A Molecular Systematic Study of the African Endemic Cycads

A molecular systematic study of the African endemic cycads

PHILIP ROUSSEAU

Dissertation submitted in fulfilment of the requirments for the degree

MAGISTER SCIENTIAE

BOTANY

in the

FACULTY OF SCIENCE DEPARTMENT OF BOTANY AND PLANT BIOTECHNOLOGY

at the

UNIVERSITY OF JOHANNESBURG

SUPERVISOR: PROF. M. VAN DER BANK CO-SUPERVISOR: DR. P.J. VORSTER

FEBRUARY 2012 Declaration

I declare that this dissertation has been composed by me and the work contained within, unless otherwise stated, is my own.

Philip Rousseau February 2012.

II Table of contents

Abstract p. V Foreword p. VI Acknowledgements p. VII List of abbreviations p. VIII Chapter 1. Introduction and aims 1.1. General introduction p. 1 1.2. Taxonomy p. 3 1.3. Distribution p. 5 1.4. Conservation status p. 6 1.5. Intra-generic concepts p. 9 1.5.1. Stangeria p. 9 1.5.2. Encephalartos p. 10 1.6. Aims and hypotheses p. 12 Chapter 2. Material and methods 2.1 Taxon sampling p. 29 2.2 DNA extraction p. 29 2.3 Polymerase chain reaction (PCR) p. 29 2.4 DNA sequencing p. 30 2.5 Sequence alignment and analysis of molecular data p. 30 2.5.1. Phylogenetic analysis p. 31 2.5.2. DNA barcoding analysis p. 32 Chapter 3. Phylogenetic analyses: Results and discussion 3.1. Molecular evolution p. 46 3.2. Combined plastid analysis p. 46 3.3. Nuclear ITS analysis p. 47 3.4. Combined molecular analyses p. 47 3.4.1. Clade A: lineage 1 p. 47 3.4.2. Clade B p. 50 Clade B: lineage 2 p. 50 Clade B: lineage 3 p. 51 Clade B: lineage 4 p. 52 Clade B: lineage 5 p. 53 Lineage 5: sub-lineage A p. 53 Lineage 5: sub-lineage B p. 54 3.4.3. Clade C p. 55 Clade C: sub-clade I: lineage 6 p. 55

III Lineage 6: sub-lineage A p. 56 Lineage 6: sub-lineage B p. 56 Clade C: sub-clade II: lineage 7 p. 57 Lineage 7: sub-lineage A p. 57 Lineage 7: sub-lineage B p. 58 Clade C: sub-clade III: lineage 8 p. 59 Clade C: sub-clade III polytomy p. 60 Encephalartos ferox p. 61 Clade C: sub-clade III: lineage 9 p. 61 Clade C: sub-clade III: lineage 10 p. 62 Encephalartos hirsutus and E. inopinus p. 63 3.5. Conclusions p. 63 Chapter 4. DNA Barcoding: Results and discussion 4.1. General introduction p. 100 4.2. Results p. 101 4.2.1. Sequencing success rates p. 101 4.2.2. Genetic variation p. 102 4.2.3. Species resolution and species concepts p. 103 4.3. Discussion p. 103 4.3.1. Success rates p. 103 4.3.2. Genetic variation p. 104 4.3.3. Species resolution p. 106 4.4. Conclusions p. 108 Chapter 5. General conclusions p. 137 Chapter 6. References p. 139 Appendix 1. Complete collection list p. 155 Appendix 2. Canadian Center for DNA Barcoding protocols p. 163

IV ABSTRACT

Africa's cycads (66 species and 2 subspecies in two endemic genera: Encephalartos and Stangeria) are extremely endangered with four species extinct in the wild and 80% threatened (CR, EN, or VU) with all included in CITES Appendix 1. Although South Africa has some of the world’s strictest cycad legislation, these plants are still under threat from illegal collection for horticulture and medicine especially where they are seized in an unidentifiable condition. Currently developed legislation demands accurate identification for permit issue. Ex situ conservation of genetic and locality based diversity is paramount. Furthermore, taxonomically many species of unknown origin are difficult to identify especially when diagnostic characters are absent. Species delimitation and numbers are uncertain with field observations often contradicting current understanding. DNA barcoding can assist with all the above-mentioned scenarios. In the current study all proposed DNA barcoding regions (matK, rbcLa, psbA-trnH, and nrITS) along with several additional regions were tested on ~350 samples from which a phylogeny of 63 of the 65 Encephalartos species was also constructed. Results show general good amplification and sequencing success of proposed barcoding regions, although a shift to specialist primers was made in several cases. Genetic variation however was extremely low as is resolution at species level, even when multi-locus barcodes were employed. Results obtained from the phylogenetic analyses show an increase in resolution at both species and higher levels compared to previous work and as such several new groupings are delimitated. Each species grouping is characterised by shared, derived morphological, ecological, and geographic characters and when compared to previous phylogenetic studies are supported to some extent. The current study provides the first step towards a much-needed monograph and revision of the entire genus Encephalartos.

