycads and their slow reproduction rate, a sur-
prising deduction was that all the species examined were derived from a common
ancestor less than a million years ago. Overall there was little genetic variation in
ch06.qxd 28/11/03 3:45 pm Page 75
the material examined, similar to results obtained through isoenzyme analysis for
Macrozamia Miquel by Sharma et al. (1998, 1999).
As part of an MSc project, M.E. Coetzer (2000, University of the Orange
Free State, Bloemfontein, South Africa, unpublished) intended to carry out a
molecular study of the whole genus by means of DNA amplification fingerprint-
ing and random amplified polymorphic DNA techniques. She had at her dispos-
al material of most species in the genus, but was not successful in extracting DNA
from all the species. Neither did she have sufficient replicates of all the species.
The results of her analysis are disappointing in that her phylogenetic tree shows
little agreement with morphological and geographical evidence.
Nevertheless, this type of evidence holds promise, and both ribosomal and
chloroplast DNA data should be gathered for all species, with sufficient replicates
from different parts of the distribution range, and using the widest possible range
of genes. However, this method should be applied only in collaboration with
botanists who understand the plant groups biology.
Geographical evidence
The non-taxonomist is often perplexed by the value accorded by taxonomists to

geographical evidence. Yet this type of evidence is an obvious reflection of the
evolutionary process, often indicative of groups of related species that are geo-
graphically grouped together. Species that became reproductively isolated from
each other relatively recently tend to occur in closer geographical proximity to
each other than species that separated reproductively in the more distant past,
reflecting the amount of time available to these species to disperse. The fact that
at present there is little evidence of any species occupying a very wide geograph-
ical range (the exceptions being Encephalartos altensteinii Lehmann/E. natalensis
R.A. Dyer & I. Verdoorn, E. villosus and E. barteri Carruthers ex Miquel), as well
as the absence of any marked geographical disjunctions (the exception being E.
laevifolius), seems to be indicative of relatively recent speciation of the living
species, which is supported by the small amount of genetic variation.
Viewed on a geographical basis, certain parts of the distribution range of the
genus contain groups of species with common morphological characteristics,
such as strongly lobed leaflets in the Eastern Cape Province of South Africa,
white wool on the emerging foliage in Zimbabwe and Mozambique, green and
cylindrical stalked female cones in central Africa, successive emergence of male
cones in tropical Africa, and a dwarf habit plus soapy green leaves in the tropi-
cal savanna. It is not unreasonable to consider these individual groups as closely
related species. In view of the large number of species involved, the difficult
accessibility in nature of many of the species, the time and expense required to
collect all the data for the phylogenetic assessment, and the unfortunate fact that
research funding is closely linked to frequent publications, it seems prudent to
divide the genus into manageable geographically/morphologically similar por-
tions which can be studied in succession. Any anomalies can be ironed out at the
ch06.qxd 28/11/03 3:45 pm Page 76
76 P. Vorster
final assessment of the whole genus. It would, however, not be wise to accept
blindly these geographical/morphological groups as phylogenetic groups and use
only one or a few representatives of each group to determine relationships
between them.
The following groups are tentatively circumscribed and their distributions
are illustrated in Fig. 6.1:
Group 1: Encephalartos brevifoliolatus Vorster, E. cycadifolius

(Jacquin) Lehmann, E. friderici-guilielmi Lehmann, E. ghellinckii
Lemaire, E. humilis I. Verdoorn, E. laevifolius Stapf & Burtt Davy
and E. lanatus Stapf & Burtt Davy. Together these species occur over a
wide area in the eastern part of South Africa, in most cases favouring relative-
ly cool and high elevations that receive frost and even snow in winter. They are
readily distinguished by their narrow (almost linear) leaflets and hairy cones.
They produce cones 6 months out of phase with the other species (in spring
rather than in autumn), and the female cones disintegrate after 3 months
instead of staying on the plants for 612 months or longer. In cultivation they
could not be hybridized with any species outside this group.
Group 2: Encephalartos arenarius R.A. Dyer, E. horridus (Jacquin)
Lehmann, E. latifrons Lehmann, E. lehmannii Lehmann, E. longi-
folius (Jacquin) Lehmann and E. trispinosus (Hooker) R.A. Dyer.
These species occur in a relatively small area in the Eastern Cape Province of
South Africa, usually in dry conditions. They carry solitary cones, and several
of them have markedly glaucous leaves and/or very prominently spiny-lobed
leaflets. None of them has prickles on the petiole.
Group 3: Encephalartos princeps R.A. Dyer. This species has a very
strong vegetative likeness to E. lehmannii, but bears multiple cones and is prob-
ably not very closely related. This is borne out by Van der Bank et al.s (2001)
molecular results.
Group 4: Encephalartos aplanatus Vorster, E. caffer (Thunberg)
Lehmann, E. cerinus Lavranos & D.L. Goode, E. ngoyanus
I. Verdoorn, E. umbeluziensis R.A. Dyer and E. villosus Lemaire.
These species, loosely referred to as the E. villosus group, occur over a long dis-
tance along the eastern part of South Africa, often close to the coast. They
share a non-emergent caudex, relatively few and initially erect leaves, exter-
nally rather poor sexual dimorphism of the cones and sporophyll apices drawn
out into drooping and often fringed lips.
Group 5: Encephalartos altensteinii Lehmann/E. natalensis R.A.
Dyer & I. Verdoorn and E. transvenosus Stapf & Burtt Davy. These
occur along the eastern coast, and the eastern and northern escarpment of
South Africa. They share a large stature and bright green leaves, and their
cones are almost indistinguishable so that they are defined in terms of their
individual diagnostic foliage morphologies.
Group 6: Encephalartos woodii Sander. This species, known from a
single gathering in northern KwaZulu-Natal in South Africa, closely resembles
ch06.qxd 28/11/03 3:45 pm Page 77
Encephalartos delucanus
E. marunguensis
E. poggei
E. schaijesii
E. schmitzii Encephalartos mackenziei
E. macrostrobilus
E. septentrionalis
Encephalartos bubalinus
E. equatorialis
Encephalartos barteri E. hildebrandtii
E. ituriensis
E. kisambo
E. sclavoi
Encephalartos laurentianus E. tegulaneus
E. whitelockii
Encephalartos inopinus
Encephalartos cupidus
E. dolomiticus Encephalartos gratus
E. dyerianus
E. eugene-maraisii
E. hirsutus Encephalartos
E. middelburgensis chimanimaniensis
E. nubimontanus E. concinnus
E. manikensis
Encephalartos brevifoliolatus E. munchii
E. cycadifolius E. pterogonus
E. friderici-guilielmii E. turneri
E. ghellinckii
E. humilis Encephalartos heenanii
E. laevifolius E. paucidentatus
E. lanatus E. relictus
Encephalartos altensteinii/natalensis Encephalartos ferox

E. transvenosus
Encephalartos aemulans
Encephalartos arenarius E. lebomboensis
E. horridus E. msinganus
E. latifrons E. senticosus
E. lehmannii
E. longifolius
E. trispinosus Encephalartos princeps Encephalartos woodii
Encephalartos aplanatus
E. caffer
E. cerinus
E. ngoyanus
E. umbeluziensis
E. villosus
Fig. 6.1. Known geographical distribution of the genus Encephalartos in Africa,

showing geographical groups of species with common characteristics.
some forms of E. natalensis. It is cryptically distinguished by its softer-textured

and bright glossy green foliage, and its juvenile-type leaflets have a character-
istic ovate profile and three to four conspicuous teeth near the base of the
upper margin, a characteristic that is not found in any other South African
species but is seen as a regular feature in a number of equatorial species such
ch06.qxd 28/11/03 3:45 pm Page 78
78 P. Vorster
as E. hildebrandtii A. Braun & Bouch, E. ituriensis Bamps & Lisowski and E.

kisambo Faden & Beentje.
Group 7: Encephalartos aemulans Vorster, E. lebomboensis
I. Verdoorn, E. msinganus Vorster and E. senticosus Vorster. These
cycads, of northern KwaZulu-Natal of South Africa, vegetatively resemble the
previous group and may be phylogenetically close to it, but are distinguished
by their individual female cone morphologies. Encephalartos aemulans and espe-
cially E. msinganus are probably the closest to the previous group on account of
their warty female cones, while E. lebomboensis and E. senticosus are more distant
but closer to each other on account of their apricot-yellow female cones with
short downy indumentum.
Group 8: Encephalartos heenanii R.A. Dyer, E. paucidentatus
Stapf & Burtt Davy and E. relictus P.J.H. Hurter. These occur on
mountains surrounding Swaziland. They share corrugately raised veins on the
abaxial leaflet surfaces, otherwise known only in E. latifrons. They have widely
differing cones. While E. heenanii and E. relictus are vegetatively similar, E. pau-
cidentatus has a quite different aspect.
Group 9: Encephalartos cupidus R.A. Dyer, E. dolomiticus
Lavranos & D.L. Goode, E. dyerianus Lavranos & D.L. Goode,
E. eugene-maraisii I. Verdoorn, E. middelburgensis Vorster,
E. nubimontanus P.J.H. Hurter and possibly the recently-
described but almost unknown E. hirsutus P.J.H. Hurter. These
species occur on the northern escarpment of South Africa, mostly at relative-
ly cool and high elevations. They share conspicuously glaucous foliage, but
there are differences in leaf morphology. The female cones of E. eugene-maraisii
and E. middelburgensis are very similar, as are those of E. nubimontanus, E. cupidus
and E. dyerianus, while those of E. dolomiticus are quite different. Encephalartos hir-
sutus was very poorly described and is still virtually unknown.
Group 10: Encephalartos inopinus R.A. Dyer. This occurs in the same
geographical area as the previous group. It is an enigma in the genus. It has an
arresting vegetative appearance quite unlike any other, with glaucous and
entire falcate drooping leaflets. Being glabrous, glaucous and having smooth
scale surfaces, the female cones are almost indistinguishable from those of E.
dyerianus. It has not been found possible to hybridize E. inopinus with any other
species.
Group 11: Encephalartos chimanimaniensis R.A. Dyer &
I. Verdoorn, E. concinnus R.A. Dyer & I. Verdoorn, E. manikensis
(Gilliland) Gilliland, E. munchii R.A. Dyer & I. Verdoorn, E. ptero-
gonus R.A. Dyer & I. Verdoorn and perhaps E. turneri Lavranos &
D.L. Goode. These occur along the ZimbabweMozambique border. They
all have relatively soft foliage, leaflets reduced to spines towards the base of the
leaf, distinct short whitish wool on the emerging fronds, and green and
glabrous cones of which the females are ovoid. Their stems tend to have a soft
texture such as is usually associated with species like those of the E. villosus
group or E. ferox.
ch06.qxd 28/11/03 3:45 pm Page 79
Group 12: Encephalartos gratus Prain. This occurs in southern Malawi

and adjacent Mozambique. It has soft-textured foliage like the previous group,
but differs by the cones being dark salmon pink instead of green. This pig-
mentation is shared by only E. ferox, which is probably not closely related. On
the other hand, the cylindrical female cones and successively emerging male
cones of E. gratus may point to an affinity with the next group.
Group 13: Encephalartos bubalinus Melville, E. equatorialis P.J.H.
Hurter, E. hildebrandtii A. Braun & Bouch, E. ituriensis Bamps &
Lisowski, E. kisambo Faden & Beentje, E. sclavoi De Luca, D.W.
Stevenson & A. Moretti, E. tegulaneus Melville (including subsp.
powysii Miringu & Beentje) and E. whitelockii P.J.H. Hurter. These
cycads are distributed on both sides of the equator. They all have hard-tex-
tured leaves, and the leaflets progressively reduce to a series of prickles so that
there is no clear petiole. Several of the species have three to four teeth grouped
together near the base of the upper leaflet margin. The cones are green to
yellow, glabrous; the female cones are cylindrical in several species but ovoid in
others, with the facets smooth and the terminal facet only moderately raised.
The male cones tend to emerge in succession. Encephalartos sclavoi is included
here for geographical reasons; but its smaller stature, and its few cones (one to
two) plus the ovoid shape of the female cone, may signify a more distant rela-
tionship to the rest of the group.
Group 14: Encephalartos macrostrobilus S. Jones & Wynants,
E. mackenziei L.E. Newton and E. septentrionalis Schweinfurth. It
seems doubtful if any botanist has seen E. septentrionalis since Schweinfurth dis-
covered it in the southern Sudan in 1869; consequently the identity of this
species is uncertain. Plants from Moyo in Uganda are ascribed to E. septentrion-
alis. Encephalartos macrostrobilus was described from north-eastern Uganda, but
so poorly that it is impossible to visualize the species. Plants from near Didinga
in south-eastern Sudan have recently been described under the name
E. mackenziei (Newton, 2002). All species in this group have leaflets reduced to
prickles towards the base of the leaf, all have three to four teeth near the base
of the upper leaflet margin, all have green and cylindrical female cones with a
sparse and ephemeral russet-brown indumentum and all have numerous male
cones emerging in succession. Except for their soft-textured leaflets, these
plants are very similar to the previous group and may well belong there.
Group 15: Encephalartos laurentianus De Wildeman. This is known
only from a small area on the AngolaCongo border. It resembles the previous
two groups by the leaflets being reduced to a series of prickles towards the base
of the leaf and the male cones emerging in succession. It differs most conspic-
uously by its gigantic stature, its soft-textured foliage with the leaflets lacking
the concentration of three to four teeth near the base of the upper margin, and
the short russet-brown indumentum on the cones (a similar but sparser indu-
mentum occurs on the group 14 species). The morphological and geographi-
cal discontinuities probably indicate a considerable evolutionary divergence
from all the other species.
ch06.qxd 28/11/03 3:45 pm Page 80
80 P. Vorster
Group 16: Encephalartos barteri Carruthers ex Miquel (including

subsp. allochrous L.E. Newton). This is geographically well isolated,
though it occurs over a wide area of Ghana, Benin and Nigeria. The plants are
somewhat nondescript. The leaves are erect and straight, with leaflets soft-tex-
tured, rather narrow, and reduced to prickles towards the base of the leaf. The
cones are ovoid, glabrous, and green to olive green, the female cones being
sessile. This is a poorly known group.
Group 17: Encephalartos delucanus Malaisse, Sclavo & Crosiers,
E. poggei Ascherson, E. marunguensis Devred, E. schaijesii
Malaisse, Sclavo & Crosiers and E. schmitzii Malaisse. These cycads
occur in the savanna of Zambia, Tanzania, Congo and Angola on fire-prone
sandy flats where there seems to be no obvious barrier against migration. They
are small, almost dwarf plants. The stems are underground or shortly emer-
gent. The leaves are deciduous during the hottest and driest part of the year,
soapy glaucous green; the leaflets are entire or less commonly with a few small
teeth and gradually reduced to spines towards the base of the leaf. The cones
emerge glaucous green and may turn yellowish at maturity. They are rather
nondescript, but the female cones are ovoid and sessile with the males shortly
stalked. Morphologically and geographically they are distinct from other
groups, but may be variants of one species.
Group 18: Encephalartos ferox Bertolini filius. This occurs on the
coast of Mozambique and the KwaZulu-Natal province of South Africa, on
white beach sand immediately next to the ocean. It is morphologically very dis-
tinct, with a shortly emergent and soft textured stem, hard-textured dull green
leaves and broad leaflets with teeth all around. The cones are conspicuously
salmon-red. This species appears to stand apart from all others and has proved
difficult to hybridize with any other species in cultivation.
Postcript
The greatest danger when compiling a phylogenetic tree is to rely, for example,
on cytological, chemical or molecular evidence and build a phylogeny on any
single type of evidence. Such a tree cannot truly be called a phylogenetic tree but
is really a cytological tree, a chemical tree or a molecular tree.
Finally, it should always be remembered that a phylogenetic tree is an
approximation only, based on the seemingly most logical interpretation of the
available evidence, with many missing components because we have only the
presently available evidence of a process that spans tens of millions of years. As
such it can never be final.
References
De Luca, P., Moretti, A., Sabato, S. and Siniscalco Gigliano, G. (1980) The ubiquity of
cycasin in cycads. Phytochemistry 19, 22302231.
ch06.qxd 28/11/03 3:45 pm Page 81
Donaldson, J.S., Nnni, I. and de Wet Bsenberg, J. (1995) The rle of insects in the pol-
lination of Encephalartos cycadifolius. In: Vorster, P. (ed.) Proceedings of the Third
International Conference on Cycad Biology. Cycad Society of South Africa, Stellenbosch,
South Africa, pp. 423434.
Du Rietz, G.E. (1930) The fundamental units of biological taxonomy. Svensk Botanisk
Tidskrift 24, 333427.
Dyer, R.A. (1956) A new cycad from the Cape Province. Journal of South African Botany 22,
14. [Encephalartos arenarius.]
Dyer, R.A. (1965) New species and notes on type specimens of South African Encephalartos.
Journal of South African Botany 31, 111121. [Encephalartos princeps.]
Giddy, C. (1974) Cycads of South Africa. Purnell, Cape Town, South Africa, 122 pp.
Goode, D. (2001) Cycads of Africa, Vol. 1. Cycads of Africa Publishers, Gallomanor, South
Africa, 351 pp.
Grilli Caiola, M. (1975) A light and electron microscopic study of blue-green algae
growing in the coralloid roots of Encephalartos altensteinii and in culture. Phycologia 14,
2533.
Grobbelaar, N. and Marshall, J. (1993) Specificity of the cycadcyanobiont symbiosis. In:
Stevenson, D.W. and Norstog, K.J. (eds) The Biology, Structure and Systematics of the
Societies of Australia Limited, Milton, Queensland, Australia, pp. 162164.
Grobbelaar, N., Scott, W.E., Hattingh, W. and Marshall, J. (1987) The identification of
the coralloid root endophytes of the southern African cycads and the ability of the
isolates to fix dinitrogen. South African Journal of Botany 53, 111118.
Hill, K.D. and Stanberg, L.C. (1999) Epicuticular waxes in the Cycadales, and their sys-
tematic implications. In: Chen, C.J. (ed.) Biology and Conservation of Cycads. Proceedings
of the Fourth International Conference on Cycad Biology. International Academic Publishers,
Beijing, China, pp. 159174.
Jones, S. and Wynants, J. (1997) Encephalartos macrostrobilus (Zamiaceae), a new cycad
species from northern Uganda. Encephalartos 50, 1316.
Koeleman, A., Robbertse, P.J. and Eicker, A. (1981) Die anatomie van die pinnas van die
Suid-Afrikaanse spesies van Encephalartos Lehm. Journal of South African Botany 47,
247271.
Marchant, C.J. (1968) Chromosome patterns and nuclear phenomena in the cycad fami-
lies Stangeriaceae and Zamiaceae. Chromosoma 24, 100134.
Marshall, J., Grobbelaar, N., Coetzee, J. and Osborne, R. (1989a) Pollen morphology of
the Cycadales with special reference to the Encephalartos species. Pollen et Spores 31,
229249.
Marshall, J., Huang, T.C. and Grobbelaar, N. (1989b) Comparative morphological and
physiological studies on cyanobionts of Encephalartos transvenosus. South African Journal
of Botany 55, 574580.
Mogford, D.J. (1981) Chromosome studies in the southern African flora: 13. Journal of
South African Botany 47, 26.
Moretti, A. (1990) Cytotaxonomy of cycads. In: Stevenson, D.W. (ed.) The Biology, Structure,
and Systematics of the Cycadales. Proceedings of the Symposium CYCAD 87. Memoirs of the New
York Botanical Garden 57, pp. 114122.
Moretti, A., Sabato, S. and Siniscalco Gigliano, G. (1983) Taxonomic significance of
methylazoxymethanol glycosides in the cycads. Phytochemistry 22, 115118.
ch06.qxd 28/11/03 3:45 pm Page 82
82 P. Vorster
Newton, L.E. (2002) A new species of Encephalartos (Zamiaceae) in Sudan. Botanical Journal
of the Linnean Society 140, 187192. [Encephalartos mackenziei.]
Norstog, K. (1987) Cycads and the origin of insect pollination. American Scientist 75,
270279.
Oberprieler, R.G. (1995) The weevils (Coleoptera: Curculionoidea) associated with
cycads. 2. Host specificity and implications for cycad taxonomy. In: Vorster, P. (ed.)
Proceedings of the Third International Conference on Cycad Biology. Cycad Society of South
Africa, Stellenbosch, South Africa, pp. 335365.
Osborne, R. and Stevens, J.F. (1996) Epicuticular waxes and glaucousness of Encephalartos
leaves. Phytochemistry 42, 13351339.
Osborne, R., Salatino, M.L.F. and Salatino, A. (1989) Alkanes of foliar epicuticular waxes
of the genus Encephalartos. Phytochemistry 28, 30273030.
Pate, J.S., Lindblad, P. and Atkins, C.A. (1988) Pathways of assimilation and transfer of
fixed nitrogen in coralloid roots of cycadNostoc symbiosis. Planta 176, 461471.
Pellmyr, O., Tang W., Groth, I., Bergstrm, G. and Thien, L.B. (1991) Cycad cone and
Angiosperm floral volatiles: inferences for the evolution of insect pollination.
Biochemical Systematics and Ecology 19, 623627.
Robbertse, P.J., Vorster, P. and Van der Westhuizen, S. (1988a) Encephalartos graniticolus
(Zamiaceae): a new species from the north-eastern Transvaal. South African Journal of
Botany 54, 363366.
Robbertse, P.J., Vorster, P. and Van der Westhuizen, S. (1988b) Encephalartos verrucosus
(Zamiaceae): a new species from the north-eastern Transvaal. South African Journal of
Botany 54, 487490.
Sharma, I.K., Jones, D.L., Forster, P.I. and Young, A.G. (1998) The extent and structure
of genetic variation in the Macrozamia pauliguilielmi complex (Zamiaceae). Biochemical
Systematics and Ecology 26, 4554.
Sharma, I.K., Jones, D.L., Forster, P.I. and Young, A.G. (1999) Low isozymic differentia-
tion among five species of the Macrozamia heteromera group (Zamiaceae). Biochemical
Spreeth, A.D. and Vorster, P. (1995) Anatomical studies on the leaflets of glaucous-leaved
Encephalartos species in South Africa [abstract]. In: Vorster, P. (ed.) Proceedings of the
Third International Conference on Cycad Biology. Cycad Society of South Africa,
Stellenbosch, South Africa, p. 263.
Stephens, D.C. and Stephen, A.M. (1988) Exudates from Encephalartos cones as chemical
taxonomic markers. South African Journal of Science 84, 263266.
Stevenson, D.W. (1981) Observations on ptyxis, phenology, and trichomes in the cycadales
and their systematic implications. American Journal of Botany 68, 11041114.
Tang, W. (1993) Heat and odour production in cycad cones and their role in insect polli-
nation. In: Stevenson, D.W. and Norstog, K.J. (eds) The Biology, Structure, and Systematics
of the Cycadales. Proceedings of the Second International Conference on Cycad Biology. Palm &
Cycad Societies of Australia Limited, Milton, Queensland, Australia, pp. 140147.
Van der Bank, F.H., Vorster, P. and Van der Bank, M. (1998) Phylogenetic relationships,
based on allozyme data, between six cycad taxa indigenous to South Africa. South
African Journal of Botany 64, 182188.
Van der Bank, H., Wink, M., Vorster, P., Treutlein, J., Brand, L., Van der Bank, M. and
Hurter, J. (2001) Allozyme and DNA sequence comparisons of nine species of
Encephalartos (Zamiaceae). Biochemical Systematics and Ecology 29, 241266.
ch06.qxd 28/11/03 3:45 pm Page 83
Vorster, P. (1990) Encephalartos aemulans (Zamiaceae), a new species from northern Natal.
South African Journal of Botany 56, 239243.
Vorster, P. (1993) Taxonomy of Encephalartos: taxonomically useful external characteristics.
In: Stevenson, D.W. and Norstog, K.J. (eds) The Biology, Structure, and Systematics of the
Vorster, P. (1995) The identity of Encephalartos lebomboensis. In: Vorster, P. (ed.) Proceedings of
the Third International Conference on Cycad Biology. Cycad Society of South Africa,
Stellenbosch, South Africa, pp. 245254.
Vorster, P. (1996a) Encephalartos msinganus (Zamiaceae): a new species from KwaZulu-Natal.
South African Journal of Botany 62, 6770.
Vorster, P. (1996b) Encephalartos venetus (Zamiaceae): a new species from the Northern
Province. South African Journal of Botany 62, 7175.
Vorster, P. (1996c) Encephalartos senticosus (Zamiaceae): a new species from northern
KwaZulu-Natal and Swaziland. South African Journal of Botany 62, 7679.
Vorster, P. and Heibloem, P. (1995) Encephalartos imbricans (Zamiaceae): a new species from
Uganda. Novon 5, 388394.
Vorster, P. and Oberprieler, R. (1999) Entomological evidence for and against taxonomi-
cal decisions in Encephalartos. In: Chen, C.J. (ed.) Biology and Conservation of Cycads.
Proceedings of the Fourth International Conference on Cycad Biology. International Academic
Publishers, Beijing, China, pp. 198207.
Vovides, A.P. (1991) Vesiculararbuscular mycorrhiza in Dioon edule Lindl. (Zamiaceae) in
its natural habitat in central Veracruz, Mexico. [Abstract of paper presented at 1991
Annual meeting of the Botanical Society of America.] American Journal of Botany 78
(6), supplement, Ecological Section, p. 76.
Walters, T.W. and Decker-Walters, D.S. (1991) Patterns of allozyme diversity in the West
Indies cycad Zamia pumila (Zamiaceae). American Journal of Botany 78, 436445.
ch06.qxd 28/11/03 3:45 pm Page 84
ch07.qxd 28/11/03 3:47 pm Page 85
Macrozamia (Zamiaceae) from
Eastern Australia
Paul I. Forster
Queensland Herbarium, Environmental Protection Agency,

Brisbane Botanic Gardens, Toowong, Queensland, Australia
Abstract
The genus Macrozamia currently comprises at least 40 species and is endemic to Australia.
Thirty-seven of these species occur in eastern Australia in Queensland and New South
Wales. Species concepts in the genus are discussed and useful character states outlined.
Descriptive examples are included and suggestions for identification tools made. Future
directions for research are outlined.
Introduction
The genus Macrozamia Miquel is one of seven extant genera in the cycad family
Zamiaceae. Species of Macrozamia are endemic to Australia with 40 (Hill et al.,
Appendix 1 this volume) or 41 recognized species (Table 7.1). The main concen-
tration of species is in eastern Australia, with at least 37 species (Hill and
Osborne, 2001; Jones et al., 2001; Forster, 2002). A single species, M. macdonnellii
(F. Mueller ex Miquel) A. de Candolle, is present in central Australia near Alice
Springs and three species are currently recognized from south-west Western
Australia.
The first modern account of Macrozamia was by Johnson (1959) but the
extremely broad concept of taxa employed therein has now been discarded.
Johnson also regarded many geographically disjunct and incomplete specimens
as representing hybrids, although no scientific evidence was ever presented to
support this pronouncement. Juvenile foliage of many species of Macrozamia can
be quite different to that found on mature individuals and misidentification of
ch07.qxd 28/11/03 3:47 pm Page 86
86 P.I. Forster
Table 7.1. Species recognized in Macrozamia and their general distribution.
Australian state
Species epithet and author of distribution
Macrozamia section Macrozamia

M. cardiacensis P.I. Forster & D.L. Jones Queensland
M. communis L.A.S. Johnson New South Wales
M. diplomera (F. Mueller) L.A.S. Johnson New South Wales
M. douglasii W. Hill ex F.M. Bailey Queensland
M. dyeri (F. Mueller) C.A. Gardner Western Australia
M. fraseri Miquel Western Australia
M. johnsonii D.L. Jones & K.D. Hill New South Wales
M. longispina P.I. Forster & D.L. Jones Queensland
M. macdonnellii (F. Mueller ex Miquel) A. de Candolle Northern Territory
M. macleayi Miquel Queensland
M. miquelii (F. Mueller) A. de Candolle Queensland
M. montana K.D. Hill New South Wales
M. moorei F. Mueller Queensland
M. mountperriensis F.M. Bailey Queensland
M. riedlei (Gaudichaud) C.A. Gardner Western Australia
M. reducta K.D. Hill & D.L. Jones New South Wales
M. serpentina D.L. Jones & P.I. Forster Queensland
Macrozamia section Parazamia
M. concinna D.L. Jones New South Wales
M. conferta D.L. Jones & P.I. Forster Queensland
M. cranei D.L. Jones & P.I. Forster Queensland
M. crassifolia P.I. Forster & D.L. Jones Queensland
M. elegans K.D. Hill & D.L. Jones New South Wales
M. fawcettii C. Moore New South Wales
M. fearnsidei D.L. Jones Queensland
M. flexuosa C. Moore New South Wales
M. glaucophylla D.L. Jones New South Wales
M. heteromera C. Moore New South Wales
M. humilis D.L. Jones New South Wales
M. lomandroides D.L. Jones Queensland
M. lucida L.A.S. Johnson Queensland, New
South Wales
M. machinii P.I. Forster & D.L. Jones Queensland
M. occidua D.L. Jones & P.I. Forster Queensland
M. parcifolia P.I. Forster & D.L. Jones Queensland
M. pauli-guilielmi W. Hill & F. Mueller Queensland
M. platyrhachis F.M. Bailey Queensland
M. plurinervia (L.A.S. Johnson) D.L. Jones New South Wales
M. polymorpha D.L. Jones New South Wales
M. secunda C. Moore New South Wales
M. spiralis (Salisbury) Miquel New South Wales
M. stenomera L.A.S. Johnson New South Wales
M. viridis D.L. Jones & P.I. Forster Queensland, New
South Wales
ch07.qxd 28/11/03 3:47 pm Page 87
Classification Concepts in Macrozamia 87
juveniles of some species (e.g. M. johnsonii D.L. Jones & K.D. Hill then known
as the New South Wales form of M. moorei F. Mueller as M. lucida L.A.S.
Johnson) also aided and abetted confusion when using his account.
Revision of the genus Macrozamia was commenced in the early 1990s by D.L.
Jones and the current author, using a variety of methods, but primarily those of
macromorphology. Accounts of new taxa, species complexes or nomenclature of
specific epithets have been published sequentially (Jones, 1991; Forster and Jones,
1992; Jones and Hill, 1992; Jones and Forster, 1994; Forster and Jones, 1998;
Forster, 1999a,b; Jones et al., 2001), but no detailed monograph has yet been pub-
lished. A Flora of Australia compilation of the majority of species was published
by Hill (1998) and a populist account derived directly from this source by Hill
and Osborne (2001). It is likely that a small number of additional species (to those
in Table 7.1) will be recognized once the ongoing revision is finalized. These will
be primarily from northern New South Wales and south-western Western
Australia.
Species Concepts
At present the genus is divided into two sections: Macrozamia section Macrozamia
and M. section Parazamia (Miquel) Miquel. These two sections are primarily dis-
tinguished by the presence or absence of mucilage canals in the leaflets, although
the presence of this character has not been determined for most species now rec-
ognized and the proposed sectional classification of at least one species (M. lucida)
remains suspect.
Species in Macrozamia have been defined in terms of morphological
discontinuity in character states and are similar to those recognized in
most vascular plant groups. All of the recognized species of Macrozamia
differ from one another in at least two character states. Some of the species
are highly distinctive and unlikely to be confused with any others [e.g.
M. macdonnellii (F. Mueller ex Miquel) A. de Candolle, M. platyrhachis F.M. Bailey].
Others form part of complexes of taxa of similar appearance and therein
potential problems lie with how many species are to be recognized. Once the
individual taxa have been defined, it is necessary to allocate them a taxonomic
rank, with decisions having to be made as to whether the use of infraspecific taxa
is justified.
The only workers to have applied infraspecific ranks in the genus Macrozamia
have been Schuster (1932) and Johnson (1959), both utilizing subspecies. While
the work of Schuster has been largely discounted, the system proposed by
Johnson was recognized for many years, even though he did not define what was
meant by a species or a subspecies. All of the subspecific taxa covered by Johnson
are now recognized at the species level. This approach of using infraspecific
ranks is fraught with supposition and is also unpopular with the end-user. If there
is morphological discontinuity in a couple of characters then little is gained in
having the classification reflect a hypothetical phylogeny by using infraspecific
ch07.qxd 28/11/03 3:47 pm Page 88
88 P.I. Forster
ranks. Inferred relationships between the taxa can be discussed when dealing
with individual species complexes.
Some species complexes in Macrozamia section Parazamia comprise very
closely related taxa [e.g. M. conferta D.L. Jones & P.I. Forster, M. cranei D.L. Jones
& P.I. Forster, M. machinii P.I. Forster & D.L. Jones, M. plurinervia (L.A.S. Johnson)
D.L. Jones (the latter two species are considered synonymous by Hill et al.,
Appendix 1 this volume, but this view is not supported by the author), M. occidua
D.L. Jones & P.I. Forster, M. viridis D.L. Jones & P.I. Forster], disjunct into adja-
cent geographical areas and on dissimilar geologies. There is relatively low
genetic dissimilarity among these taxa based on isozyme techniques (e.g. Sharma
et al., 1999b). Because of the inherent beetle-pollination system in some species
of Macrozamia and the apparent poor dispersal of the beetles between popula-
tions, disjunct populations are unlikely to have any recent genetic integration.
Their low genetic diversity and lack of difference (at least in the limited isozyme
systems studied) within or between populations tends to indicate that speciation
is relatively recent and that inbreeding within populations has occurred for some
time (Sharma et al., 1998, 1999b). A similar situation seems to be present in some
complexes of Encephalartos Lehmann in southern Africa (Van der Bank et al.,
1998, 2001), and no proposal to recognize those particular taxa as subspecies of
a single species has appeared, nor is it likely to be a popular option. In compari-
son, some taxa of Macrozamia section Macrozamia are readily separated by
isozymes (Sharma et al., 1999a; Jones et al., 2001).
Ongoing investigations into insect pollination systems in Macrozamia by Irene
Terry of the University of Utah may well shed further light on the sectional rela-
tionships of species (Mound and Terry, 2001; Terry, 2001). Species in M. section
Parazamia appear to be purely beetle-pollinated by species of Tranes, whereas
those in M. section Macrozamia are pollinated either by thrips (Cycadothrips spp.),
beetles (Tranes spp.) or a combination of the two. These differing pollination
systems are associated with marked differences in cone volatile fragrances and it
is likely that these will prove to be of taxonomic significance.
An attempt to further subdivide the taxa within Macrozamia section Parazamia
into informal groups has been undertaken by Hill (1998) and Hill and Osborne
(2001), but these hypotheses require validation using a range of anatomical and
molecular techniques as some of the groups appear artificial. By contrast, the
species in M. section Macrozamia can be easily allocated to four groups:
Group 1: Macrozamia dyeri (F. Mueller) C.A. Gardner, M. fraseri

Miquel, M. riedlei (Gaudichaud) C.A. Gardner. Restricted to Western
Australia. Large trunked cycads, leaflets amphistomatic, rachis with clear base.
Group 2: Macrozamia johnsonii D.L. Jones & K.D. Hill, M. mac-
donnellii (F. Mueller ex Miquel) A. de Candolle, M. moorei F.
Mueller. Central and Eastern Australia. Large trunked cycads, leaflets
amphistomatic, rachis with pinnacanths to base.
Group 3: Macrozamia communis L.A.S. Johnson, M. elegans K.D.
Hill & D.L. Jones, M. montana K.D. Hill, M. reducta K.D. Hill &
ch07.qxd 28/11/03 3:47 pm Page 89
D.L. Jones. Eastern Australia, in New South Wales. Medium sized cycads
with short or minimal trunk development, leaflets hypostomatic, broad-based
spines on the female sporophylls (512 mm wide).
Group 4: Macrozamia cardiacensis P.I. Forster & D.L. Jones,
M. douglasii W. Hill ex F.M. Bailey, M. longispina P.I. Forster &
D.L. Jones, M. macleayi Miquel, M. miquelii (F. Mueller) A. de
Candolle, M. mountperriensis F.M. Bailey, M. serpentina D.L.
Jones & P.I. Forster. Eastern Australia, in Queensland. Medium to large
sized cycads with short or minimal trunk development, leaflets hypostomatic,
narrow-based spines on the female sporophylls (25 mm wide).
As a result of the recent detailed account of the Macrozamia miquelii complex

(Jones et al., 2001) and an earlier account of M. section Parazamia in Queensland
(Jones and Forster, 1994), further discussion of approaches to cycad classification
and character state definition will be undertaken mainly using examples from
these groups with which the author is most familiar.
Approaches to Classification
Any proposed classification has to be user-friendly. It is of little value if the diag-

nostic characters used are so esoteric or difficult to observe that only a specialist
can utilize them. This is perhaps even more relevant with plants such as cycads.
These plants are of interest to people for a number of reasons, mainly economic
(e.g. livestock poisoning), horticultural, or legalistic (conservation legislation). It is
therefore essential that a range of end-users can identify plant material, given that
basic specimen requirements are applied. At the Queensland Herbarium (BRI)
there is a constant array of queries concerning the identity of native cycads. These
enquiries cover the gamut from poisons queries to harvesting or land-clearing
permits, horticultural identifications and nature conservation act enforcement. In
this respect the Herbarium is an integral part of the Queensland Environmental
Protection Agency and its efforts to document the floral biodiversity of the state
and to ensure its long-term protection in a sustainable manner. Staff at the
Herbarium decide which species are to be recognized for the state, and may also
propose conservation ratings for species using the IUCN categories. After external
ratification (by a non-government committee), these conservation ratings are then
included in the schedules attached to the Queensland Nature Conservation Act
and applied federally by Environment Australia for the national Rare and
Threatened list. With so much interest in the native cycads, the requirement for
modern accounts of all the species has been critical, with concurrent adaptation
and updating of protective legislation. It has been particularly important to resolve
some of the more difficult species complexes, as conservation decisions based on
omnibus species will not reflect the genetic diversity present.
The approach to revision of the genus Macrozamia has been both field- and
laboratory-based. Considerable effort was undertaken in fieldwork to determine
ch07.qxd 28/11/03 3:47 pm Page 90
90 P.I. Forster
the geographical range of taxa, to assess the full range of morphological varia-
tion present, and to create durable and informative herbarium collections. It has
often been difficult to obtain complete material of these dioecious plants that
reproduce irregularly and repeated visits to some, often remote, localities has
been necessary. It was realized early in the study that this field-based approach to
elucidation of the various taxonomic complexes was essential. Plants of
Macrozamia (or any cycad for that matter) have an abundance of measurable char-
acters; however, these are not accurately determined merely from the material
that ends up preserved in herbaria. An extensive checklist of character states in
M. section Parazamia was developed by Jones for use in the field. Up to 20 mature
plants were used to capture the data used in the descriptions published by Jones
and Forster (1994). An abridged version of this checklist follows:
1. Locality
2. Date
3. Collection number
4. Stem branching
5. Trunk emergence
6. Colour and density of trunk wool
7. Petiole wool
8. Petiole cross-section
9. Leaf number and orientation
10. Rachis twisting
11. Leaflet colour and glaucousness
12. Leaflet twist and base
13. Leaflet orientation
14. Leaflet crowding
15. Leaflet spacing along rachis
16. Stomate distribution
17. Callus colour
18. Lower leaflets form
19. Juvenile leaf colour
20. Juvenile leaf shape
21. Venation prominence on upper leaflet surface
22. Venation prominence on lower leaflet surface
23. Number of male cones/plant
24. Number of female cones/plant
25. Cone colour at maturity.
Shrinkage occurs of both reproductive material (cones or seeds) and foliage
(leaflet dimensions) during the drying process used in producing herbarium
vouchers, hence dimensions were checked on the subsequent dried material and
descriptions adjusted accordingly to reflect this. Many end-users are interested in
the diagnostic characters in identification keys. The following are the key diag-
nostic characters used in the identification key for Macrozamia section Parazamia
from Queensland (Jones and Forster, 1994):
ch07.qxd 28/11/03 3:47 pm Page 91
1. Leaf orientation/disposition
2. Leaf rachis curvature
3. Leaflet texture
4. Leaflet shape in cross-section
5. Leaflet orientation to rachis
6. Leaflet colour (wax cover often important)
7. Leaflet size
8. Cone colour (glaucous or green)
9. Microsporangiate cone spination.
In this particular identification key, nearly all use of fertile characters was
avoided, with a range of leaf characters being more useful.
With respect to the revision of the Macrozamia miquelii complex, we tried to
use characters that were readily observable, although it eventuated that a combi-
nation of vegetative (from mature plant foliage) and reproductive characters was
necessary to separate species successfully. The diagnostic characters used in the
key to species (Jones et al., 2001) were as follows:
1. Leaflet texture and appearance (e.g. thin/glossy)
2. Leaflet number
3. Leaflet width
4. Leaflets reduced/not reduced to pinnacanths at base of leaf
5. Female cone shape (ovoid or cylindrical to barrel-shaped)
6. Female cone size
7. Megasporophyll spination
8. Male cone size
9. Microsporophyll size.
For the key to work, one would need ideally to have both male and female cones
and foliage of the plant. Given that a detailed revision now exists, material that
lacks some of this information can probably still be identified by locality infer-
ence. This key requirement for fertile material is not at all unusual in the process
of plant identification and should not be considered an impediment.
A cursory examination of identification keys for species in other genera of
cycads (e.g. Ceratozamia Brongniart, Cycas Linnaeus, Dioon Lindley, Encephalartos)
revealed that the following characters were commonly used:
1. Stem development (Giddy, 1984; Hill, 1996; Wang, 1996)
2. Stem apex wooliness (Giddy, 1984; Wang, 1996)
3. Cataphyll form (Hill, 1996; Wang, 1996)
4. Cataphyll indumentum form and colour (Hill, 1996)
5. Hypodermis development (Hill, 1996)
6. Leaf orientation (De Luca and Sabato, 1979; Giddy, 1984; Hill, 1996)
7. Leaflet size (De Luca and Sabato, 1979; Giddy, 1984; Vovides et al., 1993;
Hill, 1996; Wang, 1996)
8. Leaflet orientation (Giddy, 1984; Hill, 1996; Wang, 1996)
9. Leaflet texture (Giddy, 1984; Wang, 1996)
ch07.qxd 28/11/03 3:47 pm Page 92
92 P.I. Forster
10. Leaflet colour (Giddy, 1984; Vovides et al., 1993; Hill, 1996)
11. Leaflet indumentum (Hill, 1996)
12. Leaflet lobing or division (De Luca and Sabato, 1979; Giddy, 1984; Wang,
1996)
13. Leaflets reduced to pinnacanths or not (Giddy, 1984; Hill, 1996)
14. Leaflet disposition (i.e. overlapping or not) (Giddy, 1984; Hill, 1996)
15. Leaflet number (Hill, 1996; Wang, 1996)
16. Disposition of stomata in leaflets (Hill, 1996)
17. Leaflet margin flat, recurved or revolute (Hill, 1996; Wang, 1996)
18. Cone colour (Giddy, 1984; Vovides et al., 1993)
19. Megasporophyll spination (Hill, 1996; Wang, 1996)
20. Megasporangiate cone size (Vovides et al., 1993)
21. Microsporangiate cone shape (Hill, 1996)
22. Microsporangiate cone size (Hill,1996; Wang, 1996)
23. Seed size (Wang, 1996)
24. Seed colour (Giddy, 1984; Wang, 1996)
25. Ovule/seed indumentum (Wang, 1996).
It is obvious from the above that most workers use similar morphological char-
acters to distinguish cycad species.
Irrespective of the most important diagnostic characters, it is still essential to
provide a detailed morphological description. The following example (from Jones
et al., 2001) is considered most appropriate for species of Macrozamia. Any revi-
sion of a group of cycads (or any plant for that matter) should provide similar
detailed descriptions and identification aids such as keys or discussion of impor-
tant features, preferably concentrating on vegetative characters.
Macrozamia macleayi Miquel
Caudex usually subterranean, occasionally emergent, erect, columnar or

barrel-shaped, to 35 cm tall and 30 cm diameter, unbranched. Young leaves
light green. Mature leaves elliptic-lanceolate in outline, 0.52 m long,
obliquely erect to spreading, dark green, glossy, flat in cross-section, arching
in profile, 650 in a moderately dense crown; expanded leaf base 812 cm
2.53.5 cm, covered with grey-brown, soft wool; petiole (including the
woolly expanded base) 1030 cm long, 0.81.2 cm wide at the first leaflet,
greenish, flat adaxially, convex and angular abaxially; rachis straight, or
slightly twisted, pale green, the cross-section similar to that of the petiole;
leaflets 80160, inserted at about 40 to the rachis, widely spreading, mod-
erately crowded, distal leaflets densely packed, proximal leaflets becoming
more widely spaced, narrowly linear-lanceolate, 1550 cm 69 mm, flat,
thin-textured, hypostomatic, dark green and glossy adaxially, slightly paler
beneath, contracted proximally to a whitish callous base, margins entire,
tapered to a pungent apex; lower leaflets gradually reduced to 828 pairs of
ch07.qxd 28/11/03 3:47 pm Page 93
rigid pinnacanths. Male cones 15, cylindrical, 1520 cm 3.86.5 cm,

straight or curved with age, green; peduncle 1529 cm 1.52 cm, elliptical
to round in cross-section; microsporophylls broadly cuneate, 1.52.5 cm
11.5 cm, with an erect, apical spine 0.51.5 cm long. Female cones 1 or 2,
cylindrical to barrel-shaped, 1930 cm 710 cm, green; peduncle 1541
cm 23 cm, elliptical in cross-section, furrowed; megasporophylls broadly
wedge-shaped, 1.52.5 cm 33.5 cm, with an erect apical spine 0.54.5 cm
long, the distal sporophylls with spines 2.54.5 cm long. Seeds oblong to
obovoid, 2.53.5 cm 1.52.5 cm, the sarcotesta orange to red.
Future Directions
While unambiguous diagnostic morphological characters are the backbone of

any classification, other methods such as isozymes, chromosomes (cf. Osborne,
2001), anatomy and, most importantly, DNA data have yet to be extensively
investigated in Macrozamia. Incidental observations also support further study of
seeds and seedling form, cone structure, and megasporophyll and microsporo-
phyll form. The descriptive phase of Macrozamia systematics is now nearly com-
plete. The next steps should be in determining the phylogeny of related species
and overhauling the infrageneric classification of the genus based on a range of
characters.
Acknowledgements
I thank Peter Bostock and Irene Terry for commenting on the manuscript.
References
De Luca, P. and Sabato, S. (1979) Dioon califanoi (Zamiaceae), a new species from Mexico.
Brittonia 31, 170173.
Forster, P.I. (1999a) Proposal to conserve the name Encephalartos miquelii (Zamiaceae) with
a conserved type. Taxon 48, 569570.
Forster, P.I. (1999b) Typification and application of the name Macrozamia macleayi Miq.
(Zamiaceae). Austrobaileya 5, 577.
Forster, P.I. (2002) Plantae Conifers and allies. In: Henderson, R.J.F. (ed.) Names and
Distribution of Queensland Plants, Algae and Lichens. Environmental Protection Agency,
Brisbane, Australia, pp. 201202.
Forster, P.I. and Jones, D.L. (1992) Neotypification of Macrozamia mountperriensis
(Zamiaceae) with notes on its distribution. Telopea 5, 289290.
Forster, P.I. and Jones, D.L. (1998) Macrozamia cardiacensis sp. nov., M. longispina sp. nov. In:
McCarthy, P. (ed.) Flora of Australia, Vol. 48. CSIRO Publishing, Melbourne,
Australia, p. 717.
Giddy, C. (1984) Cycads of South Africa, 2nd edn. C. Struik Publishers, Cape Town, South
Africa, 112 pp.
ch07.qxd 28/11/03 3:47 pm Page 94
94 P.I. Forster
Telopea 7, 164.
Hill, K.D. (1998) Cycadophyta. In: McCarthy, P. (ed.) Flora of Australia, Vol. 48. CSIRO
Publishing, Melbourne, Australia, pp. 597661.
Hill, K.D. and Osborne, R. (2001) Cycads of Australia. Kangaroo Press, Sydney, Australia,
116 pp.
of the Linnean Society of New South Wales 84, 64117.
Jones, D.L. (1991) Notes on Macrozamia (Zamiaceae) in Queensland with the description
of two new species in section Parazamia (Miq.) Miq. Austrobaileya 3, 481487.
[Macrozamia fearnsidei, M. lomandroides.]
Jones, D.L. and Forster, P.I. (1994) Seven new species of Macrozamia section Parazamia
(Miq.) Miq. (Zamiaceae section Parazamia) from Queensland. Austrobaileya 4,
269288. [Macrozamia conferta, M. cranei, M. crassifolia, M. machinii, M. occidua, M. parci-
folia, M. viridis.]
Jones, D.L. and Hill, K.D. (1992) Macrozamia johnsonii, a new species of Macrozamia section
Macrozamia (Zamiaceae) from northern New South Wales. Telopea 5, 3134.
Jones, D.L., Forster, P.I. and Sharma, I.K. (2001) Revision of the Macrozamia miquelii (F.
Muell.) A. DC. (Zamiaceae section Macrozamia) group. Austrobaileya 6, 6794.
Mound, L.A. and Terry, I. (2001) Thrips pollination of the central Australian cycad,
Macrozamia macdonnellii (Cycadales). International Journal of Plant Sciences 162, 147154.
Osborne, R. (2001) New research initiative in Macrozamia. Palms & Cycads No. 72, 1415.
1, pp. 1168.
Sharma, I.K., Jones, D.L., Forster, P.I. and Young, A.G. (1998) The extent and structure
of genetic variation in the Macrozamia pauliguilielmi complex (Zamiaceae). Biochemical
Sharma, I.K., Jones, D.L. and Forster, P.I. (1999a) Contribution of isozymic analysis in dif-
ferentiating Macrozamia moorei D.L. Jones and K.D. Hill from M. johnsonii F. Muell.
(Zamiaceae). Austrobaileya 5, 363365.
Sharma, I.K., Jones, D.L., Forster, P.I. and Young, A.G. (1999b) Low isozymic differenti-
ation among five species of the Macrozamia heteromera group (Zamiaceae). Biochemical
Terry, I. (2001) Thrips and weevils as dual, specialist pollinators of the Australian cycad
Macrozamia communis (Zamiaceae). International Journal of Plant Sciences 162, 12931305.
Van Der Bank, F.H., Vorster, P. and Van Der Bank, M. (1998) Phylogenetic relationships,
based on allozyme data, between six cycad taxa indigenous to South Africa. South
African Journal of Botany 64, 182188.
Van Der Bank, F.H., Wink, M., Vorster, P., Treutlein, J., Brand, L., Van Der Bank, M. and
Hurter, J. (2001) Allozyme and DNA sequence comparisons of nine species of
Encephalartos (Zamiaceae). Biochemical Systematics and Ecology 29, 241266.
Vovides, A.P., Vzquez Torres, M., Schutzman, B. and Iglesias, C.G. (1993) A new species
of Ceratozamia (Zamiaceae) from Quertaro and Hidalgo, Mexico. Novon 3, 502506.
[Ceratozamia sabatoi.]
Wang, D.Y. (1996) Taxonomy of Cycas in China. In: Wang, F. and Liang, H. (eds) Cycads in
China. Guangdong Science and Technology Press, Guangzhou, China, pp. 33142.
ch08.qxd 28/11/03 3:47 pm Page 95
Ceratozamia (Zamiaceae)
Loran M. Whitelock
4524 Toland Way, Los Angeles, California, USA
Abstract
General cycad taxonomy is reviewed, along with the problems associated with separating
species using morphology, genetics and laboratory procedures. The past and present
taxonomy of Ceratozamia is reviewed and discussed. Morphological similarities and,
more importantly, differences between Ceratozamia species and forms are listed and dis-
cussed with a view to using them as criteria in the separation of new taxa. Suggestions are
made for criteria to be used in the investigation and documentation of putative new
species.
Cycad Taxonomy
At the present time there seems to be little disagreement about the separation of
the 11 cycad genera, with the possible exception of the small genus, Chigua D.W.
Stevenson. The genera are separated on the basis of clear diagnostic features,
which leaves little room for argument as to their status. The same cannot be said
for the separation and description of taxa at the infrageneric level. At this level
there has historically been a great deal of disagreement, caused by the often
minor differences between taxa. Taxonomists hold differing concepts as to how
the delimitation of new species should be approached. At present there are no
guidelines as to how substantial the separation between species should be, and
this has caused inconsistencies from one species description to another.
It is generally accepted that taxonomic classification should reflect the results
of evolution. However, the meagre fossil evidence relating to the cycads generally
ch08.qxd 28/11/03 3:47 pm Page 96
96 L.M. Whitelock
produces more questions than answers. If this evolutionary train of thought is