V FOREWORD

ThIs dissertation is presented in six chapters. Chapter 1 focuses on the general importance of the world’s cycads with emphasis on the study group. Key areas are the taxonomy, distribution, and conservation with the aims and hypothesis stated. Importantly all species authorities are listed in Table 1.1 while generic authorities are in Table 1.2. Chapter 2 outlines the general methods used in the study. Note that all samples from the New York Botanical Garden were not submitted to GenBank, as they are the property of Dr. Damon Little. Chapter 3 presents results and a discussion of the phylogenetic relationships within the genus Encephalartos. Chapter 4 presents results of the DNA barcoding study, a discussion thereof with concluding remarks. Chapter 5 presents a general conclusion on all aspects of the study including hypotheses and aims as set out in Chapter 1. Chapter 6 contains all literature referred to in the preceeding chapters. Lastly, Appendix 1 contains a complete collection list and Appendix 2 contains all protocols as performed at the Canadian Centre for DNA Barcoding. All photographs, figure and tables, unless otherwise indicated, were done by P. Rousseau.

VI ACKNOWLEDGEMENTS

This work would not have been possible without the financial assistance of the following insitutions: National Research Fund of South Africa (NRF), Royal Society UK, University of Johannesburg, the National Science Foundation (EF-0629890), the Alfred P. Sloan Foundation at the Cullman Program for Molecular Systematics at the New York Botanical Garden. This project was also partly funded by the Government of Canada through Genome Canada and the Ontario Genomics Institute (2008-OGI-ICI-03).

Firstly, I would like to acknowledge Prof. Braam van Wyk who promted me into this project, your wisdom and vision has once more humbled me. For samples of these rare and difficult taxa, a special thanks goes to the Pretoria National Botanical Garden, the University of Johannesburg, the University of Pretoria and especially Susan Myburgh, the Lowveld National Botanical Garden and especially Karin van der Walt, Andre Cilliers and Robert Rousseau for access to their prized collections, Adolf Fanfoni for his multiple contributions to the project, and for the very rare samples donated by Dr. Xander de Kock and Prof. Nat Grobbelaar. Also for the various photographs of some of the rarest plants in the world a special thanks to the members of the cycad forum and Art Vogel and Anders Lindström. A special word of thanks is in order for Prof. Grobbelaar whose opinion was consulted and whose reference works were of undeniable value. To Prof. Dennis Stevenson, Dr. Roy Osborne and Prof. John Donaldson, whom I have had the privilege of communicating with on occasion, thanks for the advice and the motivation. This work would most certainly not be possible without the gracious help of Dr. Damon Little whose appreciation of the difficulties in doing molecular work on cycads was a welcome justification for the troubles endured and overcame with his help. At the University of Johannesburg numerous individuals have made significant contributions to the project, in particular I would like to thank Mr. Thinus Fourie for getting me started and all the administrative assistance, Mrs. Helen Long for administrative help, and Dr. Olivier Maurin for assistance with the numerous tasks demanded by the project. Also to my life long friend Wikus van Wyk for the help with the data manipulation in excel, automating hours of manual labour. To my supervisor Prof. Michelle van der Bank, thank you for your patience and commitment even though there were easier more rewarding projects, thank you for indulging a student his passion. Also for the numerous opportunities you have entrusted to me, you have contributed so much to the scientist I am today. Lastly to my co-supervisor Dr. Piet Vorster, a giant in the field, thank you so much for your genuine concern, open and friendly sharing of your wealth of knowledge and shared excitement surrounding these wonderful plants.

Finally this work is dedicated to my father Robert Rousseau whose love of these plants has shaped so much of my life. Ek het oor die twee jaar so baie van Pa in my leer sien, dankie vir wat Pa-hulle in my gekweek het.

VII

LIST OF ABBREVIATIONS

The following abbreviations are used throughout the text: ACDB: African Centre for DNA Barcoding BLAST: Basic Sequence Alignment Search Tool BOLD: Barcoding of Life Data Systems BP: Bootstrap Percentages CBOL: Consortium for the Barcode of Life CCDB: Canadian Centre for DNA Barcoding CI: Consistency Index CITES: Convention on International Trade in Endangered Species CR: Critically Endangered CTAB: Hexadecyltrimethylammonium Bromide CWOFI: Cycad World of Innovation DD: Data Deficient ddH20: Double Distilled Water DNA: Deoxyribonucleic Acid DNTP: Deoxyribonucleotide DRC: Democratic Republic of the Congo EF: Experimental Farm (UP) EN: Endangered EW: Extinct in the Wild IBOL: International Barcode of Life Project IUCN: International Union for the Conservation of Nature KZN: KwaZulu-Natal Province LBG: Lowveld National Botanical Garden LC: Least Concern matK: Maturase K MBG: Manie van der Schijff Botanical Garden (UP) Min: Minute(s) MUSCLE: Multiple Accurate and Fast Sequence Comparison by Log-Expectation MYA: Million Years Ago nrITS: Nuclear Internal Transcribed Spacer NYBG: New York Botanical Garden PAUP: Phylogenetic Analysis Using Parsimony PBG: Pretoria National Botanical Garden PCR: Polymerase Chain Reaction

VIII psbA-trnH: Spacer between the trnH and psbA genes RI: Retention Index rbcLa: Subunit ‘a’ of Ribulose 1,5-Bisphosphate Carboxylase/Oxygenase sec: Second(s) SSC: Species Survival Commission (IUCN) TBE: Ethylenediaminetetraacetic Acid Disodium Salt Dihydrate TBR: Tree-Bisection-Reconnection UJ: University of Johannesburg UP: University of Pretoria VU: Vulnerable µl: Microliter(s)

CHAPTER 1. INTRODUCTION AND AIMS

1.1 General introduction Cycads (Order Cycadales) are a group of seed bearing plants classed as gymnosperms as they have their ovules and seeds naked (not enclosed by a carpel and with seeds lacking integuments; Norstog & Nicholls 1997). They are a distinct, phylogenetically isolated group, and have been recognised as such since their first description by Linnaeus in his Species plantarum (1753). Cycads have also been identified as the most basal lineage of seed bearing plants (Doyle & Donoghue 1992; Bowe et al. 2000; Chaw et al. 1997; Chaw et al. 2000; Treutlein & Wink 2002; Ran et al. 2010). Synapomorphic characters of cycads separating them from extant seed plants include (Norstog & Nicholls 1997; Stevens 2001 but see Fisher & Marler 2006; Walters & Osborne 2004; Hermsen et al. 2006): • Apogeotropic coralloid roots with mutualistic cyanobacteria • Absence of root hairs • Coleorhiza present • Primary thickening meristem (similar to that in monocots) • Stem manoxylic (little vascular tissue, wide zones of pith and cortex) • Persistent secondary phloem • Bundles in the periphery of the pith • Axillary buds absent • Vascular tissue situated transversely in the pith (cone dome) • Transfusion tissue present • Multilacunar foliar nodes • Girdling leaf traces • Protoxylem poles changing from endarch in the stem to exarch in the leaf traces • Petiole vascular bundles an inverted omega shape • Leaf vascular bundles amphicribal • Epidermal cells with perforations • Large, pinnate euphyll leaves • Production of cataphylls • Individuals dioecious • Buffer cells surrounding the archegonium • Numerous abaxial microsporangia on sporophylls • Microsporangia in synangia dehiscing by the action of the epidermis (exothecium) • Ectexine alveolate • Megasporophylls with terminal sterile portion

• Pollen tube usually branched, growing away from the ovule, one prothallial cell, spermatogenous cells delimited by circular anticlinal wall • Seed with sarcotesta and inner fleshy layer, both vascularised • Double vasculature of the integument • Germination hypogeal, cryptocotylar • Mitochondrial nad1 intron 2 and coxIIi 3 intron present, one duplication in the PHYO gene group • Production of ß-methylamino-L-alanine and methylazoxymethanol glycosides • Mucilage in canals