followed, the conclusion is that the basic difference between species should be one
of genetics, but proving genetic differences is not an easy task. Generally, the tax-
onomy of cycad species has been based on morphology. Johnson (1959) consid-
ered a species to be a population or group of plants that was effectively isolated
in nature by any means so that they were not able to interbreed with other
related but genotypically and phenotypically different populations. These vicari-
ants (generally believed to be separated by past geological or climatic events)
gradually evolved into new species because of isolation, and often the subsequent
change in habitat. This theory works well with cycads because of their rather
inefficient means of dispersal and the fact that most cycad species are found as
small relic populations. The inability of Ceratozamia Brongniart species to be
widely dispersed results in the genus being restricted to small populations that are
not able to interbreed with other neighbouring populations. This genetic isola-
tion occurring over a period of thousands, or millions, of years can result in the
establishment of genetically distinct populations, or species.
The logical way to validate a possible new species is proving how genetical-
ly similar it is to another related species. Hybridizing the possible new
species with its closest relative and then crossing the resulting offspring to observe
the amount of variation in genetically controlled characters appearing
in the second generation could do this. With annuals this would not be a difficult
task. However, cycads take a number of years to reach sexual maturity. The tax-
onomist might die of old age before being able to assess a new species in this
manner!
In recent times considerable laboratory work has been done on the cytolog-
ical, chemical and molecular characteristics of both the genera and species of
cycads. Investigations into cycad chromosomes have produced results that
provide useful information about cycad interrelationships but this work has not
revealed much about evolutionary trends in cycad taxonomy, especially at a
species level. Chromosome numbers in most cycad genera agree in all of their
species with the exception of Zamia Linnaeus, which has some divergent kary-
otypes that have led investigators to believe that Zamia is evolving more rapidly
than the other genera (Norstog, 1980, 1981; Moretti et al., 1991; Vovides and
Olivares, 1996). Recent studies in chloroplast DNA (Caputo et al., 1991; De Luca
et al., 1995) have enabled researchers to estimate evolutionary distances between
various plant taxa but it has not been particularly helpful in the separation of
species. Other researchers have studied cycasin and macrozamin (Moretti et al.,
1981, 1983), which tend to reflect certain evolutionary relationships, but once
again the results are not applicable for separating species. Cuticular waxes of
cycads have been studied (Osborne et al., 1989) and the findings suggest that these
waxes may provide sufficient resolution for the separation of genera and species.
Although these tests provide useful information to the taxonomists in making
their determinations, at this time they merely indicate, but do not prove, the sep-
aration of species. An important consideration is that all of these tests require
laboratory facilities and trained personnel that are not always available to the
ch08.qxd 28/11/03 3:47 pm Page 97
Classification Concepts in Ceratozamia 97
taxonomist.
Taxonomic History of Ceratozamia
The taxonomic history of Ceratozamia and its 21 species has taken 156 years to
reach its present status. Adolphe Theodore Brongniart (1846) described
Ceratozamia and its type species C. mexicana Brongniart from cultivated material in
Paris, France. The following 10 years produced four more species that are still
valid: C. robusta Miquel (1848), C. latifolia Miquel (1848), C. miqueliana H.
Wendland (1854) and C. kuesteriana Regel (1857). Ceratozamia fusco-viridis D. Moore
was described in 1878. The next taxonomic change to Ceratozamia did not take
place for another 61 years, when C. matudae Lundell (1939) was described.
Another 24 years were to pass before the description of C. zaragozae Medellin-
Leal (1963). Renewed interest in the genus Ceratozamia began in 1979 with the
description of C. hildae G.P. Landry & M.C. Wilson (1979). The period from 1979
to the present has produced 12 additional species, more than doubling the
number of species that were described in the previous 133 years. These new
species included C. norstogii D.W. Stevenson (1982), C. microstrobila Vovides & J.D.
Rees (1983), C. euryphyllidia Vzquez Torres, Sabato & D.W. Stevenson (Stevenson
et al., 1986), C. sabatoi Vovides, Vzquez Torres, Schutzman & Iglesias (1993), C.
whitelockiana Chemnick & T.J. Gregory (1995), C. mixeorum Chemnick, T.J.
Gregory & Salas-Morales (1997), C. morettii Vzquez Torres & Vovides (1998), C.
alvarezii Prez-Farrera, Vovides & Iglesias (1999), C. mirandae Vovides, Prez-
Farrera & Iglesias (2001), C. zoquorum Prez-Farrera, Vovides & Iglesias (2001),
C. huastecorum S. Avendao, Vovides and Castillo-Campos (2003) and C. sp.
becerrae (A.P. Vovides, 2003 in review).
There is little doubt that Ceratozamia will have additional taxa described
over the next few years. This will be the result of renewed interest in the genus as
well as increased field studies by a number of botanists and botanical organiza-
tions. Foremost among the botanists investigating Ceratozamia are Jeffrey
Chemnick, Timothy Gregory, Miguel A. Prez-Farrera, Silvia Salas-
Morales, Mario Vzquez Torres and Andrew Vovides. Due to the efforts of these
botanists we can expect to see Ceratozamia begin to receive the critical study it
deserves.
Because all the species are closely related and have a somewhat similar
appearance, the taxonomy of Ceratozamia has been debated for many years. As
taxonomy is based on interpreting data, and taxonomists are notoriously diver-
gent in their interpretations, there have been disagreements on the justification
for some proposed new species. Taxonomists are often categorized as either
lumpers or splitters some want to see only the similarities, while others prefer
to see the differences. Whether Ceratozamia consists of ten species or 40 is basi-
cally a matter of personal opinion. The detailed relationships between all the
species have not yet been determined through DNA analysis, but even if that was
done, there would probably still not be unanimous agreement as to which species
ch08.qxd 28/11/03 3:47 pm Page 98
98 L.M. Whitelock
to recognize. If any consensus concerning the species concept develops natural-

ly, it will no doubt come slowly. It is therefore imperative that a set of criteria, that
are agreeable to all cycad taxonomists, be developed for circumscribing new
species.
Distribution of Ceratozamia
Currently, there are 21 validly described species of Ceratozamia. Their distribution

starts in a restricted narrow band that follows the mountains along the Gulf of
Mexico from Tamaulipas and continues through San Luis Potosi, Quertaro,
Hidalgo and Puebla. The direction changes slightly just west of the Isthmus of
Tehuantepec in the states of Veracruz, Tabasco and Oaxaca, where it moves
south and east through Tabasco and Chiapas, then continues from Mexico into
Guatemala, Belize and central Honduras. At the present time Ceratozamia is not
known to occur west or south of Honduras.
Speciation within Ceratozamia
For the past 40 years the author has been investigating Ceratozamia in habitat, and
studying the various populations to see how they differ. One of the most inter-
esting observations over this period of time was that most populations of
Ceratozamia are restricted in size and generally found only in a limited area (such
as a mountain top, portion of a canyon, etc.). Although forms of Ceratozamia are
not generally sympatric, it has been reported that C. microstrobila and C. latifolia
grow together (Stevenson et al., 1986; Moretti and Sabato, 1988), but the author
has never seen this personally. This separation of populations leads to speculation
that Ceratozamia was once a widespread genus that was vicariously separated into
numerous populations that have evolved over the past millions of years into indi-
vidual identifiable colonies. These small colonies are not the result of land clear-
ance or collectors, as many of them were known long before conservation issues
arose. The author has studied the morphology of many of these populations and
found them to be sufficiently distinct to be identified when growing in a collec-
tion of cultivated plants, far from their native territory. The changes in these pop-
ulations were no doubt the result of their adapting to differing climates, plant
associations, soils, light conditions, etc.
As previously noted, laboratory tests are useful primarily in identifying
species relationships; however, they have not been perfected to the point where
they can reliably distinguish between species. This makes it necessary to contin-
ue to rely mainly on morphological differences in the identification and descrip-
tion of new species. All of the variable morphological characters within
Ceratozamia must be identified and as many of these as possible should be
addressed when a new species is described. The following similarities and differ-
ences have been observed within Ceratozamia and each of these points should be
ch08.qxd 28/11/03 3:47 pm Page 99
considered individually in the circumscription of a suspected new species.
Similarities within Ceratozamia
All species and forms of Ceratozamia are similar in that they have: (i) pinnate
leaves; (ii) persistent leaf bases; (iii) entire leaflet margins; (iv) articulated leaflets;
(v) stipulate leaves and cataphylls; (vi) cone sporophylls arranged in a spiral, but
appearing to be in vertical rows; (vii) two spines or horns on the face of each
sporophyll; and (viii) a chromosome number of 2n = 16 (where determined).
Differences within Ceratozamia
Species and forms of Ceratozamia differ in the following respects:
Leaves are either upright or arching, short or long, green or brown emergent
and heavily tomentose to almost glabrous. Mature plants produce varying
numbers of new leaves, from a single leaf to over 20, depending on the species.
Petioles may be unarmed (e.g. C. microstrobila, C. zaragozae), lightly armed (e.g.
C. mexicana, C. hildae) or heavily armed (e.g. C. robusta, C. miqueliana) with stiff
spines. The spines may be restricted to the petiole or may sometimes extend up
the rachis almost to its apex.
Female cones may be erect (e.g. C. microstrobila, C. whitelockiana), decumbent (e.g.
C. mexicana, C. robusta) or pendulous (e.g. C. mixeorum, C. zaragozae). The position
of the cone is dependent on the length and diameter of the peduncle. Erect
cones have short thick peduncles, decumbent cones have somewhat longer
thinner peduncles, and pendulous cones have long, thin peduncles (Figs
8.18.3).
Male cones vary from short cylindrical to short or long conical and have sporo-
phylls that vary in size. The size, length and orientation of the horns also
varies considerably between species.
Leaflets vary from narrow (37 mm) to wide (16 cm), and short (1518 cm) to
long (2540 cm). Long narrow leaflets generally indicate a bright dry habitat,
whereas very broad leaflets indicate moist, shady conditions (Fig. 8.4).
Leaflet veins are either prominent or obscure in transmitted light. Leaflet veins
in Ceratozamia can be separated into two types: those with transparent veins (e.g.
C. zoquorum), and those with obscure veins (e.g. C. miqueliana). These two types
are easily distinguished by viewing the leaflets when held up to the light (Fig.
8.5).
Leaflet shape is generally lanceolate, oblong or linear, straight or falcate, with
the adaxial surface flat or channelled. The leaflet apex may be acute (e.g. C.
matudae), obovate (e.g. C. miqueliana) or oblanceolate (e.g. C. euryphyllidia) and
may be either equally or unequally attenuate.
Leaflet texture (or thickness) may be either coriaceous or leathery (e.g. C.
ch08.qxd 28/11/03 3:47 pm Page 100
100 L.M. Whitelock
Fig. 8.1. Ceratozamia hildae G.P. Landry &

M.C. Wilson displaying upright female cone.
Fig. 8.2. Ceratozamia mexicana Fig. 8.3. Ceratozamia zaragozae

Brongniart displaying decumbent Medellin-Leal displaying pendulous
female cone. female cone.
latifolia, C. zoquorum), papyraceous (e.g. C. miqueliana) or membranaceous (e.g. C.

euryphyllidia). Leaflet texture is generally controlled by habitat in that species
with coriaceous leaflets are generally found in seasonally dry habitats that are
often subjected to wind, those with papyraceous leaflets are found in dark con-
tinually moist habitats, and those species with membranaceous leaflets are
ch08.qxd 28/11/03 3:47 pm Page 101
Fig. 8.4. Leaflet forms of Ceratozamia, left to right: C. norstogii D.W. Stevenson; C.
matudae Lundell; C. sabatoi Vovides, Vquez Torres, Schutzman & Iglesias; C.
mexicana Brongniart; C. microstrobila Vovides & J.D. Rees; C. latifolia Miquel; C.
miqueliana H. Wendland; C. sp. becerrae and C. euryphyllidia Vzquez Torres,
Sabato & D.W. Stevenson.
restricted to shady and continually wet habitats.

Leaflet margins are either strongly or lightly revolute, or flat.
Emergent leaves vary in conformation, colour and texture of rachis and
leaflets, and amount and colour of tomentum (Figs 8.68.8).
The adaxial surface of emergent leaflets is glossy (e.g. C. mexicana, C. robusta), or
pruinose, as those of the C. miqueliana complex (e.g. C. whitelockiana, C. zoquo-
Fig. 8.5. Transparent as opposed to obscure leaflet veins as seen by transmitted

light: left, Ceratozamia sp. becerrae; right, C. miqueliana H. Wendland.
ch08.qxd 28/11/03 3:47 pm Page 102
102 L.M. Whitelock
Fig. 8.6. Emergent leaf of Ceratozamia norstogii D.W. Stevenson.
Fig. 8.7. Emergent leaf of Ceratozamia Fig. 8.8. Emergent leaves showing dif-
zoquorum Prez-Farrera, Vovides & ferences in tomentum of Ceratozamia
Iglesias. latifolia Miquel.
rum).
They may have leaflets with the veins on the abaxial surface prominently
raised or flat.
Eophylls (first seedling leaves) display differences in the leaflet shape, length
ch08.qxd 28/11/03 3:47 pm Page 103
and width, number of leaflets (Figs 8.98.11), and their colour and texture. For
example, C. zoquorum has one pair of leaflets that are coriaceous and pruinose,
C. miqueliana has two pairs of leaflets that are papyraceous and pruinose, while
C. norstogii has three pairs of leaflets that are papyraceous.
Stems are either subterranean, shortly emergent or arborescent. Ceratozamia
stems that are either subterranean or shortly emergent are generally < 20 cm
Fig. 8.9. Eophylls (first seedling leaves): left, Ceratozamia zoquorum Prez-Farrera,
Vovides & Iglesias; right, C. miqueliana H. Wendland.
Fig. 8.10. Eophylls (first seedling leaves) of Ceratozamia mexicana Brongniart.
Fig. 8.11. Eophylls (first seedling leaves) of Ceratozamia norstogii D.W. Stevenson.
ch08.qxd 28/11/03 3:47 pm Page 104
104 L.M. Whitelock
in height (e.g. C. microstrobila, C. whitelockiana) or arborescent with stems > 1 m

in height (e.g. C. mexicana, C. robusta). Stems also exhibit either smooth (e.g. C.
microstrobila) or rough (e.g. C. norstogii) exteriors. The difference in stem surface
is caused by the conformation of the cataphylls. The rough surface is pro-
duced by cataphylls that are deeply wrinkled and have a recurved apex,
whereas a smooth surface is produced by cataphylls with a more-or-less flat
exterior and an apex that is held flush with the stem (Figs 8.128.13).
Fig. 8.12. Ceratozamia sp. El Mirador displaying a more-or-less smooth surface

of short, tightly appressed cataphylls.
Fig. 8.13. Ceratozamia norstogii D.W. Stevenson displaying long, wrinkled and
divergent cataphylls.
ch08.qxd 28/11/03 3:47 pm Page 105
Leaf variation in Ceratozamia
Leaves of Ceratozamia are very plastic in response to the amount of light or shade
they receive. Plants growing in shade will have longer and broader leaves and
leaflets; conversely, plants growing in more sunlight will have shorter and nar-
rower leaves and leaflets. Seedling and immature plants of a species will also
display leaflets that are generally much broader than those on a mature plant.
These differences can be so great that if one was shown an immature specimen
and a mature specimen of the same species, they might easily be identified as two
different species (Fig. 8.14).
Fig. 8.14. Ceratozamia matudae Lundell showing the remarkable difference

between a juvenile leaf (upper left) and mature leaves.
The leaflets of Ceratozamia display a number of character differences not

related entirely to environmental conditions; these differences persist with plants
in cultivation and can be used as diagnostic characters in identifying or describ-
ing species (Table 8.1).
Suggested Criteria for New Species Descriptions

If the following criteria dealing with distribution, location, habitat, morphology
and reproduction are incorporated into future descriptions for new taxa of
Ceratozamia, there should be little doubt as to their validity (see Osborne and
Walters, Chapter 15 this volume, for a complete list of guidelines for describing
new cycad taxa).
The taxonomist must study the plant in its natural habitat and not rely entire-
ly on herbarium vouchers and the notes of others.
ch08.qxd
28/11/03
Table 8.1. Ceratozamia leaflet characteristics.
106
Abaxial New
veins Veins Veins Margins Margins Leaflets Leaflets Leaflets leaflets
Taxon raised transparent opaque revolute flat coriacious papyraceous membranaceous pruinose
3:47 pm
C. alvarezii Prez-Farrera, Vovides & Iglesias X X X X
C. euryphyllidia Vzquez Torres, Sabato & D.W. Stevenson X X X X
C. hildae G.P. Landry & M.C. Wilson X X X
C. kuesteriana Regel X X X
C. latifolia Miquel [Xilitla, San Louis Potosi] X X X X
Page 106
C. matudae Lundell X X X X
C. mexicana Brongniart [Jalapa, Veracruz] X X X
C. microstrobila Vovides & J.D. Rees X X X X
C. miqueliana H. Wendland [Tabasco] X X X X X
C. miqueliana H. Wendland [Chiapas] X X X X X
L.M. Whitelock
C. mirandae Vovides, Prez-Farrera & Iglesias X X X X
C. mixeorum Chemnick, T.J. Gregory & Salas-Morales X X X
C. morettii Vzquez Torres & Vovides X X X X
C. norstogii D.W. Stevenson X X X X
C. robusta Miquel [El Sumedero, Chiapas] X X X
C. sabatoi Vovides, Vzquez Torres, Schutzman & Iglesias X X X
C. whitelockiana Chemnick & T.J. Gregory X X X X X
C. zaragozae Medellin-Leal X X X
C. zoquorum Prez-Farrera, Vovides & Iglesias X X X X X
C. sp. Palma Sola [Veracruz] X X X X
C. sp. Santiago Tuxtla [Veracruz] X X X X
C. sp. San Gabriel Mixtepec [Oaxaca] X X X
C. sp. El Mirador [Veracruz] X X X
C. sp. Naranjal [Veracruz] X X X
C. sp. Belize X X X
C. sp. becerrae [Tabasco] X X X X X
C. sp. Lake Catemaco [Veracruz] X X X X
C. sp. Honduras X X X X X
C. sp. [Chiapas] X X X X
C. sp. Redback [Puebla] X X X
ch08.qxd 28/11/03 3:47 pm Page 107
Herbarium vouchers should be made from several plants that exhibit the full
range of variation within the population.
The exact location should be carefully noted, using a GPS if possible. If a GPS
is not available, exact distance and direction from permanent landmarks (e.g.
a bridge, a railroad crossing, geological formation) should be given.
Notes on the habitat should be as complete as possible, and include such
details as altitude, aspect, geology, soil type, climatic data and plant associa-
tions.
The presence or absence of cones and/or seed or seedlings and the extent of
regeneration should be noted. If cones are in the process of shedding pollen
they should be investigated for pollinators. When pollinators are found, speci-
mens should be collected and preserved so that they can later be identified.
Conservation considerations such as land clearance, damage by livestock, road
and dam building, introduced pests and predators, etc. should be noted.
Photographic documentation should be made of the habitat, the plants and
their reproductive structures when present.
Results from laboratory tests such as DNA sequence profiling, chromosome
counts, electrophoretic assays, etc., when available, should be included and dis-
cussed.
References
Avendao, S., Vovides, A.P. and Castillo-Campos, G. (2003) A new species of Ceratozamia
(Zamiaceae, Cycadales) from Veracruz, Mexico. Botanical Journal of the Linnaean Society
141, 395398. [Ceratozamia huastecorum.]
Brongniart, A.T. (1846) Note sur un nouveau genre de Cycades du Mexique. Annales des
Science Naturelles, series 3, 5, 510. [Ceratozamia; C. boliviana, C. mexicana.]
Caputo, P., Stevenson, D.W. and Wurtzel, E.T. (1991) A phylogenetic analysis of American
Zamiaceae (Cycadales), using chloroplast DNA restriction fragment length polymor-
phisms. Brittonia 43, 135145.
Chemnick, J. and Gregory, T.J. (1995) A new species of Ceratozamia (Zamiaceae) from
Oaxaca, Mexico, with comments on distribution, habitat, and relationships. Phytologia
79, 5157. [Ceratozamia whitelockiana.]
Chemnick, J., Gregory, T.J. and Salas-Morales, S. (1997) Ceratozamia mixeorum (Zamiaceae),
a new species of Ceratozamia from Oaxaca, Mexico, with comments on distribution,
habitat, and species relationships. Phytologia 83, 4752.
De Luca, P., Moretti, A., Siniscalco Gigliano, G., Caputo, P., Cozzolino, S., Gaudio,
L., Stevenson, D.W., Wurtzel, E.T. and Osborne, R. (1995) Molecular system-
atics of cycads. In: Vorster, P. (ed.) Proceedings of the Third International Conference on
Cycad Biology. Cycad Society of South Africa, Stellenbosch, South Africa, pp. 131137.
of the Linnaean Society of New South Wales 84, 64117.
Landry, G.P. and Wilson, M.C. (1979) A new species of Ceratozamia (Cycadaceae) from San
Luis Potos. Brittonia 31, 422424. [Ceratozamia hildae.]
Lundell, C.L. (1939) Studies of Mexican and Central American plants, VII. Lloydia 2,
7376. [Ceratozamia matudae.]
ch08.qxd 28/11/03 3:47 pm Page 108
108 L.M. Whitelock
Medellin-Leal, F. (1963) A new species of Ceratozamia from San Luis Potos. Brittonia 15,
175176. [Ceratozamia zaragozae.]
Miquel, F.A.W. (1848) Over eenige nieuwe of zeldzame Cycaden in den Hortus
Botanicus te Amsterdam. Tijdschrift voor de Wis- en Natuurkundige Wetenschappen 1,
3334, 197209. [Ceratozamia robusta, C. latifolia.]
Moretti, A. and Sabato, S. (1988) Systematics and evolution of Dioon and Ceratozamia.
Fairchild Tropical Garden Bulletin 43(1), 2228.
Moretti, A., Sabato, S. and Siniscalco Gigliano, G. (1981) Monosaccharide composition of
the mucilages in Encephalartos Lehm. (Zamiaceae). Giornale Botanico Italiano 115,
291297.
Moretti, A., Sabato, S. and Siniscalco Gigliano, G. (1983) Taxonomic significance of
methylazoxymethanol glycosides in the cycads. Phytochemistry 22, 115118.
Moretti, A., Caputo, P., Gaudio, L. and Stevenson, D.W. (1991) Intraspecific chromosome
variation in Zamia (Zamiaceae, Cycadales). Caryologia 44, 110.
Norstog, K.J. (1980) Chromosome numbers in Zamia (Cycadales). Caryologia 33, 419428.
Norstog, K.J. (1981) Karyotypes of Zamia chigua (Cycadales). Caryologia 34, 255260.
Osborne, R., Salatino, M.L.F. and Salatino, A. (1989) Alkanes of foliar epicuticular waxes
of the genus Encephalartos. Phytochemistry 28, 30273030.
Prez-Farrera, M.A., Vovides, A.P. and Iglesias, C.G. (1999) A new species of Ceratozamia
(Zamiaceae, Cycadales) from Chiapas, Mexico. Novon 9, 410413. [Ceratozamia
alvarezii.]
Prez-Farrera, M.A., Vovides, A.P. and Iglesias, C.G. (2001) New species of Ceratozamia
(Zamiaceae) from Chiapas, Mexico. Botanical Journal of the Linnean Society 137, 8185.
[Ceratozamia mirandae, C. zoquorum.]
Regel, E. (1857) Zwei neue Cycadeen, die im Botanischen Garten zu Petersburg kultivirt
werden, nebst Beitrge zu Kenntniss dieser Familie. Bulletin de la Socit Impriale des
Naturalistes de Moscou 30, 161191. [Ceratozamia kuesteriana.]
Stevenson, D.W. (1982) A new species of Ceratozamia (Zamiaceae) from Chiapas, Mexico.
Brittonia 34, 181184. [Ceratozamia norstogii.]
Stevenson, D.W., Sabato, S. and Vzquez Torres, M. (1986) A new species of Ceratozamia
(Zamiaceae) from Veracruz, Mexico with comments on species relationships, habi-
tats, and vegetative morphology in Ceratozamia. Brittonia 38, 1726. [Ceratozamia eury-
phyllidia.]
Vzquez Torres, M. and Vovides, A.P. (1998) A new species of Ceratozamia (Zamiaceae)
from Veracruz, Mexico. Novon 8, 8790. [Ceratozamia morettii.]
Vovides, A.P. and Olivares, M. (1996) Karyotype polymorphism in the cycad Zamia loddi-
gesii (Zamiaceae) of the Yucatan peninsula, Mexico. Botanical Journal of the Linnean
Society, 120, 7783.
Vovides, A.P. and Rees, J.D. (1983) Ceratozamia microstrobila (Zamiaceae), a new species from
San Luis Potos, Mexico. Madroo 30, 3942.
Vovides, A.P., Vzquez Torres, M., Schutzman, B. and Iglesias, C.G. (1993) A new species
of Ceratozamia (Zamiaceae) from Quertaro and Hidalgo, Mexico. Novon 3, 502506.
[Ceratozamia sabatoi.]
Wendland, H. (1854) Index Palmarum, Cyclanthearum, Pandanearum, Cycadearum, quae in Hortis
Europaeis Coluntur. Hahn, Hannover, Germany, pp. 4954, 68. [Ceratozamia miqueliana.]
ch09.qxd 28/11/03 3:48 pm Page 109
Relationships and 9
Phytogeography in
Ceratozamia (Zamiaceae)
Andrew P. Vovides,1 Miguel A. Prez-Farrera,2

Dolores Gonzlez1 and Sergio Avendao1
1Instituto de Ecologa
A.C., Xalapa, Veracruz, Mexico;
2Escuela de Biologa,
Universidad de Ciencias y Artes de
Chiapas (UNICACH), Tuxtla Gutirrez, Chiapas, Mexico
Abstract
Species relationships and phytogeography of Ceratozamia, an endangered neotropical

cycad genus, based on morphology, leaflet histology, DNA sequencing and ecological data,
are analysed and discussed. Seven species complexes are assigned within the genus. A
centre of origin in south-eastern Mexico, with recent speciation and immigration patterns,
is proposed.
Introduction
Ceratozamia Brongniart presents morphological variation both within and
between species. Miquel (1870) noted differences in leaflet shape and size as well
as varying sizes of cones with age of any given species of Ceratozamia but he saw
stability in the morphology of micro- and megasporophylls. Thiselton-Dyer
(18821886) commented: This variableness [sic] with age makes the separation
of nearly allied forms an all but hopeless task. Chromosome number and kary-
otype within the genus is extremely stable (Vovides, 1983, 1985; Moretti, 1990),
which contrasts with chromosome variation found in Zamia Linnaeus (Moretti
and Sabato, 1984, Vovides and Olivares, 1996). Vovides (1983, 1985) noted dif-
ferences in secondary constriction or satellite number and position within
Ceratozamia. Few anatomical and morphological characters appear to separate
related Ceratozamia species. To date, species descriptions within the genus have
ch09.qxd 28/11/03 3:48 pm Page 110
110 A.P. Vovides et al.
been based largely on gross morphological characters, strobilus indumentum,

colouring, and geographical distribution.
Phytogeography
The genus Ceratozamia presently comprises 21 species (Hill et al., Appendix 1 this
volume) and this figure will probably increase to about 26 in the near future. It is
largely distributed in the warm temperate and tropical regions of eastern and
southern Mexico, from the state of Tamaulipas at its extreme northern limit in
Mexico, to Belize and Guatemala in central America (Balduzzi et al., 1982).
Recently a taxon related to the C. miqueliana complex has been found in northern
Honduras, hence the range of the genus has been extended. The Ceratozamia
complexes in southern and south-eastern Mexico appear to coincide with floris-
tic refuges, according to Toledo (1982) and Wendt (1987). In support of the
Pleistocene refuge theory, there appears to be fossil evidence of Ceratozamia of
Miocene age from a Cenozoic floristic refuge in Oaxaca (Rzedowski and
Palacios, 1977).
Morphology and distribution
In terms of gross morphology, plants from north-eastern Mexico are generally

characterized as having small and often branching trunks with few leaves per
crown. Leaflets are usually narrow (i.e. < 2 cm wide), with the exception of those
of Ceratozamia microstrobila Vovides & J.D. Rees and another two isolated species in
northern Veracruz (under investigation). In addition, the plants from the north-
eastern areas bear relatively smaller cones than those in southern and south-
eastern Mexico.
Plants from southern and south-eastern Mexico are generally large, single
trunked plants, although branching has been observed in some taxa, e.g.
Ceratozamia alvarezii Prez-Farrera, Vovides & Iglesias from Chiapas. These south-
ern and south-eastern taxa have leaflets usually > 2 cm wide, the widest being
found in C. euryphyllidia Vzquez Torres, Sabato & D.W. Stevenson. Exceptions to
the generally wide leaflet morphology are seen in C. norstogii D.W. Stevenson, C.
mirandae Vovides, Prez-Farrera & Iglesias and C. alvarezii, all three of which have
leaflets < 1.5 cm wide.
Moretti et al. (1980) were the first to recognize species complexes within the
genus highlighting separate Ceratozamia mexicana and C. matudae complexes (Fig.
9.1). Stevenson et al. (1986) proposed two main groups in Ceratozamia, based on
leaflet morphology: (i) a C. mexicana group with long narrow leaflets; and (ii) a C.
euryphyllidia/miqueliana group with shorter and wider leaflets. However, through-
out the distribution of Ceratozamia, various other related groups or complexes can
also be found. We have confirmed the complexes identified by the earlier workers,
as well as other new complexes principally in southern and south-eastern
ch09.qxd 28/11/03 3:48 pm Page 111
Relationships and Phytogeography in Ceratozamia 111
Fig. 9.1. Distribution in Mexico and Guatemala of four Ceratozamia species com-
plexes: A = C. kuesteriana species complex; B = C. latifolia species complex; C =
C. mexicana species complex; D = C. matudae species complex.
Mexico. Our main two additional groups are a C. norstogii complex (Fig. 9.2)
(Prez-Farrera et al., 1999; Vovides et al., 2001) and a C. miqueliana complex (Fig.
9.3) (Prez-Farrera et al., 2001). The distribution of these complexes is apparent-
ly related to their habitats. For instance, C. miqueliana H. Wendland and related
species occur in evergreen tropical rainforest and are not generally found at ele-
vations > 1000 m; these plants are characterized by having both decumbent and
erect female cones, and leaves up to 2.5 m long with oblanceolate leaflets. By con-
trast, species of the C. norstogii group are largely found in oak and pine/oak forests
at elevations between 800 and 1400 m; characteristic features of these species are
branching or cylindrical and non-branching trunks, erect male and female cones,
and linear, flat or channelled leaflets with a straight or spirally twisted rachis. The
C. robusta species complex (sensu Stevenson et al., 1986) (Fig. 9.4) appears to have
a wide range of habitat type ranging over evergreen and deciduous tropical
forests, oak forests and cloud forests; plants within this group are large and robust
with leaves up to 3 m long and bear large erect or decumbent male and female
cones.
On the basis of gross morphology we now believe that there are at least
seven species complexes in the genus:
1. The Ceratozamia mexicana complex is mainly distributed on the trans-

Mexican Neovolcanic mountain range, covering central Veracruz, parts of
Puebla and Hidalgo states and possibly north-eastern Oaxaca (Fig. 9.1). Included
ch09.qxd 28/11/03 3:48 pm Page 112
Fig. 9.2. Distribution in Mexico (encircled area) of the Ceratozamia norstogii

species complex.
here is C. mexicana Brongniart var. robusta Dyer (sensu Thiselton-Dyer, 18821886;

Vovides et al., 1983) and C. mexicana var. mexicana.
2. The Ceratozamia latifolia complex, a species group to the north and
north-east of the trans-Mexican Neovolcanic mountain range, is a somewhat
Fig. 9.3. Distribution in Mexico (encircled area) of the Ceratozamia miqueliana

species complex.
ch09.qxd 28/11/03 3:48 pm Page 113
Fig. 9.4. Distribution in Mexico, Guatemala and Belize (encircled area) of the
Ceratozamia robusta species complex.
loose assemblage and may subsequently prove to comprise more than one
complex. As presently circumscribed, the group consists of C. latifolia Miquel, C.
hildae G.P. Landry & M.C. Wilson, C. microstrobila Vovides & J.D. Rees, C. huasteco-
rum S. Avendao, Vovides & Castillo-Campos, C. brevifrons Miquel (= C. mexicana
Brongniart; see Hill et al., Appendix 1 this volume) and C. morettii Vzquez Torres
and Vovides (the latter two species are on the Neovolcanic transverse mountain
range).
3. The Ceratozamia kuesteriana complex occurs in north-eastern Mexico
and includes C. kuesteriana Regel, C. sabatoi Vovides, Vzquez Torres, Schutzman
& Iglesias, and C. zaragozae Medellin-Leal. These are three taxa with narrow
lanceolate to linear leaflets. Other taxa in this region are at present under inves-
tigation.
4. The wide leaflet Ceratozamia miqueliana complex which occurs to
the south and south-east of the trans-Mexican Neovolcanic belt, comprises about
five taxa. These are C. miqueliana, C. euryphyllidia, C. zoquorum Prez-Farrera,
Vovides & Iglesias, a new taxon that will be described in the near future
(Ceratozamia sp. becerrae) and another possibly new species from Chiapas.
5. The Ceratozamia norstogii complex is largely distributed to the
south in Chiapas and eastern Oaxaca, with three apparently well-defined
species: C. norstogii, C. alvarezii and C. mirandae Vovides, Prez-Farrera &
Iglesias (see Prez-Farrera et al., Chapter 10, this volume). We are also tentative-
ly placing a new taxon, yet to be described, in this complex (Ceratozamia sp.
chimalpensis).
ch09.qxd 28/11/03 3:48 pm Page 114
6. The Ceratozamia robusta (sensu Stevenson) group is found in

Chiapas, Oaxaca, southern Veracruz, Guatemala and Belize. This is another
loose assemblage in which we place C. robusta Miquel, C. whitelockiana Chemnick
& T.J. Gregory, C. mixeorum Chemnick, T.J. Gregory & Salas-Morales, and a
further new species of Ceratozamia, yet to be described, from Oaxaca.
7. The Ceratozamia matudae complex, which is distributed from the
extreme south of Chiapas into Guatemala consists of C. matudae Lundell
and a newly discovered species yet to be described (Ceratozamia sp.).
The taxonomy of Ceratozamia is far from resolved and there is still much confu-
sion at the species level within these complexes. It is hoped that the molecular,
anatomical and phytogeographical studies at present under way on the genus will
eventually contribute to the clarification of the status of species within
Ceratozamia.
Habitats
Most species of Ceratozamia can be found within 1517.5 north and 92.598.5
west. Typical habitats for Ceratozamia in north-eastern Mexico are broadleaf decid-
uous cloud forest and pine/oak forest at elevations between 500 and 2000 m. In
southern and south-eastern Mexico, Ceratozamia can be found in lowland ever-
green tropical rainforest at elevations < 100 m, to cloud forests and pine/oak
forests from about 600 to 1400 m. Populations can be locally abundant, and in
some cases dominant in the herbaceous undergrowth of cloud forests, such
as in the cases of C. mexicana in Veracruz and C. mirandae in Chiapas. Using
multivariate analysis on taxa and their habitat types (McCune and Mefford,
1997) species with wide oblanceolate to oblong leaflets (C. euryphyllidia, C. mexicana
var. robusta, C. miqueliana, C. zoquorum and C. sp. becerrae) are confined to ever-
green tropical rainforest, whilst a number of species with narrow, linear or falcate
leaflets are frequently associated with cloud forests, oak and pine/oak forests
(Fig. 9.5).
Molecular Studies
Molecular phylogenetic analyses on New World cycads are scarce, and those so
far published have examined restriction fragment length polymorphism variation
data at generic level (Caputo et al., 1991, 1993; Moretti et al., 1993; De Luca et al.,
1995). In order to gain insight into phylogenetic relationships within the genus
Ceratozamia, we undertook a study of variation in sequences of non-coding
regions from the chloroplast and nuclear genomes. We selected two regions that
have been used in phylogenetic studies at infrageneric level for a variety of plants
(Schilling et al., 1998; Potter et al., 2000). One is the nuclear internal transcribed
spacer (ITS) region, which has proved to be an excellent source of information
ch09.qxd 28/11/03 3:48 pm Page 115
Number of species
Types of vegetation
Principal component 2
Principal component 1
Fig. 9.5. Relationships between taxa of Ceratozamia and habitat (BTSC, tropical
sub-deciduous rainforest; BTP, evergreen tropical rainforest; BMM, cloud forest;
BQ, oak forest; BC, coniferous forest; BTC, tropical deciduous forest). (A) Number
of taxa of Ceratozamia present in six vegetation types. (B) Summary of a principal
components analysis of habitat and leaflet width for taxa of Ceratozamia.
from the nuclear genome (Baldwin et al., 1995). However, only 29 characters out
of 1083 of the ITS region were informative among 24 exemplars. The second
region we examined is the chloroplast genome trnL-F non-coding region and
trnL-trnF spacer. Here, the alignment for a subsample of 16 Ceratozamia sequences
from the trnL-F non-coding region generated 998 characters, from which only
two were informative. The selection of Zamia as an outgroup was based on the
ch09.qxd 28/11/03 3:48 pm Page 116
results obtained with the analyses of the trnL-F non-coding region, where Zamia
appeared as the sister group to Ceratozamia.
The sequence data generated in this study were used to evaluate: (i) the
hypothesis that Ceratozamia is monophyletic; and (ii) the species relationships
within the genus. The results of these analyses yielded 112 trees of 51 steps; the
strict consensus of these trees is shown in Fig. 9.6A (with bootstrap values and
decay index) and one randomly chosen minimal length tree is presented in Fig.
9.6B to show average branch and distribution of the taxa. The low level of vari-
ation detected among species limits the conclusions that could be drawn from the
study; however, the cladograms did enable us to infer a tentative but interesting
scenario for the genus (Gonzlez and Vovides, 2002).
Other molecular techniques, such as random amplified polymorphic DNA
(now in progress), might help to identify the more variable regions of the genome
in order to enable us to design primers to sequence those regions in the near
future. Microsatellite techniques also appear to be promising.
Leaflet Anatomy
Leaflets from five adult plants from each of five taxa were taken from the mid-
portion of a mature leaf. Leaflets were fixed in formalin acetic alcohol then hand
microtome sections were made. These were stained in phloroglucinol HCl to
show lignified tissues and in alcoholic Sudan III and IV mixture for cutin and
other fatty acid tissues. Nineteen measurements were obtained from each indi-
vidual plant using a calibrated eyepiece scale (Table 9.1). Analysis was done using
analysis of variance (ANOVA) and multiple range tests from Statistica (StatSoft,
Inc. Tulsa, Oklahoma, USA).
Results so far from leaflet anatomy appear to be promising in our attempts
to isolate useful characters for delimiting species. Nevertheless, a homogeneity of
states has been observed in about 50% of the characters studied, reflecting a
similar situation to that found in DNA sequencing of the ITS region mentioned
previously. From a suite of 19 characters analysed using morphometric and
phenetic techniques, nine were found to be taxonomically useful and enabled us
to separate a mean of nine groups from a total of 18 operational taxonomic
units.
Discussion
Ceratozamia is largely found along the Sierra Madre Oriental, in southern and
south-eastern Mexico and through Guatemala and Belize to Honduras. The
genus presents disjunct populations in Mexico (Sabato and De Luca, 1985;
Moretti et al., 1993) with the exception of C. robusta (sensu Stevenson) where the
majority of populations are found in south-east Veracruz, north-west Chiapas
and the southern sierra of Oaxaca, Belize and Guatemala. Future work will help
ch09.qxd
28/11/03
C. morettii
3:48 pm
C. morettii
Relationships and Phytogeography in Ceratozamia
Page 117
C. mirandae C. mirandae
C. matudae C. matudae
Fig. 9.6. Phylogenetic relationships within Ceratozamia based on ITS sequences (length 51, consistency index = 0.8235, retention index
117
= 0.9151, rescaled consistency index = 0.7536) (from Gonzlez and Vovides, 2002). (A) Strict consensus tree. (B) Minimal length tree
chosen at random.
ch09.qxd 28/11/03 3:48 pm Page 118
Table 9.1. Leaflet anatomical characters used to distinguish Ceratozamia taxa.

Five plants from each of five taxa were sampled for the total 19 characters.
Characters Dimensions Number
Abaxial epiderm cells Length and width 2

Cuticles abaxial and adaxial Thickness 2
Epiderm cells abaxial and adaxial Length and width 4
Palisade parenchyma cells Length and width 2
Sclereids associated with vascular Number, length and width 3
bundles
Sclereids not associated with vascular Number, length and width 3
bundles
Bulliform-like cells Length and width 2
Stomatal pores Depth 1
Total number of characters examined 19
to resolve whether these represent a single disjunct species or a complex of

related taxa.
Stevenson et al. (1986) proposed that leaflet morphology and habitat are cor-
related, with those species of Ceratozamia having wide and thin leaflets being con-
fined to the more humid environments and taxa with narrow and more
coriaceous leaflets to the more xeric habitats. We have confirmed this with our
observations. So far, leaflet anatomy appears to group the species into nine
groups. Figures 9.7 and 9.8 show examples of two useful anatomical characters:
(i) there appear to be lower leaflet cuticle thickness differences between C.
miqueliana and C. zaragozae but no differences in stomatal pore depth; and (ii) there
are differences in the number of associated and non-associated fibres with the
vascular bundles of the leaflets of C. mexicana var. robusta and C. kuesteriana (S.
Avendao, Veracruz, 2002, personal communication). Although there appears to
be a certain correlation of the anatomical features with distribution, it is too early
to draw conclusions.
Stevenson et al. (1986) mentioned vicariance as a possible explanation for the
present distribution of Ceratozamia and that the genus is monophyletic. Prez-
Farrera (1999) suggested that at least some species could be dispersed by
mammals such as the peccari (Tayassu tajacu). Some of the present unusual distri-
bution patterns could be the result of now-extinct dispersal agents, as suggested
by Jansen and Martin (1982) for tropical trees bearing large fruits.
It appears that Ceratozamia (Cycadopodites) was present in Miocene floras in
Pichucalco, Chiapas, southern Mexico (Palacios and Rzedowski, 1993) and in
Cenozoic Engelhardtia forests in Oaxaca, where the fossil pollen spectrum is
remarkably similar to that of the modern; it is noted that Engelhardtia used to be
widely distributed over the northern hemisphere in the Tertiary (Rzedowski and
Palacios, 1977). However, owing to climatic changes during the Pleistocene
ch09.qxd 28/11/03 3:48 pm Page 119
11
10
Cuticle thickness (micron) Std. Dev.
9 Std. Error
8 Mean
2
C. miqueliana #1 C. zaragozae #1
C. miqueliana #2 C. zaragozae #2
Fig. 9.7. Significant differences in cuticle thickness of lower leaflet epidermis

between two samples of Ceratozamia miqueliana H. Wendland and two samples
of C. zaragozae Medellin-Leal.
(Burnham and Graham, 1999) some members of Ceratozamia survived in the

south of Mexico, where the greatest diversity appears to be concentrated, and
very probably an ancestral form became isolated in Honduras, a territory with a
history of geographical isolation (Coney, 1982).
Isolation and subsequent speciation in the genus appear to be relatively
recent, since DNA sequencing data show a low level of variation amongst species
(Gonzles and Vovides, 2002). However, some interesting insights into the phyto-
geography of this genus can be seen where a correlation with geographical dis-
tribution patterns appears to be evident.
The three main clades in Fig. 9.9A are consistent with distributional ranges
of the species. The two basal clades contain the species that are distributed in
southern and south-eastern Mexico, both at and south of the Neovolcanic moun-
tain range of PlioceneQuaternary (Pleistocene) age (Fig. 9.9B). Here the
Ceratozamia norstogii clade integrates well with what we consider the C. norstogii
complex. Though we consider C. zoquorum to be part of the C. miqueliana
complex, it appears to be paraphyletic to the C. miqueliana clade, but, owing to the
low number of molecular characters that generated this tree, this is a tentative
hypothesis and further research is still in progress. By and large, there is consis-
tency with the findings of Marshall and Liebherr (2000), who identified two bio-
geographic assemblages, one north of the Neovolcanic range and another to the
south. The large unresolved clade contains a group of morphologically distinct
species to the north and north-east of the Neovolcanic range which appears to be
of more recent speciation (Fig. 9.9).
The vegetational history of southern Mexico and the Maya Mountains
ch09.qxd 28/11/03 3:48 pm Page 120
af
af
Fig. 9.8. Differences in number of associated (af) and non-associated (nf) fibres
with the vascular bundles of leaflets (ue, upper epidermis; pa, palisade mesophyll;
sm, spongy mesophyll; s, bundle sclerenchyma; x, xylem; ph, phloem; le, lower
epidermis). (A) Ceratozamia mexicana Brongniart var robusta Dyer. (B)
Ceratozamia kuesteriana Regel.
(Belize) has been studied by Lundell (1939, 1945) and Miranda (1957, 1959), who
agreed that the region contains relict floral elements of great age. During the past
40,000 years, tropical forests in Mexico have been disrupted and displaced to
lower latitudes due to the onset of Pleistocene climatic changes with cycles of
cold-dry, cold-wet and warm-dry climates, thus leaving relict pockets or refuges
that protected the biota from lowering of temperature, rainfall and precipitation
(Toledo, 1982). Graham (1976), using palynological analysis, maintained that
modern tropical rainforest in Mexico is recent. However, there is a general con-
sensus on the existence of floristic and faunistic refuges of great age in southern
Mexico (Brown, 1976; Toledo, 1982, 1988), but these are apparently absent in
the areas north of the Neovolcanic mountain range. From the data presented in
this study we put forward the hypothesis that Ceratozamia has its origin in south-
ern/south-eastern Mexico. The unresolved clade of Ceratozamia species at and
north of the Neovolcanic range is very likely the result of adaptive radiation or
ch09.qxd 28/11/03 3:48 pm Page 121
C. matudae
Fig. 9.9. Relationships among species of Ceratozamia (from Gonzlez and

Vovides, 2002). (A) Phylogenetic relationships using ITS sequences. (B)
Geographical distribution in Mexico of the three major clades shown in Fig. 9.9A.
recent speciation processes. Those populations north of the range are mor-
phologically distinct, with gross morphological characters that do not change
when cultivated under similar conditions over several years; these are, however,
homogeneous in DNA sequencing. It may be argued that there are very
few mutations in those species that are resolved, but the relatively long genera-
tion times for Ceratozamia must be considered: species have a generation
time (germination to maturity to seed set) of at least 15 years under optimal
cultivation conditions and this period can safely be doubled for conditions
in the wild. Therefore, only about 300 generations of a putative Ceratozamia
ch09.qxd 28/11/03 3:48 pm Page 122
would have occurred since the end of the Pleistocene (10,000 years ago); thus any
speciation within the genus would be understandably slow. It is hoped that
anatomical, morphometric and other molecular studies will throw more light on
this matter.
Examining the cladogram (Fig. 9.9A), better-resolved clades (all south of the
Neovolcanic belt) can easily be placed into the known refuges (Fig. 9.9B). The
Ceratozamia matudae clade is found in the Soconusco refuge of Chiapas on the
Sierra Madre del Sur while C. mixeorum is from the adjacent northern mountains
of Oaxaca near the Sierra de Juarez refuge. The C. miqueliana, C. euryphyllidia, C.
sp. becerrae, C. whitelockiana and C. zoquorum clades are situated on and adja-
cent to the Los Tuxtla refuge of Toledo and the arc refuge area of Wendt
(1987), covering southern Veracruz, northern Chiapas, northern Oaxaca and
south-western Tabasco. These species comprise what might be called the C.
miqueliana complex. On the Neovolcanic range we find the C. mexicana complex
with a separate species on the Cordoba refuge of Toledo (1982, 1988). The rest
of the unresolved cladogram consists of Ceratozamia species that are north of the
Neovolcanic belt. These species appear to be much younger, suggestive of adap-
tive radiation with a northwards migratory pattern following a general warming
of climate. The C. mexicana clade (on the Neovolcanic range) appears to be the
most recently derived.
Conclusions
We put forward the hypothesis that Ceratozamia, a relict palaeoendemic, is cur-

rently in the process of active speciation. This process was probably initiated
during the early Cenozoic and was stimulated by climate changes of the late
Tertiary and Quaternary periods giving rise to the present taxa, which are appar-
ently well adapted to modern ecosystems.
From a conservation point of view, we reiterate that cycads, which are phylo-
genetically basal to the living seed plants, should be given top priority in the
agendas of countries that are rich in their cycad diversity. With 303 species (see
Hill et al., Appendix 1 this volume), cycads are relatively few worldwide; efforts for
their conservation should not be too expensive and the returns very great, consid-
ering the vast information contained in their genomes. As Norstog and Nicholls
(1997) commented, they are the Rosetta Stone of spermatophyte evolution.
References
Balduzzi, A., De Luca, P. and Sabato, S. (1982) A phytogeographical approach to the New
World Cycads. Delpinoa 2324, 185202.
Baldwin, B.G., Sanderson, M.J., Porter, J.M., Wojciechowski, M.F., Campbell, C.S. and
Donoghue, M.J. (1995) The ITS region of nuclear ribosomal DNA: a valuable source
ch09.qxd 28/11/03 3:48 pm Page 123
of evidence of angiosperm phylogeny. Annals of the Missouri Botanical Garden 82,