The cycad fossil record extends back to the Palaeozoic, possibly from the late Carboniferous (~300–325 million years ago; MYA) but at least from the early Permian (~280 MYA; Norstog & Nicholls 1997; Schneider et al. 2002). They were dominant in terms of diversity and distribution throughout the Mesozoic (Schneider et al. 2002) comprising an estimated 20% of the vegetation cover ranging from Alaska, Siberia and Greenland in the north to Antarctica in the south (Golding & Hurter 2003). As such cycads have been found in Mesozoic rocks in Alaska, Antarctica, Australia, Europe, Greenland, India, North America, South Africa, and the British Islands (Jones 2002). The cycad’s closest relatives are speculated to be the extinct, earliest group of seed plants, the ‘cycadophytes’ (Schneider et al. 2002; Brenner et al. 2003) and are ventured to have evolved from the seed ferns (Pteridospermales; Norstog & Nicholls 1997). Of note is the striking resemblance of some fossil taxa to extant species (Klavins et al. 2003) prompting cycads to be referred to as living fossils and which allows for these taxa to be phylogenetically placed (Norstog & Nicholls 1997; Klavins et al. 2003; Klavins et al. 2005; Hermsen et al. 2006). For example, Stangeria-like fossil species have been found in Lower Cretaceous deposits (70–135 MYA) in Argentina (Artabe & Stevenson 1999) and Zamiaceae-like fossils have been found in Middle Triassic to the Eocene (54–200 MYA) deposits in North America, Europe, Australia, South America, and Antarctica (Norstog & Nichols 1997), suggesting the family existed before the Pangaea split (Donaldson 2003). Extent taxa are however considered to have evolved relatively recently, with genera arising before the end of the Cretaceous period (Pant 1987 cited in Norstog & Nichols 1997) and with some species originating around the Miocene (Treutlein & Wink 2002; Schneider et al. 2002). This fossil similarity provides an almost unprecedented situation where inferences about extinct taxa and possibly the evolutionary history of others, can be made through the study of extant plants (Terry et al. 2004).

Extant species of cycads share a large number of traits, to name but a few: all cycads are dioecious bearing strikingly similar cones as reproductive structures (Figure 1.1; exception being the loosely arranged female cones in Cycas) (Norstog & Nicholls 1997); with few exceptions (Stangeria, Bowenia) have a very similar morphology and habit (Figures 1.1; 1.2); they are

generally relatively long-lived with slow growth rates (Giddy 1980; Donaldson 2008); they have late reproductive maturity though there are exceptions in American species (Jones 2002; Grobbelaar 2002; Terry et al. 2008); most of their parts are poisonous (Schneider et al. 2002; Brenner et al. 2003); have very limited individual distribution ranges and thus high levels of endemism (Norstog & Nicholls 1997; Whitelock 2002); they are highly endangered (Donaldson 2003; IUCN 2010); and have various specialist interactions such as symbioses and obligate mutualisms including obligate entomophily, associations with cyanogenic bacteria and dispersal by vertebrates (Giddy 1980; Goode 1989, 2001; Norstog & Nicholls 1997; Grobbelaar 2002; Jones 2002; Proche! & Johnson 2009; Suinyuy et al. 2009).

The above-mentioned characteristics hold true for the two African endemic genera Encephalartos and Stangeria, which represent the continent's entire cycad diversity except for a single species of Cycas, C. thouarsii (Figures 1.3–1.5). Cycas thouarsii occurs along the eastern coast of Africa in Kenya (Figure 1.3), Tanzania and Mozambique together with the oceanic islands of Madagascar, Comores, Seychelles, and Zanzibar (Goode 1989, 2001; Donaldson 2003) and is doubtfully indigenous to the continent (Goode 1989; Norstog & Nicholls 1997; Grobbelaar 2002; Treutlein & Wink 2002). It is easily distinguishable from the other African species as it has a single raised mid- vein (Figure 1.4) on its pinna (which is not fern-like) and an uncompacted female cone (Figure 1.5). The monotypic genus Stangeria was formally described in 1853 by Moore in honour of Stanger, who sent specimens to England from the KwaZulu-Natal province. The solitary species, Stangeria eriopus, was however first placed by Kunze (1839) into a fern genus as Lomaria eriopus which was amended once cones were observed. When the new genus was established Moore disregarded the previously used specific epithet eriopus and the plant was named Stangeria. This mistake was corrected by Baillon in 1892 yielding the current name Stangeria eriopus. Encephalartos, the third most specious genus of cycads, was the third cycad genus to be described in 1834 by Lehmann. Thunberg described the first species under the name Cycas caffra in 1775, which was later used by Pilger in 1926 to lectotypify Encephalartos (E. caffer (Thunb.) Lehm.) (Table 1.1). This was challenged by Stevenson (1992) who changed the lectotype to E. friderici-guilielmi Lehm. based on his opinion that the type concept had no standing by 1926 as it was only accepted at the 1930 International Botanical Congress (Briquet & Rendel 1935 cited in Vorster 2004b). However Vorster (2004b) convincingly argues that the concept, although not officially accepted or compulsory at the time, was in effect and employed in this case. Also since the rules are retroactive the first lectotypification of E. caffer should stand (McNeill et al. 2006), as it does till today (Christenhusz et al. 2011; Osborne et al. in press).

1.2 Taxonomy The currently 329 recognised species of cycad (Osborne et al. in press) accomodated in in ten genera (see Lindström 2009) are classed together at the highest taxonomic level as recognised by

various authors (Stevenson 1992, Hill & Stevenson 1998–2005; Christenhusz et al. 2011). Traditionally either three or four families have been recognised: Cycadaceae, Zamiaceae, and Stangeriaceae with a fourth family Boweniaceae recognised occasionally (Stevenson 1981). However, currently, based on lack of confidence in generic placement, only two families are recognised: Cycadaceae containing only the genus Cycas and a lumping of all other genera into Zamiaceae (Stevens 2001; Christenhusz et al. 2011). Historically Stangeriaceae contained either the monotypic genus Stangeria (S. eriopus) and the genus Bowenia or only Stangeria with Bowenia's two species placed in Boweniaceae, and Zamiaceae contained all other genera including: Zamia, Ceratozamia, Microcycas, Dioon, Lepidozamia, Macrozamia, and Encephalartos.

Before its lumping into Zamiaceae, Stangeria was classified as the sole member of the Stangeriaceae as Bowenia's inclusion did not received molecular support with many recent studies finding it to be polyphyletic (see Hill et al. 2003; Rai et al. 2003, Bogler & Fancisco-Ortega 2004; Chaw et al. 2005; Sangin et al. 2008; Zgurski et al. 2008). Autapomorphies for Stangeria and Bowenia are (Stevenson 1992, Hill & Stevenson 1998–2005): root buds present; vascularised stipules; cotyledon bundles amphivasal; leaflet traces derived from more than one rachis bundle; cataphylls that are irregularly produced; and ovules attached below megasporophyll stalks (Figures 1.6–1.7). They also share: fern-like foliage (Figures 1.8–1.11) with subterranean stems (Figures 1.10–1.11) lacking persistent leaf bases (Figures 1.12–1.13); cones terminating the apical meristem; trichomes that are unbranched and shortly curved; peltate and spineless sporophylls (Figures 1.14–1.15) (Hill & Stevenson 1998–2005; Jones 2002). However discrepancy are as follows (Norstog & Nicholls 1997; Jones 2002): Bowenioideae—Leaves bipinnately compound (Figure 1.11), cones terminal on slender branches of the main stem (Figures 1.12; 1.15), and chromosome number 2=18; Stangerioideae—Leaves pinnate (Figures 1.8–1.12), cones terminal on the main stem (Figures 1.13–1.14), and chromosome number 2n=16. Though very distinct from other genera the lack of consistency in molecular phylogenetic placement hinders classification of either of these into higher ranks.

Members of the tribe Encephalarteae (Encephalartos (Table 1.2), Lepidozamia, and Macrozamia) have consistently been resolved as a clade in both molecular and other taxonomic analyses (Stevenson 1981, 1992; Schutzman & Dehgan 1993; Klavins et al. 2003; Hermsen et al. 2006; Zgurski et al. 2008). The tribe shares relatively large cones, similar diamond shaped sporophyll faces, opening at reception in a spiral pattern; seeds attached directly to the sporophyll and notably ovules that develop to full size in the absence of pollination or fertilisation (Norstog & Nicholls 1997; Jones 2002). This latter condition, coupled to the consistent clade resolution (with molecular data), may be strong motivation for separation of these three genera into their own family.