247277.
Brown, K.S. Jr (1976) Centros de evoluo, refugios Quaternrios, e conservao de
patrimnios genticos na regio neotropical: Padres de diferenciao em Ithomiinae
(Lepidoptera: Nymphlidae). Acta Amazonica 7, 75137.
Burnham, R.J. and Graham, A. (1999) The history of neotropical vegetation: new devel-
opments and status. Annals of the Missouri Botanical Garden 86, 546589.
Zamiaceae (Cycadales) using chloroplast DNA restriction fragment length polymor-
phisms. Brittonia 43, 135145.
Caputo, P., Marquis, C., Wurtzel, T., Stevenson, D.W. and Wurtzel, E.T. (1993) Molecular
biology in cycad systematics. In: Stevenson, D.W. and Norstog, K.J. (eds) The Biology,
Structure, and Systematics of the Cycadales. Proceedings of the Second International Conference on
Cycad Biology. Palm & Cycad Societies of Australia Limited, Milton, Queensland,
Australia, pp. 213219.
Coney, P.J. (1982) Plate tectonic constraints on the biogeography of middle America and
the Caribbean region. Annals of the Missouri Botanical Garden 69, 432443.
De Luca, P., Moretti, A., Siniscalco Gigliano, G., Caputo, P., Cozzolino, S., Gaudio, L.,
Stevenson, D.W., Wurtzel, E.T. and Osborne, R. (1995) Molecular systematics of
cycads. In: Vorster, P. (ed.) Proceedings of the Third International Conference on Cycad Biology.
Cycad Society of South Africa, Stellenbosch, South Africa, pp. 131137.
Gonzlez, D. and Vovides, A.P. (2002) Low intralineage divergence in the genus
Ceratozamia Brongn. (Zamiaceae) detected with nuclear ribosomal DNA ITS and
chloroplast DNA trnL-F non-coding region. Systematic Botany 27, 654661.
Graham, A. (1976) Late Cenozoic evolution of tropical lowland vegetation in Veracruz,
Mexico. Evolution 29, 723735.
Jansen, D.H. and Martin, P.S. (1982) Neotropical anachronisms: the fruits the Gom-
photheres ate. Science 215, 1927.
Lundell, C.L. (1939) Studies of Mexican and Central American Plants,VII. Lloydia 2,
7376. [Ceratozamia matudae.]
Lundell, C. (1945) Vegetation and natural resources of British Honduras. In: Lundell,
C.L. (ed.) Plant and Plant Science in Latin America. Chronica Botanica, Waltham,
Massachusetts, pp. 270273.
Marshall, C.J. and Liebherr, J.K. (2000) Cladistic biogeography of the Mexican transition
zone. Journal of Biogeography 27, 203216.
McCune, B. and Mefford, J. (1997) Multivariate Analysis of Ecology Data. Version 3.17. MJM
Software, Gleneden Beach, Oregon, 126 pp.
Miquel, F.A.W. (1870) la connaissance des cycades. Adansonia 9, 154180.
Miranda, F. (1957) Vegetacin de la vertiente del Pacfico de la Sierra Madre de
Chiapas (Mxico) y sus relaciones florsticas. Proceedings of the Eighth Pacific Congress 4,
438452.
Miranda, F. (1959) Estudios acerca de la vegetacin. In: Beltran, E. (ed.) Los recursos natu-
rales del sureste y su aprovechamiento. D.F. Instituto Mexicano de Recursos Naturales
Renovables, Mexico, pp. 215217.
Moretti, A. (1990) Karyotypic data on North and Central American Zamiaceae
(Cycadales) and their phylogenetic implications. American Journal of Botany 77,
10161029.
Moretti, A. and Sabato, S. (1984) Karyotype evolution by centromeric fission in Zamia
(Cycadales). Plant Systematics and Evolution 146, 215223.
ch09.qxd 28/11/03 3:48 pm Page 124
Moretti, A., Sabato, S. and Vzquez Torres, M. (1980) The distribution of Ceratozamia
Brongn (Zamiaceae). Delpinoa 20, 1321.
Moretti, A., Caputo, P., Cozzolino, S., De Luca, P., Gaudio, L., Siniscalco Gigliano, G.
and Stevenson, D.W. (1993) A phylogenetic analysis of Dioon (Zamiaceae). American
Journal of Botany 80, 204214.
Palacios, C.R. and Rzedowski, J. (1993) Estudio palinolgico de las floras fsiles del
Mioceno inferior y principios del Mioceno Medio de la regin de Pichucalco,
Chiapas, Mxico. Acta Botnica de Mxico 24, 196.
Prez-Farrera, M.A. (1999) Dinmica poblacional de Ceratozamia matudae en la Reserva de
la Biosfera el Triunfo, Chiapas, Mxico. Tesis de Maestria, El Colegio de la Frontera
Sur. San Cristbal de Las Casas, Chiapas, Mexico.
(Zamiaceae, Cycadales) from Chiapas, Mexico. Novon 9, 410413. [Ceratozamia alvarezii.]
(Zamiaceae) from Chiapas, Mexico. Botanical Journal of the Linnean Society 137, 7780.
[Ceratozamia zoquorum.]
Potter, D., Luby, J.J. and Harrison, R.E. (2000) Phylogenetic relationships among species
of Fragaria (Rosaceae) inferred from non-coding nuclear and chloroplast DNA
sequences. Systematic Botany 25, 337348.
Rzedowski, J. and Palacios, C. (1977) El bosque de Engelhardtia (Oreomunnea) mexicana en la
regin de la Chinantla (Oaxaca, Mxico). Una reliquia del Cenozoico. Boletn de la
Sociedad Botnica de Mxico 36, 93123.
Sabato, S. and De Luca, P. (1985) Evolutionary trends in Dion [sic] (Zamiaceae). American
Schilling, E.E., Linder, C.R., Noyes, R.D. and Rieseberg, L.H. (1998) Phylogenetic rela-
tionships in Helianthus (Asteraceae) based on nuclear ribosomal DNA transcribed
spacer region sequence data. Systematic Botany 23, 177187.
phyllidia.]
Thiselton-Dyer, W.T. (18821886) Cycadaceae. Biologia Centrali-Americana 3, 190195.
Toledo, V.M. (1982) Pleistocene changes of vegetation in tropical Mexico. In: Prance,
G.T. (ed.) Biological Diversification in the Tropics. Proceedings of the Fifth International
Symposium of the Association for Tropical Biology. Columbia University Press, New York,
pp. 93111.
Toledo, V.M. (1988) La diversidad biolgica de Mxico. Ciencia y Desarrollo 14, 1730.
Vovides, A.P. (1983) Systematic studies on the Mexican Zamiaceae. I. Chromosome
numbers and karyotypes. American Journal of Botany 70, 10021006.
Vovides, A.P. (1985) Systematic studies on the Mexican Zamiaceae II. Additional notes on
Ceratozamia kuesteriana from Tamaulipas, Mexico. Brittonia 37, 226231.
Vovides, A.P. and Olivares, M. (1996) Karyotype polymorphism in the cycad Zamia loddi-
gesii (Zamiaceae) of the Yucatan Peninsula, Mexico. Botanical Journal of the Linnean
Society 120, 7783.
Vovides, A.P., Rees, J.D. and Vzquez Torres, M. (1983) Familia Zamiaceae. In: Sosa, V.
and Gomez-Pompa, A. (eds) Flora de Veracruz. INIREB, Xalapa, Veracruz, Mexico,
Fascculo 26.
ch09.qxd 28/11/03 3:48 pm Page 125
Vovides, A.P., Prez-Farrera, M.A. and Iglesias, C.G. (2001) Another new species of
Ceratozamia (Zamiaceae) from Chiapas, Mexico. Botanical Journal of the Linnean Society
137, 8185. [Ceratozamia mirandae.]
Wendt, T. (1987) Las selvas de Uxpanapa, Veracruz-Oaxaca, Mxico: evidencia de refu-
gios florsticos Cenozoicos. Anales del Instituto de Biologa UNAM Serie botnica 58,
2954.
ch09.qxd 28/11/03 3:48 pm Page 126
ch10.qxd 28/11/03 3:49 pm Page 127
A Morphometric Analysis of 10
the Ceratozamia norstogii
Complex (Zamiaceae)
Miguel A. Prez-Farrera,1 Andrew P. Vovides,2

Luis Hernndez-Sandoval,3 Dolores Gonzlez2
and Mahinda Martnez3
1Escuela de Biologa, Universidad de Ciencias y Artes de

Chiapas (UNICACH), Tuxtla Gutirrez, Chiapas, Mexico;
2Instituto de Ecologa A.C., Xalapa, Veracruz, Mexico;
3Facultad de Biologa, Universidad Autnoma de Quertaro,
Centro Universitario, Quertaro, Mexico
Abstract
The three species forming the Ceratozamia norstogii complex (C. alvarezii, C. miran-
dae and C. norstogii) are found in adjacent areas of the Sierra Madre de Chiapas
in Mexico. Taxonomic limits within this complex have not yet been fully defined,
but are investigated in this project. Twenty morphological variables from 90 indi-
viduals from three populations have been analysed using ANOVA and discrimi-
nant analysis. The results reveal a clear number of differences for these variables
among the three species.
Introduction
The taxonomy and distribution of the genus Ceratozamia Brongniart (Zamiaceae)

is little known due to various problems: (i) there is no clear delimitation of some
of the species, principally in the C. mexicana Brongniart, C. latifolia Miquel and C.
miqueliana H. Wendland complexes or species groups; (ii) many of the type spec-
imens on which the original descriptions were based have been lost or destroyed
and the descriptions themselves are often based on sterile or juvenile material and
ch10.qxd 28/11/03 3:49 pm Page 128
128 M.A. Perz-Farrera et al.
on plants cultivated in botanic gardens (Vovides et al., 1983); (iii) species com-
plexes are found throughout the distribution of the genus (Moretti et al., 1980);
and (iv) some neotypifications appear to be doubtful (e.g. C. robusta Miquel)
(Stevenson and Sabato, 1986).
Ceratozamia norstogii D.W. Stevenson was described in 1982 as having the dis-
tinctive characteristics of channelled leaflets and a straight rachis, but the type
specimen assigned (C.A. Purpus 6) has a twisted rachis (Stevenson, 1982). On the
basis of herbarium vouchers, Stevenson et al. (1986) assigned some populations
and forms with channelled leaflets and a straight rachis to C. norstogii under the
assumption that this species was polymorphic. However, it now appears that C.
norstogii is not polymorphic and we believe that this species belongs in a group
with two recently described related species (Prez-Farrera et al., 1999, 2001;
Vovides et al., 2001). Unfortunately the range of morphological variation within
this complex and others of the genus is not yet fully understood. This lack of
information, coupled with the difficulty in obtaining morphological characters
from fertile specimens, has contributed to taxonomic confusion in the genus.
In this study we examined the morphological variation within the Ceratozamia
norstogii complex using multivariate statistical techniques to test the morphologi-
cal differentiation between C. alvarezii Prez-Farrera, Vovides & Iglesias, C. miran-
dae Vovides, Prez-Farrera & Iglesias and C. norstogii.
Study Area
The study was done in the La Sepultura Biosphere Reserve on the western side
of the Sierra Madre de Chiapas in Mexico. This physiographical region runs
parallel to the Pacific coastal plain, from the extreme south of the Isthmus of
Tehuantepec, across Chiapas and reaches as far as Guatemala. It ranges in alti-
tude from 1000 m in the north to 4000 m (Mount Tacan) on the Guatemalan
border with Chiapas (Mllerried, 1957). The predominant vegetational types are
deciduous tropical forest, oak forest, conifer forest, cloud forest and evergreen
tropical rainforest (Rzedowski, 1978). According to De La Rosa et al. (1989) the
major part of the area is formed by Palaeozoic igneous rocks (granites and grani-
diorites) and Precambrian rocks (pink granitic gneiss and granidiorite gneiss).
Specific metamorphic substrate is present only on slopes in some areas
(Hernndez-Yaez, 1995).
Most populations within the Ceratozamia norstogii complex are found in oak
and cloud forests or intermediate zones of these two ecosystems. The cycads
grow in poor clay soils on steep slopes within an altitude range of 8001200 m.
The majority of populations are subjected to periodic fires at least once a year.
All species within this complex are characterized by erect male and female cones,
and plain or channelled leaflets.
ch10.qxd 28/11/03 3:49 pm Page 129
Ceratozamia norstogii Complex: Morphometric Analysis 129
Three natural populations were analysed (Fig. 10.1) using 30 individuals each of
Ceratozamia norstogii, C. alvarezii and C. mirandae (Figs 10.210.4) and recording data
from 14 vegetative and six reproductive morphological variables (Table 10.1).
Measurements were done using a flexible tape measure (3 m) and a precision
digital vernier (0.01 mm resolution). Data were captured and electronically
stored. Analysis of variance (ANOVA) tests were done using JMP version 3.2 sta-
tistical software (SAS Institute, Cary, North Carolina, USA) and discriminant
analyses (McCune and Mefford, 1997) with Statgraphics version 2.0 software
(Manugistics, Rockville, Maryland, USA).
Results
Univariate analysis
Tables 10.2 and 10.3 summarize the ANOVA results. Species characters do not
overlap, with the exception of intervein distance, leaf length, microsporophyll
Fig. 10.1. Location in Mexico of populations of the Ceratozamia norstogii

complex sampled in this study.
ch10.qxd 28/11/03 3:49 pm Page 130
Fig. 10.2. Species in the Ceratozamia norstogii complex: C. alvarezii

Prez-Farrera, Vovides & Iglesias.
Fig. 10.3. Species in the Ceratozamia norstogii complex: C. norstogii D.W.

Stevenson.
ch10.qxd 28/11/03 3:49 pm Page 131
Fig. 10.4. Species in the Ceratozamia norstogii complex: C. mirandae Vovides,

Prez-Farrera & Iglesias.
Table 10.1. List of morphometric variables used in the analysis of populations in

the Ceratozamia norstogii complex.
No. Abbreviation Character and measurement units (parentheses)
1 LARGTR Trunk length (cm)

2 PERMTR Trunk perimeter (cm)
3 NHOJA Number of leaves per trunk
4 LARHOJ Leaf length (cm)
5 ANCHHOJ Leaf width (cm)
6 LARESPPEC Petiole prickle length (mm)
7 LARGOPEC Petiole length (cm)
8 PERIMPEC Petiole perimeter (cm)
9 LARGORAQ Rachis length (cm)
10 NFOLIOL Number of leaflets per leaf
11 LARGOFOL Leaflet length (cm)
12 ANCHOFOL Leaflet width (cm)
13 NVENAS Number of veins per leaflet
14 DISTVEN Intervein distance (mm)
15 LARGOMICRO Microsporophyll length (mm)
16 ANCHOMICRO Microsporophyll width (mm)
17 ANCHOMEGAS Megasporophyll distal face width (mm)
18 LARGOMEGAS Megasporophyll distal face length (mm)
19 DIAMSEMI Seed diameter (mm)
20 LARGSEMI Seed length (mm)
ch10.qxd 28/11/03 3:49 pm Page 132
Table 10.2. Mean values and standard deviations (SD) for the ratios used in the
analysis of populations in the Ceratozamia norstogii complex (C. alvarezii Prez-
Farrera, Vovides & Iglesias, C. mirandae Vovides, Prez-Farrera & Iglesias and
C. norstogii D.W. Stevenson).
Species/character C. alvarezii SD C. mirandae SD C. norstogii SD
LARGTR 20.55 7.67 58.23 18.5 32.07 21.76

PERMTR 42.68 6.64 71.30 7.28 58.31 6.79
NHOJA 8.91 3.99 13.33 4.74 4.83 0.97
LARHOJ 79.64 15.5 151.87 17.9 98.60 17.35
ANCHHOJ 45.36 6.61 69.15 8.69 94.24 135.3
LARESPPEC 0.34 0.09 0.46 0.07 0.40 0.51
LARGOPEC 20.59 8.52 39.92 9.97 18.18 6.93
PERIMPEC 2.45 0.57 3.42 0.63 2.61 0.56
LARGORAQ 47.00 10.2 104.55 61.7 62.11 9.90
NFOLIOL 45.00 9.97 65.67 9.05 51.17 9.32
LARGOFOL 22.98 3.28 36.07 4.80 35.55 7.65
ANCHOFOL 0.65 0.11 0.98 0.12 0.41 0.07
NVENAS 7.55 0.93 8.67 1.56 5.31 0.71
DISTVEN 0.08 0.03 0.10 0.02 0.09 0.01
LARGOMICRO 14.90 0.83 14.55 0.99 14.32 0.99
ANCHOMICRO 6.99 0.76 7.17 0.70 7.74 0.80
ANCHOMEGAS 35.54 5.32 52.41 3.82 46.82 1.64
LARGOMEGAS 15.54 2.26 12.14 1.97 9.88 1.61
DIAMSEMI 16.0 1.34 16.58 0.97 17.55 1.37
LARGSEMI 22.36 2.73 27.45 1.12 27.13 1.30
length, petiole length and petiole prickle length. The rest of the variables
analysed were all highly significant (P < 0.0001).
Discriminant analyses
Figure 10.5 shows the scatter diagram of data derived from discriminant func-
tion analysis. The three species separate ordinately in bidimensional space and
do not present any overlapping between groups. Of the 20 variables included in
the standardized discrete canonical function, the six variables with the highest
values in factor 1 were trunk perimeter, petiole length, leaflet width, vein number,
microsporophyll length and megasporophyll width. In factor 2, the highest vari-
ables were trunk perimeter, leaf length, petiole length, leaflet length and width.
The first canonic variable showed that 70% of the variation is largely due to veg-
etative morphology. The positive correlations (Table 10.4) of all the variables
show differences between species. The Wilks lambda test was highly significant (P
< 0.0001) for the two factors (Table 10.5) thus showing that all the species were
classified correctly.
ch10.qxd 28/11/03 3:49 pm Page 133
Table 10.3. Summary of analysis of variance of the 20 morphometric characters

used in the analysis of populations in the Ceratozamia norstogii complex (R2 =
correlation coefficient, F = F value, P = probability).
Character R2 F P
LARGTR 0.47 40.04 < 0.0001

PERMTR 0.71 108.18 < 0.0001
NHOJA 0.52 43.41 < 0.0001
LARHOJ 0.76 138.79 < 0.0001
ANCHHOJ 0.06 2.81 0.06
LARESPPEC 0.02 1.02 0.34
LARGOPEC 0.06 3.02 0.05
PERIMPEC 0.37 25.63 < 0.0001
LARGORAQ 0.30 19.04 < 0.0001
NFOLIOL 0.45 36.61 < 0.0001
LARGOFOL 0.49 42.92 < 0.0001
ANCHOFOL 0.87 241.20 < 0.0001
NVENAS 0.56 55.90 < 0.0001
DISTVEN 0.01 0.77 0.46
LARGOMICRO 0.61 1.36 0.25
ANCHOMICRO 0.09 4.83 0.01
ANCHOMEGAS 0.76 145.52 < 0.0001
LARGOMEGAS 0.62 71.80 < 0.0001
DIAMSEMI 0.25 15.27 < 0.0001
LARGSEMI 0.61 68.8 < 0.0001
C. alvarezii C. norstogii
C. mirandae + centroids
5
3
Discriminant Function 2
7
6 3 0 3 6 9
Discriminant Function 1
Fig. 10.5. Scatter plot of scores derived from the functions produced by stepwise
discriminant analysis of 20 morphometric ratios from populations in the
Ceratozamia norstogii complex.
ch10.qxd 28/11/03 3:49 pm Page 134
Table 10.4. Standardized discriminate function values for each of two factors
used in the analysis of populations in the Ceratozamia norstogii complex.
Character Factor 1 Factor 2
LARGTR 0.1027 0.1943

PERMTR 0.3713 0.3334
NHOJA 0.3498 0.1579
LARHOJ 0.0975 0.3914
ANCHHOJ 0.0106 0.0358
LARESPPEC 0.0079 0.0098
LARGOPEC 0.3594 0.3624
PERIMPEC 0.1413 0.1011
LARGORAQ 0.0492 0.0142
NFOLIOL 0.0191 0.0452
LARGOFOL 0.0914 0.5259
ANCHOFOL 0.9860 0.3467
NVENAS 0.3850 0.1930
DISTVEN 0.0180 0.0047
LARGOMICRO 0.2383 0.2232
ANCHOMICRO 0.0865 0.0007
ANCHOMEGAS 0.5815 0.2458
LARGOMEGAS 0.2986 0.2986
DIAMSEMI 0.1918 0.0766
LARGSEMI 0.1804 0.2074
EIGENVALUE 15.6008 6.4400
CANONICAL CORRELATION 0.9694 0.9304
AMONG-GROUP VARIANCE 70.78% 29.22%
Table 10.5. Summary of the discriminant analysis results from the analysis of
populations in the Ceratozamia norstogii complex (2 = chi-squared factor, DF =
degrees of freedom, P = probability).
Functions derived Wilks lambda 2 DF P
Factor 1 0.00809 276.94 40 0.0000

Factor 2 0.1344 115.39 19 0.0000
Discussion and Conclusions
Amongst the three species in the complex, Ceratozamia alvarezii has the longest and
narrowest microsporophylls, and the shortest and narrowest leaves. By contrast,
C. mirandae has the longest trunks with the highest circumference and leaves
having the longest petiole with the largest circumference. The discrete analysis
ch10.qxd 28/11/03 3:49 pm Page 135
clearly showed the morphological differences between the species in the C.

norstogii complex. However, some results from the univariate analysis showed
some overlapping between species. The analysis provides evidence that delimits
the three species analysed. In the present study, the measurements of leaflets,
leaves and sporophylls were particularly well represented and it is not surprising
that the differences between these characters emerge as important discriminators
between the species.
A long interpopulational distance could probably be indicative of a long iso-
lation period if a constant rate of evolution and allopatric speciation is assumed
(Grant, 1985). This hypothesis is based on Wrights (1943) theory of isolation by
distance. According to this hypothesis, populations within a taxon with a con-
tinuous distribution pattern will show genetic differentiation proportional to the
distance between populations. It is interesting to observe that, particularly in this
complex, all the species share the same habitat (oak forest) and that distances
between species and populations are relatively close, but morphologically the
plants are distinct. Isolation barriers other than distance have had a part to play
here, namely high mountain ranges and deep canyons where vegetation types
change drastically over short distances. However, preliminary DNA sequencing
data show a homogeneity of characters in the nuclear ribosomal ITS and chloro-
plast trnL-F non-coding region (Gonzlez and Vovides, 2002), which presumes
geologically recent isolation events, perhaps the Cenozoic climatic changes prior
to and during the Pleistocene that gave rise to the floristic refuges in southern
Mexico (Toledo, 1982; Burnham and Graham, 1999).
Taxonomic Key
The following vegetative key may be used to separate the species in the
Ceratozamia norstogii complex:
1. Spirally twisted rachis ........ C. norstogii

1. Straight rachis
2. Leaflets channelled or partly channelled, 812 mm wide ... C. mirandae
2. Leaflets not channelled, 49 mm wide .... C. alvarezii
Acknowledgements
The authors thank Montgomery Botanical Center (Miami, Florida, USA) for a
student grant Taxonomic revision of the genus Ceratozamia Brongn. (Zamiaceae)
in the neotropics awarded to the first author to complete PhD studies, and
CONACYT (Grant Number 29379N) for the project Systematics, distribution
and conservation of Mexican cycads. The principal author thanks Rigoberto
Hernndez Jonapa, Jesus De La Cruz Rodriguez, Emerit Lpez Melendez, Nayeli
Martnez Melendez and Ruben Martnez Camilo for their help in the field work.
ch10.qxd 28/11/03 3:49 pm Page 136
References
Burnham, R.J. and Graham, A. (1999) The history of neotropical vegetation: new devel-
opments and status. Annals of the Missouri Botanical Garden 86, 546589.
De La Rosa, J.L.A., Eboli, M. and Dvila, S.M. (1989) Geologa del estado de Chiapas.
Subdireccin Construccin, Unidad de Estudios de Ingeniera Civil, Subjefatura de
Estudios Geolgicos, Departamento de Ecologa, CFE, Mexico, 192 pp.
Gonzlez, D. and Vovides, A.P. (2002) Low intralineage divergence in the genus
Ceratozamia Brongn. (Zamiaceae) detected with nuclear ribosomal DNA ITS and
chloroplast DNA trnL-F non-coding region. Systematic Botany 27, 654661.
Grant, V. (1985) The Evolution Process. A Critical Review of Evolutionary Theory. Columbia
University Press, New York, 587 pp.
Hernndez-Yaez, A. (1995) Propuesta para establecer el rea natural protegida (Reserva
de la Biosfera) La Sepultura, en la porcin oeste de la Sierra Madre de Chiapas,
Mxico. MSc thesis, Facultad de Biologa, Universidad Veracruzana, Xalapa,
Veracruz, Mexico.
McCune, B. and Mefford, J. (1997) Multivariate Analysis of Ecology Data.Version 3.17. MJM
Software, Gleneden Beach, Oregon, USA, 126 pp.
Moretti, A., Sabato, S. and Vzquez Torres, M. (1980) The distribution of Ceratozamia
Brongn. (Zamiaceae). Delpinoa 21, 1321.
Mllerried, F.K.G. (1957) Geologa de Chiapas. Gobierno Constitucional del Estado de
Chiapas, Tuxtla Gutirrez, Chiapas, Mexico, 180 pp.
(Zamiaceae, Cycadales) from Chiapas, Mexico. Novon 9, 410413.
Prez-Farrera, M.A., Vovides, A.P. and Iglesias, C.G. (2001) The cycad Ceratozamia norstogii
Stevenson (Zamiaceae) from southern Mexico: new information on distribution,
habitat and vegetative morphology. Botanical Journal of the Linnean Society 137, 7176.
Rzedowski, J. (1978) La Vegetacin de Mxico. Edit. LIMUSA, D.F., Mexico, 432 pp.
Stevenson, D.W. (1982) A new species of Ceratozamia (Zamiaceae) from Chiapas, Mexico.
Brittonia 34, 181184. [Ceratozamia norstogii.]
Stevenson, D.W. and Sabato, S. (1986) Typification of names in Ceratozamia Brongn., Dioon
Lindl., and Microcycas A.D.C. (Zamiaceae). Taxon 35, 578584.
phyllidia.]
Toledo, V.M. (1982) Pleistocene changes of vegetation in tropical Mexico. In: Prance, G.T.
(ed.) Biological Diversification in the Tropics. Proceedings of the Fifth International Symposium of
the Association for Tropical Biology. Columbia University Press, New York, pp. 93111.
Vovides, A.P., Rees, J.D. and Vzquez Torres, M. (1983) Zamiaceae. In: Sosa, V. and
Gomez-Pompa, A. (eds) Flora de Veracruz. INIREB, Xalapa, Veracruz, Mexico,
Fascculo 26.
Vovides, A.P., Prez-Farrera, M.A. and Iglesias, C.G. (2001) Another new species of
Ceratozamia (Zamiaceae) from Chiapas, Mexico. Botanical Journal of the Linnean Society
137, 8185. [Ceratozamia mirandae.]
Wright, S. (1943) Isolation by distance. Genetics 28, 114138.
ch11.qxd 28/11/03 3:49 pm Page 137
Hypotheses on the 11
Relationship between
Biogeography and Speciation
in Dioon (Zamiaceae)
Timothy J. Gregory1 and Jeffrey Chemnick2
1Montgomery Botanical Center, Miami, Florida, USA; 2Ganna

Walska Lotusland, Santa Barbara, California, USA
Abstract
Hypotheses are presented on the relationships between palaeoclimatic conditions during

the late Pleistocene and Holocene, and the current biogeography of cycads in the genus
Dioon. A proposal is made that Dioon populations have migrated up and down in elevation,
and north and south, with the warming and cooling climatic cycles corresponding to
glaciations. This suggests that since the end of the Wisconsin glaciation Dioon populations
have moved upward in elevation in southern Mexico with shifts in their preferred transi-
tion zone habitat. This migration has sometimes resulted in disjunction and rapid specia-
tion within river canyon systems. Mechanisms for these activities are discussed. The
chapter culminates with the proposal that the extant Dioon species are the result of rapid
evolution in a dynamic and viable group of plants instead of the classical concept that they
are slowly evolving relicts.
Introduction
The pioneering collaborative work between staff of the Orto Botanico in Naples
and the Instituto de Ecologa in Veracruz greatly expanded the understanding of
the cycad genus Dioon Lindley. Six new species were described (Dioon califanoi De
Luca & Sabato, D. caputoi De Luca, Sabato & Vzquez Torres, D. rzedowskii De
Luca, A. Moretti, Sabato & Vzquez Torres, D. holmgrenii De Luca, Sabato &
Vzquez Torres, D. merolae De Luca, Sabato & Vzquez Torres and D. tomasellii
De Luca, Sabato & Vzquez Torres) and two of the original three species (D. edule
Lindley and D. purpusii Rose) were redefined in a modern context (De Luca et al.,
ch11.qxd 28/11/03 3:49 pm Page 138
138 T.J. Gregory and J. Chemnick
1978, 1980a,b, 1981a,b, 1982, 1984; De Luca and Sabato, 1979). The collabo-
ration culminated in a summary of the characteristics of the genus and a well-
founded hypothesis on evolutionary trends in Dioon (Sabato and De Luca, 1985)
that has served as the foundation for our recent work. Since 1995 we have been
engaged in a study of the cycad populations in southern Mexico. Our objectives
have been to document populations of the known species and find new popula-
tions; to study their biology, phenology and morphology; and to attempt to
understand the biogeographic distribution and history of speciation of each pop-
ulation. The results of our work to date are: (i) the descriptions of three new
species (Chemnick and Gregory, 1995; Chemnick et al., 1997b; Gregory et al.,
2003); (ii) a proposal to raise a previously described subspecies of D. tomasellii, to
specific rank; and (iii) a dramatic range extension for D. merolae (Chemnick et al.,
1997a). We have also acquired considerable information on the biogeography of
each of the Dioon species and their populations. These populations are confined
to very restricted habitats, generally in the transition zones between major forest
types. Each species is composed of two or more well-defined populations that
share specific morphological characteristics. Several key characteristics of Dioon,
such as distribution within steep rocky habitats, separate sexes, large brightly
coloured seeds, and beetle pollination, raise interesting questions about how these
populations are established and distributed. Our attempts to answer these ques-
tions led us to the literature concerning the movement of plant species during
periods of climatic change, especially those changes since the last glaciation. The
climate changes associated with the retreat of the continental glaciers have
caused major movements of individual plant species to new habitats (Sauer,
1988). The purpose of this chapter is to summarize our field observations, cor-
relate them with what is known about the climatic changes that have occurred in
the last 35,000 years and propose a hypothesis for the mechanism by which spe-
ciation occurs within the genus Dioon.
Field Observations
Our observations in the field have confirmed and expanded those of Sabato and
De Luca (1985). Dioon populations are geographically discrete and frequently
confined to steep rocky slopes or cliffs in the transition zone between major forest
types. Dioon spinulosum Dyer occurs on kaarst hummocks in lowland tropical forest
at 70120 m. Dioon rzedowskii is found in the transition zone between lowland
tropical rainforest and premontane rainforest at 100500 m. Dioon holmgrenii and
the populations of an undetermined taxon that occur between D. holmgrenii and
D. tomasellii on the Pacific coast are found at higher elevations, 9001100 m, in
the tropical dry forest, below the oak/pine forest. The remaining Dioon species
are found primarily in the transition zone between thorn forest and oak/pine
forest at 14002000 m. The restriction to rocky transition zones suggests that
Dioon species are adapted to exist in habitats where other species are less able to
compete. These elevation/habitat associations are sufficiently consistent that we
ch11.qxd 28/11/03 3:49 pm Page 139
Dioon: Biogeography and Speciation Relationship 139
have used them to predict the existence of populations from topographic maps
and later confirm their existence in the field. We have also observed that only a
small proportion of the potentially available habitat is utilized. The patchy dis-
tributions may be related to biological and ecological constraints on the estab-
lishment of new Dioon populations, as discussed below. Most species comprise two
to eight homomorphic populations and the individual species or groups of
species tend to be restricted to particular river drainages. Those species with
numerous populations (e.g. D. edule and D. tomasellii) are currently treated as com-
plexes of populations with inadequately analysed variation. Relatively subtle leaf
morphology is the defining characteristic of Dioon species. Unfortunately, the
intraspecific variation in leaf and seedling morphology and reproductive charac-
ters of Dioon are not adequately treated in most species descriptions. Within each
population approximately 10% of the adult plants exhibit a leaf morphology at
variance with the description. Such variation is even observed between flushes of
leaves on the same plant and these atypical individuals or cohorts of leaves are
often more characteristic of other species, especially those most closely distrib-
uted geographically. The suite of characters exhibited by the vast majority of
plants is, however, consistent within the population and the species, although D.
holmgrenii is a notable exception. The variation in leaf morphology present within
populations of D. holmgrenii is extreme and leaves that could be mistaken for those
of D. merolae and D. tomasellii occur on approximately 40% of the plants. The
subtle, yet discrete, variation defining each Dioon species, the restricted habitat
occupied and the association of species with particular river valleys and canyons
have led us to consider possible mechanisms of speciation within the genus.
Dioon species possess large seeds that make their long-range dispersal prob-
lematic. It is difficult to hypothesize potential medium- or long-range dispersal
agents among the extant animals occurring in Mexico. In Dioon populations we
frequently observe that the seed sarcotesta are eaten by rodents, as evidenced by
the diagnostic teeth marks left in the dried sarcotesta and on the surface of the
sclerotesta. Occasionally the megagametophyte is eaten but the sarcotesta
appears to be the primary food element. Sometimes caches of cleaned seeds are
found near or at the base of the mother plant and individual seeds are found up
to 20 m away. However, we also frequently observe that the majority of seeds
from a cone are cleaned by rodents and yet remain in the crown of the parent
plant. The relationship between dispersal and recruitment into the population is
not known for Dioon but we observe that the vast majority of seed that germinates
near the parent plant gives seedlings that die within the first year or two. The
attrition of germinated seed and young seedlings is apparently due to desiccation.
We have observed that many of the seedlings that survive the first 2 years after
germination are those buried deeply in the ground. This suggests that successful
recruitment could be dependent on the rodents deep internment caching of seed
that then germinates before consumption. The question remains as to whether
Dioon seed is ever dispersed over distances that would allow establishment of new
populations. Dioon sarcotesta are coloured yellow to orange which suggests that
they might be attractive to a diurnal animal that can recognize colour. In many
ch11.qxd 28/11/03 3:49 pm Page 140
species the megastrobilus dehisces when the cone is held above the crown of
leaves. In these cases the female cone dehisces with the yellow sarcotesta exposed
skywards and visible for hundreds of metres. This raises the possibility that large
diurnal birds might be attracted and thus serve as longer-range dispersal agents,
though no such association has been reported for Dioon to date. Currently the
most likely local extant Mexican birds potentially capable of carrying such large
seeds to new suitable habitats are the common raven (Corvus corvax) and the mili-
tary macaw (Ara militaris). We have observed both species of bird perched in Dioon
plants in Oaxaca but have not observed any direct interaction with cones or seed.
Two possible mechanisms of population establishment present themselves
from our observations. The first is the possibility that large diurnal birds, or other
dispersal agents, transport seeds to new cliff habitats, followed by a colonization
by the pollinator, thus to establish new viable populations. The need for con-
comitant dispersal of individual plants of both sexes and the pollinating insect,
and the paucity of different populations within each species relative to the vast
amount of suitable cliff habitat, make this mechanism unlikely, but not impossi-
ble. Furthermore, such a colonization mechanism would generate a very signifi-
cant genetic bottleneck, because the founding genes would come from as few as
two individuals. The genetic repertoire of the new population would be limited
to that of the two founding parents. If the founding plants possessed characteris-
tics that are outside the norm in the parental population then the new population
would be correspondingly different. The possibility of genetic exchange back
with the parent population is extremely limited by the same impediments to the
founding of a new population. The genetic bottleneck could lead to rapid speci-
ation by the founder effect analogous to speciation that sometimes occurs on
islands. The second potential mechanism for generating new Dioon populations is
through slow local dispersal by rodents resulting in the eventual schism of an
existing population over time into new sister populations. This mechanism avoids
the problem of concomitant colonization by plants of both sexes and the polli-
nator but requires a changing environment to create new contiguous habitats into
which the population can colonize by fission. Such a mechanism also lacks the
genetic bottleneck that could drive rapid speciation. We hypothesize that both
processes are important. The fission process could be an important mechanism
for generating new populations within a species and the colonization mechanism
could be important for generating new species through a founder effect.
Very little is known about the pollination mechanism for Dioon species.
Insects may be obligate pollinators, as demonstrated in other members of the
Zamiaceae. We have collected weevils (Curculionidae), having an orange prono-
tum and black elytra, from fresh and old male cones of several Dioon species
throughout Oaxaca. Typically, these weevils are found covered with pollen and
we have recovered the remains of similar weevils from fertilized female cones.
Larvae have been recovered from dried male cones that are up to 1 year old but
not from older cones. These observations are consistent with the proposed life
history of weevils as a cycad pollinator (Norstog et al., 1986). Superficially, there
appears to be a single species that pollinates D. califanoi, D. purpusii, D. merolae and
ch11.qxd 28/11/03 3:49 pm Page 141
D. argenteum T.J. Gregory, Chemnick, Salas-Morales & Vovides, but the taxonom-
ic status of the beetles is currently under investigation. Since it appears that there
is no obvious gene flow between Dioon species, and possibly not even between
populations of the same species, it is reasonable to conclude that the pollinators
are unable to range beyond a particular population. Some fortuitous information
arose from an inadvertent experiment carried out by the people in a village near
the D. argenteum habitat. Several decades ago three male and one female D. argen-
teum plants were transplanted to the churchyard in the village. The wild popula-
tion from which the plants originated is less than 5 km from the village. On two
occasions we observed that the old male cones showed no signs of insect occu-
pation (e.g. exit holes or excavation of the microsporophylls) and there were no
signs of fertile seed around the female plant. The villagers also confirmed that
they had never observed viable seed or seedlings produced by the female plant.
The anecdotal evidence suggests that the pollinator was not able to colonize this
artificial pioneer population over at least two decades and that the range of the
pollinator is quite restricted. This confirms that any new Dioon population must
have a concomitant colonization by the pollinator to be viable and that the range
of the pollinator is so restricted that this colonization is a rare event.
We have directly observed more than 35 Dioon populations in southern
Mexico. Several of those populations that have not been directly impacted by
human activity are relatively large, consist of hundreds to thousands of individ-
uals in all life stages, and exhibit excellent recruitment. However, many of the
populations are small (20100 individuals). Some of these small populations are
apparently viable with some recruitment, but others appear to be senescent. We
define senescence as a population containing mostly older mature adults that,
despite regular coning and seed production, exhibits very little or no recruitment
of seedlings. Such senescent populations are usually located on cliff faces or talus
slopes in the tropical deciduous forest at sites where there is no oak/pine forest
above them. These plants appear stressed by drought and generally inhospitable
conditions. We propose that these Dioon populations are dying out naturally
because the habitat at their current location has changed and moved to higher
elevations with the drying and warming of the climate during the Holocene. No
higher topography exists adjacent to them to permit their upward migration with
their preferred transition zone habitat. An alternative hypothesis is that these
populations are new colonies in various phases of relative success or failure.
However, we have yet to discover similar populations at higher elevations in the
oak/pine forest, above the preferred Dioon habitat. Rather, the most robust pop-
ulations occur at sites where there is cliff habitat and oak/pine forest above. Thus
we believe that we are observing Dioon populations in various phases of maturity
to senescence possibly caused by movement of the preferred habitat to higher
elevations over time. Alternatively, the extant populations of Dioon in Mexico
might merely be scattered remnants that have persisted for millennia at the
same sites through several glacial climate cycles, somehow withstanding
generally inhospitable conditions under which they do not reproduce even in
horticulture.
ch11.qxd 28/11/03 3:49 pm Page 142
Palaeoclimatic Implications
Fossils attributable to Dioon have been found in Alaska and California (Wolfe,
1972; Axelrod, 1975). These fossils are of Eocene and Miocene age, respectively,
and suggest that the ancestors of the current Dioon species were present in
Mexico at least 1015 million years ago. The fossils demonstrate that the genus
was able to expand into northern North America during a very warm period.
The extant Dioon species are confined to Mexico and Honduras and have been
assigned to three subgeneric groupings (Sabato and De Luca, 1985).
Reproductive characters are important for distinguishing between these groups,
but within each group the species are differentiated primarily by leaf morpholo-
gy. The species within the groupings appear to be closely related and to have only
recently evolved into new taxa. We regard a Dioon species as a group of popula-
tions with consistent morphological characters sufficient to distinguish it as an
entity from its closest congeners. Implicit in this species concept is the assumption
that the closest congeners of a particular species are also some of its closest geo-
graphical neighbours. In the field we find that groups of related species are often
confined to river valleys or drainage systems. The mountains or deserts sur-
rounding the river systems lack suitable Dioon habitats and provide significant
barriers to the migration of Dioon populations. Their patchy distribution suggests
that populations of Dioon are expanding their ranges upward and becoming
increasingly isolated in the river valleys that they occupy today. The current dis-
tribution of the various taxa conforms to this model and thus we are seeking pos-
sible mechanisms that could drive this migration and speciation.
Sauer (1988) summarized data on plant migrations in the Holocene that may
be useful in explaining the current distributions of Dioon species. Palynological
data, as well as data from wood rat (Neotoma) middens from East Africa, Japan,
Colombia, northern Europe and south-eastern and south-western USA, indicate
that, following the collapse of the Wisconsin ice sheet 12,000 years ago, there
were dramatic movements of entire forests northward and to higher elevations.
The data indicate that individual tree species moved independently of each other
and not as a forest biome per se. It also appears from the pollen data that some
species advanced their ranges at mean rates of 250 m/year. Some species in the
northern hemisphere exhibited a general northward migration and colonization
of newly available habitat while species in tropical latitudes (e.g. the Colombian
Andes and the high mountains of East Africa) clearly moved upward in elevation
by 12001500 m. Neither palynological nor Neotoma midden data from Mexico
are currently available but McDonald (1993) has assembled biogeographical data
on alpine plants and Toledo (1982) has assembled a variety of biogeographical
and oceanographical data that support a similar post-glacial plant migration.
The data from alpine plants in central Mexico indicate an upward movement of
the treeline in excess of 1000 m during the past 10,000 years. During the peak
glaciation, approximately 40,000 years ago, the treeline in central Mexico
descended to 2500 m, in contrast to the current elevation of 3500 m. This
created a sub-alpine corridor extending from the southern Rocky Mountains to
ch11.qxd 28/11/03 3:49 pm Page 143
the transverse Neovolcanic belt in central Mexico. This cycle appears to repeat
with every glaciation cycle. Toledo (1982) has been cited by several authors for
his delineation of Pleistocene refugia in southern Mexico, though this aspect is
only a minor component of his work. The majority of his data are from palaeo-
climatic analysis of sediment cores from the Gulf of Mexico and geomorpho-
logical studies of mountain glaciers in central Mexico. From these data, the
pattern of warm/wet and cool/dry climate cycles over the past 40,000 years has
been discerned at the latitude of the central Caribbean. This pattern shows, not
unexpectedly, that the mean temperature at sea level during the last ice age was
67C lower than it is now and has steadily increased to the present. Toledo
(1982) then combined the temperature data with biogeographical data from the
lowland forests on the gulf coast of Mexico to recreate the history of the gulf
coastal vegetation during and since the glacial period. He suggested that there
was no rainforest on the gulf plain of Veracruz during the glaciation. Instead
there was oak/pine forest at sea level. The current rainforest elements have been
migrating northward from refugia in Chiapas and Guatamala for 12,000 years.
This provides an explanation for the relatively low species diversity and the pres-
ence of oak and pine forest remnants in the gulf coastal rainforest. During the
same time, the alpine, cloud, oak/pine and tropical deciduous forests may have
migrated to higher elevations at the same latitude and to higher latitudes in the
north. As these forest biomes moved northward and upward, the transition zone
habitats of the Dioon taxa moved in concert (i.e. to 1500 m at present in southern
Mexico). As the habitats moved, the attendant populations presumably migrated
with them up the river valleys, at an average rate of approximately 12 cm in ele-
vation per year.
Previous authors (Sabato and De Luca, 1985; Moretti et al., 1993) have
attributed the current biogeography of Dioon to events that occurred throughout
the Cenozoic. We agree that the history of the genus has certainly been shaped
by the events of the last 60 million years but also believe that the available data
suggest that the events of the most recent few hundred thousand years have much
more relevance to the current distribution of Dioon in Mexico. Since the early
Pleistocene there have been as many as 20 ice age cycles (Gould, 2001). Each of
these cycles has caused the same dramatic climatic changes in North America
and Mexico that Toledo and Sauer have described. As we suggest above, these
climatic changes have probably had profound effects on the habitats of Dioon
populations over thousands of years, not millions. The available molecular-bio-
logical (Moretti et al., 1993) and biochemical data (Siniscalco Gigliano, 1990)
imply that many of the current Dioon species are very closely related. In the
restriction fragment length polymorphism analysis of Moretti et al. (1993), the
species of their D. edule clade are so closely related that many of the sub-clades
are not statistically valid. Similarly, analyses of several biochemical characters
failed to show any distinctions between Dioon species. The close genetic and bio-
chemical relationship within the Dioon species could be interpreted either as an
unusually slow rate of evolution in a relict group over a long period of time, or
as rapid evolution and speciation in a short time frame. Moretti et al. (1993) and
ch11.qxd 28/11/03 3:49 pm Page 144
Sabato and De Luca (1985) considered both options but seemed to prefer the
former. We favour the latter as we seem to be observing the relatively rapid evo-
lution of populations into species over a period of thousands of years, not hun-
dreds of thousands or millions. We also seem to be observing a dynamic response
of a viable and successful group of plants to correspondingly rapid climatic and
environmental changes.
Examples of Interpretations of Dioon Biogeography Using

the Dynamic Habitat Hypothesis
The positions and status of several Dioon populations today may be explained by
a dynamic habitat hypothesis. Three closely related species, D. califanoi, D. pur-
pusii and D. argenteum, are found in nearly identical habitat but at different locali-
ties in the Rio Grande/Cuicatlan/Rio Salado canyon system in the upper
reaches of the Rio Santo Domingo drainage of Oaxaca. These taxa support our
species concept as each comprises two to five independent homomorphic popu-
lations but possesses a unique and diagnostic morphology. Some populations
within a species can be differentiated from each other by very subtle characters
such as size of leaves, carriage of leaves in the crown, number of prickles on the
leaflets, or amount of wax on the leaflets. All populations are found on steep
rocky slopes at altitudes of 15002000 m. Some populations are near the tops of
ridges, some are midway down canyon walls, and still others are found along
water courses. The common factor among them is their occurrence in the tran-
sition zone between tropical deciduous forest and oak/pine forest. The palaeo-
climatic data from Toledo (1982) suggest that the floor of the canyon, at
approximately 1100 m, was occupied by high-altitude pine or cloud forest
10,00012,000 years ago. If this were the case, then these three distinct species,
if they existed at that time, would likely have been forced to retreat and converge
their ranges either in the lower part of the Rio Santo Domingo canyon or out on
the coastal plane in a refugium. It seems unlikely that they could have maintained
their genetic uniqueness under such conditions. Rather it seems more likely that
these taxa have evolved recently from a common ancestor that colonized the
canyon system in the Holocene following warming after the collapse of the
Wisconsin ice sheet. We propose that the individual populations within each of
these species have evolved by vicariance. We also propose that the species them-
selves could have evolved more rapidly from each other by the type of moderate-
range dispersal discussed above in which a founder effect could have facilitated
the establishment of a new cohort with slightly different morphology (i.e. more
or less keeled leaves). If the putative common ancestor found refuge in the lower
Rio Santo Domingo canyon during the ice age then the most distant populations
of D. purpusii, D. califanoi and D. argenteum have moved into the smaller river valleys
at an average rate of approximately 9 m per year over the past 12,000 years. This
rate is feasible by either rodent or bird dispersal.
Dioon caputoi illustrates a dilemma that can potentially be resolved by the
ch11.qxd 28/11/03 3:49 pm Page 145
dynamic habitat hypothesis. It is currently found only in remnant oak forest on

the peaks of mountains on the south-eastern side of the Tehuacan valley. Its pop-
ulations are located on mountain islands surrounded by a sea of high desert
and in a very unusual geographical location relative to other Dioon species.
Archaeological evidence suggests that there has been a dramatic drying of the
Tehuacan valley over at least the last 7000 years (MacNeish, 1961) which could
have pushed D. caputoi into its current geographical position. Dioon caputoi does not
appear to be closely related to its nearest neighbours, D. califanoi, D. purpusii and
D. argenteum, by either morphological or molecular characters (Stevenson, 1992;
Moretti et al., 1993). Instead it appears to be most closely related to an unde-
scribed taxon found to the south-east in the Mixteca region of Oaxaca. This
cycad, in turn, appears to be most closely related to D. holmgrenii and D. tomasellii,
both of which are found on the Pacific coast. The Mixteca area is part of the
upper drainage of the Rio Balsas that drains to the Pacific Ocean. Thus D. caputoi
populations may have shared a common ancestral lineage with the species on the
Pacific coast. Quite possibly, that common ancestor migrated with climatic
changes up the Rio Balsas valley to the low mountains on the south-eastern wall
of the Tehuacan valley. Subsequently, the Tehuacan population may have
become isolated on mountain islands of habitat in the high desert of south-
eastern Puebla following the end of the last ice age.
The locations and characteristics of several populations of the Dioon edule
complex present biogeographical dilemmas that can potentially be resolved by
the above hypothesis. Most of the disjunct forms of D. edule are found in similar
habitats within drainages in central Veracruz, San Luis Potosi, Quertaro and
Hidalgo, such as the Rio Pescado and Rio Monctezuma valleys. One unusual dis-
junct population occurs near sea level on the top of a high sand dune on the
central coast of Veracruz. This population is restricted to a remnant patch of
tropical deciduous forest but consists of hundreds of individuals within its small
range. Perhaps this population is a remnant of the original populations that were
at sea level during the ice age. A population with similar morphology occurs at
1100 m on the ancient volcanic mountain directly above the dune population on
the coast. Perhaps this illustrates an upward migration by one population in
response to climatic warming while the dune population remained behind. This
would occur as the range of suitable habitat expanded with the warming condi-
tions. Another observation, recorded by Sabato and De Luca (1985), is that the
northern populations of D. angustifolium Miquel possess smaller leaves, narrower
leaflets and smaller seeds. These populations are among the largest known for
Dioon and become a major component of the vegetation in the states of San Luis
Potosi, Tamaulipas and Nuevo Leon. They give the impression of being more
vigorous and ecologically aggressive than their more restricted southern relatives.
The smaller leaves and leaflets of these northern populations have been proposed
(Sabato and De Luca, 1985) as an adaptation to the lower temperatures and drier
conditions experienced by these plants. We propose that the significantly smaller
seeds, possibly more easily dispersed, together with adaptations for cooler and
drier conditions, have allowed these northern populations rapidly and successfully
ch11.qxd 28/11/03 3:49 pm Page 146
to colonize new habitat in a northward extension from their ice age refugium.
The D. edule complex could be interpreted as a group of recently evolved species
or subspecies that have differentiated by migration and isolation within river
valleys and by colonizing newly available habitat to the north as conditions
became sufficiently favourable since the last glaciation.
Dioon rzedowskii is another illustrative example. This species consists of two
populations within a narrow range on the limestone cliffs of the lower canyon of
the Rio Santo Domingo. It is closely related to D. spinulosum, which occurs over a
wider range in a different habitat along the limestone hills at the base of the
mountains where the Rio Santo Domingo flows out on to the coastal plain.
According to the dynamic habitat hypothesis, D. rzedowskii could be colonizing
new habitat made available by post-glacial warming, and in so doing migrating
up the river canyon. This migration and disjunction is allowing D. rzedowskii to
evolve independently of D. spinulosum.
The examples presented above seem to demonstrate the utility of the
dynamic habitat model as a means of framing interpretations of Dioon bio-
geography. If Dioon species are rapidly and dynamically evolving in response to
climatic changes caused by glacial cycles, instead of the traditional view that they
are slowly evolving and relictual, then the preservation and study of individual
populations should be of paramount importance. Supra- and subspecific naming
conventions should be considered to reflect the interplay between these process-
es and speciation in the genus.
Conclusions
Dioon species currently occur chiefly in restricted habitats primarily in transition

zones between major forest biomes where they presumably can best compete.
The establishment of new populations probably occurs primarily by expansion
and fission of existing populations, or vicariance. The establishment of new pop-
ulations may also occur through medium-range dispersal requiring concomitant
colonization by at least two individuals (i.e. one of each sex) and their pollinator.
This mechanism would generate a genetic bottleneck that could drive rapid spe-
ciation. The morphological data and the available molecular data both support
a proposal that many Dioon species are closely related and have recently evolved.
Several lines of palynological and palaeoclimatic evidence suggest that there
have been dramatic climate changes since the collapse of the last ice age and, by
inference, during each of the many ice age cycles of the past 2 million years.
These climatic changes suggest that the restricted Dioon habitats have moved
upward in elevation by 1500 m and northward by hundreds of kilometres in just
the past 12,000 years. Similarly this suggests that Dioon species have moved with
their respective habitats and that this movement has resulted in colonization of
river valleys as the species move to higher elevations. The secondary isolation of
populations within and between these river drainages has resulted in the rapid
evolution of the majority of the extant species. During the cooling phases of the
ch11.qxd 28/11/03 3:49 pm Page 147
glacial cycles, the Dioon habitats and their attendant species move down in eleva-
tion and southward. This phenomenon could result in a contraction of ranges
and a genetic remixing as well as possible extinctions. We offer the speculations
presented above as a working hypothesis that raises many testable questions con-
cerning the mechanism(s) of dispersal in Dioon, the mechanisms and possible
directionality of recruitment in Dioon, the degree to which the species are related
at the molecular level, and ultimately the rate at which they are evolving.
Acknowledgements
The authors would like to thank Satie Aram for very helpful input on the man-
uscript. We would also like to thank Silvia Salas-Morales, Leo Schibli and the
staff of SERBO in Oaxaca for collaboration and friendship in the field.
References
Axelrod, D.I. (1975) Evolution and biogeography of Madrean-Tethys sclerophyll vegeta-
tion. Annals of the Missouri Botanical Garden 62, 280334.
Chemnick, J. and Gregory, T.J. (1995) A new species of Ceratozamia (Zamiaceae) from
Oaxaca, Mexico with comments on distribution, habitat, and relationships. Phytologia
79, 5157. [Ceratozamia whitelockiana.]
Chemnick, J., Gregory, T.J. and Salas-Morales, S. (1997a) A revision of Dioon tomasellii
(Zamiaceae) from western Mexico, a range extension of D. merolae, and clarification
of D. purpusii. Phytologia 83, 16.
Chemnick, J., Gregory, T.J. and Salas-Morales, S. (1997b) Ceratozamia mixeorum
(Zamiaceae), a new species of Ceratozamia from Oaxaca, Mexico with comments on
distribution, habitat, and species relationships. Phytologia 83, 4752.
De Luca, P. and Sabato, S. (1979) Dioon califanoi (Zamiaceae), a new species from Mexico.
De Luca, P., Sabato, S. and Vzquez Torres, M. (1978) Dioon purpusii Rose (Zamiaceae), a
misknown species. Delpinoa 20, 3135.
De Luca, P., Sabato, S. and Vzquez Torres, M. (1980a) Dioon caputoi (Zamiaceae), a new
species from Mexico. Brittonia 32, 4346.
De Luca, P., Moretti, A., Sabato, S. and Vzquez Torres, M. (1980b) Dioon rzedowskii
(Zamiaceae), a new species from Mexico. Brittonia 33, 225229.
De Luca P., Sabato, S. and Vzquez Torres, M. (1981a) Dioon merolae (Zamiaceae), a new
species from Mexico. Brittonia 33, 179185.
De Luca, P., Sabato, S. and Vzquez Torres, M. (1981b) Dioon holmgrenii (Zamiaceae), a
new species from Mexico. Brittonia 33, 552555.
De Luca, P., Sabato, S. and Vzquez Torres, M. (1982) Distribution and variation of Dioon
edule (Zamiaceae). Brittonia 34, 355362.
De Luca, P., Sabato, S. and Vzquez Torres, M. (1984) Dioon tomasellii (Zamiaceae), a new
species with two varieties from western Mexico. Brittonia 36, 223227.
Gould, S.J. (2001) The Book of Life. W.W. Norton & Co. New York, 256 pp.
Gregory, T.J., Chemnick, J., Salas-Morales, S. and Vovides, A.P. (2003) A new species in
ch11.qxd 28/11/03 3:49 pm Page 148
the genus Dioon (Zamiaceae) from north-central Oaxaca, Mexico. Botanical Journal of
the Linnean Society 141, 471476. [Dioon argenteum.]
MacNeish, R.S. (1961) First Annual Report of the Tehuacan Archaeological-Botanical Project. R.S.
Peabody Foundation for Archaeology, Andover, Massachusetts, 115 pp.
McDonald, J.A. (1993) Phytogeography and history of the alpine-subalpine flora of north-
eastern Mexico. In: Ramamoorthy, T.P., Bye, R., Lot, A. and Fa, J. (eds) Biological
Diversity of Mexico, Origins and Distribution. Oxford University Press, New York, pp.
681703.
and Stevenson, D.W. (1993) A phylogenetic analysis of Dioon (Zamiaceae). American
Norstog, K.J., Stevenson, D.W. and Niklas, K.J. (1986) The role of beetles in the pollina-
tion of Zamia furfuracea. Biotropica 18, 300306.
Sabato, S. and De Luca, P. (1985) Evolutionary trends in Dion [sic] (Zamiaceae). American
Sauer, J.D. (1988) Plant Migration, the Dynamics of Geographic Patterning in Seed Plant Species.
University of California Press, London, 282 pp.
Siniscalco Gigliano, G. (1990) Chemotaxonomic significance of MAM glycosides and
mucilages in cycads. In: Stevenson, D.W. (ed.) The Biology, Structure, and Systematics of the
Cycadales. Proceedings of the Symposium CYCAD 87. Memoirs of the New York Botanical Garden
57, pp. 123131.
Toledo, V.M. (1982) Pleistocene changes in vegetation in tropical Mexico. In: Prance, G.T.
(ed.) Biological Diversification in the Tropics. Proceedings of the Fifth International Symposium of
the Association for Tropical Biology. Columbia University Press, New York, pp. 93111.
Wolfe, J.A. (1972) An interpretation of Alaska tertiary floras. In: Graham, A. (ed.) Floristics
and Paleofloristics of Asia and Eastern North America. Elsevier Publishing Co., Amsterdam,
The Netherlands, pp. 210233.
ch12.qxd 28/11/03 3:50 pm Page 149
Molecular Phylogeny of Zamia 12