Generic delimitations have recieved good taxonomic support, with exceptions such as the recent

transfer of Chigua to Zamia (Lindström 2009) and Microcycas possibly included in Zamia (Piet Vorster pers. comm.1). Relationships between genera however has seen some conflict between molecular and more traditional lines of evidence (Table 1.2 and Figure 1.16 vs Figure 1.17). Zgurski et al. (2008) presents a single molecular tree (Figure 1.17) that is in agreement with the recent molecular studies of Hill et al. (2003) and Chaw et al. (2005) and the morphological classification by Hermsen et al. (2006; excluding extinct taxa), which is also consistent with Bogler & Francisco-Ortega's (2004) results. They found that the genera Bowenia, Stangeria and Dioon all have uncertain phylogenetic placements and state that: “Our plastid gene survey includes a substantial fraction of the plastid genome (approximately a ninth of the non-repetitive total) but still does not provide enough variation to rule out several alternative placements for Bowenia, Dioon and Stangeria”. Confidence levels for the various clades however are high with: all cycad genera strongly supported as monophyletic; Cycas considered ancestral through molecular (Treutlein & Wink 2002; Hill et al. 2003; Rai et al. 2003; Bogler & Francisco-Ortega 2004; Chaw et al. 2005; Sangin et al. 2008; Zgurski et al. 2008) and morphological support (Johnson 1959; Stevenson 1992; Schutzman & Dehgan 1993; Walters & Osborne 2004; Hermsen et al. 2006), Ceratozamia, Microcycas and Zamia form a clade, as does Encephalartos, Lepidozamia and Macrozamia. An important distinction between morphological and molecular classifications is that Encephalartos and Lepidozamia are a separate sister clade based on molecular data (Treutlein & Wink 2002; Hill et al. 2003; Rai et al. 2003; Bogler & Francisco-Ortega 2004; Chaw et al. 2005; Sangin et al. 2008; Zgurski et al. 2008) while morphological and other classical taxonomic evidence resolves Macrozamia to be of closer relation to Encephalartos (Johnson 1959; Stevenson 1992; Schutzman & Dehgan 1993; Walters & Osborne 2004).

1.3 Distribution Current distribution patterns of Cycadales are remnants of their once greater distribution (Donaldson 2008), though distribution of some of the actively speciating groups may be due to secondary colonisation. Today cycads are centred around the tropics and subtropics of the formerly united supercontinents Laurasia and Gondwana, including four centres of diversity (Figure 1.18): the Americas (new world/central and south America), Africa, southern Asia, and Australia (Donaldson 2003). Species only occur a few degrees above and below the tropics of Cancer and Capricorn (Figure 1.18) being confined between 35º south and north of the equator (Grobbelaar 2002; Whitelock 2002). Genera are landmass bound and continent endemics except for Cycas, which is found in Asia, Australia, and Africa (Jones 2002). Also high levels of species endemism occurs as over 70% of all known species occur in just five countries: South Africa, Australia, Mexico, China, and Vietnam (Donaldson 2003).

Encephalartos, which is endemic to Africa, occurs in 16 sub-Saharan countries (Figure 1.19)

1 Piet Vorster, Encephalartos taxonomist. University of Stellenbosch, Department of Botany and Zoology.

predominantly in the south and east continuing across central Africa to the western countries of Angola, Nigeria, Ghana and Benin (Jones 2002; Donaldson 2003; Golding & Hurter 2003; Donaldson 2008). Notably conspicuous gaps in distribution occur (Figure 1.19) where cycads seem to be absent from entire countries and yet found in neighbouring ones. This is speculated to be due to poor botanical exploration in these regions (Norstog & Nicholls 1997; Goode 2001; Donaldson 2003; Golding & Hurter 2003) but may in fact be a true reflection (Piet Vorster pers. comm.1). Within South Africa (Figure 1.20) species occur from the northern Limpopo Province south to the Eastern Cape occurring only in provinces that border the Indian Ocean coast (except Gauteng) as disjunct populations in a continuous range. Some species only occur on single isolated outcrops (Donaldson 2008; e.g. E. dyerianus) and even when they have wide distributions these are made up of small probably isolated colonies (e.g. E. laevifolious; Ezemvelo KZN Wildlife 2004). Stangeria eriopus also follows this trend as it is endemic to an area slightly inland of the South African east- coast with multiple possibly isolated populations (Donaldson 2003). Notably cycads are absent from the western, winter rainfall areas of South Africa (Figures 1.20). This is congruent with other genera as no cycad occurs naturally in a Mediterranean winter rainfall climate (Whitelock 2002). Encephalartos longifolius though does receive winter rain where it occurs in the Fynbos biome, but this is also coupled to at least some summer rainfall (Goode 2001; Grobbelaar 2002; Jones 2002; Whitelock 2002). This poses interesting questions as to their absence in predominantly winter rainfall areas, which is seemingly not physiologically based as they are readily cultivated in such climates.