(Zamiaceae)
Paolo Caputo,1 Salvatore Cozzolino,1 Paolo De

Luca,1 Aldo Moretti1 and Dennis Wm.
Stevenson2
1Dipartimento di Biologia Vegetale, Universit degli Studi di

Napoli Federico II, Napoli, Italy; 2Institute of Systematic
Botany, New York Botanical Garden, Bronx, New York, USA
Abstract
The preliminary results of a phylogenetic analysis of genus Zamia, based on the sequences
of the internal transcribed spacer 2 of the nuclear ribosomal DNA combined with a mor-
phological data set, are reported. The consensus tree among the six equally parsimonious
cladograms obtained shows Zamia divided in various, often poorly resolved, clades. The
most inclusive dichotomy defines two groups: one composed of mainly Central American
species, i.e. Z. acuminata, Z. neurophyllidia, Z. obliqua, Z. pseudoparasitica and Z. skinneri; and the
other including the rest of the investigated species. The latter clade basally shows Z. inermis
and, more internally, a small unit composed of one Mexican and two West Indian taxa (Z.
fischeri, Z. portoricensis and Z. pumila). This group is in a sister group relationship with a large,
poorly resolved clade which has Z. standleyi at the base. The internal clade includes Z.
soconuscensis and two groups, one made of North American species and the other of mainly
South American species. The first group includes Z. furfuracea, Z. loddigesii, Z. paucijuga, Z.
polymorpha and Z. verschaffeltii. The other group includes mainly South American taxa (i.e.
Chiqua restrepoi, Zamia muricata, Z. boliviana, Z. leicontei, Z. manicata, Z. roezlii and Z. wallisii).
The results indicate that morphological resemblance in the genus does not correspond to
the pattern of phylogenetic relationships, whereas the latter pattern is broadly congruent
with geographical distribution.
Introduction
Within the Zamiaceae, Zamia Linnaeus, with nearly 60 species (Hill et al.,
Appendix 1 this volume) from south-eastern USA, through Central America and
ch12.qxd 28/11/03 3:50 pm Page 150
150 P. Caputo et al.
the Caribbean to Brazil, and belonging to subfamily Zamioideae (Stevenson,

1992), is probably the most studied genus. This is largely on account of its infra-
generic (Marchant, 1968; Norstog, 1980; Moretti, 1990a,b; Moretti et al., 1993b;
Caputo et al., 1996) and infraspecific (Norstog, 1980, 1981; Moretti and Sabato,
1984; Moretti et al., 1991) karyotype variation, both characteristics being unique
among cycads.
The botanical relevance of cycads has caused their inclusion in virtually all
the studies dealing with large-scale phylogenetic relationships in higher plants;
moreover, the endangered status, the commercial value and the beauty of many
species have fostered a large number of investigations on cycad biology. In spite
of this, few of the modern studies focus on the phylogeny of cycads at lower tax-
onomic circumscription. Among them, Stevenson (1990a) dealt with the phy-
logeny of the whole order using morphological traits; Caputo et al. (1991) focused
on the phylogeny of neotropical Zamiaceae using chloroplast DNA characters;
Moretti et al. (1993a) investigated the phylogeny of Dioon Lindley using chloro-
plast DNA; while Caputo et al. (1996), in a study on the karyology of
Meso-American Zamia species, published a preliminary cladogram on the rela-
tionships of the species of Zamia based mainly on morphological and karyologi-
cal characters.
The authors of the latter study showed that Zamia is divided in two major
clades, one of which includes mainly North American and Caribbean species
and the other containing mainly Central and South American species. In the
same cladogram, Zamia appeared as a strictly monophyletic unit, and the small
genus Chigua D.W. Stevenson from Colombia (Stevenson, 1990b) was its sister
group.
In this contribution, we discuss ongoing studies on the phylogeny of genus
Zamia, based on the sequence of the internal transcribed spacer 2 (ITS2) of the
nuclear ribosomal DNA in combination with a morphological data set that rep-
resents an expansion of that used in Caputo et al. (1996). The ITS sequences are
regarded as suitable for inferring phylogeny at the level of genus or below
(Baldwin et al., 1995; Andreasen et al., 1999).
Taxa
A wide selection of the species present in cultivation at the Orto Botanico

(Naples Botanical Garden), as well as at the New York Botanical Garden, was
employed. In particular, for this chapter the following species of Zamia were
investigated (Table 12.1): Z. acuminata Oersted ex Dyer, Z. boliviana (Brongniart)
A. de Candolle, Z. fischeri Miquel, Z. furfuracea Linnaeus filius, Z. inermis Vovides,
J.D. Rees & Vzquez Torres, Z. lecointei Ducke, Z. loddigesii Miquel, Z. manicata
Linden ex Regel, Z. muricata Willdenow, Z. neurophyllidia D.W. Stevenson, Z. obliqua
A. Braun, Z. paucijuga Wieland, Z. polymorpha D.W. Stevenson, A. Moretti &
ch12.qxd
28/11/03
Table 12.1. Species, genebank accession numbers and origin of material (NY, New York Botanical Garden; FTG, Fairchild Tropical
Garden; NAP, Orto Botanico di Napoli) used in the study. The length of sequences (bp, base pairs) and percentage of bases that are
guanine (G) and cytosine (C) are indicated for each species.
3:50 pm
Genebank
Species accession number Origin of specimen ITS2 length (bp) GC%
Chigua
Chigua restrepoi
restrepoi D.W.
D.W. Stevenson
Stevenson AJ287354
AJ287354 NY
NY specimen260
specimen 64 260 64
Zamia
Zamia acuminata
acuminata Oersted
Oersted ex ex Dyer
Dyer AJ287324
AJ287324 FTG
FTG 89-159
89-159 260
260 64
64
Page 151
Z. boliviana
Z. boliviana (Brongniart)
(Brongniart) A. A. de
de Candolle
Candolle AJ287325
AJ287325 NAP, s.n.
NAP, s.n. 261
261 62
62
Z. fischeri
Z. fischeri Miquel
Miquel AJ287328
AJ287328 NAP537/1-1
NAP537/1-1 260
260 67
67
Molecular Phylogeny of Zamia

Z. furfuracea
Z. furfuracea Linnaeus
Linnaeus filius
filius AJ287364
AJ287364 NAP501/8-1
NAP501/8-1 261
261 64
64
Z. inermis
Z. inermis Vovides,
Vovides, J.D.
J.D. Rees
Rees & & Vzquez
Vzquez Torres
Torres AJ287331
AJ287331 NAP540/1-1
NAP540/1-1 260
260 66
66
Z. lecointei
Z. lecointei Ducke
Ducke AJ287339
AJ287339 NAP943/22-1
NAP943/22-1 261
261 65
65
Z. loddigesii
Z. loddigesii Miquel
Miquel AJ287333
AJ287333 NAP539/0-1
NAP539/0-1 260
260 65
65
Z. manicata
Z. manicata Linden
Linden ex
ex Regel
Regel AJ287334
AJ287334 NAP ZZ M204
NAP M204 1984
1984 260
260 66
66
Z. muricata
Z. muricata Willdenow
Willdenow AJ287362
AJ287362 NAP533/10-1
NAP533/10-1 260/261
260/261 65/64
65/64
Z. neurophyllidia
Z. neurophyllidia D.W.
D.W. Stevenson
Stevenson AJ287335
AJ287335 FTG 93-816
FTG 93-816 260
260 63
63
Z. obliqua
Z. obliqua A.A. Braun
Braun AJ287336
AJ287336 FTG 89-163
FTG 89-163 260
260 65
65
Z. paucijuga
Z. paucijuga Wieland
Wieland AJ287337
AJ287337 NAP928/74-1
NAP928/74-1 260
260 65
65
Z. polymorpha
Z. polymorpha D.W.D.W. Stevenson,
Stevenson, A. A. Moretti
Moretti &
& Vzquez
Vzquez Torres
Torres AJ287342
AJ287342 NAP2285/0-1
NAP2285/0-1 260
260 65
65
Z. portoricensis
Z. portoricensis Urban
Urban AJ287343
AJ287343 NAP528/0-1
NAP528/0-1 260
260 66
66
Z. pseudoparasitica
Z. pseudoparasitica Yates
Yates in in Seemann
Seemann AJ287344
AJ287344 Stevenson 1149
Stevenson 1149 261
261 66
66
Z. pumila
Z. pumila Linnaeus
Linnaeus AJ287323
AJ287323 NAP531/3-1
NAP531/3-1 260
260 66
66
Z. roezlii
Z. roezlii Linden
Linden AJ287346
AJ287346 FTG 71-535
FTG 71-535 260
260 65
65
Z. skinneri
Z. skinneri Warszewicz
Warszewicz ex ex A.A. Dietrich
Dietrich AJ287347
AJ287347 FTG 76-609
FTG 76-609 260
260 65
65
Z. soconuscensis Schutzman,
Z. soconuscensis Schutzman, Vovides Vovides && Dehgan
Dehgan AJ287360
AJ287360 NAP939/1-1
NAP939/1-1 260
260 67
67
Z. standleyi
Z. standleyi Schutzman
Schutzman AJ287349
AJ287349 From the
From the collection
collection
of J. P. Sclavo, France
of J. P. Sclavo, France 260
260 66
66
Z. verschaffeltii
Z. verschaffeltii Miquel
Miquel AJ287350
AJ287350 NAP535/6-1
NAP535/6-1 260
260 65
65
Z. wallisii A. Braun AJ287361 FTG Stevenson
Stevenson et et al.
al. 582
582 260 65
151
Z. wallisii A. Braun AJ287361 FTG 260 65
ch12.qxd 28/11/03 3:50 pm Page 152
Vzquez Torres, Z. portoricensis Urban, Z. pseudoparasitica Yates in Seemann, Z.

pumila Linnaeus, Z. roezlii Linden, Z. skinneri Warszewicz ex A. Dietrich, Z. soconus-
censis Schutzman, Vovides & Dehgan, Z. standleyi Schutzman, Z. verschaffeltii
Miquel, Z. wallisii A. Braun, Chigua restrepoi D.W. Stevenson, Microcycas calocoma
(Miquel) A. de Candolle, as well as representatives of other zamiaceous genera as
outgroups. Voucher specimens of the plants examined are deposited at Orto
Botanico di Napoli (NAP) or at New York Botanical Garden (NY).
Morphological and micromolecular data
The majority of the morphological and micromolecular characters employed

have already been published (Stevenson, 1980, 1981, 1987, 1988, 1990a,b, 1993;
Dehgan and Dehgan, 1988; Stevenson and Siniscalco Gigliano, 1989;
Richardson, 1990; Norstog, 1993; Meurer-Grimes and Stevenson, 1994; Caputo
et al., 1996; Stevenson et al., 1996). These comprise stem characters (e.g. epigeous
vs. hypogeous), leaf and petiole characters (e.g. presence/absence of prickles,
stipules, terminal pinnae, leaflet plication, stomata guard cell shape), sporophyll
characters (e.g. cone shape, pubescence, sporophyll vasculature, sporophyll shape,
colour, microsporangia position), pollen and sperm characters and secondary
compound characters (e.g. mucilage monosaccharide patterns).
Molecular data
Extraction of total DNA was carried out following the method described in
Caputo et al. (1991) and ITS2 was amplified according to Aceto et al. (1999) and
double-strand sequenced in both directions using a modification of the Sanger
dideoxy method (Sanger et al., 1977) in a double-strand DNA cycle sequencing
system with fluorescent dyes. Sequence reactions were then loaded into a 373A
Applied Biosystems Automated DNA sequencer (Applied Biosystems, Foster City,
California, USA).
Some sequencing experiments had to be repeated to solve all uncertainties.
In several cases this approach was unsuccessful and the purified PCR product
was ligated into a pUC18 vector (Farmacia Biotech, Uppsala, Sweden) and then
automatically sequenced in the same manner as above by using universal M13
primers.
Data analysis
The optimal alignment was searched by using CLUSTAL W version 1.74

(Thompson et al., 1994) and employing the same approach described in Aceto et
al. (1999), i.e. all the gap opening and extension parameters of the CLUSTAL W
software were varied across every alignment with unit increments from 4 to 15
ch12.qxd 28/11/03 3:50 pm Page 153
and from 4 to 10, respectively. This imposed the constraint that gap opening
parameters should always be greater than, or at least equal to, gap extension
parameters within any single alignment, while all the other parameters were set
at the default values. After cladistic analysis of all the alignments obtained, the
optimal alignment was chosen as that with the highest values of consistency index
(CI) retention index (RI) values, i.e. that giving the highest rescaled consisten-
cy index (RC) (Farris, 1989).
This alignment was combined with the morphological data set, and the
resulting matrix was investigated by using the cladistic software environment
WINCLADA (Nixon, 1999), running Nona (Goloboff, 19931999) as a daughter
process, with the following parameters: hold 100,000; hold/100; mult*100; max.
The resulting cladograms were investigated with WINCLADA, which was also used
to calculate branch support (up to trees five steps longer).
Results
The selected alignment, when analysed in combination with the morphological

data set, gave six equally parsimonious cladograms (length = 345 steps, CI = 0.80
and RI = 0.89), the consensus tree of which is shown in Fig. 12.1. Zamia appears
divided in several incompletely resolved clades. The most inclusive dichotomy
defines two clades, one composed of mainly central American species (the only
partial exception being Z. obliqua, distributed from southern Panama to the south-
ern Choc in Colombia), i.e. Z. acuminata, Z. neurophyllidia, Z. obliqua, Z. pseudopar-
asitica and Z. skinneri, referred to as the pseudoparasitica clade, and the other
including the rest of the investigated species. The latter clade basally shows Z.
inermis, a species endemic to Veracruz, Mexico. More internally, a small unit com-
posed of one Mexican species (Z. fischeri, endemic to San Luis Potosi and neigh-
bouring Quertaro, Mexico) and two West Indian taxa (Z. portoricensis, distributed
in central Cuba and the Dominican Republic, and Z. pumila, from western Puerto
Rico) is in a sister group relationship with a large, poorly resolved clade that has
Z. standleyi (endemic to Honduras) at the base. The internal clade includes Z.
soconuscensis (endemic to Sierra Madre de Soconusco, Chiapas, Mexico) and two
groups, one made of North American species and the other of mainly South
American species. The first group includes Z. furfuracea, Z. loddigesii, Z. paucijuga,
Z. polymorpha and Z. verschaffeltii (this group we loosely refer to as the loddigesii
clade). The other group, here indicated as the Chigua clade has two unresolved
South American taxa (C. restrepoi and Z. muricata) at the base and then a clade
including Z. boliviana and Z. lecointei, which is in turn unresolved compared with
another clade including Z. manicata, Z. roezlii and Z. wallisii.
Branch support for the great majority of the clades is low. Among the main
clades, only the pseudoparasitica clade, the loddigesii clade and the Chigua clade
show a support greater than one.
ch12.qxd 28/11/03 3:50 pm Page 154
Fig. 12.1. Consensus tree out of six maximum parsimony cladograms obtained for
the combined data set (matrix consensus length 378 characters, 272 of which rep-
resent ITS2; length = 345 steps, CI = 0.80 and RI = 0.89). Numbers above branch-
es represent synapomorphies; numbers below indicate branch support (up to five
steps longer trees). Taxa of Zamia are indicated with acronyms derived from the
first letter of the genus and the first three letters of the specific epithet. Chigua
restrepoi D.W Stevenson and Microcycas calocoma (Miquel) A. de Candolle are
indicated with acronyms derived from the first four letters of the genus.
Discussion
The results of this investigation offer new insights on the phylogeny of Zamia, as
they do not match in entirety with any previous hypothesis on the relationships
within the genus. However, the consensus tree in Fig. 12.1 seems to show good
correlations with geographical distribution. Among the major clades that can be
observed, the pseudoparasitica clade is composed of mainly Central American
species; a second (the loddigesii clade) comprises mainly North American species,
and a third (the Chigua clade) includes Central and South American species.
This latter group basally shows a poorly resolved assemblage of taxa (C. restrepoi,
ch12.qxd 28/11/03 3:50 pm Page 155
Z. boliviana, Z. lecointei, Z. muricata), the members of which (excluding Chigua D.W.

Stevenson) have been informally included by Norstog and Nicholls (1997) in the
eastern South American Zamias. A fully resolved CentralSouth American
clade is the fourth major group in the analysis.
Caribbean Zamia species (represented here by Z. portoricensis and Z. pumila)
form a clade, which, together with Z. fischeri, is in a sister group relationship with
a group basally showing Z. standleyi, and including both the loddigesii clade and
the Chigua clade.
Chigua is strongly nested in Zamia, in spite of the presence of a midrib, as well
as the male cone peduncle which is scarcely narrower than the fertile portion
(characters that have been used to segregate it from Zamia); these are simple albeit
evident species-level autapomorphies. Previous work based on chloroplast DNA
(Caputo et al., 1991) and karyology, supplemented by morphological characters
(Caputo et al., 1996), tentatively indicated that Chigua was the sister group of
Zamia. However, in the first case (Caputo et al., 1991), the study, aiming at a
genus-level phylogeny of American Zamiaceae, used only Z. fischeri and Z. skinneri
as representative of genus Zamia. Also in the latter paper (Caputo et al., 1996) the
Central and South American species, which belong to the Chigua clade (Fig.
12.1), were under-represented, as Z. boliviana, Z. lecointei and Z. muricata were
missing from that analysis.
Interestingly enough, various of the morphological characters that have
been thought to define synapomorphic units in the genus (e.g. strong leaflet pli-
cation, possession of a trunk, presence of adaxial microsporangia) seem to rep-
resent the consequence of pervasive parallel evolution which besets this taxon.
Indeed, Zamia itself is hardly definable in terms of morphological synapomor-
phies, as the only morphological character which acts as a synapomorphy at
genus level (distinct male cone facets) reverts more than once within the genus.
No trait, among the morphological character used, seems to have evolved only
once in the genus. As a consequence, various species that show great morpho-
logical resemblance have to be interpreted as the product of similar selective
pressures on quite different lineages. For example, the Aulacophyllum Regel body
plan (here represented by Z. neurophyllidia, Z. roezlii, Z. skinneri and Z. wallisii) seems
to have evolved at least twice in the genus.
Chromosome characters were not included in our analysis, because they, as
well as karyotypes, show infraspecific variation in several taxa (Norstog, 1980,
1981; Moretti and Sabato, 1984; Moretti et al., 1991). Plotting the chromosome
number on to the phylogenetic tree of Fig. 12.1 under unambiguous optimiza-
tion reconstructs the plesiomorphic chromosome number for the genus as 2n =
16, and reconstructs the plesiomorphic number for the loddigesii clade as 2n =
18. This would imply that centric fission occurred more than once in the genus,
i.e. in the pseudoparasitica clade, the loddigesii clade and the Chigua clade.
The fact that morphological resemblance does not always correspond to pat-
terns of descent, and that also chromosomal rearrangement seems to have devel-
oped more than once, is not surprising given the comparative antiquity of a
genus whose stem lineage dates back at least to the early Cenozoic. Long times
ch12.qxd 28/11/03 3:50 pm Page 156
under similar selective pressures have probably caused morphological conver-

gence in lineages that are quite separate from a phylogenetic point of view.
Further study, including a totality of species in the genus, will probably show that
the above-mentioned cases of parallelism are even more pervasive than shown in
this contribution.
References
Aceto, S., Caputo, P., Cozzolino, S., Gaudio, L. and Moretti, A. (1999) Phylogeny and evo-
lution of Orchis and allied genera based on ITS DNA variation: morphological gaps
and molecular continuity. Molecular Phylogenetics and Evolution 13, 6776.
Andreasen, K., Baldwin, B.G. and Bremer, B. (1999) Phylogenetic utility of the nuclear
ITS region in subfamily Ixoroideae (Rubiaceae): comparison with cpDNA rbcL
sequence data. Plant Systematics and Evolution 217, 119135.
Baldwin, B.G., Sanderson, M.J., Porter, M.J., Wojciechowski, M.F., Campbell, C.S. and
Donoghue, M.J. (1995) The ITS region of nuclear ribosomal DNA: a valuable source
of evidence in angiosperm phylogeny. Annals of the Missouri Botanical Garden 82,
247277.
Zamiaceae (Cycadales) using chloroplast DNA restriction fragment polymorphisms.
Caputo, P., Cozzolino, S., Gaudio, L., Moretti, A. and Stevenson, D.W. (1996) Karyology
and phylogeny of some Meso-American species of Zamia (Zamiaceae). American
Farris, J.S. (1989) The retention index and the rescaled consistency index. Cladistics 5,
417419.
Goloboff, P.A. (19931999) Nona. Instruction manual. Distributed by the author.
Marchant, C.J. (1968) Chromosome patterns and nuclear phenomena in the cycad fami-
lies Stangeriaceae and Zamiaceae. Chromosoma 24, 100134.
Meurer-Grimes, B. and Stevenson, D.W. (1994) The biflavones of the Cycadales revisited:
biflavones in Stangeria eriopus, Chigua restrepoi and 32 other species of Cycadales.
Biochemical Systematics and Ecology 22, 595603.
Moretti, A. (1990a) Cytotaxonomy of cycads. In: Stevenson, D.W. (ed.) The Biology,
Structure, and Systematics of the Cycadales. Proceedings of the Symposium CYCAD 87. Memoirs
of the New York Botanical Garden 57, pp. 114122.
Moretti, A. (1990b) Karyotypic data on North and Central American Zamiaceae
(Cycadales) and their phylogenetic implications. American Journal of Botany 77,
10161029.
Moretti, A. and Sabato, S. (1984) Karyotype evolution by centromeric fission in Zamia
(Cycadales). Plant Systematics and Evolution 146, 215223.
Moretti, A., Caputo, P., Gaudio, L. and Stevenson, D.W. (1991) Intraspecific chromosome
variation in Zamia (Zamiaceae, Cycadales). Caryologia 44, 110.
and Stevenson, D.W. (1993a) A phylogenetic analysis of Dioon (Zamiaceae). American
ch12.qxd 28/11/03 3:50 pm Page 157
Moretti, A., Caputo, P., Cozzolino, S. and Gaudio, L. (1993b) Karyotypes of New World
Cycads. In: Stevenson, D.W. and Norstog, K.J. (eds) The Biology, Structure, and Systematics
of the Cycadales. Proceedings of the Second International Conference on Cycad Biology. Palm &
Cycad Societies of Australia Limited, Milton, Queensland, Australia, pp. 263269.
Nixon, K.C. (1999) WinClada (BETA), Version 0.9.9. Published by the author, Ithaca,
New York.
Norstog, K.J. (1993) Spermatogenesis in Microcycas: evolutionary significance of male
gametes of seed plants. In: Stevenson, D.W. and Norstog, K.J. (eds) The Biology,
Structure, and Systematics of the Cycadales. Proceedings of the Second International Conference on
Cycad Biology. Palm & Cycad Societies of Australia Limited, Milton, Queensland,
Australia, pp. 270278.
Richardson, P.M. (1990) Flavonoid chemistry and the taxonomy of the cycads. In:
Stevenson, D.W. (ed.) The Biology, Structure, and Systematics of the Cycadales. Proceedings of
the Symposium CYCAD 87. Memoirs of the New York Botanical Garden 57, pp. 132141.
Sanger, F., Niklen, S. and Coulson, A.R. (1977) DNA sequencing with chain terminating
inhibitors. Proceedings of the National Academy of Sciences, USA 74, 54635467.
Stevenson, D.W. (1980) Radial growth in the Cycadales. American Journal of Botany 67,
465475.
Stevenson, D.W. (1981) Observations on ptyxis, phenology, and trichomes in the
Cycadales and their systematic implications. American Journal of Botany 68,
11041114.
Stevenson, D.W. (1987) Comments on character distribution, taxonomy, and nomencla-
ture of the genus Zamia L. in the West Indies and Mexico. Encephalartos 9, 37.
Stevenson, D.W. (1988) Strobilar ontogeny in the Cycadales. In: Leins, P., Tucker, P.C. and
Endress, P.K. (eds) Aspects of Floral Development. G. Fischer, Stuttgart, Germany, pp.
205224.
Stevenson, D.W. (1990a) Morphology and systematics of the Cycadales. In: Stevenson,
Stevenson, D.W. (1990b) Chigua, a new genus in the Zamiaceae with comments on its bio-
geographical significance. In: Stevenson, D.W. (ed.) The Biology, Structure, and Systematics
of the Cycadales. Proceedings of the Symposium CYCAD 87. Memoirs of the New York Botanical
Garden 57, pp. 169172.
Stevenson, D.W. (1993) The Zamiaceae in Panama with comments on phytogeography
and species relationships. Brittonia 45, 116.
Stevenson, D.W. and Siniscalco Gigliano, G. (1989) The systematic value of the mono-
saccharide composition and distribution pattern of cycad mucilages. Biochemical
Stevenson, D.W., Norstog, K.J. and Molsen, D. (1996) Midribs of cycad pinnae. Brittonia
48, 6774.
Thompson, J.D., Higgins, D.G. and Gibson, T.J. (1994) CLUSTAL W: improving the sen-
sitivity of progressive multiple sequence alignment through sequence weighting, posi-
tions-specific gap penalties and weight matrix choice. Nucleic Acids Research 22,
46734680.
ch12.qxd 28/11/03 3:50 pm Page 158
ch13.qxd 28/11/03 3:50 pm Page 159
Systematics of Meso-American 13
Zamia (Zamiaceae)
Bart Schutzman
Environmental Horticulture Department, University of Florida,

Gainesville, Florida, USA
Abstract
The authors work includes morphological, geographical and phenetic analyses of

Meso-American Zamia representatives and is a continuation of earlier Masters and
Doctoral studies. This research has led to the publication of six new species descriptions;
in addition, several others are in progress, several preliminary species groups have been
recognized and recommendations made for future work. However, an infrageneric classi-
fication has not yet been included. Such a classification will depend on the culmination of
alpha-taxonomic work in the genus, thorough morphological and geographical character-
ization of all known species, and finally, cladistic analyses, including those arising from
molecular systematic investigation. Examination of possible introgression and hybrid
species complexes is also necessary for a thorough understanding and development of an
infrageneric classification in Zamia.
Introduction
In the past, taxonomy of the genus Zamia Linnaeus has been based primarily on
leaflet number, size and shape. Newell (1985) first pointed out the predominance
of leaflet morphology in taxonomic treatments and demonstrated that it could be
environmentally influenced. Leaf and leaflet features, unfortunately, are highly
plastic, being affected by environment and juvenility, and data commonly overlap
between species. This is best illustrated in Eckenwalders (1980) work with the
Caribbean zamias. He did not consider other sources of data such as reproduc-
tive features before lumping all the Caribbean species into Z. pumila Linnaeus
ch13.qxd 28/11/03 3:50 pm Page 160
160 B. Schutzman
subspecies pumila and subspecies pygmaea (Sims) Eckenwalder (= Z. pygmaea Sims;

see Hill et al., Appendix 1 this volume). Stevenson (1987) looked at strobilar fea-
tures and revised the Caribbean species to six, but this number of taxa still does
not separate the Florida species (one, two, or three, depending on the systematist)
from other Caribbean taxa. The authors approach in the study of
Meso-American species (Schutzman, 1982, 1998) was Adansonian and called for
the examination of as many characters as possible. This approach will be neces-
sary to resolve other groups of cycad taxa, e.g. the Caribbean Zamia species.
Plant material was collected by the author and others during field expeditions to
Mexico, Honduras, Panama and Costa Rica. Additional material was obtained
from private collectors and botanical gardens, including the Montgomery
Botanical Center, Fairchild Tropical Garden, Marie Selby Botanical Gardens
and the Huntington Botanical Gardens. Herbarium specimen loans were
obtained from 32 institutions in the USA and elsewhere.
Gross morphology of eophylls (first seedling leaves) and the progression of
juvenile to adult leaflets were compared. Scanning electron micrographs were
made of vegetative and reproductive features, including leaflets, microsporophylls
and microsporangia. Light microscopy was used to examine leaf epidermal clear-
ings after treatment with Jeffreys (1917) solution. Fourier morphometrics and phe-
netic analyses of leaflet shape were conducted on several groups of specimens.
These included comparisons of leaflets from the same plant, from different plants
in the same population, from different populations, and from different species.
Results and Discussion
Results by character
Eophyll (first leaf) morphology

Eophyll morphology was useful in identifying individual species in most cases,
and species groups in others, but was not useful at all in some cases. Variation
depended on species; most significant was the number of leaflets per eophyll.
Morphological features that were observed to vary in a meaningful way included
leaflet number and shape. Examples shown in Fig. 13.1 include Zamia cremnophila
Vovides, Schutzman & Dehgan, Z. vazquezii D.W. Stevenson, Sabato, A. Moretti
& De Luca, Z. sp. maritima and Z. spartea A. de Candolle.
Gross morphology of microsporophylls

Examples of microsporophylls are shown in Fig. 13.2, and include Zamia
fairchildiana L.D. Gmez, Z. pseudoparasitica Yates in Seemann, Z. spartea and
Z. vazquezii. Apomorphic species were easily identified. The characteristics most
ch13.qxd
28/11/03
3:50 pm
Page 161
Systematics of Meso-American Zamia
Fig. 13.1. Example of eophyll shape. (A) Zamia cremnophila Vovides, Schutzman & Dehgan. (B) Zamia vazquezii D.W. Stevenson,
161
Sabato, A. Moretti & De Luca. (C) Zamia sp. maritima. (D) Zamia spartea A. de Candolle.
ch13.qxd
28/11/03
162
3:50 pm
Page 162
B. Schutzman
Fig. 13.2. Example of microsporophyll shape (a, abaxial; b, adaxial). (A) Zamia fairchildiana L.D. Gmez. (B) Zamia pseudoparasitica
Yates in Seemann. (C) Zamia spartea A. de Candolle. (D) Zamia vazquezii D.W. Stevenson, Sabato, A. Moretti & De Luca.
ch13.qxd 28/11/03 3:50 pm Page 163
Systematics of Meso-American Zamia 163
Fig. 13.3. Diagram illustrating from where standardized scanning micrographs of

microsporophyll surfaces were taken (a, area where pubescence generally ends; b,
wing area on adaxial surfae).
useful in identifying species were: (i) shape of microsporophyll, including pres-

ence or absence of narrowing to a pedicel; (ii) presence or absence of microspo-
rangia in the interstitial region; (iii) shape and size of the apex; and (iv) surface
features of the distal microsporophyll surface.
Scanning electron microscopic examination of microsporophylls

Microsporophylls were examined under SEM and two primary useful regions
were identified, as illustrated in Fig. 13.3 (see regions a and b). Examples of
one region (a in Fig. 13.3) are shown in Fig. 13.4 for Zamia cremnophila, Z. loddi-
gesii Miquel, Z. pseudoparasitica and Z. soconuscensis Schutzman, Vovides & Dehgan.
Useful characteristics included shape, size and frequency of trichomes, frequen-
cy of stomata, and visibility of cellular outlines.
Scanning electron microscopic examination of microsporangia

Microsporangial features were visible under low SEM magnification and were
rich in morphological variation. Examples are shown in Fig. 13.5 and include
Zamia neurophyllidia D.W. Stevenson, Z. standleyi Schutzman, Z. vazquezii and Z.
tuerckheimii Donnell Smith. Useful characteristics observed were ornamentation of
the suture (line of dehiscence), stomatal frequency and position, and visibility of
cellular outlines. The usefulness of these features was also mentioned by
Schutzman and Dehgan (1988) and Dehgan et al. (1993).
ch13.qxd 28/11/03 3:50 pm Page 164
164 B. Schutzman
Fig. 13.4. Example of microsporophyll variation; scanning micrograph of area a

in Fig. 13.3. (A) Zamia cremnophila Vovides, Schutzman & Dehgan. (B) Zamia
loddigesii Miquel. (C) Zamia pseudoparasitica Yates in Seemann. (D) Zamia
soconuscensis Schutzman, Vovides & Dehgan.
Leaflet shape
While leaflet shape and size did not always resolve taxa, the fine resolution of
Fourier analysis was a better quantifier of shape than length:width ratios or
absolute measurements, as also discussed by Schutzman and Dehgan (1993). If
enough Fourier coefficients are used, shape may be satisfactorily reconstructed. It
ch13.qxd 28/11/03 3:50 pm Page 165
Fig. 13.5. Example of microsporangial variation. (A) Zamia neurophyllidia D.W

Stevenson. (B) Zamia standleyi Schutzman. (C) Zamia vazquezii D.W. Stevenson,
Sabato, A. Moretti & De Luca. (D) Zamia tuerckheimii Donnell Smith.
was found that 32 coefficients gave sufficient details in the shape reconstruction,
and could be plugged into phenetic analyses to allow resolution of taxa in many
cases. Figures 13.6 and 13.7 exemplify the results of a Fourier analysis of leaflet
shapes. In the example, Zamia spartea, Z. furfuracea Linneaus filius and their artifi-
cial hybrid were analysed and principal component analysis was used to display
the results graphically. The hybrid fell between the parents, but was closer to the
Z. spartea parent, consistent with the resemblance between the hybrid and that
parent.
ch13.qxd 28/11/03 3:50 pm Page 166
166 B. Schutzman
Fig. 13.6. Example of phenetic analysis of leaflet shape Fourier transforms from
Zamia furfuracea Linnaeus filius (filled circle), Z. spartea A. de Candolle (filled
square) and their hybrid [Norstog hybrid] (open circle), respectively. The coding
symbols shown correspond to the principal component analysis of these leaflet
shapes shown in Fig. 13.7.
Results by taxa
Species concept
The widely used morphogeographical species concept has been employed in the
authors studies of Zamia. Species are recognized by distinct morphological gaps
between groups correlated with geographical distribution patterns.
Species described
Descriptions of six new species have been published: Zamia splendens Schutzman
(1984) (not synonymous with Z. verschaffeltii Miquel; Hill et al., Appendix 1 this
ch13.qxd 28/11/03 3:50 pm Page 167
10
8 Z. spartea
Principal component 2 (21.0%)
Z. furfuracea
6 Z. (Norstog hybrid)
4
6 4 2 0 2 4 6 8 10
Principal component 1 (51.8%)
Fig. 13.7. Principal component analysis of leaflet shapes from Zamia furfuracea
Linnaeus filius, Z. spartea A. de Candolle and their hybrid [Norstog hybrid]
described in Fig. 13.6.
volume), Z. standleyi Schutzman (1989), Z. soconuscensis Schutzman, Vovides &

Dehgan (1988), Z. cremnophila Vovides, Schutzman & Dehgan (Schutzman et al.,
1988), Z. lacandona Schutzman & Vovides (1998) and Z. elegantissima Schutzman,
Vovides & R.S. Adams (1998). Species descriptions in progress include Zamia
species nova #1 (from Guatemala), Z. species nova #2 (from Guatemala) and Z. species
nova (from Honduras).
Remaining Taxonomic Problems in Meso-American Zamias

Studies of living plants and herbarium material have identified several taxonom-
ic problems that exist amongst the Meso-American species, including the follow-
ing examples.
Misapplied species names
Zamia acuminata Oersted ex Dyer

The species was described from the San Juan River in Nicaragua, but cultivated
plants from El Valle, Panama, are mislabelled as this species and in reality are
undescribed. We are acquiring materials of the Nicaraguan species to make
detailed comparisons with the Panamanian plants.
ch13.qxd 28/11/03 3:50 pm Page 168
168 B. Schutzman
Zamia fischeri Miquel / Z. vazquezii

At first the author believed two varieties to be present, but is now leaning toward
the existence of two species. The commonly cultivated plants range from
Veracruz to San Luis Potosi and Hidalgo, Mexico, and belong to Z. vazquezii,
while Z. fischeri is narrowly endemic to a small area in western San Luis Potosi.
Some believe that the large San Luis Potosi plant belongs to Z. fischeri while only
the large plant from Veracruz belongs to Z. vazquezii.
Zamia fairchildiana / Z. pseudomonticola L.D. Gmez / Z. elegantissima

The first of these was described using Costa Rican material from the Osa
Peninsula, and includes plants considered Z. pseudomonticola. The morphogeo-
graphically well-separated central Panamanian plants, once whimsically referred
to as Z. pseudo-pseudoparasitica, then later known as Z. fairchildiana, are cor-
rectly described as Z. elegantissima, but are still frequently referred to as Z.
fairchildiana.
Possible hybrid species complexes
Zamia loddigesii /Z. sp. maritima / Z. furfuracea / Z. paucijuga Wieland

There are multiple forms of Z. loddigesii that appear distinct, exhibiting variation
in strobilar features, but certain populations seem to be introgressing into the
coastal species referred to as Z. sp. maritima (Schutzman, 1998). The coastal
species was never named but the putative introgressive hybrid was used by
Stevenson and Sabato (1986) to lectotypify Z. furfuracea (Schutzman, 1998).
Zamia polymorpha D.W. Stevenson, A. Moretti & Vzquez Torres /

Z. variegata Warszewicz
The variegated species from the east coast of Guatemala appear to intergrade
with the interior species found there and in Belize. Plants range from enormous
specimens, with prominent variegation and 4-metre long leaves, to diminutive
plants less than a metre tall with only a few dull yellowish spots on their leaflets.
The name Z. polymorpha has been applied to populations from Yucatan, Mexico,
Belize and the Guatemalan interior (Maya Mountains) but some of these
Guatemalan plants exhibit the leaf spotting, indicating possible contact with Z.
variegata.
Zamia neurophyllidia / Z. skinneri Warszewicz ex A. Dietrich

The former is described from the vicinity of Panamas border with Costa Rica,
and interior Costa Rican plants quite unlike the type are mislabelled Z. neurophyl-
lidia. Several distinct and non-overlapping morphological groups of populations
are all lumped under the name Z. skinneri, and these should be resolved into dif-
ferent taxa.
ch13.qxd 28/11/03 3:50 pm Page 169
Discussion
Species delimitation
Reasons for difficulties in Zamia species delimitation are several and are enumer-
ated here:
Scant herbarium materials

Herbarium specimens often consist only of vegetative parts, because reproduc-
tive material in Zamia is very rarely found in the field. Rarity of fertile specimens
in the field has therefore commonly limited the use of reproductive characters in
species descriptions.
Cycads as survivors
Under unfavourable conditions (situations in which other plants would simply not
be found alive), many cycads are often be found in a highly reduced form, often
reverting to a juvenile condition. This frequently happens in regrowth circum-
stances, after fires or other catastrophic events, under low soil fertility, or when
plant cover has been stripped away from shade-loving species. A corollary to this
is that plants of a given species may easily outgrow their descriptions under cul-
tivation. Entire populations of plants under unfavourable conditions often
survive and can even reproduce, but vegetative and reproductive structures may
not reach the sizes that would be found under optimal growing conditions. This,
combined with cycads juvenile-to-adult transition in vegetative features, the
unique ability of small plants of many apparently immature zamias to produce
equally small mega- and microstrobili, and the fact that plants may never
outgrow juvenile vegetative characters, could easily have resulted in overlapping
species descriptions.
Typological thinking
This is a belief that a member of a species should conform to a predefined model
or type. Typological thinking was employed early in the evolution of systemat-
ic thought, prior to the idea that species could be polythetic (members of a group
sharing multiple, but not necessarily all, features). Typology attempted to
straight-jacket a species to agree with its type specimen or description; anything
that did not fit the type must certainly be another species. A corollary of typo-
logical thinking is that a published species name must be associated with a taxon;
this results in an approach to taxonomic revision in which a groups nomencla-
ture is dealt with prior to the determination of taxa; this modus operandi brings to
mind the expression putting the cart before the horse.
Use of unicates to describe new species

Because it is now realized that taxa can be and most often are polythetic, the
collection of a single specimen and the description of a taxon from that specimen
is the least accurate way to delimit taxa. Unfortunately, this is the situation
ch13.qxd 28/11/03 3:50 pm Page 170
170 B. Schutzman
inherited from 18th and 19th century taxonomists; the rule of nomenclatural pri-
ority forces us to deal with these poor collections and ambiguous names rather
than by extirpating dubious ones. Many treatments simply list a name as nomen
dubium but this decreases nomenclatural stability because the question may later
be resolved and newer names reduced to synonymy in favour of the older ones.
Conservation of species names, now an acceptable practice, is a stabilizing influ-
ence in nomenclature that eliminates the need to use the term nomen dubium.
Misunderstanding of plasticity of leaves and leaflets

Since leaves were the primary characteristic used in descriptions, their extreme
variability was the predominant factor in the misunderstanding of Zamia species.
Production of small mega- and microstrobili by small sexually mature

plants, and successively larger strobili by older, larger plants
The diameter of successive strobili and length of sporophylls of individual plants
appear to reach a stable size earlier than the lengths of those strobili, until they,
too, reach some maximum size range for that individual, population or species.
Nearsighted local floras with taxa overlapping other areas

Regional geographical areas of taxonomic study can often produce misleading
results for two major reasons: taxa are not always confined to the areas of study;
and related taxa that could shed light on many systematic questions are often not
considered. This has been an especially prevalent problem in Zamia taxonomy,
because of nearsighted local floras that duplicated naming of species spanning
several political divisions. This has most notably occurred in the Caribbean area,
where the cycad taxonomy still remains to be elucidated to the satisfaction of most
cycadologists. Selection of Meso-America as the geographical range of the
authors project was in full cognizance of these facts; as he worked with the plants,
he attempted to keep the Caribbean and South American taxa in perspective.
Bijan Dehgan (Florida, 2002, personal communication) believes that cycad
taxonomists have become carried away with new species descriptions and are
using minute differences to differentiate new taxa.
Infrageneric classification
Successful completion of an infrageneric classification in Zamia will depend on

several factors:
To increase the potential accuracy of a phylogenetic analysis, the alpha-taxo-

nomic stage of classification must include as many species as possible in the
analysis.
Nomenclatural difficulties must be resolved. These include the resolution of
species complexes and the role of hybridization, old descriptions that lack geo-
graphical information, and/or type specimens that are difficult to identify.
ch13.qxd 28/11/03 3:50 pm Page 171
Thus far, cladistic characters, other than DNA sequence data, are few in
number, and further study must be completed to amass sufficient data to enu-
merate cladistic characters fully.
Future use of molecular systematics
Molecular biological applications to systematics, specifically the use of DNA

sequences to uncover phylogeny, are currently very popular. The successful use of
molecular phylogenetics in cycads will unfortunately be delayed by alpha-taxon-
omy, because species of Zamia, Ceratozamia Brongniart and Dioon Lindley are still
being described. This is especially true for the genus Zamia, which is estimated to
include a dozen or more undescribed species in Meso-America alone. The situa-
tion in South America is most probably similar. Successful species delimitation
relies on the totality of characters, most importantly the successful interpretation
of morphogeographical variation. Ultimately, the use of cladistic methods to
produce an evolutionary hypothesis will be likely to be more accurate the closer
species delimitation is to completion.
Species sampling problems are mentioned by Wendel and Doyle (1998) as
one of five organism-level processes that can cause conflicting phylogenetic
hypotheses in molecular systematics. Incomplete alpha-taxonomy would almost
certainly result in this type of problem. Also listed among these causes are rapid
morphological evolution or convergence, hybridization and/or introgression and
rapid diversification.
Conclusions
While DNA sequence data have proved invaluable as a tool in the reconstruction
of infrageneric relationships, the value of an Adansonian approach to taxonomy
that utilizes all available data cannot be overstressed. The relative value of the
wide variety of data, both vegetative and reproductive, uncovered in the authors
work in Meso-American Zamia remains to be determined in the context of an
infrageneric classification of the genus. It is the challenge of continuing and
future systematic studies to establish a balance between molecular and morpho-
logical data that will result in the most accurate classification. Additionally, the
value of phenetic studies must not be dismissed merely because of the possibili-
ty of parallel or convergent features; these data can be compared and will provide
additional insight into relationships within the genus.
References
Dehgan, B., Schutzman, B. and Almira, F. (1993) Utilization of scanning electron
microscopy in the study of surface features in Cycadales. In: Stevenson, D.W. and
ch13.qxd 28/11/03 3:50 pm Page 172
172 B. Schutzman
Norstog, K.J. (eds) The Biology, Structure, and Systematics of the Cycadales. Proceedings of the
Second International Conference on Cycad Biology. Palm & Cycad Societies of Australia
Limited, Milton, Queensland, Australia, pp. 228235.
Eckenwalder, J.E. (1980) Taxonomy of the West Indian cycads. Journal of the Arnold
Arboretum 61, 701722.
Jeffrey, E.C. (1917) The Anatomy of Woody Plants. University of Chicago Press, Chicago,
Illinois, 478 pp.
Newell, S. (1985) Intrapopulational variation in leaflet morphology of Zamia pumila L. in
relation to microenvironment and sex. American Journal of Botany 72, 217221.
Schutzman, B. (1982) Preliminary systematic and taximetric studies of Meso-American
Zamia L. (Zamiaceae). MSc thesis, University of Florida, Gainesville, Florida.
Schutzman, B. (1984) A new species of Zamia (Zamiaceae, Cycadales) from Chiapas,
Mexico. Phytologia 55, 299304. [Zamia splendens.]
Schutzman, B. (1989) A new species of Zamia (Zamiaceae, Cycadales) from Honduras.
Systematic Botany 14, 214219. [Zamia standleyi.]
Schutzman, B. (1998) Revisionary studies of Meso-American Zamia L. (Zamiaceae,
Cycadales). PhD Dissertation. University of Florida, Gainesville, Florida.
Schutzman, B. and Dehgan, B. (1988) Microsporophylls and microsporangia of
Cycadales: comparative morphology and systematic implications. American Journal of
Botany 75, 204205.
Schutzman, B. and Dehgan, B. (1993) Computer assisted systematics in the Cycadales. In:
Stevenson, D.W. and Norstog, K.J. (eds) The Biology, Structure, and Systematics of the
Schutzman, B. and Vovides, A.P. (1985) Phenetic and other systematic studies of the Zamia
loddigesii / Z. furfuracea complex. Abstract of paper presented at 36th annual AIBS
meetings, University of Florida, Gainesville, Florida 1115 August 1985.
Schutzman, B. and Vovides, A.P. (1998) A new Zamia (Zamiaceae, Cycadales) from eastern
Chiapas, Mexico. Novon 8, 441446. [Zamia lacandona.]
Schutzman, B., Vovides, A.P. and Dehgan, B. (1988) Two new species of Zamia
(Zamiaceae, Cycadales) from Southern Mexico. Botanical Gazette 149, 347360.
[Zamia cremnophila, Z. soconuscensis.]
Schutzman, B., Vovides, A.P. and Adams, R.S. (1998) A new Zamia (Zamiaceae,
Cycadales) from Central Panama. Phytologia 85, 137145. [Zamia elegantissima.]
Stevenson, D.W. (1987) Again the West Indian Zamias. Fairchild Tropical Garden Bulletin 42,
2327.
Stevenson, D.W. and Sabato, S. (1986) Typification of names in Zamia L. and Aulacophyllum
Regel (Zamiaceae). Taxon 35, 134144.
Wendel, J.F. and Doyle, J.J. (1998) Phylogenetic incongruence: window into genome
history and molecular evolution. In: Soltis, D.E., Soltis, P.S. and Doyle, J.J. (eds)
Molecular Systematics of Plants II: DNA Sequencing. Kluwer Academic Publishers, Boston,
Massachusetts, pp. 265296.
ch14.qxd 28/11/03 4:26 pm Page 173
Zamiaceae of Bolivia, Ecuador 14

and Peru
Dennis Wm. Stevenson
Institute of Systematic Botany, New York Botanical Garden,

Bronx, New York, USA
Abstract
Ten previously known species of Zamia from Bolivia, Ecuador and Peru are discussed and
a new species from Peru, Zamia macrochiera, is described. Keys to all 11 species are given as
well as complete descriptions, synonymy, types, exsiccata, distributional data and conser-
vation status as used in the 1997 IUCN Red List of Threatened Plants. Zamia boliviana is found
only in Bolivia and one collection from contiguous Brazil; Z. gentryi is endemic to Ecuador,
while Z. macrochiera and Z. urep are endemic to Peru. The other seven species are more
widespread, with Z. poeppigiana and Z. ulei found in both Ecuador and Peru; Z. disodon
known from only one locality in Peru, in an area disjunct from northern Colombia; Z.
hymenophyllidia found only in north-eastern Peru and contiguous south-eastern Colombia;
and Z. roezlii found only in north-western Ecuador and contiguous south-western
Colombia. Although widespread in the Amazon basin of Colombia, Venezuela and
Brazil, Z. lecointei and Z. amazonum are each known from only one or two collections,
respectively, in north-eastern Peru.
Introduction
The last treatment of the neotropical species of Zamia Linnaeus was that of
Schuster (1932). As discussed in Sabato (1990), Stevenson (1990, 1991, 2001) and
Norstog and Nicholls (1997), the nomenclature and species descriptions in
Schusters work are inadequate, mainly because of a paucity of collections and a
lack of field experience.
This treatment of the cycad genus Zamia is intended to complement previ-
ous treatments of the neotropical cycads that have appeared in the past 10 years
ch14.qxd 28/11/03 4:26 pm Page 174
174 D. Wm. Stevenson
for the Floras of the Guianas (Stevenson, 1991), Panama (Stevenson, 1993),
Colombia (Stevenson, 2001) and the checklists for Peru (Brako, 1993) and
Ecuador (Stevenson, 1999). Information from recent collections and fieldwork is
collated here to augment data presented on the conservation status of those
species of Zamia listed by Walter and Gillett (1998) in the 1997 IUCN Red List of
Threatened Plants.
Key to the species of Zamia in Bolivia, Ecuador and Peru
1. Leaflets with a distinct petiolule and a gland-like collar

at the base of the lamina ............................... 7. Z. macrochiera
1. Leaflets sessile and without a gland-like collar
2. Leaflets deeply grooved above between veins,
appearing plicate .......................................................... 9. Z. roezlii
2. Leaflets smooth to striate, not appearing plicate
3. Leaflets membranous, almost translucent
4. Leaflet margins coarsely toothed, at least
some teeth biserrate ........................... 3. Z. disodon
4. Leaflet margins entire or with only a few
uniserrate teeth apically 5. Z. hymenophyllidia
3. Leaflet papyraceous to coriaceous
5. Leaflets entire or rarely with a few minute teeth apically
6. Petiole smooth, leaflet margins
strongly revolute ...................... 2. Z. boliviana
6. Petiole with at least a few prickles, leaflet margins flat
7. Petiole and rachis densely prickled throughout,
prickles often branched ........ 1. Z. amazonum
7. Petiole sparsely to densely prickled in lower half,
prickles unbranched, rachis smooth
8. Leaflet surfaces striolate, veins
obvious .................... 4. Z. gentryi
8. Leaflet surfaces smooth, veins
obscure .................. 6. Z. lecointei
5. Leaflets obviously and variously toothed
9. Leaflets long linear-lanceolate, marginal teeth
aculeate and nearly at right angles, strongly falcate
basally ..................................... 8. Z. poeppigiana
9. Leaflets elliptic to oblong-elliptic, marginal teeth at
an acute angle, leaflets straight or only slightly sub-
falcate basally
10. Petiole with prickles, leaflet surface smooth
veins obscure ............ 10. Z. ulei
10. Petiole smooth, leaflet surface striolate, veins
obvious ...................... 11. Z. urep
ch14.qxd 28/11/03 4:26 pm Page 175
Zamiaceae of Bolivia, Ecuador and Peru 175
Fig. 14.1. Distribution map of Zamia amazonum D.W. Stevenson, Z. bolivana

(Brongniart) A. de Candolle, Z. disodon D.W. Stevenson & Sabato, Z. gentryi
Dodson, Z. hymenophyllidia D.W. Stevenson and Z. lecointei Ducke in Bolivia,
Ecuador and Peru.
Fig. 14.2. Distribution map of Zamia macrochiera D.W. Stevenson, Z. poeppigiana

Martius & Eichler, Z. roezlii Linden, Z. ulei Dammer and Z. urep Wallnfer in
Bolivia, Ecuador and Peru.
ch14.qxd 28/11/03 4:26 pm Page 176
1. ZAMIA AMAZONUM D.W. Stevenson (Fig. 14.3), (2001) Flora de

Colombia 21, 33 and Fig. 3
TYPE: BRAZIL. Amazonas, Confluence of Rio Tauri and Rio Icana, D.

Stevenson 886 (HOLOTYPE: INPA; ISOTYPES: COL, MO, NY, U)
Stem subterranean, 38 cm in diameter. Cataphylls triangular basally, linear-

lanceolate apically, 38 cm long, 12 cm wide. Leaves 26, 0.52.5 m long, oval
to elliptic; petiole 0.51 m long, often ferrugineous, armed with small to stout
branched prickles; rachis 0.51 m long, with 10-30 subopposite pairs of leaflets,
armed with prickles in the lower third. Leaflets chartaceous to papyraceous,
oblong-lanceolate to lanceolate, acuminate at apex, denticulate in upper half,
1520 cm long, 24 cm wide. Pollen strobili 26, cylindrical, brown, 610 cm long,
12 cm in diameter; peduncle 815 cm long. Ovulate strobili usually solitary, dark
Fig. 14.3. Zamia amazonum D.W. Stevenson. (A) Habit. (B) Habit. (C) Detail of a
leaflet and rachis. (D) Ovulate strobilus.
ch14.qxd 28/11/03 4:26 pm Page 177
red-brown, 1015 cm long, 35 cm in diameter; peduncle 58 cm long. Seeds with

a red outer fleshy layer, ovoid, 1 cm long, 0.5 cm in diameter.
DISTRIBUTION A widespread species found in Amazonian areas of Brazil,

Colombia, Ecuador, Peru (Fig. 14.1), Venezuela and disjunct in the Choc of
Colombia.
ECUADOR. NAPO: Limoncocha, Maddison et al. 5419 (AAU, F, FTG, K,
QCA, SEL, US); Jatun Sacha, D. Neill 9461 (MO). PERU. LORETO: Fundo
Ciudadela, Rio Itaya, J. Schunke 14240 (F, NY): Iquitos, R. Tryon & A. Tryon 5185
(F).
ETYMOLOGY The specific epithet refers to the wide distribution of this species
throughout the upper Amazon basin.
DISTINGUISHING FEATURES Zamia amazonum is a variable species that most closely

resembles Z. lecointei Ducke and Z. muricata Willdenow. However, it can be distin-
guished from these two species on the basis of the rachis being densely tomentose
when young and being densely prickled with the prickles often branched. The
seed cone is smaller, being < 15 cm, including the peduncle, than the seed cones
of Z. lecointei and Z. muricata, which are longer than 15 cm.
CONSERVATION STATUS Fairly common locally throughout its range. However,

Z. amazonum is known from only two collections in Peru and two collections in
Ecuador and is not known from Bolivia, but is to be expected in the Amazonian
region of the latter. Not yet listed in the 1997 IUCN Red List of Threatened Plants
but the status throughout its range would be R,II,R.
2. ZAMIA BOLIVIANA (Brongniart) A. de Candolle (1868) Prodromus 16(2),

540
TYPE: BOLIVIA. Coteaux de Sn. Xavier, A. dOrbigney 663 (LECTOTYPE: P;

ISOLECTOTYPES: P, W; fide Stevenson and Sabato, 1986).
Ceratozamia? boliviana Brongniart (1846) Annales des Sciences Naturelles, Series 3,

Volume 5, 9.
Z. brongniartii Weddell (1849) Annales des Sciences Naturelles, Series 3, Volume 13, 249
and Planche III, nomen illegitimum, superfluous name for Ceratozamia boliviana
Brongniart.
Stem subterranean and tuberous, 310 cm in diameter. Cataphylls from 12

cm long, sheathing at first, with a pair of inconspicuous stipules. Leaves 25,
0.30.7 m long; petiole smooth; rachis bearing 520 pairs of opposite to sub-
opposite leaflets. Leaflets linear-lanceolate to lanceolate, 1225 cm long
and 0.51.5 cm wide, margins entire and strongly revolute, rarely with 210
ch14.qxd 28/11/03 4:26 pm Page 178
indistinct teeth in the upper third, somewhat pungent apically. Pollen strobili
usually solitary, cylindrical, slender, pedunculate, tan to brown, 0.51.5 cm long
and 0.60.8 cm in diameter, peduncle 24 cm long. Ovulate strobili oblong, acumi-
nate apically, pedunculate, brown, 1215 cm long and 35 cm in diameter,
peduncle 810 cm long. Seeds with a red to orange-red sarcotesta, ovoid, 12 cm
long.
DISTRIBUTION Known only from Beni, Cochabamba and Santa Cruz in Bolivia
(Fig. 14.1) and a single collection from adjacent Mato Grosso of Brazil, where it
grows in sandy well-drained soils. It grows essentially under conditions similar to
those for Caribbean species of Zamia.
BOLIVIA. BENI: U. Patino s.n. (GH); El Porvenir, C. Paz & E. Polanco 205
(MO); J.C. Solomon 6244 (MO); Trinidad, E. Werdermann 2486 (MO); Lake
Rogoaguado, C.E. White 1543 (GH, K, NY, P, PH). COCHABAMBA: Campero,
C. Antezana 1029e (BOLV). SANTA CRUZ: Lomerio, J.R. Abbott 16303 (USZ);
Cerro Puquio Norte, J.R. Abbott & A. Jardim 16630 (USZ); Huanchaca II, A.
Carrinet et al. 406 (USZ, MO); M. de Castelnau s.n. (P); San Matias, R. Guilln et al.
2249 (USZ); Porongo, A. Henderson & M. Morales 767 (NY); Nuflo de Chavez, M.
Hopkins 117 (NY), 175 (NY), 204 (NY); Velasco, M. Nee 41141 (MO, NY), 41468
(NY); Nuflo de Chavez, T. Killeen 1164 (F, MO, NY), 2223 (F, MO, NY); Nuflo de
Chavez, A. Krapovickas & A. Schinini 32209 (F, G, MO); S. Moore 389 (BM);
Aserradero Cerro Pelao, M. Saldas et al. 2965 (USZ); San Ignacio, E. Schmidt 183
(M); J. C. Solomon & S. Urcullo 14152 (MO); Lomerio, Marisol, Toledo et al. 526
(USZ); Mato Grosso, Weddell 3331 (P). BRAZIL. MATO GROSSO: Bantel et al.
1342 (MG).
ETYMOLOGY Based upon its presumed endemism in Bolivia at the time of its
description.
DISTINGUISHING FEATURES This species is very similar in appearance to

Caribbean species of Zamia, particularly Z. integrifolia Linnaeus filius in Aiton and
Z. portoricensis Urban. It may be distinguished from these two species by its more
or less sharp leaflet tips, strongly revolute leaflet margins and brown cones.
SYSTEMATIC AND NOMENCLATURAL HISTORY Zamia boliviana was originally described

as Ceratozamia boliviana by Brongniart, most probably because he had no cones of
the species at the time of his description. In fact, he questioned the assignment of
this species to his newly described genus, Ceratozamia. Consequently, when cones
became available, Weddell redescribed the species as Zamia brongniartii, while at
the same time citing Ceratozamia boliviana Brongniart as a synonym as well as citing
its type. Because the specific epithet was available at that time for transfer to
Zamia, Weddell created a superfluous name. This was later corrected by A. de
Candolle. Also, some material cited by Weddell as Z. brongniartii was thought by
him to be from the Mato Grosso of Brazil but is in fact from Bolivia (Sabato,
1990). However, a single collection, Bantel et al. 1342, of this species from the
ch14.qxd 28/11/03 4:26 pm Page 179
Mato Grosso of Brazil is in the collectionsof the Museo Goeldi (MG).
CONSERVATION STATUS This species is commonly collected but there is no in-for-

mation on the size or extent of the populations and thus it is hard to assess
its status. It may well be collected simply because it is a novelty. This leads
one to think that it is very rare. 1997 IUCN Red List of Threatened Plants Category
R,II,R.
3. ZAMIA DISODON D.W. Stevenson & Sabato (Fig. 14.4), (2001) Flora de
TYPE: COLOMBIA. Antioquia, D. Restrepo et al. s.n. (HOLOTYPE: COL;

ISOTYPE: NY).
Fig. 14.4. Zamia disodon D.W. Stevenson & Sabato. (A) Habit. (B) Leaflets and
rachis.
ch14.qxd 28/11/03 4:26 pm Page 180
Stem to 50 cm, 58 cm in diameter. Cataphylls triangular basally, linear-lance-

olate apically, 36 cm long, 12 cm wide. Leaves 24, erect or slightly recurved,
oblong, 50 cm long; petiole terete, to 25 cm long, sparsely armed with small
prickles; rachis terete, usually unarmed, to 25 cm long, 35 pairs of leaflets.
Leaflets membranous, elliptic, subfalcate, acute at apex, margins serrate and
always with some doubly serrate teeth near the apex, the median ones 1220 cm
long, 610 cm wide. Strobili unknown.
DISTRIBUTION Known only from the type locality and another nearby popula-
tion in Antioquia, Colombia, and disjunct to a single locality in Peru (Fig. 14.1).
PERU. HUNUCO: Pachitea, Llullapichis, R. Dressler 4938, 4939 (NY);
Puerto Inica, R. Foster 8688 (MO).
ETYMOLOGY The specific epithet is derived from the doubly serrate leaflets.
DISTINGUISHING FEATURES Zamia disodon is the only cycad with doubly toothed
leaflet margins. In addition, it is the only Zamia besides Z. hymenophyllidia D.W.
Stevenson with transparent leaflets. However, Z. hymenophyllidia has entire to
minutely toothed margins. In many respects, Z. disodon resembles an acaulescent
to short trunked Z. obliqua A. Braun with transparent biserrate leaflets.
CONSERVATION STATUS Because Zamia disodon has been so infrequently collected

it is assumed that the species is very rare. However, it is apparently available in
local markets in Colombia (Ian Turner, Zimbabwe, 1996, personal communica-
tion). Because cones have not been seen, reproduction is assumed to be limited.
Rodrigo Bernal (Colombia, 2002, personal communication) believes that the
habitat in Colombia is severely threatened. In Peru, very little is known about this
species other than the fact that two sterile collections have been made. More
fieldwork is needed to assess the situation. Not yet listed in the 1997 IUCN Red
List of Threatened Plants but the status throughout its range would be I,II,I.
4. ZAMIA GENTRYI Dodson (1998) Novon 8, 1214 and Fig. 1af
TYPE: ECUADOR. Esmeraldas, C.H. Dodson & A.H. Gentry 17520A (HOLO-
TYPE: QCNE; ISOTYPES: AAU, MO, NY, RPSC, SEL, U).
Stems up to 1.5 m long, generally prostrate when terrestrial or U-shaped

when growing on fallen logs, 515 cm in diameter, smooth. Cataphylls up to 12 cm
long and of two forms, either elongate triangular or with a long acuminate apex.
Leaves 59, 0.52.5 m long; petiole 4590 cm long, very densely prickled; rachis
with 1525 (less in juvenile plants) pairs leaflets, densely prickled in lower half.
Leaflets long-lanceolate and subfalcate, chartaceous to membranous, margins
entire, attenuate basally, acute apically, the larger median ones 1040 cm long,
3.56 cm wide. Pollen strobili 17, wine-red, cylindrical, 3040 cm long, 34 cm in
ch14.qxd 28/11/03 4:26 pm Page 181
diameter; peduncle 2025 cm long, decumbent. Ovulate strobili singular, red-

brown, narrowly ovoid to cylindrical, 2030 cm long, 812 cm in diameter,
peduncle short. Seeds obovoid, 3.5 cm long, 1.7 cm in diameter, pink to red.
DISTRIBUTION Endemic to Ecuador (Fig. 14.1), where it grows at altitudes from

300 to 800 m in extremely wet premontane (cloud) forest on compact clay soils.
Although having a trunk, Zamia gentryi is generally prostrate along the soil surface.
Contrary to some reports, this species is not epiphytic.
ECUADOR. CARCHI: Canton Tulcan, G. Tipaz et al. 1878 (QCNE).
ESMERALDAS: Lita to San Lorenzo, B. llgaard et al. 98726 (AUU, QCNE); C.
Dodson et al. 17107 (MO, QCNE), 19098 (QCNE); H. Luther et al. 1235 (SEL).
ETYMOLOGY Named in honour of Alwyn H. Gentry, intrepid botanist and

botanical explorer of the neotropics who had an early death as the result of an
aeroplane crash in Amazonian Peru.
DISTINGUISHING FEATURES The smooth prostrate trunks, combined with several

leaves up to 2.5 m long bearing several narrow entire leaflets, serve to distinguish
Zamia gentryi from other species of Zamia. This species appears most similar in
many aspects to Z. chigua Seemann from the Choc of Colombia and north-
western Panama. Most plants of the latter in fact have semi-prostrate trunks with
numerous adventitious roots on the bottom surface and are without any subter-
ranean stem. Both species have very thin leaflets with entire margins and appear
slightly grooved, but not to the extent of Z. skinneri Warszewicz ex A. Dietrich and
its allies. The two species differ in that Z. gentryi has neither as many leaflets nor
imbricate leaflets as in Z. chigua. Also, the pollen cones of the former are larger
than those of the latter and the leaflets are much broader. Juvenile plants of both
species are quite difficult to distinguish from one another.
CONSERVATION STATUS This species is known only from an area that is very
poorly known botanically and is presumed to have a wider distribution. For the
present time it is considered very rare pending further data. Not listed in the 1997
IUCN Red List of Threatened Plants.
5. ZAMIA HYMENOPHYLLIDIA D.W. Stevenson (Fig. 14.5), (2001) Flora de

TYPE: COLOMBIA. Amazonas, D. Cardenas et al. 10089 (HOLOTYPE: COAH;

ISOTYPE: NY).
Stem subterranean and tuberous, 24 cm in diameter. Cataphylls triangular

basally, linear-lanceolate apically, membranous, 23 cm long, 0.51 cm wide. Leaves
25, erect, oblong, 3070 cm long; petiole terete, to 35 cm long, armed with very
small prickles; rachis terete, usually unarmed, to 2035 cm long, 410 pairs of
ch14.qxd 28/11/03 4:26 pm Page 182
Fig. 14.5. Zamia hymenophyllidia D.W Stevenson. (A) Habit. (B) Leaflet. (C) Pollen
strobilus. (D) Microsporophyll, abaxial view. (E) Microsporophyll, adaxial view. (F)
Ovulate strobilus.
leaflets. Leaflets membranous, elliptic to elliptic-lanceolate, long acuminate apically,

cuneate to oblique basally, margins entire or rarely with a few diminutive apical
teeth, 1215 cm long, 12 cm wide. Pollen strobili brown, ovoid, 13 cm long, 0.10.3
cm in diameter; peduncle 1015 cm long; microsporophylls with both abaxial and
a few (13) adaxial sporangia. Ovulate strobili dark red-brown, cylindrical to ovoid-
cylindrical, to 5 cm long, 34 cm in diameter; peduncle to 15 cm long, 0.5 cm in
diameter. Seeds red to orange-red, 11.2 cm long, 0.50.8 cm in diameter.
DISTRIBUTION Only four small populations are known, three in south-eastern

Amazonian Colombia and one in adjacent north-eastern Peru (Fig. 14.1).
PERU. LORETO: Maynas, Pebas, Brillo Nuevo, M. Balick et al. 1035 (NY);
Plowman et al. 6773, 7065, 7255, 7256 (F, GH).
ETYMOLOGY The specific epithet refers to the extremely thin and almost trans-
parent leaflets.
DISTINGUISHING FEATURES Zamia hymenophyllidia is most similar to the Colombian

species Z. melanorrhachis D.W. Stevenson. Both species are unique among the
cycads in having extremely long thin peduncles (up to 30 cm long and 13 mm
in diameter) terminated by very small pollen cones (13 cm long and 13 mm in
ch14.qxd 28/11/03 4:26 pm Page 183
diameter). Zamia hymenophyllidia has membranous, elliptic, long acuminate leaflets

in contrast to the papyraceous, lanceolate, generally acute leaflets of Z. melanor-
rhachis. The leaflet margins of Z. hymenophyllidia are generally entire and, if
toothed, the teeth are inconspicuous and apical. In contrast, Z. melanorrhachis has
well-developed marginal teeth. Living plants of Z. melanorrhachis have a dark
purple to black rachis, a feature not found in Z. hymenophyllidia.
CONSERVATION STATUS Zamia hymenophyllidia was discovered only in the past 3

years in three small populations. The plants are reproductive and produce seeds
and seedlings. Little is known about its biology and distribution but the paucity
of collections indicates that it is rare. Not yet listed in the 1997 IUCN Red List of
Threatened Plants but the status throughout its range would be R,II,R.
6. ZAMIA LECOINTEI DUCKE (Fig. 14.6), (1915) Archivos do Jardim Botnico

do Rio de Janeiro 1, 910 and Tabula 1
TYPE: BRAZIL. Par, Rio Erepecur, 2 October 1913, Ducke MG 15027

(HOLOTYPE: MG). Z. obidensis Ducke, (1922) Archivos do Jardim Botnico do Rio de
Janeiro 3, 20.
TYPE: BRAZIL. Par, Rio Branco de Obidos, March 1918, Ducke s.n. (Museu
Goeldi 17015) (HOLOTYPE: MG; ISOTYPE: F). Z. ulei subsp. lecointei (Ducke)
Ducke, (1935), Archivos do Instituto Biologia Vegetale do Rio de Janeiro 2, 2728. Z. jiri-
jirimensis R.E. Schultes, (1953) Mutisia 15, 25 and Fig. p. 5 s.n.
TYPE: COLOMBIA. Amazonas: Rio Apaporis, Raudal de Jirijirimo, March
1951, R.E. Schultes 12101 (HOLOTYPE: GH2 sheets)
Stem subterranean and tuberous, 58 cm in diameter. Cataphylls triangular

basally, linear-lanceolate apically, 36 cm long, 12 cm wide. Leaves 24, erect or
slightly curved, oblong, 12 m long; petiole terete, to 75 cm long, sparsely armed
with stout prickles; rachis terete, usually unarmed, to 1 m long, 3040 pairs of
leaflets. Leaflets subcoriaceous to coriaceous, linear-lanceolate, subfalcate, long
acuminate at apex, margins entire or rarely with 23 indistinct teeth near the
apex of lower margin, the median ones 3037 cm long, 12 cm wide. Pollen stro-
bili 26, cylindrical, cream to light brown, 610 cm long, 12 cm in diameter;
peduncle 1020 cm long. Ovulate strobili usually solitary, brown, 1015 cm long,
35 cm in diameter; peduncle 58 cm long. Seeds with a red outer fleshy layer,
ovoid, 3 cm long, 2 cm in diameter. 2n = 16 (Caputo et al., 1996).
DISTRIBUTION In Amazonas and Bolivar, Venezuela, and Par, Brazil, north of

the Amazon River to central Colombia and Brazil. For Bolivia, Ecuador and
Peru, known from only a single collection in Peru (Fig. 14.1).
PERU. LORETO: Ucayali, J. Schunke 14269 (F, NY).
ETYMOLOGY The specific epithet honours Paul LeCointe who lived in Obidos,
Brazil and who accompanied Adolf Ducke when the species was discovered.
ch14.qxd 28/11/03 4:27 pm Page 184
Fig. 14.6. Zamia lecointei Ducke. (A) Habit. (B) Leaflet. (C) Ovulate strobilus.
DISTINGUISHING FEATURES Zamia lecointei is most similar to Z. muricata. The two

taxa share a common distribution pattern but are not sympatric and prefer dif-
ferent habitats. Zamia lecointei grows in open dry areas of sandy to pure sand soils
whereas Z. muricata grows as an understorey plant in mesic conditions in wet clay
soils. Consequently, Z. lecointei has subcoriaceous to coriaceous, linear-lanceolate,
subfalcate leaflets that are entire or with only 13 indistinct apical teeth. By con-
trast, Z. muricata has chartaceous to papyraceous, oblong-lanceolate to lanceolate,
ch14.qxd 28/11/03 4:27 pm Page 185
inequilateral leaflets that are clearly denticulate in the upper half of both
margins. In the open in direct sun, or as an understorey in secondary vegetation,
the leaflets often become narrower and quite coriaceous with entire margins, but
in more shaded conditions the leaflets are wider, thinner and show well-devel-
oped serrations in the upper third. Plants growing in extremely dry, sandy soil
were described as Z. jirijirimensis.
SYSTEMATIC AND NOMENCLATURAL HISTORY Ducke (1915) described Zamia lecointei

and mentioned another probable new Zamia, which he described as Z. cupatiensis
in 1922. The type, illustration (Ducke, 1915, tabula 1) and description of Z. lecoin-
tei all show linear-lanceolate leaflets. Duckes (1915) concept of Z. ulei is present-
ed in tabula 2 of the same paper; this bears little resemblance to the type or
description (Dammer, 1907) of Z. ulei Dammer, which has ovate-lanceolate to
elliptic leaflets. As a result of his misconception, Ducke (1922) described Z. cupa-
tiensis Ducke. The description and photograph of Z. cupatiensis (Ducke 1922,
tabula 2) match the photograph and isotypes of Z. ulei. In the same paper, Ducke
(1922) also described Z. obidensis Ducke which was intermediate between his Z.
lecointei and his Z. cupatiensis. Because Ducke mistakenly thought that Z. ulei had
linear-lanceolate leaflets, he (Ducke, 1935) considered Z. lecointei to be a sub-
species, Z. ulei subsp. lecointei. This problem was compounded by Schuster (1932).
As a result, Schultes (1953) redescribed Z. lecointei as Z. jirijirimensis. However, the
type specimens for these two names are identical. Thus, Z. cupatiensis is a synonym
of Z. ulei, and Z. jirijirimensis is a synonym of Z lecointei. Moreover, Z. obidensis is a
shade and/or juvenile form of Z. lecointei.
CONSERVATION STATUS Zamia lecointei is most common in Bolivar, Venezuela, in

both pristine and disturbed habitats. Seed set appears high and the populations
are not at present threatened by development or destruction. Collections in
Colombia indicate that it is locally common with good seed set. It does well in
minor disturbed situations. However, because Z. lecointei is known from only a
single collection from Peru, it is assumed that the species is very rare in Peru and
does not occur in either Ecuador or Bolivia even though suitable habitat does
occur in the latter two countries. Not listed in the 1997 IUCN Red List of Threatened
Plants but the status throughout its range would be R,II,R.
7. ZAMIA MACROCHIERA D.W. Stevenson, species nova (Fig. 14.7)
TYPE: PERU. LORETO: Maynas, Pebas, Rio Amiyacu, D. Stevenson 1160

(HOLOTYPE: NY; ISOTYPES: AMAZ, FTG, NY, U)
Zamia manicata Regel praesertim collo glanduliformi ad foliolorum basi sito oblongo-
elliptico e basi cuneato spiraliter adscendenti ultra medium serrato, distaliter concavo inter
alia diversa.
Stem subterranean, subcylindric to cylindric, 1020 cm in diameter. Cataphylls

triangular basally, long acuminate apically, 38 cm long, 12 cm wide. Leaves 13,
ch14.qxd 28/11/03 4:27 pm Page 186
Fig. 14.7. Zamia macrochiera D.W. Stevenson. (A) Habit. (B) Leaflets. Gland-like
structure at the juncture of petiole and lamina: (C) adaxial view, (D) abaxial view,
(E) apical view and (F) basal view. (G) Ovulate strobilus.
0.53.5 m long; petiole 0.22 m long, slightly to densely prickled; rachis often
with prickles in lower third, with 1030 pairs of leaflets. Leaflets with distinct peti-
olule and abaxial gland-like collar or flap that curves upward to touch the lamina
forming a tunnel at the juncture, oblong to long-elliptic, margins serrate in upper
third, cuneate basally, acute to acuminate apically, the larger median leaflets
2045 cm long, 515 cm wide. Pollen strobili cream to tan, cylindrical, 46 cm
long, 11.5 cm in diameter; peduncle 1530 cm long. Ovulate strobili wine-red to
dark red-brown, cylindrical to ovoid-cylindrical, 1015 cm long, 47 cm in diam-
eter. Seeds red, 11.5 cm long, 0.50.8 cm in diameter. 2n = 18 (Aldo Moretti,
Italy, 1993, personal communication).
DISTRIBUTION Zamia macrochiera occurs in rainforest and secondary forest from

100 m to 300 m in a limited area of Peru (Fig. 14.2).
PARATYPES: PERU. LORETO: Maynas, Rio Napo, M. Mathias s.n. (MO,
NY); Pebas, T. Plowman et al. 6937 (F, GH), 7254 (F, GH, K, NY, U), 7275 (F, GH,
K).
ch14.qxd 28/11/03 4:27 pm Page 187
ETYMOLOGY The specific epithet refers to the large gland-like collar at the
leaflet base.
DISTINGUISHING FEATURES The petiolule and gland-like collar occur among

cycads only in Zamia manicata Linden ex Regel (Stevenson, 1990, 2001) and Z.
macrochiera. However, these structures are not present in seedling leaves, are only
diminutively present in juvenile leaves but become distinct in adult leaves of both
species. Also, in transplanted adult plants, the structures may not be produced in
the first set of leaves after transplanting. The most obvious difference between the
two species is in the morphology of the gland-like collar. The collar in Z. manica-
ta is a rim-like structure, in contrast to that of Z. macrochiera where it is a well-
developed flap of tissue that curves up to meet the lamina forming a tunnel (Fig.
14.7). In addition, both the leaves and ovulate strobili of Z. macrochiera are much
larger than those of Z. manicata. This is the undescribed species discussed by
Wrinkle (1993).
CONSERVATION STATUS Zamia macrochiera has been collected only four times in the
past 25 years and all collections have been near the type locality. Although some
seedlings were seen at the type locality, only one seed cone has been observed.
However, the species appears to be threatened by continual habitat disturbance.
The type locality was being converted from secondary forest into a communal
garden. Not yet listed in the 1997 IUCN Red List of Threatened Plants.
8. ZAMIA POEPPIGIANA Martius & Eichler, (1863) Flora Brasiliensis 4(1),

414416 and Tabula 109
TYPE: PERU. Maynas Alto, Toache River, 1830, Poeppig s.n. (LECTOTYPE: F
ex Herbario Musei Palatino Vindobensis; ISOLECTOTYPE: GH ex Herbario
Musei Palatino Vindobensis; fide Stevenson and Sabato, 1986).
Z. lindenii Regel ex Andr, (1875) LIllustration Horticole 22, 23 and Planche 195.
LECTOTYPE: Planche 195 in LIllustration Horticole 23, (1875) (fide Stevenson and
Sabato, 1986).
Aulacophyllum lindenii (Regel ex Andr) Regel, (1876) Gartenflora 25, 141.
Z. baraquiniana Regel, (1876) Acti Horti Petropolitani 4(4), 308309.
TYPE: ex Horto Petropolitano, Regel s.n. (HOLOTYPE: LE).
Z. wielandii Schuster, (1932) Das Pflanzenreich IV.1, Heft 99, 149, nomen illegitimum,
superfluous name for Z. baraquiniana Regel.
Stem arborescent, to 3 m tall and 25 cm in diameter. Cataphylls cuneate

basally and acuminate apically, to 2 cm wide and 4 cm long. Leaves 1015, 13 m
long; petiole with numerous small prickles, 3070 cm long; rachis with prickles in
lower third, 2040 pairs of leaflets. Leaflets long-lanceolate, falcate basally, acumi-
nate to acute apically, margins strongly spinulose in upper third with serrations at
nearly 90 and 0.51 cm apart; median leaflets 1540 cm long, 24 cm wide.
ch14.qxd 28/11/03 4:27 pm Page 188
Pollen strobili cream to tan, long cylindrical, 2030 cm long, 35 cm in diameter;

peduncle 58 cm long; microsporophylls with sterile tip composed of six steeply
inclined facets surrounding a centrally depressed terminal facet. Ovulate strobili tan
to brown, cylindrical to ovoid-cylindrical, 2540 cm long, 1015 cm in diameter.
Seeds red, oblong, distinctly flattened, 11.5 cm long, 0.50.8 cm in diameter. 2n
= 18 (Norstog, 1980).
DISTRIBUTION Known from the coastal plains and foothills of the Cordillera
Occidental in Ecuador and rich humus soils of Acre, Brazil and south-western
Colombia and eastern Peru (Fig. 14.2).
ECUADOR. E. Andr 3687 (F, GH, K, NY, P, US); F. Barclay 709 (US); H.
Eggers 14034 (GH); C. Horn s.n. (US); F. Lehmann 658 (G); A. Rimbach 85, (F), 252
(US). AZUAY: Huigra, J.N. Rose & G. Rose 22613 (US); Saraguro, J. Steyermark
52786 (F, NY, P). EL ORO: Balslev et al. 62526 (AAU, COL, QCA). BOLIVAR:
Balsapamba, T. Lockwood 825 (GH). ESMERALDAS: Fila de Bilsa, A. Gentry et al.
72867 (MO); Coronel C. Concha, P. Maas et al. 2920 (U); Esmapaidas, P. Maas et
al. 2020 (U); Rio Mira, M.G. Patwa s.n. (US). GUAYAS: Naranjal, G. Harling & L.
Andersson 19417 (GB); F. Lehmann 5292 (K). LOS RIOS: Jaunche, C. Dodson & A.
Gentry 12698, 7991 (MO, SEL); C. Dodson et al. 7991 (MO, SEL); Rio Palenque,
A. Gentry & C. Dodson 18045 (MO); K. Norstog 805 (FTG). PICHINCHA:
Congoma Grande, L. Kvist 40203 (AAU, MO, NY); Manab, G. Harling & L.
Andersson 24751 (GB); Toache-Las Pampas, C. Dodson & A. Gentry 9715, 13694
(MO, SEL). PERU. AMAZONAS: Bagua, T. Plowman 5536 (GH). HUNUCO:
Tingo Maria, T. Plowman et al. 7575 (F, GH, NY), 11232 (F, GH, INPA, NY);
LORETO: Lower Rio Huallaga, L. Williams 5373 (F, US); Maynas, L. Williams
3794 (F). MADRE DE DIOS: Tambopata, M. Alexiades 1322 (NY); Cocha Cashu
Station, R. Foster 5852 (F). SAN MARTIN: Campanilla, J. Schunke 4200 (F);
Tarapoto, F. Montes 57 (F, NY); T. Plowman 7601 (F); M. Rimachi Y. 5185 (NY),
5356 (MO, NY); Tocache Neuvo, A. Gentry et al. 25710 (MO, NY); T. Plowman &
J. Schunke 11543 (F). TUMBES: Tumbes, C. Daz et al. 5117 (MO); Matapalo, C.
Daz et al. 7500 (MO); D. Simpson & J. Schunke 428 (COL, F, GH, NY, US).
ETYMOLOGY The specific epithet honours Eduard Poeppig, who first collected
the species in Peru.
DISTINGUISHING FEATURES Zamia poeppigiana is readily identified by its falcate

linear leaflets with evenly spaced teeth at nearly right angles to the margin. Zamia
poeppigiana and Z. lindenii Regel ex Andr are here considered conspecific. The
types of both are identical but the former is from the lower Pacific Andean slopes
of Ecuador whereas the latter is from the lower Amazonian slopes of Peru and
Colombia. The seeds of Z. poeppigiana are unique within Zamia because of their
somewhat flattened oblong shape. By contrast, the seeds of Z. lindenii are report-
ed to be unflattened and oval in shape, as in all other known species of Zamia, but
observations on both fresh seeds and herbarium material show a flattened oblong
shape.
ch14.qxd 28/11/03 4:27 pm Page 189
CONSERVATION STATUS Zamia poeppigiana is widely but sporadically distributed.

Local populations exhibit good seed set and seedling establishment. Unlike Z.
obliqua, Z. poeppigiana appears to produce seed cones when the stems are not yet
arborescent, thus reducing the threat from habitat destruction. Not yet listed in
the 1997 IUCN Red List of Threatened Plants.
9. ZAMIA ROEZLII Linden, (1873) Linden Catalogue des Plantes Nouvelles

No. 90, 10
NEOTYPE: Planches 133134 in LIllustration Horticole 20 of 1873, (fide Stevenson

and Sabato, 1986).
Aulacophyllum roezlii (Linden) Regel, (1876) Gartenflora 25, 141.
Stem arborescent, to 7 m tall. Leaves 510, 13 m long; petiole 0.51 m long,

sparsely to densely prickled; rachis with 1020 pairs of leaflets, occasionally with
a few prickles in the lower third. Leaflets linear-lanceolate, grooved between the
veins on the adaxial surface, cuneate basally, acute apically, margins entire, 3050
cm long, 1215 cm wide. Pollen strobili cream to tan, cylindrical to elongate-cylin-
drical, 2030 cm long, 46 cm in diameter. Ovulate strobili brown, short peduncu-
late, cylindrical to ovoid-cylindrical, 3060 cm long, 1020 cm in diameter. Seeds
red, ovoid, 1.52.5 cm in diameter. 2n = 22, 24, 25, 26 (Norstog, 1981).
DISTRIBUTION Along coastal Choc, Colombia, and contiguous coastal

Ecuador (Fig. 14.2).
ECUADOR. ESMERALDAS: Zapollo Grande, A. Barfod & F. Skov 60116
(AAU, NY, QCA, QNA); San Lorenzo, C. Dodson 19622 (RPSC); C. Dodson & A.
Gentry 19048 (RPSC); C. Dodson & T. Neudecker 19083 (QCNE, RPSC); H.E. Luther
s.n. (MO); David Neill et al. 11792 (MO, QCNE); F Nicolalde & J. Palacios 234
(QCNE); Reserva Awa, G. Tipaz et al. 2242 (0MO, QCNE); Cotacachi Cayapas,
G. Tipaz et al. 2255 (MO, QCNE); M. Tirado et al. 646 (QCNE); Communa
Corriente Grande, A. Yanez 1413 (QCNE). NAPO: Jatun Sacha, cultivated from
seed from Reserva Awa, Esmeraldas, D. Neill 11160 (MO, QCNE).
ETYMOLOGY The specific epithet honours Benedict Roezl, prodigious plant

collector who travelled throughout Mexico, Central America and South America
in the latter half of the 19th century and who first collected the species.
DISTINGUISHING FEATURES Zamia roezlii is the largest of the South American

Zamia plants and is unmistakable with its massive trunk, leaves and strobili. This
species appears to be part of a complex composed of Z. dressleri D.W. Stevenson,
Z. neurophyllidia D.W. Stevenson and Z. skinneri in Panama, Z. amplifolia hort. Bull
ex Masters and Z. roezlii in the Choc of Colombia and adjacent Ecuador, and
Z. wallisii A. Braun in northern Antioquia, Colombia. This grouping is based
upon the common feature of a plicate appearance of the leaflets. This character,
ch14.qxd 28/11/03 4:27 pm Page 190
along with simultaneous rather than sequential leaf production, prompted Regel
(1876) to erect the segregate genus Aulacophyllum. However, no other synapomor-
phies have been found to support Aulacophyllum as a genus (Sabato, 1990).
Moreover, the distinction between simultaneous and sequential leaf production
in Zamia seems problematic at best. However, Z. roezlii is readily distinguishable
from other species with deeply grooved leaflets by its falcate, linear and entire
leaflets.
CONSERVATION STATUS Zamia roezlii is locally abundant and produces very large
seed cones and numerous seeds and seedlings. However, seed cones are produced
only by arborescent plants and, under continual disturbance, seed production will
be severely diminished, resulting in high vulnerability for the species. 1997 IUCN
Red List of Threatened Plants Category R,II,V.
10. ZAMIA ULEI Dammer, (1907) Verhandlungen des Botanischen Vereins

der Provinz Brandenburg 47, 117118
TYPE: BRAZIL. Cachoeira oberer Rio Jura, May 1901, E. Ule 5523
(LECTOTYPE: HBG; ISOLECTOTYPES: L, G, MG, K, F-fragment; fide
Stevenson and Sabato, 1986).
Z. cupatiensis Ducke, (1922) Archivos do Jardim Botnico do Rio de Janeiro 3, 20 and
Planche 1.
LECTOTYPE: Planche 1 as above.
Stem subterranean and tuberous, tapering distally, 46 cm in diameter. Leaves

24, 11.5 m long, broadly ovate; petiole to 1 m long, terete, generally armed
with numerous prickles; rachis to 50 cm long, with 36 pairs of subopposite
leaflets, often armed with prickles in the lower third. Leaflets sessile, papyraceous
to subcoriaceous, oblong-elliptic to elliptic-ovate, sometimes slightly falcate,
obtuse and constricted at base, acute to short-acuminate at apex, 1215 teeth on
each margin in the upper one-half, the median ones 1220 cm long, 610 cm
wide. Pollen strobili usually 25, tan, cylindrical, 610 cm long, 12 cm in diame-
ter; peduncle 68 cm long. Ovulate strobili usually solitary, brown, cylindrical,
1820 cm long, 46 cm in diameter; peduncle 810 cm long. Seeds red, ovoid to
oblong, 15 mm long, 8 mm in diameter. 2n = 26 (Caputo et al., 1996).
DISTRIBUTION North of the Amazon River in Par and Rio Branco, Brazil and
disjunct to the bordering regions of western Brazil and eastern Colombia,
Ecuador and Peru (Fig. 14.7).
ECUADOR. NAPO: Pastaza, D. Neill et al. 11136 (QCNE). PERU.
LORETO: A. Gentry et al. 77429 (MO); Maynas, I. Cabrera 3351 (GH); T. Croat
18675 (MO); D. Daly et al. 6189 (AMAZ, NY); D. Simpson & J. Schunke 800 (F, GH,
US); R. Schultes 8366 (GH); J. Schunke 14029 (F); R. Vsquez 1536, 17594 (MO), R.
Vsquez & N. Jarmillo 15955 (MO); D. Stevenson 1161 (AMAZ, FTG, MO, NY, U).
ch14.qxd 28/11/03 4:27 pm Page 191
MADRE DE DIOS: Iberia, R. Seibert 2158 (US); Tambopata, V.P. Baca et al. 143
(MO).
ETYMOLOGY The specific epithet honours Eduard Ule, an Amazonian plant

collector of the latter half of the 19th century, who first collected the species in
Brazil.
DISTINGUISHING FEATURES Zamia ulei is similar to Z. obliqua but the latter always
has obliquely inserted and basally falcate leaflets while Z. ulei always has sym-
metrical leaflets. Moreover, adult plants of Z. obliqua are arborescent in contrast
to the acaulescent Z. ulei.
CONSERVATION STATUS Zamia ulei is fairly common in sporadic populations

throughout its range. Seed set and seedling establishment appear to be high. Not
yet listed in the 1997 IUCN Red List of Threatened Plants but the status throughout
its range would be R,II,R.
11. ZAMIA UREP Wallnfer, (1996) Linzer Biologische Beitrge 28(2),

10561058 and Fig. 1
TYPE: PERU. HUNUCO: Pachitea, Llullapichis, B. Wallnfer 11215788

(HOLOTYPE: W; ISOTYPES: LZ, USM).
Stem subterranean, to 20 cm long, 2.55 cm in diameter. Cataphylls narrowly

lanceolate, 34 cm long, 58 mm wide. Leaves 14, 4070 cm long; petiole
smooth, 2040 (48) cm long, canaliculate adaxially at the base and flat distally;
rachis without prickles, 815 cm long, 25 pairs of leaflets. Leaflets broadly lance-
olate to elliptic, coriaceous, obliquely cuneate basally, acuminate apically,
margins sharply serrulate in upper third, median leaflets 1020 cm long, 36 cm
wide. Pollen strobili 12, tan to light brown, cylindrical, 36 cm long, 0.61 cm in
diameter; peduncle 1822 cm long, 23 mm in diameter; microsporophylls with
sterile tip composed of six steeply inclined facets surrounding a centrally
depressed terminal facet, sporangia present on both the adaxial and abaxial sur-
faces of the fertile region. Ovulate strobili tan to light brown, cylindrical to ovoid-
cylindrical, 3.55 cm long, 1.52.5 cm in diameter; peduncle 1316 cm long, 35
mm in diameter. Seeds unknown.
DISTRIBUTION Zamia urep is endemic to Peru (Fig. 14.2) and known only from
the vicinity of the type locality where it occurs, from 250 to 750 m on stony hill-
sides with thin soils.
PERU. HUNUCO: Pachitea, Llullapichis, Listabarth 1111589, 1211589,
141293 (USM, W); B. Wallnfer 15241188 (LZ, USM, W).
ETYMOLOGY An anagram for Peru.