Golding & Hurter (2003) found that 45 Encephalartos taxa (70%) are endemic to one country while a further 18 are found in only two (Table 1.3). Southern Africa hosts well over half of the diversity and South Africa has twice as many species as any other country (Golding & Hurter 2003; Osborne et al. in press). Whether this is the true state of affairs or simply due to higher taxonomic and botanical activity (e.g see species descriptions Vorster 1996a–e, Hurter & Glen 1996) is debatable. Over ten, mostly northerly, countries also seem to have only a single cycad species within its borders, which again seems to be congruent with the division at large (Donaldson 2003) and may be a reflection of topographical diversity and meteorological history (Piet Vorster pers. comm.1).

1.4 Conservation status The cycads are among the most threatened groups of plants worldwide with over half included in the IUCN Red List as threatened (Critically Endangered, Endangered, or Vulnerable) (Donaldson 2003). In terms of the four centres of diversity Donaldson (2003) tallied all known species by their respective status under the IUCN (Figure 1.21). Specifically in Africa, four species are Extinct in the Wild, 16 are classified as Critically Endangered (CR) (~24%), eight are classified as Endangered (EN) with a further 19 (28%) classified as Vulnerable (VU), thus a total of 80% of Encephalartos species are considered threatened (CR, EN, or VU; IUCN 2010). Some tropical African species

however have not been fully assessed (IUCN 2010; Osborne et al. in press), which, along with speculative undiscovered tropical Africa species (Goode 1989, 2001; Donaldson 2003; Golding & Hurter 2003), could inflate these numbers even more. This has been found in part as the most recent review on African cycads' Red List status yielded five undescribed taxa, all endemic and at least Critically Endangered (Golding & Hurter 2003). Stangeria eriopus on the other hand is listed as Near Threatened (Donaldson 2003; Golding & Hurter 2003). Based on their status all Encephalartos species are listed under CITES as Appendix 1 (the strictest control group; Golding & Hurter 2003) to avoid identification errors and to allow all newly discovered species to automatically enter the categorisation (Golding & Hurter 2003).

Encephalartos is one of the most endangered cycad genera and its southern African species are thought to be the most vulnerable of all cycad taxa (IUCN/ SSC Cycad Specialist Group 2003). The greatest threat to African cycads is over collection for either horticultural or medicinal purposes (Donaldson 2003, Ezemvelo KZN Wildlife 2004), though tropical African species do suffer from habitat loss (Miringu 1999; Golding & Hurter 2003). Strangely while wildlife often suffers during socio-political unrest, cycads have been protected by the inability of collectors to access them due to the ongoing conflicts in the countries where they occur (Goode 1989). All four centres of endemism (Mexico, Australia, southern Africa (specifically South Africa) and China with Vietnam and Thailand) have been found to be affected by illegal trade (Golding & Hurter 2003) though CITES trade data from 24 years (1977–2002) (IUCN/ SSC Cycad Specialist Group 2003) indicates artificially propagated specimens are by far the most traded material. Most trade in Encephalartos however does not occur over international borders (nullifying CITES) but is domestic and often focussed on rare species or specimens from specific locations due to their collectors value (IUCN/ SSC Cycad Specialist Group 2003). Many of these specimens must be of wild origin, as for example with E. hirsutus, nature conservation officials have never issued permits for the harvest of plants or seeds (IUCN/ SSC Cycad Specialist Group 2003), yet they are readily available.

Another trend in trade that has seen a rise of late is the trade in large, arborescent or otherwise attractive, but often common species, for use in landscaping (IUCN/ SSC Cycad Specialist Group 2003). Due to their slow growing nature a limited amount of large plants are available in cultivation, opening the market for plants of wild origin which results in illegal collection such as the ca. 400 E. altensteinii in 1995 (IUCN/ SSC Cycad Specialist Group 2003), a common species, which would hold no appeal to collectors. Donaldson (2008) states that: “almost all populations of threatened Encephalartos species have declined over the past 20 years” and is corroborated by many others (Giddy 1980; Goode 1989, 2001; Golding & Hurter 2003). In a study by Donaldson and Bösenberg (1999) repeat photographs showed that of the 130 populations compared, 67% had declined based on adult plant loss, most probably due to illegal collection and a further 5% due to medicinal harvesting. It is considered (IUCN/ SSC Cycad Specialist Group 2003) that Encephalartos species'