ch14.qxd 28/11/03 4:27 pm Page 192
DISTINGUISHING FEATURES Zamia urep can be distinguished by the combination of

symmetrical almost perfectly elliptic to broadly lanceolate leaflets with sharply
serrated margins, acuminate apices and unarmed petioles and rachis. In general
appearance, Z. urep is most similar to Z. obliqua and looks like a smaller version of
the latter. It differs from Z. obliqua in being a smaller plant with subterranean
stems, lacking prickles on the petiole and rachis, and having more prominently
toothed leaflet margins. Similarly, the cones are much smaller and with longer
peduncles. Zamia urep also resembles Z. ulei in having elliptic leaflets and a sub-
terranean stem, but differs from that species in having longer cone peduncles and
no prickles on the petiole or rachis.
CONSERVATION STATUS Zamia urep is known from only seven collections, all from
essentially the same locality, even though over the past 50 years many expeditions
have collected in the area where it is found. Neither seeds nor seedlings of this
species have been seen so nothing is known of its reproductive capacity. Thus it
is assumed that the species is quite rare. Not yet listed in the 1997 IUCN Red List
of Threatened Plants.
Conclusions
As yet there are very few collections of Zamia from Peru. Consequently, of the
eight species known from Peru, six (Z. amazonum, Z. disodon, Z. hymenophyllidia, Z.
lecointei, Z. macrochiera and Z. urep) are each known basically from single localities or
small areas. Of these six species, two (Z. amazonum and Z. lecointei) are widespread
in the upper Amazon basin in Brazil, Colombia and Venezuela. It is expected that
they are much more widespread in Amazonian Peru. Zamia machrochiera and Z. urep
are endemic to Peru and from poorly collected areas in terms of general botani-
cal exploration. Although Z. hymenophyllidia is known also from Colombia, it is in a
contiguous relatively unexplored area botanically and is found in only four small
populations from a relatively small area. Again, it is expected that with more
exploration this species will be found to have a wider distribution than current
data indicate. The presence of Z. disodon in Peru represents a trans-Andean dis-
junct from the Darien of Colombia. It has been collected only twice in Peru and
both collections are from the same locality, and perhaps, population. In Colombia,
Z. disodon is known from only two collections from nearby localities. Because of the
paucity of material from both Colombia and Peru, along with the disjunct distri-
bution of Z. disodon, it is difficult to know if the materials from Colombia and Peru
are truly conspecific. Until further material is collected, including cone material
which is unknown at present, this species will remain enigmatic.
The conservation status of Zamia disodon, Z. urep, Z. macrochiera and Z. hymeno-
phyllidia is difficult to assess because of the lack of collections and population
information. Moreover, there are very few plants of these species in cultivation so
that information from this source concerning coning and growth patterns is not
currently available.
ch14.qxd 28/11/03 4:27 pm Page 193
At present, Zamia boliviana is the only species of the genus known from
Bolivia, where it is widely distributed in the Beni, Cochabamba and Santa Cruz
Provinces. Not only are there numerous collections of Z. boliviana but many of
these have cones, indicating that reproduction in these localities is good. It is
expected that Z. poeppigiana and Z. ulei will be found in Bolivia in the poorly
collected areas contiguous with Brazil and Peru where both species are well
known.
Ecuador has been more extensively explored in terms of cycad collections
than either Bolivia or Peru. As a result more is known about population sizes, dis-
tribution and coning frequencies of Zamia species in Ecuador; much of this
knowledge is the result of the interest and work of Dodson (1994, 1998). From
his work we have a concept of successful reproduction and seedling establishment
of Z. roezlii, Z. gentryi and Z. poeppigiana in Ecuador (Z. poeppigiana as treated here
includes Z. lindenii). Zamia poeppigiana occurs on both sides of the Andes in
Ecuador and Peru. The type for Z. poeppigiana is from the eastern side of the
Andes whereas the type for Z. lindenii is from the western side of the Andes. Both
types are of sterile material. However, in terms of leaf and leaflet morphology,
all material from both sides of the Andes is identical and, in fact, unique within
Zamia in having very uniformly and spreading marginal teeth on distinctly basally
falcate leaves. Cones, as far as known, also appear identical in all plants assigned
here to Z. poeppigiana. On the other hand, Calaway Dodson (USA, 2000, person-
al communication) has noted differences in seed shape, with seeds of eastern
Andean plants being distinctly flattened and elongate in shape as compared with
the oval to more or less spherical shape of the seeds of western Andean plants.
There is a paucity of reproductive material, particularly seeds, in herbarium col-
lections. The seed morphology question needs further study but could provide
support, along with molecular sequence data, for the recognition of two species.
If so, this would most likely represent a vicariant trans-Andean species pair,
because the cycads were there long before the Andes.
Acknowledgements
I offer my sincere gratitude to my cycad colleagues, without whose help this work
would not have been possible. In particular, I would like to thank those who
accompanied me in the field: Aldo Moretti, Luciano Gaudio and Jean-Pierre
Sclavo. I also thank Rupert Barneby for help with the Latin diagnosis. I am
deeply grateful to Amy Melson and Cynthia Armstrong for their excellent illus-
trations of the taxa included here. This work was made possible in part as a result
of National Science Foundation Grants BSR-8607049 and BSR-8796279
awarded to the author.
This chapter is dedicated to the memory of my friend, colleague and
mentor, Knut J. Norstog.
ch14.qxd 28/11/03 4:27 pm Page 194
References
Brako, L. (1993) Cycadaceae. In: Brako, L. and Zarucchi, J. (eds) Checklist of the Plants of
Peru. Missouri Botanical Garden Press, St Louis, Missouri, p. 1.
Caputo, P., Cozzolino, S., Gaudio, L., Moretti, A. and Stevenson, D.W. (1996) Karyology
and phylogeny of some Meso-American species of Zamia (Zamiaceae). American
Dammer, U. (1907) Cycadaceae. Verhandlungen des Botanischen Vereins der Provinz Brandenburg
47, 117118.
Dodson, C. (1994) The Zamias of Equador. The Cycad Newsletter 18(3), 25.
Dodson, C. (1998) A new species of Zamia (Zamiaceae) from Ecuador. Novon 8, 1214.
[Zamia gentryi.]
Ducke, A. (1915) Plantes nouvelles ou peu connues de la rgion Amazonienne.
Cycadaceae. Archivos do Jardim Botnico do Rio de Janeiro 1, 910.
Ducke, A. (1922) Plantes nouvelles ou peu connues de la rgion Amazonienne.
Cycadaceae. Archivos do Jardim Botnico do Rio de Janeiro 3, 20.
Ducke, A. (1935) Plantes nouvelles ou peu connues de la rgion Amazonienne. Archivos do
Jardim Botnico do Rio de Janeiro, Series 8a, 2, 2728.
Regel, E. (1876) Die Cycadeen, deren Gattungen und Arten. Gartenflora 25, 140144.
Sabato, S. (1990) West Indian and South American cycads. In: Stevenson, D.W. (ed.) The
Biology, Structure, and Systematics of the Cycadales. Proceedings of the Symposium CYCAD 87.
Memoirs of the New York Botanical Garden 57, pp.179180.
Schultes, R. (1953) Notes on Zamia in the Colombian Amazon. Mutisia 15, 16.
1, pp. 1168.
Stevenson, D.W. (1990) Morphology and systematics of the Cycadales. In: Stevenson,
Stevenson, D.W. (1991) Zamiaceae. Flora of the Guianas, Series A, Fascicle 9, 711.
Stevenson, D.W. (1993) The Zamiaceae in Panama with comments on phytogeography
and species relationships. Brittonia 45, 116.
Stevenson, D.W. (1999) Zamiaceae. In: Jorgensen, P.M. and Len-Ynez, S. (eds) Catalogue
of the Vascular Plants of Ecuador. Missouri Botanical Garden Press, St Louis, Missouri,
pp. 189190.
Stevenson, D.W. (2001) Cycadales. Flora de Colombia 21, 192.
Stevenson, D.W. and Sabato, S. (1986) Typification of names in Zamia L. and Aulacophyllum
Regel (Zamiaceae). Taxon 35, 134144.
Walter, K. and Gillett, H. (1998) 1997 IUCN Red List of Threatened Plants. IUCN The
World Conservation Unit, Gland, Switzerland.
Wrinkle, G. (1993) A new species of Zamia from amazonian Peru. Encephalartos 36, 2022.
ch15.qxd 28/11/03 3:52 pm Page 195
In Search of the True Tree: 15

Guidelines for Classification
Roy Osborne1 and Terrence Walters2
1PO Box 244, Burpengary, Queensland, Australia;

2Montgomery Botanical Center, Miami, Florida, USA
Abstract
The central goal in cycad systematics is to uncover the true tree that correctly represents
the hierarchy of relationships within the Cycadales. Consensus reached at the Cycad
Classification Concepts Workshop in Miami in April 2002 is that the cycads are a mono-
phyletic group, that a species is the smallest practical unit which should be used at the
present time, that species descriptions should be field-based with as wide a suite of char-
acters as possible and that taxonomic descriptions of the highest standard are essential to
avoid past deficiencies in the literature. The identification of facilities, the sharing of
resources and a vigorous interaction between all cycad systematics will collectively facili-
tate progress towards the true tree. Data in this respect are presented in the form of a
number of schedules.
Introduction
This chapter is based on discussions by participants at the Cycad Classification

Concepts (CCC) Workshop held at the Montgomery Botanical Center on 79
April 2002. It commences with a summary of the core philosophy held by all par-
ticipants, as has been more fully detailed in Chapter 1 of this volume, and then
progresses to specific guidelines recommended during the workshop sessions.
CAB International 2004. Cycad Classification: Concepts and Recommendations 195

(eds T. Walters and R. Osborne)
ch15.qxd 28/11/03 3:52 pm Page 196
196 R. Osborne and T. Walters
Further work in cycad systematics is constrained by the availability of resources

and a number of schedules listing various facilities available to cycad research
workers have therefore been included.
Core Philosophy
There exists an underlying coherent natural order within the Cycadales. The
objective of those working in cycad systematics is to construct hypotheses that will
ultimately reveal the phylogeny of the group, i.e. the true tree for the Cycadales.
This tree comprises a hierarchy of recognizable categories that can be arranged
systematically in an order descending towards the smallest taxonomic unit. At the
present time, the smallest unit that is practically useful is our concept of a species,
this being a geographically and morphologically distinct entity that is reproduc-
tively isolated from any other such entity. Variations in interpretation of such
entities will inevitably occur and are recognized as part of the process in striving
towards stability.
The work presented in Chapters 214 of this volume shows that consider-
able progress has been made in understanding relationships between the cycads
as presently circumscribed, but it is clear that a wealth of data remains to be cap-
tured and further botanical exploration is necessary. The data exist in a broad
range of forms that encompass much more than classical taxonomic descriptions
based solely on plant morphology. The continuous exchange of information and
sharing of resources is vital to ensure further meaningful progress.
Guidelines for Taxonomic Descriptions
Inadequate taxonomic descriptions for cycads (and other plants) in the 17th to
20th centuries have led to much confusion and have inhibited progress in cycad
systematics (references throughout, Chapters 214 this volume). It is vital that
new taxonomic treatments are of the highest standard and, while concise,
provide a sound basis for critical evaluation and elucidation of relationships
between taxa. The requirements for naming species are provided in the princi-
ples, provisions, articles and recommendations given in the codes of botanical
nomenclature, the current of which is the St Louis Code (Greuter et al., 2000).
Apart from the mandatory requirements, a proposal from the CCC Workshop is
that authors should, as far as possible, include the following in their treatments,
and that manuscript reviewers should be cognisant of the following guidelines.
Full details, qualitative and quantitative where appropriate, of vegetative and

reproductive morphological characters should be presented in all texts.
Geographical data, such as climatic data, distribution range and profile, GPS lat-
itude and longitude coordinates (except where such information may prejudice
conservation of the taxon) and altitude, should be included in the descriptions.
ch15.qxd 28/11/03 3:52 pm Page 197
Guidelines for Classification 197
Ecological data, including edaphic and topographical data, habitat details and
plant associations and symbiotic associations, should be given.
Life history aspects, such as coning phenology, pollinating and seed dispersal
agents, plant predators, plant abundance and population dynamics, and evi-
dence of possible hybridization, should be included.
Micromorphological and anatomical details should be given where significant.
Cytological details, such as chromosome number and morphology and details
of chromosome satellites, should be provided if available.
Chemical data, such as details of primary and secondary plant metabolites,
isoenzyme analyses and DNA investigations, should be included where possi-
ble.
Ethnobotanical commentary, especially as may relate to past distribution pat-
terns, local names and human usage of plant material, should be given.
Ex situ experimental results (e.g. where trials have been made to access the plas-
ticity of discriminatory characters) should be provided.
Guidelines for Presenting Reports
In addition to the above features, which relate directly to the comprehensiveness

of the taxonomic description for any particular taxon, the following ancillary
guidelines are proposed by the CCC Workshop participants.
Species are recognized as the basic evolutionary unit; until cycad classification
becomes more stable, the designation of infraspecific units (subspecies, vari-
eties, etc.) is discouraged.
Taxonomic descriptions should be submitted only to widely recognized, inter-
national and refereed journals.
A distribution map (or descriptive text) should be provided to show the extent
of the known range and that of associated and/or closely related taxa.
Taxonomic keys using discriminatory characters should be included where
appropriate.
The conservation status of the taxon should be given, or an appropriate con-
servation status proposed (see Donaldson, Chapter 2 this volume).
Where data are derived from ex situ plants, the source of such material should
be clearly identified.
Epithets chosen for species names should be simple and relevant, and an ety-
mology should be provided.
Authors names associated with plant species should follow internationally
recognised guidelines (e.g. Brummit and Powell, 1992; see also the IPNI
website listed in Schedule 15.9).
Cycad taxonomists are encouraged to send their draft publication to peers for
critical evaluation before submission for publication.
Major advances in cycad classification will follow from in-depth monographic
treatments based on geographical zones or monophyletic groupings.
ch15.qxd 28/11/03 3:52 pm Page 198
Guidelines for Collection of Field Data
Much of the past information about individual cycad taxa has been based on
single typical specimens. The necessity for data that best represent a plant pop-
ulation, rather than single specimens, has only recently been recognized as being
paramount. Recommended guidelines for fieldworkers are as follows.
Sample size is critical in data collection and must be appropriate both for the
variation within the population and for the nature of the analysis.
Cognisance must be made of changes that occur on a seasonal basis, or with
the passage of time, which may affect the data collected.
Photographic evidence should be used to complement field notes.
Standardized data collection sheets (see Hill, Chapter 3 this volume) are
extremely useful in providing rigour and structure to field data collection; a
uniform character checklist for each cycad genus is highly desirable.
Guidelines for Herbarium Specimen Collecting and

Processing
The need for quality herbarium specimens goes hand in hand with the concept of
quality taxonomic publications. Many of the older cycad specimens are sadly
inadequate in terms of both material and collector annotations (references
throughout, Chapters 214 this volume). Guidelines for fieldworkers are as follows.
Herbarium specimens should include both vegetative and reproductive mate-

rial where possible.
A type specimen should be carefully selected so that it best represents the pop-
ulation being described. Since cycads are dioecious, the type will necessarily
represent only a single male or female plant. Additional vouchers should be
made to represent the other sex.
The herbarium sheets should be informatively documented.
Photographs should be used to illustrate points that will not be evident in pre-
served specimens.
Duplicate herbarium sheets should be deposited in at least three institutions.
Availability of Resources
In parallel with the distribution patterns of the extant cycads, workers in cycad tax-
onomy and systematics are geographically dispersed and often isolated. Their work
is often inhibited by limited financial and infrastructural resources. Since sharing of
resources is an obvious way to eliminate duplication of effort and to increase effi-
ciency, it is appropriate to list resources that may be useful to cycad workers.
ch15.qxd 28/11/03 3:52 pm Page 199
Personnel resources. Schedule 15.1 gives the names and addresses of workers
actively engaged in cycad studies.
Facilities. Schedules 15.215.6 give lists of cycad-orientated herbaria, ex situ
plant collections, fossil collections, insect collections and molecular laborato-
ries.
Software packages for cladistics. See Schedule 15.7.
Cycad societies and magazines. See Schedule 15.8.
Electronic information. Schedule 15.9 gives a list of other cycad-oriented web-
sites.
Remaining Issues to be Resolved
As would be expected in any long-term multi-disciplinary scientific project, there

are several issues that can be identified but which could not be resolved at the
CCC Workshop. Among these are the following.
Terminology. A need has been identified for the standardization of cycad ter-
minology. Examples of inconsistency are seen at present in the use of terms
such as leaflet vs. pinna, the distinction between spines, pinnacanths (as mod-
ified leaflets) and prickles (as epidermal structures), the use of the terms rachis
(rachis) vs. leaf stalk (petiole), cone vs. strobilus, seed vs. omnule etc. It is rec-
ommended that an authoritative taxonomic glossary to be used as a guideline
for cycad workers is prepared and distributed; the glossary developed for this
volume (see Osborne and Walters, Appendix 2 this volume) may be useful as
a starting point in this project.
Archiving of data. A great deal of information resides in personal literature
collections, plant material, photographs/slides, field notebooks, etc. which
often represent a lifetimes work in cycad studies. No provision exists for the
archiving of such material for posterity.
International Conferences on Cycad Biology. Commenced in 1987, these
immensely valuable meetings have been organized every 3 years on an ad hoc
basis, with no provision for continuity. It is recommended that a coordinating
role in this planning is given to the Cycad Specialist Group of the IUCN.
The World List of Cycads. Several versions of a cycad world list have been
published in paper and electronic forms with the format and choice of accept-
able species in the hands of the authors. A single and authoritative world list,
timeously updated, is a vital data source for numerous and diverse users. It is
recommended that the Cycad Specialist Group of the IUCN be given a mon-
itoring responsibility for this list, with a brief to secure its long-term continua-
tion and to resolve points of conflict if necessary. A listing of the cycad flora
of the world, excluding synonyms, was compiled in 2003 by three of the par-
ticipants as another reference source for this volume (see Hill et al., Appendix
1 this volume).
ch15.qxd 28/11/03 3:52 pm Page 200
Conclusions
The revelation of a single tree, which correctly and fully represents the phylogeny
of all Cycadales, remains as an ultimate, but probably unattainable, objective in
cycad biology. During the foreseeable future, it is important for cycad workers to
strive towards resolution of at least an approximate tree for the extant cycads.
Adherence to the proposed guidelines in this chapter, diligent use of resources,
quality of workmanship, timely dissemination of information and participation in
international meetings are seen as key factors in making progress towards this goal.
Acknowledgements
This chapter is based on the input from all who attended the Cycad Classification
Concepts Workshop held at Montgomery Botanical Center on 79 April 2002.
Participants were invited to review this text during its preparation, and the
authors believe that the contents of this chapter correctly represent the consen-
sus view of all Workshop participants.
Schedules
Schedule 15.1: Active cycad workers
This list comprises names and addresses of workers at present active in cycad-
oriented research projects. It is drawn up from the lists of regular participants in
the International Conferences on Cycad Biology, those who attended the Cycad
Classification Concepts Workshop in Miami in 2002, and authors of recently
published cycad work. We are aware that this list may be incomplete and that
address details may change with the passage of time.
Avendao, Sergio
Instituto de Ecologa A.C., Apdo Postal 63, Xalapa, Veracruz 91000,
Mexico
avenda@ecologia.edu.mz
Beentje, Henk
Royal Botanic Gardens, Kew, Richmond, Surrey TW9 3AE, UK
h.beentje@rbgkew.org.uk
Bonta, Mark
Division of Social Sciences, 203B Kethley Hall, Delta State University,
Cleveland, Mississippi 38733, USA
markabonta@yahoo.com
Broome, Tom
The Cycad Jungle, PO Box 325, Polk City, Florida 33868-0325, USA
cycadjung@aol.com
ch15.qxd 28/11/03 3:52 pm Page 201
Calonje, Alvaro
Carrera 28 5B-102, Cali, Colombia
alcalonje@hotmail.com
Calonje, Michael
770 East Saginaw Way, Fresno, California 93704, USA
caleno@collectorseeds.com
Capela, Pedro
Plantas de Mozambique, Apartado No. 293, Chimoio, Mozambique
mozaplant@teledata.mz
Caputo, Paolo
Dipartimento di Biologia Vegetale, Universit degli Studi di Napoli
Federico II, Via Foria 223, Napoli 80139, Italy
pacaputo@unina.it
Chaves, Ramiro
Calle 46 No. 904, Playa, La Habana, CP 11300, Cuba
rjc66@yahoo.com
Chavez, Victor M.
Universidad Nacional Autnoma de Mexico, Jardin Botanico del Instituto
de Biologia, DF 04510, Mexico
victorm@ibiologia.unam.mx
Chaw, Sha-Miaw
Institute of Botany, Academia Sinica, #128 Academy Road Sec. 2, Taipei
11529, Taiwan
bochaw@sinica.edu.tw
Chemnick, Jefferey
Lotusland Foundation, 695 Ashley Road, Santa Barbara, California
93108, USA
jeffchemnick@cox.net
Chen, Chia-Jui
Institute of Botany, Chinese Academy of Sciences, 20 Nanxincum,
Xiangshan, Beijing 100093, China
chenjiar@95777.com
Clos, Lynne M
1185 Claremont Drive, Boulder, Colorado 80303, USA
lmclos@netone.com
Connell, Stephen
3 Glenbank Crescent, Kalloroo, Western Australia 6025, Australia
Cozzolino, Salvatore
cozzolin@unina.it
Dalzell, Chris
Durban Botanic Gardens, PO Box 3720, Durban 4000, South Africa
dalzellc@prcsu.durban.gov.za
ch15.qxd 28/11/03 3:52 pm Page 202
De Laubenfels, David J.
107 Will-O-Wind Drive, Jamesville, New York 13078, USA
Dehgan, Bijan
Horticultural Systematics Laboratory, Department of Environmental
Horticulture, University of Florida, Gainesville, Florida 32611, USA
bdehgan@mail.ifas.ufl.edu
Donaldson, John
Kirstenbosch Research Center, National Botanical Institute, Private Bag
X7, Claremont 7735, South Africa
donaldso@nbict.nbi.ac.za
Forster, Paul I.
Queensland Herbarium, Environmental Protection Agency, Brisbane
Botanic Gardens, Mt Coot-tha Road, Toowong, Queensland 4066, Australia
paulforster@uq.net.au
Francisco-Ortega, Javier
Florida International University/Fairchild Tropical Garden, 11935 Old
Cutler Road, Miami, Florida 33156, USA
ortega@fiu.edu
Gaudio, Luciano
Dipartimento di Genetica, Biologia Generale e Molecolare, Facolt di
Scienze, Universit degli Studi di Napoli Federico II, Napoli, Italy
gaudio@biol.dgbm.unina.it
Gonzles, Dolores
Mexico
dolores@ecologia.edu.mx
Gonzlez-Geigel, Lutgarda
Jardn Botnico Nacional, Universidad de La Habana, Carretera El
Roco, Km 31/2, Calabazar, Boyeros, CP 19230, C. Habana, Cuba
hajb@ceniai.inf.cu
Goode, Douglas
10 Hudson Bennett Crescent, Gillitts 3610, South Africa
Gorelick, Root
Department of Biology, Arizona State University, Tempe, Arizona 85287-
1501, USA
cycad@asu.edu
Gregory, Timothy J.
Montgomery Botanical Center, 11901 Old Cutler Road, Miami, Florida
33156, USA
gregory.tim@gene.com
Grobbelaar, Nathanal
PO Box 15357, Lynn East 0039, South Africa
natgrob@hotmail.com
ch15.qxd 28/11/03 3:52 pm Page 203
Hall, John
Department of Entomology, University of Queensland. Correspondence to
11/26 Lang Parade, Auchenflower, Queensland 4066, Australia
cycadguy@hotmail.com
Hayes, Virginia
Lotusland Foundation, 695 Ashley Road, Santa Barbara, California 93108,
USA
vahayes@lotusland.org
Haynes, Jody
33156-4242, USA
jodyhaynes@bellsouth.net
Hill, Ken D.
Royal Botanic Gardens, Mrs Macquaries Road, Sydney, New South Wales
2000, Australia
ken.hill@rbgsyd.nsw.gov.au
Hurter, P. Johan H.
Lowveld National Botanic Gardens, PO Box 1024, Nelspruit 1200, South
Africa
herbarium@glow.co.za
Jones, David
13 Saville Close, Melba, ACT 2615, Australia
david.jones@pi.csiro.au
Kennedy, Paul
21 Sierra Road, Engadine, New South Wales 2233, Australia
Kokubugata, Goro
Tsukuba Botanical Garden, National Science Museum, Tokyo, Amakubo
4, Tsukuba, Ibaraki 305-0005, Japan
gkokubu@kahaku.go.jp
Ladd, Philip G.
School of Environmental Science, Murdoch University, Murdoch, Western
Australia 6150, Australia
ladd@essun1.murdoch.edu.au
Lazcano, Julio
Jardn Botnico Nacional, Universidad de La Habana, Carretera
El Roco, Km 31/2, Calabazar, Boyeros, CP 19230, C. Habana,
Cuba
hajb@ceniai.inf.cu
Lindstrm, Anders
Nong Nooch Tropical Garden, 34/1 Sukhumvit Highway, Najomtien,
Sattahip, Chonburi 20250, Thailand
kampon@loxinfo.co.th
Liu, Nian
South China Botanical Garden, Academia Sinica, Guangzhou 510650,
Guangdong, China
ch15.qxd 28/11/03 3:52 pm Page 204
Litz, Richard E.
University of Florida, Tropical Research and Education Center, 18905 SW
280 Street, Homestead, Florida 33031, USA
rel@mail.ifas.ufl.edu
Loc, Phan Ke
Department of Botany, University of Science, Vietnam National
University, Hanoi, Vietnam
pkeloc@yahoo.com
Meerow, Alan W.
USDA-ARS-SHRS, National Germplasm Laboratory, 13601 Old Cutler
Road, Miami, Florida 33158, USA
miaam@ars-grin.gov
Moon, Pamela A.
University of Florida, Tropical Research and Education Center, 18905 SW
280 Street, Homestead, Florida 33031, USA
pamoon@mail.ifas.ufl.edu
Moretti, Aldo
moretti@unina.it
Musial, Kathy
Huntington Botanical Gardens, 1151 Oxford Road, San Marino,
California 91108, USA
Nan, Li
Fairy Lake Botanical Garden, Lian-Tang, Shenzhen, Guangdong 518004,
China
liandrea1963@yahoo.com
Nguyen, Tien Hip
Institute of Ecology and Biological Resources, Nghia Do, Cau Giay,
Hanoi, Vietnam
ntienhiep@hn.vnn.vn
Nicholls, Trevor J.
University of Bristol, Bristol, UK
Oberprieler, Rolf
CSIRO Entomology, GPO Box 1700, Canberra, ACT 2601, Australia
rolfo@ento.csiro.au
OBrien, Charles
Center for Biological Control, 105 [South] Perry-Paige Bldg,
Florida A&M University, Tallahassee, Florida 32307-4100,
USA
charles.obrien@famu.edu or biocontrol@nettally.com
Osborne, Roy
PO Box 244, Burpengary, Queensland 4505, Australia
cycad@iprimus.com.au
ch15.qxd 28/11/03 3:52 pm Page 205
Pea, Esperanza
Calle 2 No. 464, Plaza, CP 10400, La Habana, Cuba
esperanza@ish.co.cu
Prez-Farrera, Miguel Angel
Escuela di Biologa, Universidad de Ciencias y Artes de Chiapas
(UNICACH), Calzada Samuel Len Brindis 151, Tuxtla Gutirrez,
Chiapas 29000, Mexico
miguelangel@chiapas.net or perezfarreram@yahoo.com.mx
Richardson, Peter (Mick)
Missouri Botanical Garden, PO Box 299, St Louis, Missouri 63166-0299,
USA
mick.richardson@mobot.org
Salas-Morales, Silvia
Sociadad para el Estudio de Recursos Biticos de Oaxaca (SERBO), Calle
Porfirio Diaz, No. 211, Centro, Oaxaca, Mexico
serbo@prodigy.net.mx
Seawright, Alan A.
National Research Center for Environmental Toxicology, University
of Queensland, PO Box 594, Archerfield, Queensland 4108,
Australia
a.seawright@mailbox.uq.oz.au
Schutzman, Bart
University of Florida, Environmental Horticulture Department, 1525
W.M. Fifield Hall, Gainesville, Florida 32611-0670, USA
bschutzman@mail.ifas.ufl.edu
Scott-Shaw, C. Rob
Biodiversity Research Division, KwaZulu-Natal Conservation Service, PO
Box 13053, Cascades 3202, South Africa
robss@kznwildlife.com
Siniscalco Gigliano, Gesualdo
Stevenson, Dennis Wm.
Institute of Systematic Botany, New York Botanical Garden, Bronx, New
York 10458, USA
dws@nybg.org
Tang, William
Fairchild Tropical Garden, 11935 Old Cutler Road, Miami, Florida
33156, USA
william.tang@aphis.usda.gov
Taylor, Alberto
Universidad de Panama, 6-4957 El Dorado, For Z6, Panama
sidney@cwp.net.pa
Terry, Irene
Department of Biology, University of Utah, 257 South 1400 East, Salt
ch15.qxd 28/11/03 3:52 pm Page 206
Lake City, Utah 84112, USA

terry@biology.utah.edu
Thompson, Craig
PO Box 140, Sylvania, New South Wales 2224, Australia
Vzques Torres, Mario
Centro de Investigaciones Biologicas, Universidad Veracruzana, Apdo
Postal 294, Xalapa, Veracruz 91000, Mexico
mvazquez@bugs.invest.uv.mx
Visser, Jan-Maarten
Hortus Botanicus Amsterdam, Plantage Middenlaan 2a, 1018 DD,
Amsterdam, The Netherlands
jan-maarten.visser@dehortus.nl
Vorster, Piet
Botany Department, University of Stellenbosch, Private Bag X1,
Matieland 7602, South Africa
pjvor@sun.ac.za
Vovides, Andrew P.
Mexico
vovidesa@ecologia.edu.mx
Walters, Terrence
33156-4242, USA
walters@fiu.edu
Whitelock, Loran M.
4524 Toland Way, Los Angeles, California 90041, USA
cycadgardens@juno.com
Wilson, Gary W.
Department of Tropical Plant Sciences, James Cook University, Smithfield,
Cairns, Queensland 4870, Australia
Gary.Wilson@jcu.edu.au
Wink, Michael
Institut fr Pharmazeutische Biologie, Universitt Heidelberg, Im
Nuwnhwimwe Feld 364, D-69120 Heidelberg, Germany
wink@uni-hd.de
Wijnants, Jeff
42 Oudstrijderstraat, Merchtem 1785, Belgium
Schedule 15.2: Major herbaria with cycad accessions
A number of herbaria with important cycad accessions are listed below. Many
of the European herbaria have vouchers from the days of colonial exploration;
more recent collections will generally be found in herbaria in the USA and in
the countries of origin of the material. As names and addresses may change in
time, a useful website for more permanent access is the Index Herbariorum,
http://www.nybg.org/bsci/ih
ch15.qxd 28/11/03 3:52 pm Page 207
Australian National Herbarium (CANB)

Center for Plant Biodiversity Research, GPO Box 1600, Canberra, ACT
2601, Australia
Website: http://www.anbg.gov.au/cpbr
Contact: Jim Croft, jrc@anbg.gov.au
Botanischer Garten und Botanisches Museum Berlin-Dahlem (B)
Zentraleinrichtung der Freien Universitt Berlin, Knigin Luise Strasse
6-8, D-14191 Berlin, Germany
Contact: Director, wg@zedat.fu-berlin.de
Botany Department (BM)
The Natural History Museum, Cromwell Road, London SW7 5BD,
UK
Website: http://www.nhm.ac.uk
Contact: R. Huxley, r.huxley@nhm.ac.uk
Chinese National Herbarium, Institute of Botany, Chinese Academy of
Sciences (PE)
20 Nanxincun, Xiangshan, Beijing 100093, China
Website: http://www.ibcas.ac.cn
Contact: Hai-Ning Qin, hainingqin@ns.ibcas.ac.cn
Conservatoire et Jardin Botaniques de la Ville de Genve (G)
Case postale 60, CH-1292 Chambsy/Genve, Switzerland
Contact: Fernand Jacquemoud, fernand.jacquemoud@cjb.ville-ge.ch
Compton Herbarium (NBG)
National Botanical Institute, Private Bag X7, Claremont 7735, South
Africa
Website: http://www.nbi.ac.za
Contact: Koos Roux, roux@nbict.nbi.ac.za
Fairchild Tropical Garden (FTG)
11935 Old Cutler Road, Miami, Florida 33156, USA
Website: http://www.virtualherbarium.org
Contact: Gerald Guala, stinger@fairchildgarden.org
Field Museum of Natural History (F)
1400 south Lake Shore Drive, Chicago, Illinois 60605-2496,
USA
Contact: Chair, gmueller@fmnh.org
Gray Herbarium (G) and Arnold Arboretum (A)
Harvard University Herbaria, 22 Divinity Avenue, Cambridge,
Massachusetts 02138, USA
Contact: Emily Wood, ewood@oeb.harvard.edu
Herbario, Departamento de Botnica, Instituto Nacional de Biodiversidad
(INB)
Apdo Postal 22-3100, Santo Domingo de Heredia, Costa Rica
Website: http://www.inbio.ac.cr
Contact: Nelson Zamora, nzamora@inbio.ac.cr
ch15.qxd 28/11/03 3:52 pm Page 208
Herbario, Instituto de Ecologa (XAL), A.C.

Apdo Postal 63, 2.5 Km sobre carretera antigua a Coatepec, Cong El
Haya 351, Xalapa, Veracruz 91000, Mexico
Website: http://www.ecologia.edu.mx
Contact: Francisco G. Lorea-Hernndez, loreaf@ecologia.edu.mx
Herbario, Instituto de Investigao Cientfica Tropical (LISC) Travessa Conde
da Ribeira 9, P-1300-142, Lisbon, Portugal
Contact: Maria Adlia Diniz, cbotn@iict.pt
Herbarium Jardin Botanique National de Belgique (BR)
Domein van Bouchout, B-1860 Meise, Belgium
Website: http://www.br.fgov.be
Contact: Elmar Robbrecht, robbrecht@br.fgov.be
Herbarium Neapolitanum (NAP)
Dipartimento di Biologia Vegetale Universit degli Studi di Napoli
Federico II, Via Foria 223, 80139 Napoli, Italy
Contact: Annalisa Santangelo, santange@cds.unina.it
Herbarium, University of Coimbra (COI)
Acros do Jardim, P-3001-401, Coimbra, Portugal
Contact: Ftima Sales, fsales@ci.uc.pt
Missouri Botanical Garden Herbarium (MO)
PO Box 299, St Louis, Missouri 63166-0299, USA
Website: http://www.mobot.org
Contact: James Solomon, jim.solomon@mobot.org
Musum national dhistoire naturelle, Herbier (P)
16 rue Buffon, F-75005 Paris, France
Contact: Philippe Morat, morat@mnhn.fr
Nationaal Herbarium Nederland, Leiden University Branch
(Rijksherbarium) (L)
Postbus 9514, 2300 RA Leiden, The Netherlands
Website: http://www.nationaalherbarium.nl
Contact: Pieter Baas, baas@nhn.leidenuniv.nl
Nationaal Herbarium Nederland, Utrecht University Branch (U),
Heidelberglaan 2, 3584 CS Utrecht, The Netherlands. Website:
http://www.bio.uu.nl/~herba
Contact: Paul Maas, p.j.m.maas@bio.uu.nl
National Herbarium (PRE)
National Botanical Institute, Private Bag X101, Pretoria 0001, South
Africa
Contact: Marinda Koekemoer, mk@nbipre.nbi.ac.za
National Herbarium of New South Wales (NSW)
Royal Botanic Gardens, Mrs Macquaries Road, Sydney 2000,
Australia
Website: http://www.rbgsyd.gov.au
Contact: Barry Conn, barry.conn@rbgsyd.nsw.gov.au
ch15.qxd 28/11/03 3:52 pm Page 209
National Herbarium of Zimbabwe (SRGH)

Botanic Garden, PO Box CY 550, Causeway, Harare, Zimbabwe
Contact: Nozipo Nobanda, srgh@icon.co.zw
New York Botanical Garden Herbarium (NY)
Bronx, New York 10458-5126, USA
Website: http://www.nybg.org
Contact: Barbara Thiers, bthiers@nybg.org
Northern Territory Herbarium (DNA)
Parks and Wildlife Commission of the Northern Territory, PO Box 496,
Palmerston, Northern Territory 0831, Australia
Contact: Gregory Leach, greg.leach@nt.gov.au
Queensland Herbarium (BRI)
Brisbane Botanic Gardens, Mt Coot-tha Road, Toowong, Queensland
4066, Australia
Contact: Gordon Guymer, queensland.herbarium@env.qld.gov.au
Royal Botanic Gardens Herbarium (K)
Kew, Richmond, TW9 3AE, UK
Website: http://www.rbgkew.org.uk
Contact: Simon Owens, s.owens@lion.rbgkew.org.uk
South China Institute of Botany (IBSC)
Academia Sinica, Guangzhou 510650, Guangdong, China
Contact: Hai-Shan Chen, chenhs@scib.ac.cn
United States National Herbarium (US)
Botany Section, MRC-166, Smithsonian Institution, PO Box 37012,
Washington, DC 20013-7012, USA
Website: http://nmnh.si.edu/sysbiology
Contact: Collections Manager, russell.rusty@nmnh.si.edu
V.L. Komarov Botanical Institute Herbarium (LE)
Prof. Popov Street 2, Saint Petersburg 197376, Russia
Contact: Rudolf Kamelin, binrag@glas.apc.org
Schedule 15.3: Major ex situ cycad collections
Darwin Botanic Gardens

Northern Territory Parks and Wildlife Commission, Geranium Street,
Stuart Park, Darwin, PO Box 496, Palmerston, Northern Territory 0831,
Australia
Website: http://www.nt.gov.au/paw
Contact: Greg Leach, greg.leach@nt.gov.au
Durban Botanic Gardens
PO Box 3720, Durban 4000, South Africa
Contact: Chris Dalzell, dalzellc@prcsu.durban.gov.za
Fairchild Tropical Garden
ch15.qxd 28/11/03 3:52 pm Page 210
Website: http://www.fairchildgarden.org
Contact: Jack Fisher, fisherj@fiu.edu
Fairy Lake Botanical Garden
Lian-Tang, Shenzhen, Guangdong 518004, China
Website: http://www.szbg.org/eng/cont.htm
Contact: Li Nan, liandrea1963@yahoo.com
Ganna Walska Lotusland Foundation
695 Ashley Road, Santa Barbara, California 93108, USA
Website: http://www.lotusland.org/general.html
Contact: Virginia Hayes, vahayes@lotusland.org
Hortus Botanicus Amsterdam
Plantage Middenlaan 2a, 1018 DD, Amsterdam, The Netherlands
Contact: Jan-Maarten Visser, jan-maarten.visser@dehortus.nl
Huntington Botanical Gardens
1151 Oxford Road, San Marino, California 91108, USA
Website: http://www.huntington.org
Contact: Kathy Musial
Jardn Botnico FCO J. Clavijero, Instituto de Ecologa, A.C.
Apdo Postal 63, 2.5 Km sobre carretera antigua a Coatepec, Cong El
Haya 351, Xalapa, Veracruz 91000, Mexico (Mexican National Cycad
Collection)
E-mail: jbclavij@ecologia.edu.mx
Kirstenbosch National Botanical Garden
Private Bag X7, Claremont 7735, South Africa
Contact: Philip le Roux, leroux@nbict.nbi.ac.za
Lowveld National Botanic Gardens
PO Box 1024, Nelspruit 1200, South Africa
Website: http://www.nbi.ac.za/lowveld/mainpage.htm
Contact: Johan Hurter, herbarium@glow.co.za
Montgomery Botanical Center
11901 Old Cutler Road, Miami, Florida 33156-4242, USA
Website: http://www.montgomerybotanical.org
Contact: Jody Haynes, jodyhaynes@bellsouth.net
Nong Nooch Tropical Garden
34/1 Sukhumvit Highway, Najomtien, Sattahip, Chonburi 20250,
Thailand
Contact: Anders Lindstrm, kampon@loxinfo.co.th
Orto Botanico
Contact: Salvatore Cozzolino, cozzolin@unina.it
Pretoria National Botanical Garden
National Botanical Institute, Private Bag X101, Pretoria 0001, South
Africa
ch15.qxd 28/11/03 3:52 pm Page 211
Website: http://www.nbi.ac.za/pretoria/mainpage.htm
Contact: Gideon Smith, gfs@nbipre.nbi.ac.za
Quail Botanic Gardens
230 Quail Gardens Drive, Encinitas, California 92024, USA
Website: http://www.qbgardens.com
Contact: Julian Duval, qbgardens@aol.com
Schedule 15.4: Major cycad fossil collections
Details of a limited number of institutions with important fossil cycad collec-

tions are given below. A useful generic website is:
http://uni-wuerzburg.de/mineralogie/palbot1.html
Bernard Price Institute for Palaeontology

University of Witwatersrand, Private Bag X3, Johannesburg 2050, South
Africa
Contact: Marion Bamford, 106mab@cosmos.wits.ac.za
Burke Museum of Natural History and Culture
PO Box 353010, University of Washington, Seattle, Washington 98195,
USA
Contact: Wes Wehr, wwehr@u.washington.edu
Denver Museum of Nature and Science
2001 Colorado Blvd., Denver, Colorado 80205, USA.
Website: http://www.dmns.org
Contact: Kirk Johnson, kjohnson@dmns.org
Florida Museum of Natural History
Museum Road and Newell Drive, University of Florida, Gainesville,
Florida 32611-7800, USA
Website:
http://www.flmnh.ufl.edu/natsci/paleobotany/paleobotany.htm
Contact: Steven Manchester, steven@flmnh.ufl.edu
Herbarium of Paleobotany
Institute of Botany, the Chinese Academy of Sciences, Xiangshan, Beijing
100093
Website: http://www.ibcas.ac.cn
Contact: Yu-Fan Hu.
Instituto de Geologa, Universidade Nacional Autnoma de
Mxico
Ciudad Universitaria, 04510 Mexico, DF, Mexico
Contact: Reinhard Weber, Weber@servidor.unam.mx
Museo Argentino de Ciencias Naturales B. Rivadavia
Av. A. Gallardo 470, 1405 Buenos Aires, Argentina
Contact: Sergio Archangelsky, sarcang@nsl.overnet.com.ar
ch15.qxd 28/11/03 3:52 pm Page 212
Museo de La Plata
Paseo del Bosque s/n. 1900, La Plata, Argentina
Contact: Analia Artabe
Museo Paleontolgico Egidio Feruglio
9100 Trelew, Chubut, Argentina
Contact: Ruben Cneo, mef004@infovia.com.ar
Museum Victoria
PO Box 666e, Melbourne, Victoria 3001, Australia
Website: http://www.museum.vic.gov.au
Contact: Dermot Henry, dhenry@museum.vic.gov.au
National Herbarium Fossil Collection
National Botanical Institute, Private Bag X101, Pretoria 0001, South Africa
Contact: John Anderson, jma@nbipre.nbi.ac.za
Natural History Museum
Department of Palaeontology, Cromwell Road, London
SW7 5BD, UK
Website: http://www.nhm.ac.uk/palaeontology
Contacts: Paul Kenrick, pauk@nhm.ac.uk and
Paul Davis, P.Davis@nhm.ac.uk
Peabody Museum of Natural History
Yale University, PO Box 208118, 170 Whitney Avenue, New Haven,
Connecticut, USA
Website: http://www.peabody.yale.edu
Contact: Leo Hickey, leo.hickey@yale.edu
Petrified Forest National Park
PO Box 2277, Petrified Forest, Arizona 86028, USA
Website: http://www.nps.gov/pefo/index.htm
Smithsonian National Museum of Natural History (Department of
Palaeobiology)
Website: http://www.mnh.si.edu
Contact: Scott Wing, wing.scott@nmnh.si.edu
Schedule 15.5: Insect collections with significant cycad-associated

specimens
Albany Museum
Somerset Street, Grahamstown 6139, South Africa
Website: http://www.ru.ac.za/departments/am
Contact: F.W. Gess, f.gess@ru.ac.za
Australian National Insect Collection
CSIRO Entomology, GPO Box 1700, Canberra, ACT 2601, Australia
Website: http://www.ento.csiro.au/research/natres/natres.html
Contact: Rolf Oberprieler, rolfo@ento.csiro.au
ch15.qxd 28/11/03 3:52 pm Page 213
Charles W. OBrien (Private) Collection

Florida A&M University, Tallahassee, Florida 32307-4100, USA
Contact: Charles OBrien, charles.obrien@famu.edu or
biocontrol@netally.com
Iziko Museums of Cape Town (South African Museum)
PO Box 61, Cape Town 8000, South Africa
Website: http://www.nhm.org/~lorquin/evans/a2collections.html
Contact: Simon van Noort, svannoort@samuseum.ac.za
National Collection of Insects
Plant Protection Research Institute, Private Bag X134, Pretoria 0001,
South Africa
Website: http://www.sciref.org/netc/netc1-2/dit2c.htm
Contact: Riaan Stals, vrehrs@plant5.aric.za
The Natural History Museum
Department of Entomology, Cromwell Road, London SW7 5BD, UK
Website: http://www.nhm.ac
Contact: Chris Lyal, c.lyal@nhm.ac.uk
Schedule 15.6: Molecular laboratories carrying out cycad-related work
The list below gives details of some laboratories that are active in molecular
analyses. It must be remembered that this work is time-consuming and
expensive fees may be requested on a per sample basis for any work
done. However, most universities with botany, biology or biochemistry depart-
ments would have access to equipment for carrying out molecular studies and
the possibility of a joint project is worth considering; such projects are usually
dependent on the interest of students and supervisors and on funding
availability.
Cold Spring Harbor Laboratories

PO Box 100, Cold Spring Harbor, New York 11724, USA
Website: http://www.cshl.org
Dipartimento di Biologia Vegetale
Universit degli Studi di Napoli Federico II, Via Foria 223, Napoli 80139,
Italy
Contact: Paolo Caputo, pacaputo@unina.it
Institut fr Pharmazautische Biologie
Universitt Heidelberg, Im Neuenheimer Feld 364, D-69120, Heidelberg,
Germany
Website: http://Pharm-Biologie.uni-hd.de
Contact: Michael Wink, wink@uni-hd.de
Institute of Botany, Academia Sinica
#128 Academy Road Sec. 2, Taipei 11529, Taiwan
Contact: Shu-Miaw Chaw, bochaw@sinica.edu.tw
ch15.qxd 28/11/03 3:52 pm Page 214
Institute of Systematic Botany, New York Botanical Garden

Bronx, New York 10458, USA
Website: http://www.nybg.org
Contact: Dennis Stevenson, dws@nybg.org
Instituto de Ecologa, Sistemtica Vegetal
Apdo Postal 63, Xalapa, Veracruz 91000, Mexico
Contact: Andrew Vovides, vovides@ecologia.edu.mx
Jodrell Laboratory
Royal Botanic Gardens, Kew, Richmond, Surrey, TW9 3DS, UK
Website: http://www.rbgkew.org.uk
Contact: Mark Chase, m.chase@rbgkew.org.uk
Fairchild Tropical Garden
Website: http://www.fiu.edu/~biology/hmpgs/jfo.html
Contact: Javier Francisco-Ortega, ortegaj@fiu.edu
Fairy Lake Botanical Garden
Lian-Tang, Shenzhen, Guangdong 518004, China
Website: http://www.szbg.org/eng/cont.htm
Contact: Li Nan, liandrea1963@yahoo.com
National Herbarium of New South Wales
Royal Botanic Gardens, Mrs Macquaries Road, Sydney 2000,
Australia
Website: http://www.rbgsyd.nsw.gov.au
Contact: Ken Hill, ken.hill@rbgsyd.nsw.gov.au
USDA-ARS-SHRA National Germplasm Repository
13601 Old Cutler Way, Miami, Florida 33158, USA
Contact: Alan Meerow, miaam@ars-grin.gov
Schedule 15.7: Software packages for cladistic analysis

A large number of software packages is available as tools for cladistic analyses,
the more popular of which amongst cycad systematists are listed below.
Particularly useful generic information is found on the Phylogeny Programs
website: http://evolution.genetics.washington.edu/phylip/software.html.
BIOSYS2
Author: W.C. Black. A modified version of Swofford and Selanders BIOSYS
program for the analysis of allelic variation in genetics.
File transfer site: ftp://lamar.colostate.edu/pub/wcb4
HENNIG86
Author: J.S. Farris. Current version 1.5. Reference manual and software
published by the author, Port Jefferson Station, New York, USA.
Website: http://www.cladistics.org
ch15.qxd 28/11/03 3:52 pm Page 215
MACCLADE
Authors: D. and W. Maddison. Current version 4. Published by Sinauer
Associates.
Website: http://phylogeny.arizona.edu/macclade/macclade.html
NONA
Author: P.A. Goloboff. Current version 2.0. Instruction manual distributed
by the author, INSUE Fundacin e Instituto Miquel Lillo, Miquel Lillo
205, 4000 S. M. de Tucumn, Argentina.
Website: http://www.cladistics.com
PAUP Phylogenetic Analysis Using Parsimony
Author: D.L. Swofford. Current version 4.0. Smithsonian Institute,
Washington, DC, USA.
Website: http://www.paup.csit.fsu.edu
PHYLIP Phylogeny Inference Package
Author: J. Felsenstein. Current version 3.5c. Distributed by the author,
Department of Genetics, University of Washington, Seattle, USA.
Website: http://evolution.genetics.washington.edu/phylip/software.html
WINCLADA
Author: K.C. Nixon. Current version 1.00.08. Published by the author,
Ithaca, New York, USA.
Website: http://www.cladistics.com
Schedule 15.8: Cycad societies and magazines
The Cycad Society (USA)

c/o Montgomery Botanical Center, 11901 Old Cutler Road, Miami, FL
33156, USA
Magazine: The Cycad Newsletter
Website: http://www.cycad.org
The Cycad Society of China
c/o Institute of Botany, Chinese Academy of Sciences, 20 Nanxincum,
Xiangshan, Beijing 100093, China
Contact: Chia-Jui Chen, chenjiar@95777.com
The Cycad Society of South Africa
PO Box 1790, Groenkloof, Pretoria 0027, South Africa
Magazine: Encephalartos
Website: http://www.cycadsociety.org
Contact: Guillaume Theron, GTheron@nsnper1.up.ac.za
Palm and Cycad Societies of Florida, Inc. (PACSOF)
9525 Jamaica Drive, Miami, Florida 33189, USA
Contact: Jody Haynes, webmaster@plantapalm.com
The Palm and Cycad Societies of Australia Limited (PACSOA)
PO Box 1134, Milton, Queensland 4064, Australia
Magazine: Palms & Cycads
ch15.qxd 28/11/03 3:52 pm Page 216
Website: http://www.pacsoa.org.au
E-mail: pacsoa1@ozemail.com.au
The Palm and Cycad Society of New Zealand
PO Box 3871, Auckland, New Zealand
Quarterly magazine
E-mail: nzpalmcycad@yahoogroups.com
Contact: Gary Coleman, aries@splurge.net.nz
West Coast Cycad Society
PO Box 754, Bonsall, California 92028-0754, USA
Website: http://www.home.earthlink.net/~wccshome/index/htm
Contact: Bruce Ironmonger, bruceironmonger@msn.com
Schedule 15.9: Other cycad-associated websites
Gymnosperm Database
Christopher J. Earle (ed.)
Website: http://www.geocities.com/RainForest/Canopy/2285
Harvard University Herbaria
A site including the Gray Herbarium Index, the International Plant Names
Index and other useful information
Website: http://www.huh.harvard.edu/databases
Qtaxa University of California at Riverside
Website: http://maya.ucr.edu/pril/PRIL.html
The Convention on International Trade in Endangered Species of Wild Fauna
and Flora (CITES)
Website: http://www.cites.org
The Cycad Pages
Ken Hill (ed.), Royal Botanic Gardens, Sydney, Australia. This includes the
current World List of Cycads
Website: http://plantnet.rbgsyd.nsw.gov.au/PlantNet/cycad
The International Association for Plant Taxonomy (IAPT)
This gives details of the journal Taxon and the Codes of Botanical
Nomenclature
Website: http://www.bybm.org/iapt/default/htm
The International Plant Names Index (IPNI)
This gives names and authors to plant species
Website: http://www.ipni.org/index.html
The International Union for the Conservation of Nature and Natural
Resources (IUCN)
Website: http://www.iucn.org
The IUCN Red List of Threatened Species
Website: http://www.redlist.org
The Species Survival Commission (SSC)
Including details of the various specialist groups
Website: http://www.iucn.org/theme/ssc
ch15.qxd 28/11/03 3:52 pm Page 217
The University of California, Los Angeles (Berkeley)

Website: http://www.ucmp.berkeley.edu/seedplants/cycadophyta/cycads
Virtual Cycad Encyclopedia
Jody Haynes (ed.), Palm and Cycad Societies of Florida
Website: http://www.plantapalm.com
References
Brummit, R.K. and Powell, C.E. (1992) Authors of Plant Names. Royal Botanic Gardens,
Kew, England, 732 pp.
ch15.qxd 28/11/03 3:52 pm Page 218
X00app1.qxd 28/11/03 3:52 pm Page 219
Appendix 1: The World List of

Cycads
Ken D. Hill,1 Dennis Wm. Stevenson2 and Roy

Osborne3
1Royal Botanic Gardens, Sydney, Australia; 2New York

Botanical Garden, Bronx, New York, USA; 3PO Box 244,
Burpengary, Queensland, Australia
Introduction
Updates of the World List of extant cycad taxa have been published regularly,
generally appearing as part of the Proceedings of the various International
Cycad Conferences on Cycad Biology (see References). These reports list taxa
accepted by the compilers as currently valid, country of occurrence, dates,
authors and journals of publication, often together with much useful commen-
tary about recent taxonomic changes, synonymity and anticipated new species.
The list is also now available as a periodically updated searchable database on the
internet (http://plantnet.rbgsyd.nsw.gov.au/PlantNet/cycad).
This Appendix lists the names of all extant cycads (considered valid as at
April 2003), authors, dates and publication journals. Legitimately published
species names have been excluded when, in our judgement, they represent
species considered synonymous with those having priority under the
International Code of Botanical Nomenclature. Compilation is alphabetical by
genus. Authors names are abbreviated in accordance with Brummit and Powell
(1992) and the International Plant Names Index (IPNI website:
http://www.ipni.org/index.html). Journals are abbreviated according to the
Botanico-Periodicum-Huntianum standard (Lawrence et al., 1968); other refer-
ence material follows the style of Stafleu and Cowan (19761983) and page
numbers are given as the first and last numbers of the taxonomic treatment
within the relevant publication. The type species for each genus is indicated by
an asterisk.
The total number of species, validly published or known to be in press, in
our list, not including synonyms, species dubia, or taxa below specific rank, is 303.
219
X00app1.qxd 28/11/03 3:52 pm Page 220
220 K.D. Hill et al.
Compared with the 130 species recorded in the first such list (Osborne and
Hendricks, 1985), this is testimony to the very considerable progress that has been
made in cycad taxonomy over the past two decades.
X00app1.qxd
28/11/03
The World List
BOWENIA Hook. ex Hook. f. (1863) (2 species, Australia) Bot. Mag. 89: sub t. 5398
3:52 pm
B. serrulata (W. Bull) Chamb. Australia (Qld) 1912 Bot. Gaz. 54 : 419
*B. spectabilis Hook. ex Hook. f. Australia (Qld) 1863 Bot. Mag. 89: sub t. 5398
Appendix 1: The World List of Cycads

CERATOZAMIA Brongn. (1846) (21 species; Mexico, Guatemala & Belize) Ann. Sci. Nat. Bot., ser. 3, 5: 79, t. 1
Page 221
C. alvarezii Prez-Farr., Vovides & Iglesias Mexico (Chiapas) 1999 Novon 9(3): 410413
C. becarrae Prez-Farr., Vovides & Schutzman Mexico (Tabasco) xxxx In review
C. euryphyllidia Vzq. Torres, Sabato & D. W. Mexico (Oaxaca, Veracruz) 1986 Brittonia 38(1): 1726
Stev.
C. fusco-viridis D. Moore Mexico 1878 Sci. Proc. Roy. Dublin Soc., ser. 2: 113114
C. hildae G.P. Landry & M.C. Wilson Mexico (Quertaro, San Luis Potosi) 1979 Brittonia 31(3): 422424, fig. 1
C. huastecorum S. Avendao, Vovides & Cast.- Mexico (Veracruz) 2003 Bot. J. Linn. Soc. 141: 395398
Campos
C. kuesteriana Regel Mexico (Tamaulipas) 1857 Bull. Soc. Imp. Naturalistes Moscou 30:
186188, t. 3, fig. 6, t. 4. fig. 22
C. latifolia Miq. Mexico (Hidalgo, Quertaro, San 1848 Tijdschr. Wis-Natuurk. Wetensch.
Luis Potosi) Eerste Kl. Kon. Ned. Inst. Wetensch. 1: 197209
C. matudae Lundell Guatemala, Mexico (Chiapas, Oaxaca) 1939 Lloydia 2(2): 7576
*C. mexicana Brongn. Mexico (Hidalgo, Puebla, Veracruz) 1846 Ann. Sci. Nat. Bot., ser. 3, 5: 79, t. 1
C. microstrobila Vovides & J.D. Rees Mexico (San Luis Potosi) 1983 Madroo 30(1): 3942
C. miqueliana H. Wendl. Mexico (Chiapas, Tabasco, Veracruz) 1854 Index Palm.: 4954, 68
C. mirandae Vovides, Prez-Farr. & Iglesias Mexico (Chiapas) 2001 Bot. J. Linn. Soc. 137(1): 8185, fig. 12, as
221
C. mirandai
continued
X00app1.qxd
222
28/11/03
C. mixeorum Chemnick & T.J. Greg. & Salas- Mexico (Oaxaca) 1997 Phytologia 83(1) (publ. June 1998):
Morales 4752
C. morettii Vzq. Torres & Vovides Mexico (Veracruz) 1998 Novon 8(1): 8790, fig. 1
C. norstogii D.W. Stev. Mexico (Chiapas, Oaxaca) 1982 Brittonia 34: 181184
C. robusta Miq. Belize, Guatemala, Mexico (Chiapas, 1848 Tijdschr. Wis-Natuurk. Wetensch.
3:52 pm
Oaxaca, Veracruz) Eerste Kl. Kon. Ned. Inst. Wetensch. 1: 3344
C. sabatoi Vovides, Vzq. Torres, Schutzman Mexico (Hidalgo, Quertaro) 1993 Novon 3(4): 502504
& Iglesias
C. whitelockiana Chemnick & T.J. Greg. Mexico (Oaxaca) 1995 Phytologia 79(1) (publ. 1996): 5157
Page 222
C. zaragozae Medellin-Leal Mexico (San Luis Potosi) 1963 Brittonia 15: 175176
C. zoquorum Prez-Farr., Vovides & Iglesias Mexico (Chiapas) 2001 Bot. J. Linn. Soc. 137(1): 7780 , fig. 1
K.D. Hill et al.

CHIGUA D.W. Stev. (1990) (2 species, South America) Mem. New York Bot. Gard. 57: 169172
C. bernalii D.W. Stev. Colombia 1990 Mem. New York Bot. Gard. 57: 169172
*C. restrepoi D.W. Stev. Colombia 1990 Mem. New York Bot. Gard. 57: 169172
CYCAS L. (1753) (98 species, Asia, Australia, Indian and SW Pacific Ocean countries) Sp. Pl.: 1188.
C. aculeata K.D. Hill & Hip N. Vietnam xxxx Mem. New York Bot. Gard. (in press)
C. angulata R. Br. Australia (NT, Qld) 1810 Prodr. 1: 348
C. apoa K.D. Hill New Guinea 1994 Austral. Syst. Bot. 7: 553554, fig. 9
C. arenicola K.D. Hill Australia (NT) 1993 Telopea 5(2): 419422
C. armstrongii Miq. Australia (NT) 1868 Arch. Nerl. Sci. Exact. Nat. 3(5): 235236
C. arnhemica K.D. Hill Australia (NT) 1994 Telopea 5(4): 693696, fig. 1
subsp. muninga Chirgwin & K.D. Hill Australia (NT) 1996 Telopea 7(1): 4446, fig. 20
subsp. natja K.D. Hill Australia (NT) 1996 Telopea 7(1): 4647, fig. 21
C. badensis K.D. Hill Australia (Qld) 1996 Telopea 7(1): 2021, fig. 9
C. balansae Warb. China (Guangxi), N. Vietnam 1900 Monsunia 1: 179
X00app1.qxd
C. basaltica C.A. Gardner Australia (WA) 1923 Bull. Woods Forests Dept., Western Australia 32:
28/11/03
31
C. beddomei Dyer India (Andhra Pradesh) 1883 Trans. Linn. Soc. London, Bot., ser. 2, 5:
8586, pl. 17
C. bifida (Dyer) K.D. Hill China (Guangxi), N. Vietnam xxxx Mem. New York Bot. Gard. (in press)
C. bougainvilleana K.D. Hill Bougainville, New Britain, Solomon 1994 Austral. Syst. Bot. 7: 557560, fig. 11
3:52 pm
Islands
C. brachycantha K.D. Hill, Hip & P.K. Loc N. Vietnam xxxx Mem. New York Bot. Gard. (in press)
C. brunnea K.D. Hill Australia (NT, Qld) 1992 Telopea 5(1): 200201, fig. 15

C. cairnsiana F. Muell. Australia (Qld) 1876 Fragm. 10: 63, 121
Page 223
C. calcicola Maconochie Australia (NT) 1978 J. Adelaide Bot. Gard. 1(3): 175178, fig. 1
C. campestris K.D. Hill New Guinea 1994 Austral. Syst. Bot. 7: 538540
C. canalis K.D. Hill Australia (NT) 1994 Telopea 5(4): 698700, fig. 4a-d
subsp. carinata K.D. Hill Australia (NT) 1994 Telopea 5(4): 699700, fig. 4e-g
C. candida K.D. Hill Australia (Qld) xxxx Telopea (in press)
C. chamaoensis K.D. Hill Thailand 1999 Brittonia 51(1): 58, fig. 6
C. changjiangensis N. Liu China (Hainan) 1998 Acta Phytotax. Sin. 36(6): 552554, fig. 1
C. chevalieri Leandri C. Vietnam 1931 in Lecompte, Fl. Indo-Chine 5(10): 1092
*C. circinalis L. India (Andhra Pradesh, Karnataka, 1753 Sp. Pl.: 1188
Kerala, Maharashtra, Tamil Nadu)
C. clivicola K.D. Hill Malaysia, Thailand 1999 Brittonia 51(1): 6263, fig. 8a-d, g-h
subsp. lutea K.D. Hill Cambodia, Thailand, S. Vietnam 1999 Brittonia 51(1): 64, fig. 8e,f,i
C. collina K.D. Hill, Hip & P.K. Loc N. Vietnam xxxx Mem. New York Bot. Gard. (in press)
C. condaoensis K.D. Hill & S.L. Yang S. Vietnam xxxx Mem. New York Bot. Gard. (in press)
C. conferta Chirgwin ex Chirgwin & Wigston Australia (NT) 1993 J. Adelaide Bot. Gard. 15(2): 147
C. couttsiana K.D. Hill Australia (Qld) 1992 Telopea 5(1): 197198, fig. 13
C. cupida P.I. Forst. Australia (Qld) 2002 Austrobaileya 6: 153
C. curranii (J. Schust.) K.D. Hill Philippines (Palawan) 1995 Proc. Third Int. Conf. Cycad Biol.: 150
223
C. debaoensis Y.C. Zhong & C.J. Chen China (Guangxi) 1997 Acta Phytotax. Sin. 35(6): 571
continued
X00app1.qxd
224
C. desolata P.I. Forst. Australia (Qld) 1995 Austrobaileya 4(3): 345352, fig. 15
28/11/03
C. diannanensis Z.T. Guan & G.D. Tao China (Yunnan) 1995 Sichuan Forestry & Design 1995: 1
C. dolichophylla K.D. Hill, Hip & P.K. Loc China (Yunnan), N. Vietnam xxxx Mem. New York Bot. Gard. (in press)
C. edentata de Laub. Philippines (Sulu) 1998 in De Laub. & Adema, Blumea 43: 372
C. elephantipes A. Lindstrm & K.D. Hill Thailand xxxx Brittonia (in press)
C. elongata (Leandri) D. Yue Wang S. Vietnam 1996 Cycads China: 51, as C. elonga
3:52 pm
C. fairylakea D. Yue Wang China (Guangdong) 1996 Cycads China: 54
C. falcata K.D. Hill Indonesia 1999 Kew Bull. 54(1): 209
C. ferruginea F.N. Wei China (Guangxi) 1994 Guihaia 14: 300
C. fugax K.D. Hill, Hip & P.K. Loc N. Vietnam xxxx Mem. New York Bot. Gard. (in press)
Page 224
C. furfuracea W. Fitzg. Australia (WA) 1918 J. & Proc. Roy. Soc. Western Australia 3: 108
C. glauca hort. ex Miq. East Timor, Indonesia 1840/41 Comm. Phytog.: 127
C. guizhouensis K.M.Lan & R.F. Zou China (Guanxi, Guizhou, Yunnan) 1983 Acta Phytotax. Sin. 21(2): 209210
K.D. Hill et al.

C. hainanensis C.J. Chen China (Hainan) 1975 Acta Phytotax. Sin. 13(4): 82, t. 2, fig. 56
C. hoabinhensis K.D. Hill, Hip & P.K. Loc N. Vietnam xxxx Mem. New York Bot. Gard. (in press)
C. hongheensis S.Y. Yang & S.L. Yang China (Yunnan) 1996 in D.Y. Wang, Cycads China: 62
C. inermis Lour. S. Vietnam 1793 Fl. Cochinch., ed. 2, 2: 776777
C. javana (Miq.) de Laub. Indonesia 1996 in D.Y. Wang, Cycads China: 65
C. lane-poolei C.A. Gardner Australia (WA) 1923 Bull. Woods Forests Dept., Western Australia 32:
3032, fig. E
C. lindstromii S.L. Yang, K.D. Hill & Hip S. Vietnam 1997 Novon 7: 213215, fig. 1
C. litoralis K.D. Hill Malaysia, Myanmar, Sumatra 1999 Brittonia 51(1): 7072, fig. 11
Thailand, S. Vietnam
C. maconochiei Chirgwin & K.D. Hill Australia (NT) 1996 Telopea 7(1): 4849, fig. 22
subsp. lanata K.D. Hill Australia (NT) 1996 Telopea 7(1): 4951
subsp. viridis K.D. Hill Australia (NT) 1996 Telopea 7(1): 5152. fig. 23
C. macrocarpa Griff. Malaysia, Thailand 1854 Not. Pl. Asiat. 4: 1114, pl. 362, fig. 2
C. media R. Br. Australia (Qld) 1810 Prodr. 1: 348
subsp. banksii K.D. Hill Australia (Qld) 1996 Telopea 7(1): 1516
subsp. ensata K.D. Hill Australia (Qld) 1996 Telopea 7(1): 1415, fig. 6
X00app1.qxd
C. megacarpa K.D. Hill Australia (Qld) 1992 Telopea 5(1): 188189, fig. 5
28/11/03
C. micholitzii Dyer E. Laos, Central Vietnam 1905 Gard. Chron., ser. 3, 38: 142144, fig. 4849
C. micronesica K.D. Hill Micronesia (Marianas Islands) 1994 Austral. Syst. Bot. 7: 554557, fig. 10
C. multipinnata C.J. Chen & S.Y. Yang China (Yunnan) 1994 Acta Phytotax. Sin. 32: 239, 480481
C. nathorstii J. Schust. Sri Lanka 1932 Pflanzenr. 99: 76, fig. 10E
C. nongnoochiae K.D. Hill Thailand 1999 Brittonia 51(1): 6062, fig. 7
3:52 pm
C. ophiolitica K.D. Hill Australia (Qld) 1992 Telopea 5(1): 190191, fig. 7
C. orientis K.D. Hill Australia (NT) 1994 Telopea 5(4): 696697, fig. 3
C. pachypoda K.D. Hill S. Vietnam xxxx Mem. New York Bot. Gard. (in press)

C. panzhihuaensis L. Zhou & S.Y. Yang China (Sichuan, Yunnan) 1981 Acta Phytotax. Sin. 19(3): 335, t. 10, fig. 16,
Page 225
t.11, fig. 110
C. papuana F. Muell. New Guinea 1876 Descr. Notes Papuan Pl. 4: 7172
C. pectinata Buch.-Ham. China (Yunnan), India, Laos, 1826 Mem. Wern. Nat. Hist. Soc. 2(2): 322
Myanmar, Thailand 323, fig. 3,5
C. petraea A. Lindstrm & K.D. Hill Thailand xxxx Brittonia (in press)
C. platyphylla K.D. Hill Australia (Qld) 1992 Telopea 5(1): 193194, fig. 9
C. pranburiensis S.L. Yang, W. Tang, K.D. Hill Thailand 1999 Brittonia 51(1): 4447, fig. 1
& Vatacharakorn
C. pruinosa Maconochie Australia (WA) 1978 J. Adelaide Bot. Gard. 1(3): 177178, fig. 2.
C. revoluta Thunb. Japan (Ryukyu Islands) 1782 Verh. Holl. Maatsch. Weetensch. Haarlem 20(2):
424, 426427
C. riuminiana Porte ex Regel Philippines (Luzon) 1863 Gartenflora 12: 1617
C. rumphii Miq. S. Borneo, E. Indonesia, NE Java, 1839 Bull. Sci. Phys. Nat. Nerl. 2: 45
Moluccan Islands, New Guinea, Sulawesi
C. saxatilis K.D. Hill & A. Lindstrm Philippines (Palawan) xxxx In preparation
C. schumanniana Lauterb. New Guinea 1901 Fl. Schutzger. Sudsee (publ. Nov. 1900):
154155
C. scratchleyana F. Muell. New Guinea 1885 Victorian Naturalist 2(2): 1819
225
continued
X00app1.qxd
226
C. seemannii A. Braun SW Pacific islands 1876 Sitzungdber. Ges. Naturf. Freunde Berlin: 114115
28/11/03
C. segmentifida D. Yue Wang & C.Y. Deng China (Guanxi, Guizhou, Yunnan) 1995 Encephalartos 43: 1114
C. semota K.D. Hill Australia (Qld) 1996 Telopea 7(1): 2325, fig. 11
C. sexseminifera F.N. Wei China (Guangxi), Vietnam 1996 Guihaia 16: 1
C. siamensis Miq. Cambodia, Laos, Myanmar, Thailand, 1863 Bot. Zeitung Berlin 21: 334
S. Vietnam
3:52 pm
C. silvestris K.D. Hill Australia (Qld) 1992 Telopea 5(1): 181182, fig. 1
C. simplicipinna (Smitinand) K.D. Hill China (Yunnan), Laos, Myanmar, 1995 Proc. Third Int. Conf. Cycad Biol.: 150
Thailand, N. Vietnam
C. sphaerica Roxb. India (Orissa) 1832 Fl. Ind.: 747
Page 226
C. sundaica K.D. Hill & A. Lindstrm Indonesia xxxx In preparation
C. szechuanensis C.Y. Cheng, W.C. Cheng China (Fijian, Guangdong) 1975 Acta Phytotax. Sin. 13(4): 81, t. 1, fig. 78
& L.K. Fu
K.D. Hill et al.

C. taitungensis C.F. Shen, K.D. Hill, C.H. Taiwan 1994 Bot. Bull. Acad. Sin. 35: 135138
Tsou & C.J. Chen
C. taiwaniana Carruth. China (Guangdong) 1893 J. Bot. 31: 13, pl. 331
C. tanqingii D. Yue Wang China (Yunnan) 1996 Cycads of China: 134
C. tansachana K.D. Hill & S.L. Yang Thailand 1999 Brittonia 51(1): 6566
C. thouarsii R. Br. ex Gaudich. E. Africa, Indian Ocean Islands 1829 Voy. Uranie, Bot: 434
C. tropophylla K.D. Hill & P.K. Loc N. Vietnam xxxx Mem. New York Bot. Gard. (in press)
C. tuckeri K.D. Hill Australia (Qld) 1996 Telopea 7(1): 2023, fig. 10
C. wadei Merrill Philippines (Culion) 1936 Phillipp. J. Sci. 60(3): 234236, pl. 14
C. xipholepis K.D. Hill Australia (Qld) 1996 Telopea 7(1): 3233, fig. 15
C. yorkiana K.D. Hill Australia (Qld) 1996 Telopea 7(1): 1819, fig. 8
C. zeylanica (Schuster) K.D. Hill & Sri Lanka xxxx Novon (in press)
A. Lindstrm
X00app1.qxd
DIOON Lindl. (1843) (13 species; Honduras, Mexico, Nicaragua) Edwards Bot. Reg. 29: misc. 5960
28/11/03
D. angustifolium Miq. Mexico (Nuevo Leon, Tamaulipas) 2003 Biol. J. Linn. Soc. 80 (in press)
D. argenteum T.J. Greg., Chemnick, Salas-Mor. Mexico (Oaxaca) 2003 Bot. J. Linn. Soc. 141: 471476
& Vovides
D. califanoi De Luca & Sabato Mexico (Oaxaca, Puebla) 1979 Brittonia 31(1): 170172
3:52 pm
D. caputoi De Luca, Sabato & Vzq. Torres Mexico (Oaxaca, Puebla) 1980 Brittonia 32(1): 4446
*D. edule Lindl. Mexico (Veracruz) 1843 Edwards Bot. Reg. 29: misc. 5960
D. holmgrenii De Luca, Sabato & Vzq. Torres Mexico (Oaxaca) 1981 Brittonia 33(4): 552554

D. mejiae Standl. & L.O. Williams N. Honduras, Nicaragua 1950 Ceiba 1: 3638
Page 227
D. merolae De Luca, Sabato & Vzq. Torres Mexico (Oaxaca, Chiapas) 1981 Brittonia 33(2): 180184
D. purpusii Rose Mexico (Oaxaca) 1909 Contr. U.S. Natl. Herb. 12(7): 260261
D. rzedowskii De Luca, A. Moretti, Sabato & Mexico (Oaxaca) 1980 Brittonia 32(2): 225229
Vzq. Torres
D. sonorense (De Luca, Sabato & Vzq. Torres) Mexico (NW coast) 1997 Phytologia 83(1) (publ. 1998): 16
Chemnick, T.J. Greg. & Salas-Mor.
D. spinulosum Dyer Mexico (Oaxaca, Veracruz) 1883 in Eichler, Gart. Zeit. 2: 411413, fig. 80
D. tomasellii De Luca, Sabato & Vzq. Torres Mexico (SW coast) 1984 Brittonia 36: 225227
ENCEPHALARTOS Lehm. (1834) (65 species, Africa) Pugill. 6: 14
E. aemulans Vorster South Africa (KwaZulu-Natal) 1990 S. African J. Bot. 56(2): 239243
E. altensteinii Lehm. South Africa (E. Cape) 1834 Pugill. 6: 11, t. 4 & 5
E. aplanatus Vorster Swaziland 1996 S. African J. Bot. 62: 5760
E. arenarius R.A. Dyer South Africa (E. Cape) 1956 J. S. African Bot. 22(1): 14
E. barteri Carruth. ex Miq. Benin, Ghana, Nigeria 1868 Arch. Nerl. Sci. Exact. Nat. 3: 243
subsp. allochrous L.E. Newton Nigeria 1978 J. Linn. Soc. Bot. 77: 125129
E. brevifoliolatus Vorster South Africa (Limpopo) 1996 S. African J. Bot. 62(1): 6164
227
E. bubalinus Melville Kenya, Tanzania 1957 Kew Bull. 1957: 252
continued
X00app1.qxd
228
E. caffer (Thunb.) Lehm. South Africa (E. Cape) 1834 Pugill. 6: 14
28/11/03
E. cerinus Lavranos & D.L. Goode South Africa (KwaZulu-Natal) 1989 Durban Mus. Novit. 14: 153156
E. chimanimaniensis R.A. Dyer & I. Verd. Mozambique, Zimbabwe 1969 Kirkia 7: 147158
E. concinnus R.A. Dyer & I. Verd. Zimbabwe 1969 Kirkia 7: 147158
E. cupidus R.A. Dyer South Africa (Mpumalanga) 1971 Bothalia 10(2): 379383
E. cycadifolius (Jacq.) Lehm. South Africa (E. Cape) 1834 Pugill. 6: 13
3:52 pm
E. delucanus Malaisse, Sclavo & Crosiers Tanzania 1992 Ann. Gembloux 98: 153
E. dolomiticus Lavranos & D.L. Goode South Africa (Limpopo) 1988 Bull. Jard. Bot. Belg. 58: 219224
E. dyerianus Lavranos & D.L. Goode South Africa (Limpopo) 1988 Bull. Jard. Bot. Belg. 58: 219224
E. equatorialis P.J.H. Hurter Uganda 1995 S. African J. Bot. 61(4): 226229
Page 228
E. eugene-maraisii I. Verd. South Africa (Limpopo) 1945 J. S. African Bot. 11(1): 13
E. ferox Bertol. f. Mozambique, South Africa 1851 Mem. Reale Accad. Sci. Ist. Bologna 3: 264
(KwaZulu-Natal)
K.D. Hill et al.

*E. friderici-guilielmi Lehm. South Africa (E. Cape, KwaZulu-Natal) 1834 Pugill. 6: 8, T. 13
E. ghellinckii Lem. South Africa (E. Cape, KwaZulu-Natal) 1867 Ill. Hort. 14, misc.: 80
E. gratus Prain Malawi, Mozambique 1916 Bull. Misc. Inform.: 181
E. heenanii R.A. Dyer South Africa (Mpumalanga), Swaziland 1972 Bothalia 10(4): 539546
E. hildebrandtii A. Braun & Bouch Kenya, Tanzania 1874 Ind. Sem. Hort. Berol.: 8
E. hirsutus P.J.H. Hurter South Africa (Limpopo) 1996 S. African J. Bot. 62(1): 4648, fig. 1
E. horridus (Jacq.) Lehm. South Africa (E. Cape) 1834 Pugill. 6: 14
E. humilis I. Verd. South Africa (Mpumalanga) 1951 Bothalia 6: 220241, pl. 3
E. inopinus R.A. Dyer South Africa (Limpopo) 1964 Bothalia 8(2): 169170
E. ituriensis Bamps & Lisowski Dem. Rep. of Congo, Uganda 1990 Mem. New York Bot. Gard. 57: 152155
E. kisambo Faden & Beentje Kenya 1989 Utafiti 2(1): 710
E. laevifolius Stapf & Burtt Davy South Africa (E. Cape, KwaZulu-Natal, 1926 Fl. Transvaal & Sw. 1: 40, 99
Limpopo Province, Mpumalanga),
Swaziland
E. lanatus Stapf & Burtt Davy South Africa (Mpumalanga) 1926 Fl. Transvaal & Sw. 1: 40, 90, Fig. 4D
E. latifrons Lehm. South Africa (E. Cape) 183738 Tijdschr. Natuurl. Gesch. Physiol. 4: 424, T. 9
X00app1.qxd
E. laurentianus De Wild. Angola, Dem. Rep. of Congo 1903 Etudes Fl. Bas et Moy Congo 1: 10, T. 25
28/11/03
E. lebomboensis I. Verd. Mozambique, South Africa 1949 Fl. Pl. Africa 27: pl. 10781079
(KwaZulu-Natal), Swaziland
E. lehmannii Lehm. South Africa (E. Cape) 1834 Pugill. 6: 14
E. longifolius (Jacq.) Lehm. South Africa (E. Cape) 1834 Pugill. 6: 14
E. mackenziei L.E.Newton Sudan 2002 Bot. J. Linn. Soc. 140: 187192, fig. 16
3:52 pm
E. macrostrobilus S. Jones & Wynants N. Uganda 1997 Encephalartos 50: 1317, figs. 1, 2
E. manikensis (Gilliland) Gilliland Mozambique, Zimbabwe 1939 Proc. Rhodesia Sci. Assoc. 37: 133134
E. marunguensis Devred Dem. Rep. of Congo 1958 Bull. Soc. Roy. Bot. Belgique 91: 10410

E. middelburgensis Vorster South Africa (Mpumalanga) 1989 S. African J. Bot. 55: 122126
Page 229
E. msinganus Vorster South Africa (KwaZulu-Natal) 1996 S. African J. Bot. 62(2): 6770
E. munchii R.A. Dyer & I. Verd. Mozambique 1969 Kirkia 7: 147158
E. natalensis R.A. Dyer & I. Verd. South Africa (KwaZulu-Natal) 1951 Bothalia 6(1): 205211, pl. 13
E. ngoyanus I. Verd. South Africa (KwaZulu-Natal), 1949 Fl. Pl. Africa 27: pl. 1053, 1054
Swaziland
E. nubimontanus P.J.H. Hurter South Africa (Limpopo) 1995 Phytologia 78(6): 409410, fig. 1
E. paucidentatus Stapf & Burtt Davy South Africa (Mpumalanga), Swaziland 1926 Fl. Transvaal & Sw. 1: 40 & 99, fig. 4A
E. poggei Asch. Dem. Rep. of Congo 1878 Verh. Bot. Vereins Prov. Brandenburg 20: 3536
E. princeps R.A. Dyer South Africa (E. Cape) 1965 J. S. African Bot. 31(2): 111112, pl. 19
E. pterogonus R.A. Dyer & I. Verd. Mozambique 1969 Kirkia 7: 147158
E. relictus P.J.H. Hurter Swaziland 2001 Bothalia 31(2): 197199
E. schaijesii Malaisse, Sclavo & Crosiers Dem. Rep. of Congo 1993 Bull. Jard. Bot. Belg. 62: 215219
E. schmitzii Malaisse Dem. Rep. of Congo, Zambia 1969 Bull. Jard. Bot. Belg. 39: 401406
E. sclavoi De Luca, D.W. Stev. & A. Moretti Tanzania 1989 Delpinoa 2930: 35
E. senticosus Vorster South Africa (KwaZulu-Natal), 1996 S. African J. Bot. 62(2): 7679
Swaziland
E. septentrionalis Schweinf. Sudan, Uganda 1871 Bot. Zeitung Berlin 29: 334
E. tegulaneus Melville Kenya 1957 Kew Bull. 1957: 249
229
subsp. powysii Miringu & Beentje Kenya 1999 J. East African Nat. Hist. 88: 35
continued
X00app1.qxd
230
E. transvenosus Stapf & Burtt Davy South Africa (Limpopo) 1926 Fl. Transvaal & Sw. 1: 40, 99, fig. 4B
28/11/03
E. trispinosus (Hook.) R.A. Dyer South Africa (E. Cape) 1965 J. S. African Bot. 31(2): 112116, pl. 20
E. turneri Lavranos & D.L. Goode Mozambique 1985 Garcia de Orta, sr. Bot. 7: 1114
E. umbeluziensis R.A. Dyer Mozambique, Swaziland 1951 Fl. Pl. Africa 28: pl. 1100
E. villosus Lem. South Africa (E. Cape, KwaZulu-Natal), 1867 Ill. Hort. 14, misc.: 79; 15: T. 557
Swaziland
3:52 pm
E. whitelockii P.J.H. Hurter Uganda 1995 Phytologia 78(6): 410411, fig. 3
E. woodii Sander South Africa (KwaZulu-Natal) 1908 Gard. Chron.: 257
(extinct in the wild)
Page 230
LEPIDOZAMIA Regel (1857) (2 species, Australia) Bull. Soc. Imp. Naturalistes Moscou 30: 182
L. hopei Regel Australia (Qld) 1876 Gartenflora 25: 36
K.D. Hill et al.

*L. peroffskyana Regel Australia (NSW, Qld) 1857 Bull. Soc. Imp. Naturalistes Moscou 30: 184
MACROZAMIA Miq. (1842) (40 species, Australia) Monogr. Cycad.: 35
M. cardiacensis P.I. Forst. & D.L. Jones Australia (Qld) 1998 Fl. Australia 48: 717
M. communis L.A.S. Johnson Australia (NSW) 1959 Proc. Linn. Soc. New South Wales 64: 98
M. concinna D.L. Jones Australia (NSW) 1998 Fl. Australia 48: 718
M. conferta D.L. Jones & P.I. Forst. Australia (Qld) 1994 Austrobaileya 4: 271273, fig. 1
M. cranei D.L. Jones & P.I. Forst. Australia (Qld) 1994 Austrobaileya 4: 273275, fig. 2
M. crassifolia P.I. Forst. & D.L. Jones Australia (Qld) 1994 Austrobaileya 4: 275276, fig. 3
M. diplomera (F. Muell.) L.A.S. Johnson Australia (NSW) 1959 Proc. Linn. Soc. New South Wales 84: 98
M. douglasii W. Hill ex F.M. Bailey Australia (Qld) 1883 Syn. Qld Fl. 500
M. dyeri (F. Muell.) C.A. Gardner Australia (WA) 1930 Enum. Pl. Austral. Occid.: 3
M. elegans K.D. Hill & D.L. Jones Australia (NSW) 1998 Fl. Australia 48: 718
M. fawcettii C. Moore Australia (NSW) 1884 J. & Proc. Roy. Soc. New South Wales 17: 120
M. fearnsidei D.L. Jones Australia (Qld) 1991 Austrobaileya 3: 481
X00app1.qxd
M. flexuosa C. Moore Australia (NSW) 1884 J. & Proc. Roy. Soc. New South Wales 17: 121
28/11/03
M. fraseri Miq. Australia (WA) 1842 Monogr. Cycad.: 37
M. glaucophylla D.L. Jones Australia (NSW) 1998 Fl. Australia 48: 718
M. heteromera C. Moore Australia (NSW) 1884 J. & Proc. Roy. Soc. New South Wales 17: 122
M. humilis D.L. Jones Australia (NSW) 1998 Fl. Australia 48: 719
M. johnsonii D.L. Jones & K.D. Hill Australia (NSW) 1992 Telopea 5(1): 31
3:52 pm
M. lomandroides D.L. Jones Australia (Qld) 1991 Austrobaileya 3: 483
M. longispina P.I. Forst. & D.L. Jones Australia (Qld) 1998 Fl. Australia 48: 717
M. lucida L.A.S. Johnson Australia (NSW, Qld) 1959 Proc. Linn. Soc. New South Wales 84: 102

M. macdonnellii (F. Muell. ex Miq.) A. DC. Australia (NT) 1868 Prodr. 16(2): 537
Page 231
M. macleayi Miq. Australia (Qld) 1868 Arch. Nerl. Sci. Exact. Nat. 3(5): 250
M. miquelii (F. Muell.) A. DC. Australia (Qld) 1868 Prodr. 16(2): 535
M. montana K.D. Hill Australia (NSW) 1998 Fl. Australia 48: 717
M. moorei F. Muell. Australia (Qld) 1881 Austral. Chem. Drugg. 4: 84
M. mountperriensis F.M. Bailey Australia (Qld) 1886 Syn. Qld Fl., suppl. 1: 50
M. occidua D.L. Jones & P.I. Forst. Australia (Qld) 1994 Austrobaileya 4: 278, fig. 5
M. parcifolia P.I. Forst. & D.L. Jones Australia (Qld) 1994 Austrobaileya 4: 279281, fig. 6
M. pauli-guilielmi W. Hill & F. Muell. Australia (Qld) 1859 Fragm. 1: 86
M. platyrhachis F.M. Bailey Australia (Qld) 1898 Qld Agric. J. 3: 356
M. plurinervia (L.A.S. Johnson) D.L. Jones Australia (NSW, Qld) 1991 Austrobaileya 3: 484
M. polymorpha D.L. Jones Australia (NSW) 1998 Fl. Australia 48: 718
M. reducta K.D. Hill & D.L. Jones Australia (NSW) 1998 Fl. Australia 48: 718
*M. riedlei (Gaudich.) C.A. Gardner Australia (WA) 1930 Enum. Pl. Austral. Occid.: 3
M. secunda C. Moore Australia (NSW) 1884 J. & Proc. Roy. Soc. New South Wales 17: 122
M. serpentina D.L. Jones & P.I. Forst. Australia (Qld) 2001 Austrobaileya 6(1): 9092, fig. 2021
M. spiralis (Salisb.) Miq. Australia (NSW) 1842 Monogr. Cycad.: 36
M. stenomera L.A.S. Johnson Australia (NSW) 1959 Proc. Linn. Soc. New South Wales 84: 106
M. viridis D.L. Jones & P.I. Forst. Australia (Qld) 1994 Austrobaileya 4: 281283, fig. 7
231
continued
X00app1.qxd
232
MICROCYCAS (Miq.) A. DC. (1868) (1 species, Cuba) Prodr. 16(2): 538
28/11/03
*Microcycas calocoma (Miq.) A. DC. W. Cuba 1868 Prodr. 16(2): 538
STANGERIA T. Moore (1853) (1 species, Southern Africa) Hook. Kew J., misc. 5: 228229
3:52 pm
*Stangeria eriopus (Kunze) Baill. South Africa (E. Cape, kwaZulu-Natal) 1892 Hist. Pl. 12: 68
ZAMIA L. (1763) (58 species, South, Central & North America) Sp. Pl., ed. 2, 2: 1659
Page 232
Z. acuminata Oerst. ex Dyer Costa Rica, Nicaragua, N. Panama 1884 in Hemsl., Biol. Cent.-Amer., Bot. 3: 190195
Z. amazonum D.W. Stev. Brazil, Colombia, (Amazonas, Choc), 2001 Fl. Colombia 21: 33
Ecuador, Peru, S. Venezuela
K.D. Hill et al.

Z. amblyphyllidia D.W. Stev. Cuba, Jamaica, Puerto Rico 1987 Fairchild Trop. Gard. Bull. 42(3): 26
Z. amplifolia hort. Bull ex Mast. Colombia (Valle) 1878 Gard. Chron., ser. 2, 10(261): 810
Z. angustifolia Jacq. Bahamas 1789 Collectanea 3: 263265, pl. 636
Z. boliviana (Brongn.) A. DC. N. Bolivia, Brazil. 1868 Prodr. 16(2): 540
Z. chigua Seem. Colombia (Choc), Panama (Chiriqui) 1854 Bot. Voy. Herald 6: 201203, t. 43
Z. cremnophila Vovides, Schutzman & Dehgan Mexico (Tabasco) 1988 Bot. Gaz. 149(3): 351354, fig. 45
Z. cunaria Dressler & D.W. Stev. Panama 1993 Brittonia 45(1): 56
Z. disodon D.W. Stev. & Sabato Colombia (Antioquia), Peru (Hunuco) 2001 Fl. Colombia 21: 38
Z. dressleri D.W. Stev. Panama 1993 Brittonia 45(1): 6
Z. elegantissima Schutzman, Vovides & Panama 1998 Phytologia 85(3) (publ. Apr. 2000): 137145
R.S. Adams
Z. encephalartoides D.W. Stev. Colombia (Santander) 2001 Fl. Colombia 21: 40
Z. fairchildiana L.D. Gmez Costa Rica, Panama 1982 Phytologia 50: 401404
Z. fischeri Miq. Mexico (Quertaro, San Luis Potosi) 1845 in Lem., Hort. Vanhoutt. 1: 20
Z. furfuracea L. f. Mexico (Veracruz) 1789 in Aiton, Hortus Kew. 3: 477
Z. gentryi Dodson Ecuador 1998 Novon 8(1): 1214, fig. 1a-f
X00app1.qxd
Z. herrerae Caldern & Standl. El Salvador, Guatemala, Honduras, 1924 Proc. Wash. Acad. Sci. 14(4): 9399, fig. 1
28/11/03
Mexico (Chiapas)
Z. hymenophyllidia D.W. Stev. Colombia (Amazonas), Peru (Loreto) 2001 Fl. Colombia 21: 43
Z. inermis Vovides, J.D. Rees & Vzq. Mexico (Veracruz) 1983 Flora de Veracruz 26: 2224
Torres
Z. integrifolia L. f. Bahamas, Cayman Islands, Cuba, 1789 in Aiton, Hortus Kew. 3: 477479
3:52 pm
USA (Florida, Georgia)
Z. ipetiensis D.W. Stev. Panama 1993 Brittonia 45(1): 79
Z. kickxii Miq. Cuba 1842 Monogr. Cycad.: 71, t. 8

Z. lacandona Schutzman & Vovides Mexico (Chiapas) 1998 Novon 8(4): 441446, fig. 13
Page 233
Z. lecointei Ducke Brazil, Colombia (Amazonas, 1915 Arch. Jard. Bot. Rio de Janeiro 1: 910,
Antioquia), Peru (Loreto), S. Venezuela pl. 12
Z. loddigesii Miq. Guamemala, Mexico (Hidalgo, 1843 Tijdschr. Natuurl. Gesch. Physiol. 10: 7273
Oaxaca, Veracruz)
Z. lucayana Britton Bahamas (Long Island) 1907 Bull. New York Bot. Gard. 5(18): 311318
Z. macrochiera D.W. Stev. Peru 2004 [see Stevenson, Chapter 14 this volume]
Z. manicata Linden ex Regel Colombia (Antioquia), Panama (Darien) 1876 Trudy Imp. S.-Petersburgsk. Bot. Sada 4: 310, t.
926, fig. e
Z. melanorrhachis D.W. Stev. Colombia (Amazonas, Cordoba, 2001 Fl. Colombia 21: 55
Santander)
Z. montana A. Braun Colombia (Antioquia) 1875 Monatsber. Knigl. Preuss. Akad. Wiss. Berlin:
376
Z. monticola Chamb. Guatemala 1926 Bot. Gaz. 81: 219, 223, fig. 13
Z. muricata Willd. Colombia (Boyac, Guajira, Meta, 1806 Sp. Pl. ed. 4, 4: 847848
Santander),Venezuela (Guarico)
Z. neurophyllidia D.W. Stev. Costa Rica, S. Nicaragua, N. Panama 1993 Brittonia 45(1): 10, fig. 5
Z. obliqua A. Braun Colombia (Antioquia, Choc, Valle), 1875 Monatsber. Knigl. Preuss. Akad. Wiss.
S. Panama Berlin: 376
233
Z. paucijuga Wieland Mexico 1916 American Fossil Cycads 2: 212, fig. 86
continued
X00app1.qxd
234
Z. picta Dyer Guatemala, Mexico (Chiapas) 1884 in Hemsl., Biol. Cent.-Amer., Bot. 3: 194
28/11/03
Z. poeppigiana Mart. & Eichler Brazil, Colombia (Huila, Meta, 1863 Fl. Bras. 4(1): 414416, pl. 108109
Tolima), Ecuador, Peru
Z. polymorpha D.W. Stev., A. Moretti & Vzq. Belize, Mexico (Campeche, Quintana 199596 Delpinoa n.s. 3738 (publ. 1998): 38, fig. 1
Torres Roo, Yucatan)
Z. portoricensis Urban Puerto Rico 1899 Symb. Antill. 1: 291
3:52 pm
Z. prasina Bull Belize 1881 Retail List: 20
Z. pseudomonticola L.D. Gmez W. Costa Rica 1982 Phytologia 50(6): 401404
Z. pseudoparasitica Yates in Seem. N. Panama 1854 Bot. Voy. Herald 6: 201203, 253
*Z. pumila L. Cuba, Dominican Republic, Puerto Rico 1763 Sp. Pl., ed. 2, 2: 1659
Page 234
Z. purpurea Vovides, J.D. Rees & Vzq. Torres Mexico (Oaxaca, Veracruz) 1983 Flora de Veracruz 26: 2831
Z. pygmaea Sims Cuba (W. Cuba, Isla de la Juventud) 1815 Bot. Mag. 43: t. 1741
Z. roezlii Linden Colombia (Amazonas, Choc, Nario, 1873 Catalogue des Plantes Nouvelles 9: 10
K.D. Hill et al.

Valle), Ecuador
Z. skinneri Warsz. ex A. Dietrich C. Panama, N. Panama 1851 Allg. Gartenzeitung 19: 146, fig. s.n.
Z. soconuscensis Schutzman, Vovides & Dehgan Mexico (Chiapas) 1988 Bot. Gaz. 149(3): 347351
Z. spartea A. DC. Mexico (Oaxaca) 1868 Prodr. 16(2): 539
Z. standleyi Schutzman Honduras 1989 Syst. Bot. 14(2): 214219
Z. tuerckheimii Donn. Sm. Guatemala 1903 Bot. Gaz. (Crawfordsville) 35(1): 8, pl. 1
Z. ulei Dammer W. Brazil, Colombia, Ecuador, Peru 1907 Verh. Bot. Vereins Prov. Brandenburg 47:
117118
Z. urep Walln. Peru 1996 Linz. Biol. Beitr. 28(2): 1056
Z. variegata Warsz. Guatemala, Mexico (Chiapas) 1845 Allg. Gartenz. 32: 252253
Z. vazquezii D.W. Stev., Sabato, A. Moretti & Mexico (Veracruz) 199596 Delpinoa n.s. 3738 (publ. 1998): 917
De Luca
Z. verschaffeltii Miq. Mexico (Chiapas) 1870 Verh. Kon. Ned. Akad. Wetensch., Afd. Natuurk.
2(4): 31
Z. wallisii A. Braun Colombia (Antioquia) 1875 Monatsber. Knigl. Preuss. Akad. Wiss. Berlin:
376
X00app1.qxd 28/11/03 3:52 pm Page 235
Appendix 1: The World List of Cycads 235
References and Further Reading

Brummit, R.K. and Powell, C.E. (1992) Authors of Plant Names. Royal Botanic Gardens,
Kew, UK, 732 pp.
Hill, K.D. and Stevenson, D.W. (2004) The world list of cycads. In: Stevenson, D.W. (ed.)
Proceedings of the Fifth International Conference on Cycad Biology. Memoirs of the New York
Botanical Garden (in press).
Lawrence, G.H.M., Buchheim, A.F.G., Daniels, G.S. and Dolezal, H. (eds) (1968) Botanico-
Periodicum-Huntianum. Hunt Institute for Botanical Documentation, Pittsburgh,
Pennsylvania, 1063 pp.
Osborne, R. and Hendricks, J. (1985) A world list of cycads. Encephalartos 3, 1317.
Osborne, R. and Hendricks, J. (1986) A world list of cycads supplement. Encephalartos
5, 27.
Osborne, R. Stevenson, D.W. and Hill, K.D. (1999) The world list of cycads. In: Chen,
C.J. (ed.) Biology and Conservation of Cycads. Proceedings of the Fourth International Conference
on Cycad Biology. International Academic Publishers, Beijing, China, pp. 224239.
Stafleu, F.A. and Cowan, R.S. (1976, 1979, 1981, 1983) Taxonomic Literature, Second
Edition, Vols 14. (Regnum Vegetabile 94, 98, 105, 110).
Stevenson, D.W. and Osborne, R. (1993a) The world list of cycads. In: Stevenson, D.W.
and Norstog, K.J. (eds) The Biology, Structure, and Systematics of the Cycadales. Proceedings of
the Second International Conference on Cycad Biology, Palm & Cycad Societies of Australia
Limited, Milton, Queensland, Australia, pp 354364.
Stevenson, D.W. and Osborne, R. (1993b) The world list of cycads. Encephalartos 33:
1925.
Stevenson, D.W., Osborne, R. and Hendricks, J. (1990) A world list of cycads. In:
Stevenson, D.W. (ed.) The Biology, Structure, and Systematics of the Cycadales. Proceedings of
the Symposium CYCAD 87. Memoirs of the New York Botanical Garden 57, pp. 200206.
Stevenson, D.W., Osborne, R. and Hill, K.D. (1995) The world list of cycads. In: Vorster,
P. (ed.) Proceedings of the Third International Conference on Cycad Biology. Cycad Society of
South Africa, Stellenbosch, South Africa, pp. 5564.
X00app1.qxd 28/11/03 3:52 pm Page 236
X00app2.qxd 28/11/03 3:55 pm Page 237
Appendix 2: Glossary of Terms

Encountered in Cycad
Systematics
Roy Osborne1 and Terrence Walters2
1PO Box 244, Burpengary, Queensland, Australia;

2Montgomery Botanical Center, Miami, Florida, USA
Introduction
A large number of terms is encountered in the literature relating directly or indi-

rectly to cycad systematics. Many are technical terms found generally in botani-
cal descriptions but some (e.g. omnule, omnel, corruptule, corruptel) are exclusive
to cycad literature. Other terms (e.g. bulb, endosperm, frond) are used common-
ly but incorrectly with reference to cycads. Most terms are understood by readers
in context, but many have a precise definition that is not readily found in general
reference texts. Many terms encountered do not relate to plant descriptions per se,
but to techniques of statistical and chemical analysis and to the relation of cycads
to their environment. This glossary attempts to list and define all such terms, with
cycad-specific examples where appropriate. Common acronyms, abbreviations
and some Latin phrases are included.
Glossary
Notes: Primary entries and relevant cross-references are indicated by bold type.
Terms of the opposite meaning are indicated by Cf.
abaxial. Side of an organ facing away from a central axis, e.g. the lower side of
a leaf or leaflet. Cf. adaxial.
abscission. Process in which a corky cell layer forms across an axis, cutting off
water and nutrient supply to the distal portion and resulting in its loss, as in
237
X00app2.qxd 28/11/03 3:55 pm Page 238
loss of a leaf, leaflike organ or section of stem.

acaulescent. Without any evident trunk or stem. Cf. caulescent.
acropetal. Progressing in direction from promixal to distal zone; from base
to apex, e.g. as for the pollen dehiscence sequence in most cycad cones. Cf.
basipetal.
actinomorphic. Of a radially symmetric structure that can be halved in more
than one plane to give two halves that are mirror images of each other; of
seeds, radiospermic. Cf. zygomorphic.
actran optimization. Optimization technique used in cladistics, based on
accelerated transformation and favouring acquisition of characters, with
subsequent homoplasy accounted for by reversal. Cf. deltran optimiza-
tion.
aculeate. Having sharp prickles, as for the petiole of Cycas aculeata.
acuminate. Tapering to a protracted point, with sides somewhat concave, as for
the leaflets of Zamia acuminata. See also acute.
acute. Narrowing to end in a sharp point, as for the leaflets of many cycad
species. See also acuminate.
Adansonian. Using as many unweighted characters as possible to arrive at a
natural classification; 18th century naturalist Michel Adansons philosophy
predated modern computer-aided numerical taxonomy.
adaxial. Side of an organ facing towards a central axis, e.g. the upper side of a
leaf or leaflet. Cf. abaxial.
adventitious bud. Vegetative growth arising on a cycad stem, giving rise to
suckers at the base or to offsets on the stem.
affinis, affin., aff. Akin or allied to; used in reference to a specimen similar to
a known species, but showing some differences.
allele. One of two or more alternative forms of a gene at a given locus.
allopatric. Of two or more taxa not overlapping in their distribution. Cf. sym-
patric.
alpha-taxonomy. Descriptive taxonomy, based exclusively on morphological
parameters.
amphistomatic. Of leaves and leaflets with stomata on both surfaces. Cf.
hypostomatic.
anatomy. Study of the internal structure of organisms and their component
parts.
anemophily. Wind-effected pollen transfer. Cf. entomophily.
angle(s) of insertion. See leaflet angle(s) of insertion.
ANOVA. Analysis of variance, a tool used in statistics to apportion observed vari-
ance into probable causes.
antherozoid (more commonly spermatozoid). Motile male gamete of lower
plants, cycads, Ginkgo, etc.
apex. Tip; proximal or distal end of an organ such as a shoot, leaf or root.
apical meristem. Zone of actively dividing but as yet undifferentiated tissue
at a shoot or root apex.
apogeotropic, ageotropic, apogravitropic. Negatively geotropic; nega-
X00app2.qxd 28/11/03 3:55 pm Page 239
Appendix 2: Glossary 239
tively gravitropic; developing in an orientation contrary to gravitational

force, as with cycad coralloid roots.
apomixis. Production of viable seeds without any apparent fertilization process.
apomorphy. Derived character or character state. See also autapomorphy,
homology, plesiomorphy, symplesiomorphy, synapomorphy.
arborescent. Tree-like in habit.
armed. Having spines or prickles. Cf. inerm.
articulate. Separated by a node or joint; e.g. leaflets of Ceratozamia, Chigua,
Microcycas and Zamia are articulated at their base.
auctoris, auctorum, auct. Obsolete term previously used in taxonomy when
an author had applied an incorrect name.
autapomorphy. Derived character state unique to a terminal taxon in a par-
ticular data set. Note that an autapomorphy at a given hierarchical level may
be a synapomorphy at a less inclusive level. See also apomorphy, homolo-
gy, plesiomorphy, symplesiomorphy, synapomorphy.
author, authority. Person who first publishes a valid name for a taxon. The
authors or authors name accompanies the taxonomic name, usually in
abbreviated form, e.g. Carl Peter Thunberg in Cycas revoluta Thunb.
autonym. Taxonomic name automatically established when a subdivision of a
taxon such as a genus or species is published. The infrageneric or infraspe-
cific taxon containing the type of the genus or species carries the same name
or epithet as the respective genus or species, e.g. the publication of varieties
within Dioon edule established the autonym Dioon edule var. edule.
axis. Central line of development of a plant; the main stem.
basionym. Combination of a name in the same rank as first validly published,
e.g. genus and specific epithet in case of a species name, or generic name and
infrageneric epithet in case of an infrageneric name. Basionyms are cited
only when there is a recombination of genus name and epithet, e.g. the
basionym in the publication of Encephalartos caffer (Thunb.) Lehm. is Zamia
caffra Thunb.
basipetal. Progressing in direction from distal to proximal; from apex to
base. Cf. acropetal.
bifid. Deeply notched or cleft for more than half the length.
bifurcate. Forked in a Y-shaped manner, usually applying to leaves, leaflets or
stems.
binomial. Genus name and specific epithet for any species.
biome. Broad vegetational subdivision of some biogeographical realm, e.g.
forest, grassland, desert.
bipinnate. Twice pinnate; of a compound leaf having both first order and
second order divisions, i.e. pinnae and pinnules, as in both Bowenia species
and some Cycas species. See also multipinnate.
BMAA. -n-methylamino-L-alanine, a neurotoxin found in cycad tissues.
bootstrap value. Statistical estimate of confidence in a dendrogram or phy-
logenetic tree, obtained from repeated tree calculations by randomly elim-
inating selected characters while duplicating others to keep the total number
X00app2.qxd 28/11/03 3:55 pm Page 240
of characters constant.
bract. Leaf-like structure subtending an axillary bud or shoot, usually with a
protective function. Occasionally misapplied to cycad cataphylls.
bulb. Storage stem of limited longitudinal growth enveloped in fleshy leaf bases;
misapplied to the bulbous trunks of some Cycas species; occasionally mis-
applied to cycad suckers.
bulbous. Swollen to an almost spherical shape, as for the trunk base of Cycas
pachypoda and several other Asian cycads.
bulla (plural bullae). Bubble, blister or vesicle. Commonly also used in refer-
ence to the expanded shield-like distal portion of some cycad sporophylls
such as Encephalartos megasporophylls. Although the derivation is technically
incorrect, this term has become widely entrenched in cycad literature and its
continued use is recommended.
caducous. Deciduous at an early stage or prematurely.
caespitose. Turf-shaped and forming a clump, as in many suckering cycads,
e.g. Encephalartos cupidus.
callous (adjective). Of the distinctive swollen tissue, often coloured, formed at
the point of insertion of Macrozamia leaflets on to the rachis. Often confused
with callus.
callus (noun). Mass of hardened, thickened or undifferentiated parenchymatous
tissue, e.g. as formed at the base of a cutting prior to root formation; undif-
ferentiated cellular mass arising in tissue culture. See also callous.
cataphyll. Modified leaf, much reduced and thickened, serving to protect the
apical meristem in cycads and usually produced in flushes preceding the
emergence of cones or leaves.
caudex. Thick stem or trunk, often at least partially subterranean.
caulescent. Having a trunk or stem. Cf. acaulescent.
centromere. Region of a chromosome where spindle microtubules are
attached during nuclear division. See also satellite.
chalaza. Proximal end of a seed; its point of attachment. Cf. micropyle.
channelled. With raised longitudinal edges to form a channel, as in the adaxial
petiole surface of many cycads.
character. Identifiable and hereditable trait which can be used in comparing
one taxon with another.
character states. Changeable attribute of any given character, e.g. if sarcotes-
ta colour is a character, then red and yellow are character states. Many char-
acter states are simply recorded as present or absent.
chartaceous. Papery in texture. See also membranous, papyraceous.
chlorophyll. Green plant pigment in the cells of some bacteria and in plant
chloroplasts, which captures energy from sunlight; an electron donor in
photosynthesis.
chloroplast. Plant cellular organelle in which photosynthesis occurs. See also
chlorophyll.
chromatid. One of the pair of threadlike forms of each chromosome.
chromosome. Submicroscopic filamentous strand of DNA and associated pro-
X00app2.qxd 28/11/03 3:55 pm Page 241
teins in the nucleus of all cells, by which hereditary information is transmit-

ted from generation to generation. See also centromere, chromatid.
CI. Consistency index, a measure of the amount of homoplasy for a charac-
ter in a cladogram. See also RC, RI.
circinate. Rolled in a coil-like manner with the apex innermost, as in leaflets of
emerging leaves of Cycas species, hence C. circinalis. See also conduplicate,
inflexed, ptyxis, reflexed, vernation.
CITES. Convention on International Trade in Endangered Species, a United
Nations treaty which sets out a conservation regulatory process between the
signatory countries.
clade. One particular monophyletic branch in a cladogram.
cladistics. Method of classification that groups taxa hierarchically and parsi-
moniously into nested sets according to their synapomorphies; the output
conventionally presented in the form of a cladogram.
cladogram. Tree diagram based on parsimony analysis showing taxa grouped
hierarchically in nested sets according to their synapomorphies. A clado-
gram has no connotation of ancestry and has no implied time axis. See also
dendrogram, phenogram, phylogram.
classification. Grouping of taxa or taxonomic groups into categories accord-
ing to an overall plan.
clone. Set of genetically identical individuals produced vegetatively from the
same progenitor.
coevolution. Concurrent evolution of two different but interdependent organ-
isms, as in the case of a cycad and its insect pollinator.
collar. In cycads, a coloured or textured band at base of a rachis, e.g. as seen in
Encephalartos lehmannii leaves; a glandular swelling at the base of a leaflet, e.g.
as seen in Zamia manicata leaves.
combinatio nova, comb. nov. Nomenclatural new combination usually
made by transferring a specific epithet from one genus to another, e.g.
Dyerocycas micholitzii (Dyer) Nakai was a combinatio nova from Cycas micholitzii
Dyer.
community. Total of all living species in a particular habitat.
conduplicate. Folded together lengthwise, usually in two equal halves, as in
leaflets of emergent Stangeria leaves. See also circinate, inflexed, ptyxis,
reflexed, vernation.
cone. Reproductive structure of gymnosperms; organized collection of sporo-
phylls on a central axis. See also strobilus, megasporangiate and
microsporangiate strobili.
confer, cf. Latin, meaning compare.
congeneric. Belonging to the same genus; e.g. it is now thought that Epicycas is
congeneric with Cycas.
consensus tree. Cladogram representing the clades found in all the most
parsimonious trees of an analysis, often from a large number of possible res-
olutions.
conspecific. Belonging to the same species; e.g. some workers believe that
X00app2.qxd 28/11/03 3:55 pm Page 242
Encephalartos altensteinii and E. natalensis are conspecific.

contiguous. Touching or neighbouring, as in a common border between two
countries.
contractile. Of roots, and occasionally stems, which contract so as to pull the
stem apex lower into the ground.
convergence. Evolutionary process where dissimilar organs or organisms show
strong superficial similarities.
coralloid root. Club-shaped apogravitropic (apogeotropic) root with the
potential for hosting symbiotic Cyanobacteria.
coriaceous. Leathery in texture, as for the leaflets of many cycad species.
corruptel. Kernel of a corruptule (Grobbelaar, 2002).
corruptule. Unfertilized ovule that is superficially indistinguishable from a true
seed (Grobbelaar, 2002). See also omnule.
CSG. Cycad Specialist Group, a conservation-based group within the IUCN.
cuneate. Wedge-shaped and attached at the narrow end.
Curculionidae. Weevil family of insects.
cuticle. In plants, the outer waxy layer of an epidermis, comprising mainly
cutin with lesser amounts of overlaying waxes and polysaccharides.
cutin. Complex polymeric mixture of fatty acids and phenolic compounds; the
main component of a plant cuticle.
Cyanobacteria. Group of bacteria capable of photosynthesis, previously
known as blue-green algae, found in coralloid roots of cycads.
Cycad Action Plan. Series of proposals for cycad conservation drafted by the
CSG.
Cycadales. Gymnosperm plant order containing all extant and extinct cycads.
cycads. Members of the plant order Cycadales.
cycasin. Toxic MAM glycoside found only in cycad tissue; methyl-
azoxymethanol--D-glucopyranoside. See also macrozamin.
cymbiform. Boat-shaped.
cytology. Study of the structure, physiology and reproduction of cells.
DAF. DNA amplification fingerprinting, a technique used in DNA analysis.
decumbent. Of stems, lying along the ground but turning upwards distally. See
also procumbent, prostrate.
decurrent. Extending basipetally from the point of insertion, as for Lepidozamia
leaflets where the leaflet base extends downwards along the rachis.
dehiscence. Rupturing process where the contents of a plant structure are
released, as in pollen shedding.
deltran optimization. Optimization technique used in cladistics, based on
delayed transformation and favouring independent gains rather than acqui-
sition and reversal. Cf. actran optimization.
dendrogram. Generic term for any kind of tree diagram, including clado-
gram, phenogram and phylogram.
dentate. Having sharp indentations or teeth along the edge of a structure and
perpendicular to its margin. See also serrate.
denticulate. Finely dentate. See also serrate, serrulate.
X00app2.qxd 28/11/03 3:55 pm Page 243
determinate. With growth of an axis ceasing at a particular stage. Cf. inde-

terminate.
dichotomous. Branching into two equal parts, as in forked branching of stems
of Cycas elongata, leaflets of Macrozamia stenomera or megasporophyll lobes of
Cycas segmentifida. Also used in reference to branching patterns in dendro-
grams.
dimorphic. Having two different forms. See also homomorphic, monomor-
phic, polymorphic.
dioecious. Having male and female reproductive parts on different plants. All
extant cycads are dioecious.
diploid. Having a pair of chromosomes of each kind. Cf. haploid.
discriminant analysis. Statistical technique used in separating discrete sets of
objects.
disjunct. Separated geographically, pertaining to populations of the same taxon
occurring in more than one geographical area.
distal. Furthest away from the centre of a structure or, more commonly, furthest
from its point of attachment; terminal; apical. Cf. proximal.
DNA. Deoxyribonucleic acid; desoxyribonucleic acid, the molecule in which an
organisms genetic definition is determined by the sequence of nitrogenous
bases projecting along a backbone of sugar moieties linked by phosphodi-
ester bonds.
ecology. Study of relationships of living organisms to each other and to their
physical and chemical environment.
ecosystem. Sum of all biological, chemical and physical components of an
area and their interaction.
edaphic. Relating to the soil environment, e.g. soil quality, pH.
electrophoresis. Analytical technique in which an electrical gradient is used to
separate compounds, especially proteins, according to their charge and
molecular mass.
elliptic. Of a two-dimensional structure that is widest near the middle but nar-
rowed towards each rounded end. See also obovate, ovate.
elongate. Lengthened or drawn out, e.g. as for the megasporophyll apical spines
of Cycas elongata.
embryo. In seeds, diploid tissue arising from the zygote, until the time of ger-
mination.
embryology. Study of the morphology and development of embryos,
pollen, megagametophytes and microgametophytes.
endemic. Restricted in occurrence to a particular area; e.g. Microcycas calocoma is
endemic to Cuba.
endocarp. Internal layer of a fruit wall; misapplied to the endotesta in cycads.
endosperm. Triploid (or more) nutritive tissue within angiosperm seeds; mis-
applied to the megagametophyte of cycad seeds.
endotesta. Inner layer of a seed coat.
ensiform. Sword-shaped, as approximated by the leaflets of some cycads, e.g.
Cycas media subspecies ensata.
X00app2.qxd 28/11/03 3:55 pm Page 244
entire. With a continuous margin; not toothed or lobed; without incisions of any
kind.
entomophily. Pollen transfer brought about by insects. Cf. anemophily.
eophyll. First leaf produced by a seedling. See also euphyll.
epidermis. Outermost primary cellular layer of an organism. Cf. hypoder-
mis.
epigeous. Occurring above soil level, as for the stems of arborescent cycads.
Cf. hypogeous.
epiphytic. Growing on another plant non-parasitically, or on some other ele-
vated support.
epithet. Second word of a botanical binomial; the specific or subspecific
epithet.
etymology. Dealing with the origin of words. Species descriptions commonly
give the etymology for the specific epithet.
eukaryote. Organism having cells with a true nucleus, as for all plants and
animals.
euphyll. Conventional foliage leaf, not modified in any way. See also eophyll.
ex situ. Of plants or plant collections in cultivation outside their natural habitat.
exclamation mark (!). Used in taxonomic literature to signify that a particu-
lar herbarium specimen has been examined by the writer. Cf. non visus.
exsiccatum (plural exsiccata). Dried specimen material; plantae exsic-
catae refers to herbarium specimens.
extant. Existing at the present time. Cf. extinct.
extinct. No longer existing; e.g. Encephalartos woodii is believed to be extinct in the
wild. Amongst the very many extinct cycad genera are: Ceratozamites,
Crossozamia, Dioonites, Eostangeria, Palaeocycas, Pseudoctenis, Ticoa and Zamites. Cf.
extant.
facet. Flattened terminal, median or lateral section of a sporophyll bulla
defined by ridges; in cycads the terminal facet is often rhombic in outline.
falcate. Curved in a sickle shape, as for the leaflets of Cycas falcata.
family. Taxonomic rank below order but above genus. Cycad families com-
prise the Cycadaceae, Stangeriaceae and Zamiaceae, with some authors sep-
arating Boweniaceae as a fourth family.
farinaceous. Having the texture of flour or similar starchy material; sometimes
referring to a dusty covering.
fasciculate. Arranged in a whorl, as for the leaflets of the common form of
Ceratozamia hildae.
ferrugineous. Rusty brown in appearance, as for the tomentum at the stem
apex of Cycas ferruginea.
fertilization. Union of male and female gametes resulting in a zygote.
fide. According to; by the assurance of. See also sensu.
filius, fil.or f. Son of, used in reference to father-and-son authors of taxa.
flabellate. Fan-shaped, as in the megasporophyll blade of some Cycas species.
flavonoid. Large range of plant secondary metabolites comprising specific phe-
nolic compounds, usually occurring as glycosides, often coloured as in many
X00app2.qxd 28/11/03 3:55 pm Page 245
plant pigments.
flexuose. In a zig-zag shape, as used in allusion to the much twisted rachis of
Macrozamia flexuosa leaves.
floccose. Bearing soft, uneven hairs.
foliiform. Leaf-like.
Fourier transformation. Mathematical process converting state space to fre-
quency space, usually applied to time series data to find periodic signals.
frond. Having the form of a fern leaf; commonly misapplied to other pinnately
compound leaves, as in palms and cycads.
funiculus (plural funiculi). Attachment stalk of an ovule.
furfuraceous. Covered with bran-like scales or powder, as for the emergent
foliage of Cycas furfuracea and Zamia furfuracea.
fusiform. Spindle-shaped; narrowed at each end and swollen centrally.
gamete. Mature male (sperm) or female (egg) cell able to take part in reproduc-
tion.
gametophyte. Haploid structure or tissue; that component of the life cycle
which produces gametes.
gene. Specific part of the DNA molecule which comprises the basic unit of
inheritance, each prescribing a code for the synthesis of a specific protein.
genealogy. Line of descent from an ancestor through its derivatives; the pedi-
gree of an organism. See also lineage.
genetics. Study of genes and genetic processes.
genome. Entire genetic complement of an organism or clone, as defined by its
haploid chromosome complement. See also genotype.
genotype. Genetic constitution of an organism or clone. See also genome. Cf.
phenotype.
genus (plural genera). Taxonomic rank below family but above species.
Extant cycad genera comprise Bowenia, Ceratozamia, Chigua, Cycas, Dioon,
Encephalartos, Lepidozamia, Macrozamia, Microcycas, Stangeria and Zamia.
Aulacophyllum is now considered congeneric with Zamia while Dyerocycas and
Epicycas are congeneric with Cycas.
genus novum, gen. nov. Citation at the time a new genus is first described. See
also species nova.
glabrous. Of a smooth surface, without hair of any kind.
glaucous. Of a surface covered by a bluish grey waxy or powdery bloom, as for
the foliage of Macrozamia glaucophylla. See also pruinose.
GPS. Global Positioning System; worldwide electronic satellite-linked system for
establishing latitude, longitude and altitude.
gymnosperms. Loosely-related (polyphyletic) group of seed-bearing but
non-flowering plants, including cycads, conifers, Ephedra, Ginkgo, Gnetum,
Welwitschia together with various extinct taxa. All bear ovules, later seeds,
without any enveloping pericarp.
habit. Growth form of an organism.
habitat. Environment where a plant or animal exists naturally.
haploid. Having only one set of chromosomes. Cf. diploid.
X00app2.qxd 28/11/03 3:55 pm Page 246
hirsute. Covered with short coarse hairs, as for the leaves of Encephalartos hirsu-
tus. See also lanate, pilose, pubescent, sericeous, tomentose.
histology. Study of biological tissues.
holotype, holo. Single herbarium specimen or illustration of the type collec-
tion used or designated by the author of the name. See also isotype, lecto-
type, neotype, paratype.
homology. Features having a common origin but not necessarily the same func-
tion, e.g. cycad leaves, cataphylls and sporophylls. See also apomorphy,
autoapomorphy, homoplasy, plesiomorphy, symplesiomorphy,
synapomorphy.
homomorphic. Of a population, uniform in morphology. See also dimorphic,
monomorphic, polymorphic.
homoplasy. Mistaken homology; superficial similarity between characters or
character states due to convergent or parallel evolution or by reversal; e.g. the
occurrence of bipinnately compound leaves in Bowenia and Cycas.
hybrid. Natural or artificially produced plant resulting from a cross of geneti-
cally dissimilar parents, commonly between two different species.
hypocotyl. Embryonic axis at the time of seed germination; that part from
which the first leaf emerges. See also radicle.
hypodermis. Cellular layer immediately internal to an epidermis.
hypogeous. Occurring below soil level; subterranean, as for the stems of
Stangeria eriopus, many Zamia species and most Macrozamia Section Parazamia
species. Cf. epigeous.
hypostomatic. Of leaves and leaflets with stomata on the abaxial surfaces
only. Cf. amphistomatic.
ICBN. International Code of Botanical Nomenclature, e.g. St Louis Code of
2000.
idioblast. Specialized cell with inclusions, in cycads storing toxins possibly as a
herbivore deterrent; a cell without known function.
imbricate. Overlapping, as for leaflets of many Encephalartos species. See also
incubous, succubous.
imparipinnate. Of a leaf or leaflet where the rachis or rachilla terminates in
a solitary pinna or pinnule. Cf. paripinnate.
incertae sedis. Of uncertain placement in a classification system.
incubous. Arrangement in which a leaflet partially shields the leaflet above (the
next distal leaflet) when viewed from above (adaxially). Cf. succubous.
incurved. With apex bent or curving adaxially. Cf. recurved.
indeterminate. With growth of the axis continuing indefinitely. Cf. determi-
nate.
indumentum. Covering of trichomes or scales. See also tomentum.
ineditus, ined. Unpublished, often in reference to an anticipated new species
description.
inerm. Without spines or prickles; unarmed, as for the leaves of Zamia inermis.
Cf. armed.
inflexed. Bent longitudinally inwards (adaxially) as in emerging leaves of many
X00app2.qxd 28/11/03 3:55 pm Page 247
Zamia species. See also circinate, conduplicate, inflexed, ptyxis,

reflexed, vernation.
insertion angle(s). See leaflet angle(s) of insertion.
integument. Outer covering of an ovule, forming the layers of seed coat at
maturity. See also sarcotesta, sclerotesta.
interstitial region. Intervening zone between areas of defined structure.
involute. With margins rolled inwards (adaxially). Cf. revolute.
isoenzymes, isozymes. Differing molecular forms of an enzyme that serve
the same function.
isotype, iso. Herbarium specimen that is a duplicate of (having the same col-
lection details as) the holotype. This term also has other applications, e.g.
in immunology. See also lectotype, neotype, paratype.
ITS. Internal transcribed spacer region of a gene. ITS2 is often used for cycad
DNA analysis.
IUCN. The World Conservation Union. Originally, the International Union for
Conservation of Nature and Natural Resources, a United Nations organiza-
tion.
juvenile. Not fully developed; not yet capable of sexual reproduction. Cf.
mature.
karyology. Dealing with the characteristics of chromosomes, commonly in
terms of their number and morphology.
lamina (plural laminae). Flattened and expanded portion of a leaf, sporophyll
or similar structure; blade.
lanate. Woolly with long, intertwined curly hairs, as for the stem apex, emergent
leaves and cones of Encephalartos lanatus. See also hirsute, pilose, pubes-
cent, sericeous, tomentose.
lanceolate. Lance-shaped, much longer than broad, with a wide base, tapered
apex and widest below the centre. See also oblanceolate.
lateral. At or on the side of an organ.
leaf stalk. Stalk of a leaf; petiole.
leaflet. Primary division of a compound leaf; pinna.
leaflet angle(s) of insertion. Pinnapinna (pp) angle: angle on adaxial
side of a compound leaf subtended by leaflets on opposite sides of the rachis.
Pinnarachis (pr) angle: distal angle between leaflet and rachis.
Shielding (s) angle: angle by which a leaflet is transversely twisted often
resulting in overlapping. If leaflets overlap incubously, the s-angle is nomi-
nally positive; if leaflets overlap succubously, the s-angle is nominally nega-
tive (Grobbelaar, 2002).
lectotype, lecto. Herbarium specimen chosen from the original material to
replace a missing holotype or isotype. A duplicate of a lectotype is an
isolectotype (isolecto.). See also isotype, neotype, paratype.
Lepidoptera. Order of insects that includes moths and butterflies.
lignified. Of cell walls impregnated with lignin.
lignin. Complex insoluble polysaccharide mixture serving to strengthen and
protect cell walls. See also xylem.
X00app2.qxd 28/11/03 3:55 pm Page 248
lineage. Line of descent of an organism; ancestry. See also genealogy.

linear. Long and narrow, the sides parallel or nearly so.
lit(t)oral. Of the coast or shore, as for the habitat for Cycas litoralis.
lobe. Deeply incised segment of a flattened organ such as a leaf, leaflet or sporo-
phyll.
macrozamin. Toxic MAM glycoside found only in cycad tissue; methyl-
azoxymethanol--primeveroside. See also cycasin.
MAM. Methylazoxymethanol, the toxic principle found as its glycoside in
cycasin, macrozamin and structurally related compounds in cycad
tissues.
manoxylic. Having a small amount of secondary xylem or woody tissue with
relatively abundant parenchyma, as in cycad stems.
masting. Simultaneous reproductive activity by all or most plants in a particu-
lar area, typically seasonal and separated by long periods of low reproduc-
tive activity.
mature. Capable of sexual reproduction. Cf. juvenile.
megagametophyte. Mass of haploid cellular tissue surrounding the embryo
in a gymnosperm seed; analogous in function but not in origin to the
endosperm of angiosperm seeds.
megasporangiate strobilus (less correctly megastrobilus). Female stro-
bilus; female cone; seed cone, ovulate cone. Cf. microsporangiate
strobilus.
megaspore. Spore that gives rise to a female gametophyte. Cf. microspore.
megasporophyll. Sporophyll bearing one or more ovules or later seeds, or
potentially so. Cf. microsporophyll.
membranous (less commonly membranaceous). Thinly textured, as in a
membrane. See also chartaceous, papyraceous.
micropyle. Orifice in the integuments and later seedcoats, at the distal end
of a cycad ovule, through which the pollen or pollen tube enters. Cf.
chalaza.
microsporangiate strobilus (less correctly microstrobilus). Male stro-
bilus; male cone; pollen cone. However, in Ceratozamia microstrobila the allu-
sion is to the small size of cones. Cf. megasporangiate strobilus.
microsporangium (plural microsporangia). Structure on the abaxial
surface of microsporophylls containing microspores; pollen sacs.
microspore. Spore that gives rise to a male gametophyte; pollen grain. Cf.
megaspore.
microsporophyll. Sporophyll bearing microsporangia. Cf. megasporo-
phyll.
monomorphic. Having only one form. See also dimorphic, homomorphic,
polymorphic.
monophyletic. Derived from a single ancestor. See also paraphyletic, poly-
phyletic.
monospecific. Of a genus with only one species, as for Microcycas and Stangeria.
monothetic. Of a group sharing all features. See also polythetic.
X00app2.qxd 28/11/03 3:55 pm Page 249
monotypic. Of a family with only one genus, or a genus with only one species.
Cycadaceae is a monotypic family while Microcycas and Stangeria are mono-
typic (and monospecific) genera.
montane. Of mountains or high places, as for the localities for Macrozamia
montana and Zamia montana.
morphogeographic. Combining aspects of shape and distribution, i.e. mor-
phology and geography.
morphology. Study of the external architecture of any entity.
morphometric. Involved with the measurement of morphological characters.
mucilage canal. Passage within an organ which allows for the transport of
mucilage.
mucilage. Gelatinous substance; in cycads comprising complex water-soluble
carbohydrates and produced in response to stress, e.g. to wounding.
mucronate. Ending abruptly in a sharp point or spur known as a mucro.
multipinnate. Of a compound leaf, having more than two orders of division,
as in Cycas multipinnata. See also bipinnate.
multivariate analysis. Simultaneous statistical analysis of two or more vari-
ables.
mycorrhiza (plural mycorrhizae). Web of root-like structures arising from a
symbiotic association of a fungus and a plant, and facilitating nutrient
uptake by the host plant.
neotype, neo. New material designated to replace a missing holotype when
no original material remains in a herbarium collection. See also isotype, lec-
totype, paratype.
nomen dubium, nom. dub. Name of doubtful taxonomic validity, e.g. the
name Encephalartos tridentatus (Willdenow) Lehmann (Pugillus 6, 1834) is a
nomen dubium which may refer to several species of Encephalartos or
Macrozamia.
nomen illegitimum, nom. illeg. Name published in contravention of the
rules of nomenclature, e.g. Zamia brongniartii Weddell is a nomen illegitimum,
being a superfluous name for Ceratozamia boliviana Brongniart.
nomen novum, nom. nov. New name designated when a name cannot be
used for nomenclatural purposes and no type or original material exists.
nomen nudum, nom. nud. or nomen solum, nom. sol. Avowed new
name unaccompanied by a description or diagnosis, e.g. Lindens name Cycas
neocaledonica (LIllustration Horticiole 28, 1881).
nomenclature. Assignment of names to taxa; in botany in accordance with the
International Code of Botanical Nomenclature.
non visus, n.v. In reference to an item not seen, as in the case of a herbarium
specimen that could not be examined. Cf. exclamation mark (!).
nucleus. Core part of any eukaryotic cell, a membrane-encased organelle con-
taining the genetic material.
oblanceolate. Lance-shaped, much longer than broad, with a wide apex,
tapered base and widest above the centre. See also lanceolate.
obligate. Restricted to only one taxon or activity, as in an obligate pollinator.
X00app2.qxd 28/11/03 3:55 pm Page 250
obovate. Of a two-dimensional structure that is egg-shaped in outline but

broadest above the middle. See also elliptic, ovate.
obovoid. Of a three-dimensional structure that is obovate in longitudinal
section. See also ovoid.
obtuse. Blunt or rounded at the apex.
offset. Vegetative axis arising from an adventitious bud laterally on a cycad
trunk. See also sucker.
omnel. Kernel of an omnule; a cleaned cycad seed (Grobbelaar, 2002).
omnule. Collective term for mature cycad seeds and fully expanded unfertilized
ovules that are externally indistinguishable; cycad seed (Grobbelaar, 2002).
ontogeny. Developmental cycle or life history of an individual from its incep-
tion until its sexual maturity.
opere citado, op. cit. In reference to a publication already cited. This phrase
is no longer used in modern scientific writing.
order. Taxonomic rank below division or class but above family.
orthostichy (plural orthostichies). Row created by intersection of contact
parastichies. Sporophylls in Ceratozamia and Zamia cones show well-
defined orthostichies in a near-vertical arrangement.
OTU. Operational taxonomic unit. Any convenient taxonomic unit used in
cladistics and phenetics.
outgroup. In cladistics, an organism or group, closely related to but not
included within the group under analysis, and used for comparative purpos-
es with respect to character polarity determination. See also sister group.
ovate. Of a two-dimensional structure that is egg-shaped in outline but broad-
est below the middle. See also elliptic, ovate.
ovoid. Of a three-dimensional structure that is ovate in longitudinal section. See
also obovoid.
ovulate cone/strobilus. Female cone; megasporangiate strobilus.
ovule. Female reproductive structure; in seed plants comprising a megaspore,
or megaspore-derived tissue, one or more integuments and an attachment
stalk (funiculus).
pachycaulous. Thick-stemmed and columnar but without any substantial sec-
ondary wood, as for the stems of arborescent cycads.
paleoendemic, palaeoendemic. Referring to a taxon that is a geographical
remnant of a taxon formerly of much wider geographical distribution.
paleontology, palaeontology. Study of organisms from former geological
periods, typically fossilized plants or animals or their parts.
palynology. Study of all aspects of pollen from extant and extinct plants.
papyraceous. Papery in texture. See also chartaceous, membranous.
paraphyletic. Of an artificial group comprising a single ancestor and some,
but not all, of its descendants. See also monophyletic, polyphyletic.
parastichy (plural parastichies). Spiral arrangement of leaves on an axis (or
sporophylls in a cone). Encephalartos, Lepidozamia and Macrozamia cones show
well-defined clockwise and anticlockwise parastichies. See also orthostichy.
paratype, para. Herbarium specimen cited in a description together with the
X00app2.qxd 28/11/03 3:55 pm Page 251
holotype and any isotypes. See also lectotype, neotype.

parenchyma. Thin-walled cells in storage tissues such as the cortex zone
between the epidermis and vascular tissue. The starch-rich stems of Stangeria
eriopus and similar cycads are mainly parenchymatous.
paripinnate. Even-pinnate; of a leaf or leaflet where the rachis or rachilla does
not terminate in a solitary pinna or pinnule. Cf. imparipinnate.
PCR. Polymerase chain reaction, a technique for duplicating a small amount of
DNA into a large number of DNA fragments of identical sequence, selec-
tively controlled by a chosen DNA primer. See also RAPD.
pectinate. Comb-like; with closely spaced, narrow segments, as in the megas-
porophyll of Cycas pectinata.
pedicel. Supporting stalk of a flower or gymnosperm sporophyll. See also
peduncle.
peduncle. Supporting stalk of an inflorescence or gymnosperm cone. See also
pedicel.
peltate. With a stalk positioned centrally on the surface of a lamina or sporo-
phyll, as for most cycad bullae.
pendent, pendulous. Hanging downwards, as for the female cones of many
Dioon species.
petiole. Stalk of a leaf; in a compound leaf, that part of the axis below the low-
ermost leaflets, pinnacanths or spines. See also rachis.
petiolule. Stalk of a leaflet, that part of the axis below the leaflet blade, as seen
clearly in Zamia manicata. See also rachilla.
phenetic. Condition of an overall similarity of chosen characters between taxa
but without regard to whether the characters are derived or due to conver-
gence.
phenology. In plants, the study of recurrent phenomena, such as episodes of
reproduction, in relation to seasonal patterns.
phenogram. Tree diagram showing taxa arranged hierarchically on the basis
of phenotypic similarities without any regard for ancestry. See also clado-
gram, dendrogram, phylogram.
phenotype. Sum total of morphological or other characters defining an organ-
ism or group. Cf. genotype.
pheromone. Volatile chemical substance emitted by an organ and acting as a
signal to other organisms, as in cycad cone emissions attracting insect polli-
nators.
photosynthesis. Fundamental process in biosynthesis, by which carbon
dioxide and water are converted to simple sugars using sunlight as an energy
source. See also chlorophyll.
phylogeny. Evolutionary history of an organism or group.
phylogram, phylogenetic tree. Cladogram where branch lengths are pro-
portional to the number of changes that occur at each branch. See also den-
drogram, phenogram.
physiographic. Relating to physical geography.
physiology. Study of the functioning of organisms and their parts.
X00app2.qxd 28/11/03 3:55 pm Page 252
phytochemistry. Study of plant chemistry, particularly in relation to second-

ary metabolites.
phytogeography. Study of plants and their distribution in relation to geo-
graphical factors.
pilose. With soft hairs. See also lanate, hirsute, pubescent, sericeous,
tomentose.
pinna (plural pinnae). Primary division of a compound leaf; leaflet.
pinnacanth. Sharply pointed structure intermediate between a leaflet and a
spine, usually green and hence photosynthetic, as seen in on the rachises of
some species of Dioon, Encephalartos and Macrozamia. See also prickle, tooth.
pinnate. Feather-shaped; typically a leaf with leaflets or pinnae on either side
of a rachis. See also bipinnate, multipinnate, pinnule.
pinnule. Secondary division of a compound leaf. See also pinna.
planche. French term used in some of the older botanical literature to denote
illustrative plate. See also tabula.
platyspermic. Of seeds, flattened so as to be not radially symmetrical, as in
Cycas seeds. See also actinomorphic, zygomorphic. Cf. radiospermic.
plesiomorphy. Ancestral or underived character or character state. Note that
a plesiomorphy at a given hierarchical level may be an apomorphy at a
more inclusive level. See also autapomorphy, homology, symple-
siomorphy, synapomorphy.
plicate. Pleated or folded longitudinally to give a corrugated appearance, as for
the leaflet surface of Zamia roezlii.
plumose. Feathery in a whorled appearance.
pollen. Fine powdery material (microspores) shed from the microsporan-
gia of seed plants; in cycads sometimes referred to as prepollen;
microspores containing a male gametophyte (microgametophyte).
pollen cone/strobilus. Male cone; microsporangiate strobilus.
pollination. Transfer of pollen towards ovules. See also anemophily, ento-
mophily.
polymorphic. Persistently variable in form, as for populations of Macrozamia
polymorpha and Zamia polymorpha. See also dimorphic, homomorphic,
monomorphic.
polyphyletic. Of an artifical group comprising members that have originated
independently from more than one ancestor. See also monophyletic, para-
phyletic.
polythetic. Referring to a group sharing many, but not all, features. See also
monothetic.
population. Group of individuals of a species occupying a particular area.
prepollen. Microspore with a proximal aperture but without a distal aper-
ture. Cycad pollen is considered an intermediate form between prepollen
and conventional pollen.
prickle. Small sharp protuberance of epidermal origin, usually green, usually
somewhat irregularly distributed, as seen on the petioles of most Ceratozamia
and Zamia species. Pinnacanths in Encephalartos have also been called prick-
X00app2.qxd 28/11/03 3:55 pm Page 253
les, but they are not epidermal in origin. See also spine, tooth.
procumbent, prostrate. Trailing or lying along the ground but not rooting;
typically referring to stems. See also decumbent.
proximal. Nearest to the point of attachment of a structure; basal. Cf. distal.
pruinose. With surface covered by a waxy bloom, as for the foliage of Cycas pru-
inosa. See also glaucous.
ptyxis. Manner of folding of a leaf and leaflets at emergence. See also circi-
nate, conduplicate, inflexed, reflexed, vernation.
pubescent. Densely covered with fine short hairs. See also lanate, hirsute,
pilose, sericeous, tomentose.
pungent. Terminating in a stiff, sharp point.
r(h)achilla (plural r(h)achillae). A diminutive of rachis; a secondary axis,
in particular, in the grasses or sedges, the axis that bears the florets.
Sometimes used in reference to the axis along which pinnules are attached
in cycads having compound leaflets, e.g. Bowenia.
r(h)achis (plural r(h)achides, r(h)achises). That portion of the axis of a
compound leaf where leaflets are attached and excluding the petiole, as
alluded to in the flattened rachis of Macrozamia platyrachis and the dark purple
rachis of Zamia melanorrhachis.
radicle. Embryonic root; often misapplied to the emergent hypocotyl in ger-
minating cycad seeds.
radiospermic. Of seeds, radially symmetric as for seeds of cycad genera
except Cycas. See also actinomorphic, zygomorphic. Cf. platyspermic.
RAPD. Random amplified polymorphic DNA; a technique in DNA analysis
based on patterns obtained in electrophoresis after PCR amplification
using randomly selected primers. See also RFLP.
RC. Rescaled consistency index, the product of the consistency index (CI) and
the retention index (RI) for a character in a cladogram.
recruitment. Increase in a population due to migration, vegetative prolifera-
tion or reproduction from seed.
recurved. Bent or curving abaxially. See also reflexed. Cf. incurved.
Red List. IUCN-published listing of plant and animal taxa in terms of per-
ceived threatened status.
reflexed. Abruptly recurved or bent sharply abaxially, as in the emerging
leaves of some Dioon species. See also circinate, conduplicate, inflexed,
ptyxis, vernation.
relictual. Remaining relatively small portion of a previously larger population
or taxon, as for Encephalartos relictus.
revolute. With margins rolled downwards (abaxially), as in leaflets of Cycas rev-
oluta and Encephalartos ghellinckii. Cf. involute.
RFLP. Restriction fragment length polymorphism; a technique used in DNA
analysis based on pattern of bands obtained in electrophoresis of DNA
fragments produced after digestion of sample material by restriction
endonuclease enzymes. See also RAPD.
RI. Retention index, a measure of the amount of similarity in a character that
X00app2.qxd 28/11/03 3:55 pm Page 254
can be interpreted as a synapomorphy in a given cladogram. See also CI,

RC.
ribosome. Cellular organelle in which protein synthesis occurs.
sarcotesta (plural sarcotestae). Fleshy outer layer of the integument of a
cycad seed, brightly coloured in many cycad genera. See also sclerotesta.
satellite. Minute chromosome portion separated from the main body by the
centromere or primary constriction.
sclerotesta (plural sclerotestae). Hard or stony inner layer of the integu-
ment of a cycad seed. See also sarcotesta.
scurfy. Covered with small papery scales.
section, subsection, series. Taxonomic ranks used within some genera, e.g.
the genus Macrozamia comprises two sections, Macrozamia and Parazamia,
while the genus Cycas has the four sections Asiorientales, Stangerioides, Indosinensis
and Cycas, the last having one or more subsections within which may be one
or more series.
seed. Fully mature ovule after fertilization, with an embryo, storage tissue and
all integuments. See also omnule, corruptule.
SEM. Scanning electron microscope (instrument) or micrograph (image pro-
duced by the instrument).
senescence. Natural ageing processes leading to the death of an organ or
organism.
sensu. According to; in the sense of. See also fide.
sensu lato. In a broad or all-encompassing sense.
sensu stricto. In the narrow or restricted sense.
sericeous. With fine hairs giving a silky texture. See also lanate, hirsute,
pilose, pubescent, tomentose.
series. See section.
serrate. With a margin of saw-toothed, sharply tipped protrusions pointing for-
wards apically. See also dentate, tooth, serrulate.
serrulate. Minutely serrate, as for leaf margins in Bowenia serrulata.
sessile. Without any apparent stalk, as for the cones of many cycads; also used
in reference to cycad leaves without petioles, e.g. Encephalartos villosus.
sine numero, s.n. Of a herbarium or other specimen accession that is without
any collectors accession number.
sister group. In cladistics, the group most closely related to a specific indi-
vidual or group. See also outgroup.
sorus (plural sori). Cluster of fern sporangia; term misapplied to groups of
microsporangia sharing a common vascular supply in some cycads.
speciation. Evolutionary process in which taxa accumulate sufficient genetic
change to be recognized as individual species.
species, sp. (plural species, spp.). Basic taxonomic rank; taxonomic rank
below genus but above subspecies and varietas. A working definition
for cycads is one or more populations where individuals are morphological-
ly similar, interfertile, but sometimes geographically and hence reproduc-
tively isolated from other such populations. [See also the discussions of species
X00app2.qxd 28/11/03 3:55 pm Page 255
concept in this volume.]

species complex. Variable group of closely related members known or sus-
pected to represent different species but often difficult to circumscribe;
usually resolved as a discreet clade in taxonomic analysis; e.g. the
Ceratozamia norstogii complex comprises C. norstogii, C. alvarezii and C. mirandae.
species nova, sp. nov. Citation at the time a new species is first described, e.g.
Zamia macrochiera D.W. Stevenson species nova (Chapter 14 this volume). (Species
novum, whilst often seen in the literature, is grammatically incorrect). See also
genus novum.
spermatophyte. Any seed-bearing plant, including all gymnosperms and
angiosperms.
spermatozoid (sometimes antherozoid). Motile male gamete of lower
plants, cycads, Ginkgo, etc.
spine. Hard, sharp and non-photosynthetic protuberance with vascular con-
nections and not of epidermal origin. Spines on many cycad rachises and
sporophylls are reduced leaflets. See also pinnacanth, prickle, tooth.
spinescent. Terminating in a sharp point.
spinulose. With many small spines, as incorrectly alluded to in reference to the
leaflet margins of Dioon spinulosum which are in fact teeth.
sporophyll. Modified leaf bearing reproductive structures. See megasporo-
phyll, microsporophyll.
SSC. Species Survival Commission, an agency of the IUCN.
staminate cone/strobilus. Incorrectly derived term for male cone; pollen
cone; microsporangiate strobilus.
status novus, stat. nov. New status or rank. For example, Encephalartos
manikensis (Gilliland) Gilliland was a status novus for the taxon previously
named E.gratus Prain var. manikensis Gilliland.
stipule. Appendage at the base of a petiole. In Stangeria, a hood-like stipule
subtends each leaf base.
stoma (sometimes stomate, plural stomata). Orifice allowing gaseous
exchange across a plant epidermis, structurally defined by guard and sub-
sidiary cells.
striate. With longitudinal lines, grooves or ridges.
striolate. Finely striate.
strobilus (plural strobili). Reproductive structure of gymnosperms and some
cryptogams such as Selaginella; organized collection of sporophylls on a
central axis. See also cone, megasporangiate strobilus, microsporan-
giate strobilus.
subspecies, subsp. Taxonomic rank immediately below species; group of
individuals which differ morphologically from another group but insuffi-
ciently so as to justify separate specific status, e.g. Cycas media subsp. banksii.
Subspecies are reproductively compatible but are reproductively isolated
from each other. The usefulness of infraspecific taxonomic categories in
cycads is questioned in this volume. See also varietas.
succubous. Arrangement in which a leaflet partially shields the leaflet below
X00app2.qxd 28/11/03 3:55 pm Page 256
(the next proximal leaflet) when viewed from above (adaxially). Cf.
incubous.
sucker. Vegetative axis arising from an adventitious bud at the base of a
cycad trunk. See also offset.
sulcate. Having a longitudinal groove or furrow. Cycad pollen grains are char-
acteristically monosulcate.
suture. Line of opening or dehiscence of a closed structure such as a cycad
microsporangium.
symbiont. Organism living in a symbiotic relationship with another, e.g. the
cyanobacteria in cycad coralloid roots.
sympatric. Of two or more taxa which exist naturally together in the same geo-
graphical area. Cf. allopatric.
symplesiomorphy. Ancestral or underived character state shared by several
members of a monophyletic group that does not define a monophyletic
subset of that more inclusive group and has not experienced reversal. See also
apomorphy, autapomorphy, plesiomorphy, synapomorphy.
synapomorphy. Shared derived character state that unites two or more
members of a monophyletic group. For example, the presence of a glan-
dular collar at the leaflet base is a synapomorphy for a monophyletic group
comprising Zamia macrochiera and Z. manicata. A synapomorphy at one level is
an autoapomorphy at a more inclusive level. See also apomorphy,
plesiomorphy, symplesiomorphy.
systematics. Description and classification of life forms and the study of their
relationships.
tabula. Latin term used in some of the older French botanical literature to
denote a black and white plate. See also planche.
taxon (plural taxa). Any group of individuals, in any rank, having characteris-
tics in common and of the same evolutionary origin.
taxonomy. Circumscription, classification and naming of organisms; in plants,
systematic botany.
terete. Solid structure which is circular in transverse section; cylindrical or
nearly so.
thermogenesis. Self-heating through respiratory activity, as in cones, especial-
ly male cones, of many cycads.
threatened status. Perceived degree of threat, determined by conservation
agencies, to the continued natural existence of individual taxa, e.g. in cate-
gories such as Critically Endangered, Endangered and Vulnerable.
tomentose. Densely woolly in a finely matted fashion. See also lanate, hirsute,
pilose, pubescent, sericeous.
tomentum. Covering of fine hairs. See also indumentum.
tooth. Sharply tipped protrusion along a leaf or leaflet (or analogous structure)
margin pointing away at an angle of 90o. See also serrate.
topographic. Relating to physical features, usually of the landscape but also
used in reference to the surfaces of structures such as pollen grains, leaves
etc.
X00app2.qxd 28/11/03 3:55 pm Page 257
trichome. Small hair or scale of epidermal origin.

trnL-F. Spacer region of the maternally inherited chloroplast genome.
truncate. With an abrupt ending, as though cut off terminally.
t-test, Students t-test. Statistical tool to access if the mean values from two
sets of data are the same or not.
tuberculate. Covered with small raised protuberances, as for the bulla surface
of some cycad megasporophylls. See also verrucose.
type, T. Herbarium specimen or other element (e.g. illustration) to which the
name of a taxon is permanently attached, whether as a correct name or a
synonym. The type is not necessarily the most typical or representative
element of a taxon. See also holotype, isotype, lectotype, neotype or
paratype.
undulate. With wavy margins, usually of leaflets.
variegated. Having two or more colours in a blotched or mottled pattern, as
for the leaflets of Zamia variegata.
varietas, variety, var. Taxonomic rank below species; group of individuals
which differ morphologically from another group but insufficiently so to
justify separate specific status, e.g. Dioon edule var. angustifolia. Varieties are
reproductively compatible, not reproductively isolated from each other and
may occur together in mixed stands. The usefulness of infraspecific taxo-
nomic categories in cycads is questioned in this volume. See also subspecies.
vascular plant. Plant having phloem or xylem as conducting tissues, i.e. any
Pteridophyte (ferns and their allies) or Spermatophyte (gymnosperms and
angiosperms).
venation. Pattern of veins in a leaf or leaflet.
vernation. Manner of folding of a leaf prior to emergence. See also circinate,
conduplicate, inflexed, reflexed, ptyxis.
verrucose. Warty in appearance, as for the bulla surface of many cycad sporo-
phylls. See also tuberculate.
Wilks lambda test. Statistical test comparing the centroids of a distribution of
means; a multivariate analysis of variance. See also ANOVA.
xylem. Water-conducting tissue in vascular plants, comprising vessels and tra-
cheids, dead at maturity and having a secondary layer of lignin.
zygomorphic. Of a bilaterally symmetric structure that can be halved only in
one plane to give two halves that are mirror images of each other; of seeds,
platyspermic. Cf. actinomorphic.
zygote. Cell resulting from the fusion of male and female gametes; the first cell
of an embryo.
Acknowledgements
We thank Giancarlo Contrafatto, Root Gorelick, Nat Grobbelaar, Lou Randall

and contributors to this volume for much helpful assistance in the compilation of
this glossary.
X00app2.qxd 28/11/03 3:55 pm Page 258
References and Further Reading

Grobbelaar, N. (2002) Cycads with Special Reference to the Southern African Species. Published
by the author, Pretoria, South Africa, 331 pp.
Jackson, B.D. (1971) A Glossary of Botanical Terms with their Derivation and Accent, 4th edn.
Gerald Duckworth & Co., London, 481 pp.
Jones, D.J. (2002) Cycads of the World Ancient Plant in Todays Landscape, 2nd edn. Reed New
Holland, Sydney, Australia, 456 pp.
Lawrence, G.H.M. (1989) Taxonomy of Vascular Plants. MacMillan, New York, 823 pp.
Schuh, R.T. (2000) Biological Systematics Principles and Applications. Cornell University Press,
Stearn, W.T. (1992) Botanical Latin, 4th edn. Timber Press, Portland, Oregon, 546 pp.

Cicadas, Clasificacion, Conceptos y Recomendaciones

Uploaded by

Document Information

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

Cicadas, Clasificacion, Conceptos y Recomendaciones

Uploaded by

Copyright:

Available Formats

00prelims.

qxd 28/11/03 3:40 pm Page i

CABI Publishing is a division of CAB International

CABI Publishing CABI Publishing

Tel: +44 (0)1491 832111 Tel: +1 617 395 4056

CAB International 2004. All rights reserved. No part of this

A catalogue record for this book is available from the British

Library of Congress Cataloging-in-Publication Data

ISBN 0 85199 741 4

Typeset by MRM Graphics Ltd, Winslow, Bucks

1. We Hold these Truths 1

2. Saving Ghosts? The Implications of Taxonomic Uncertainty

3. Character Evolution, Species Recognition and Classification

4. Morphological Characters Useful in Determining Species

5. Comments on Cycas, Dyerocycas and Epicycas (Cycadaceae) 57

6. Classification Concepts in Encephalartos (Zamiaceae) 69

7. Classification Concepts in Macrozamia (Zamiaceae) from

8. Classification Concepts in Ceratozamia (Zamiaceae) 95

9. Relationships and Phytogeography in Ceratozamia

10. A Morphometric Analysis of the Ceratozamia norstogii

11. Hypotheses on the Relationship between Biogeography and

12. Molecular Phylogeny of Zamia (Zamiaceae) 149

13. Systematics of Meso-American Zamia (Zamiaceae) 159

14. Zamiaceae of Bolivia, Ecuador and Peru 173

15. In Search of the True Tree: Guidelines for Classification 195

Appendix 1: The World List of Cycads 219

Appendix 2: Glossary of Terms Encountered in Cycad

Martnez, M., Facultad de Biologa, Universidad Autnoma de Quertaro, Centro

About the Editors

Terrence Walters is the executive director of Montgomery Botanical Center,

x About the Editors

Terrence Walters and Roy Osborne

The directors, members, staff and volunteers of Montgomery Botanical Center

We Hold these Truths 1

1Montgomery Botanical Center, Miami, Florida, USA; 2PO

Cycad Classification Concepts Workshop

We Hold these Truths 3

Cycad Taxonomy and Systematics

Taxonomy is the process of circumscribing and assigning scientific names to the

We Hold these Truths 5

logical entities to humanity; therefore, most cycad systematists undertake studies

We Hold these Truths 7

Cycad Phylogenetics Uncovering Genealogical

Cycad systematists use phylogenetic trees to try to produce a phylogenetic

CCC Workshop Beliefs

The CCC Workshop participants enumerated their beliefs concerning cycad

Beliefs about biological relationships

We believe there is value in the biological world.

We Hold these Truths 9

We believe that the natural order is based on genealogical relationships.

Beliefs about what plant systematists can and should do in order to

Beliefs about what cycad systematists can and should do in order to

We believe that the Cycadales forms a distinct monophyletic lineage, and

We Hold these Truths 11

Saving Ghosts? The 2

Kirstenbosch Research Centre, National Botanical Institute,

Red Listing and Conservation Planning 15

Taxonomy and the Threatened Status of Cycads

Red Listing and Conservation Planning 17

Pre-1994 Rare Post-1994 Vulnerable

post-1994 criteria, there is also an increase in the number of threatened taxa

Red Listing and Conservation Planning 19

The influence of higher-level classification on conservation planning

Red Listing and Conservation Planning 21