current highly endangered status is possibly due firstly to the domestic market in South Africa, which allows wild harvested plants to be introduced to the artificial market through gardens that also then pass between borders with neighbouring countries, and secondly wild collection, which was legal in the early 1900's where possibly large scale activity left some populations already depleted. Most nurseries stock the more common species while an increasing number of specialist nurseries deal in threatened species, including large specimens (Donaldson 2008). Nature conservation agencies in South Africa estimated that more than a million indigenous cycads occurred in private gardens within the country (IUCN/ SSC Cycad Specialist Group 2003). However determining how much of this constitutes illegally obtained specimens is impossible (IUCN/ SSC Cycad Specialist Group 2003), though case and point is the various species that have become extinct in the wild to name but a few are: E. brevifoliolatus (IUCN 2010); E. relictus, which was described as Extinct in the Wild (Hurter & Glen 1996); E. nubimontanus (IUCN 2010) and various other local extinctions. Also populations where very few individuals are left are for all purposes already extinct as no seedling recruitment occurs (Goode 1989, 2001) and pollinators are thought to be extinct (Donaldson 1993b).

In terms of ethnobotanical use cycads are used as food (seeds and stems), starch (stems), for ceremonies and decoration (leaves), basket work (leaves), and medicinally or magically (stems, roots, bark; Norstog & Nicholls 1997). The practise of making flour from stem tissue is no longer in use in southern Africa (Donaldson 2008). Three quarters of Zulu Stangeria usage however is for magical purposes (Osborne et al. 1994). The harvesting of sections of bark (Donaldson 2008), or whole stems in the case of smaller species or individuals for sale (Cousins et al. 2011), is the usual practice (Figure 1.22, 1.23). This occurs mainly in the South African provinces of the Eastern Cape and KwaZulu-Natal (KZN), where this material is sold on the local muthi markets as far away as Johannesburg (Cousins et al. 2011). Several populations have suffered from this practice such as E. friderici-guilielmi, several of E. natalensis populations and even Critically Endangered species such as E. latifrons (Donaldson 2008). The harvesting of S. eriopus and other smaller Encephalartos or even juveniles has a greater impact as it results in the death of individuals. Exact numbers are noteworthy: Osborne et al. (1994) found that over 3000 plants were traded in the Durban area per month; Cunningham found that 30% of the 54 muthi shops investigated in KZN sold Encephalartos averaging a 50kg-size bag annually (Unpublished data cited in Cousins 2009); 50 shops surveyed on the Witwatersrand (Williams et al. 2001) and Johannesburg areas (Williams 2003) showed trade in Encephalartos in 4% and 6% of shops respectively; Mander (1998 cited in Cousins 2009) found Encephalartos to be ranked 12th in terms of importance according to traders; Cousins (2009) found that Encephalartos spp. were sold by 26.4% and 13.2% of traders at Faraday and Warwick muthi markets respectively, with the estimated total quantity of annually traded Encephalartos being approximately 9 tonnes.

1.5 Intra-generic concepts Intra-generic concepts here refers to the diversity within Encephalartos and Stangeria and our understanding thereof. The importance of applying consistent intra-generic classification stretches far beyond taxonomy but affects all other botanical aspects and importantly conservation (Donaldson 2003; Golding & Hurter 2003; Walters & Osborne 2004). Taxonomic authorities on the genera have consistently resolved to discourage the creation of intra-specific taxa (Dyer 1965; Vorster 2004a) as this would infer closer affinities, which cannot be substantiated (Vorster 2004a). As such only two subspecies are currently recognised (Osborne et al. in press) E. barteri subsp. allochrous and E. tegulaneus subsp. powsii (Table 1.1), both of which should be elevated to specific rank (Piet Vorster pers. comm.1). Mention is made below of the variation found, species concepts and delimitations, along with taxonomic problems.

1.5.1 Stangeria The taxonomy of Stangeria has been relatively stable since its inception, though there has always been an informal recognition of variation within S. eriopus (Chamberlain 1915, 1935). Dyer (1965) recognised two forms namely a grassland and forest form based on their morphology and habits. Various sources however are not always in agreement, mainly on numerical issues (e.g. leaf lengths and leaf numbers; Dyer 1965; Vorster & Vorster 1974; McLellan & Ndamase 1995; Goode 1989, 2001; Grobbelaar 2002; Jones 2002; Whitelock 2002).

The grassland form has shorter (250–600mm), pale, tough leathery leaves as it grows in full sun and is exposed to frequent fires (Figure 1.9). Leaves are held almost vertical by stiff rachises (McLellan & Ndamase 1995). Around 1–3 leaves are borne (Goode 1989), which may increase to over 10 in cultivation (Jones 2002). Leaves are replaced periodically (mostly annually) through the production of 2–3 leaves often after fire (Goode 2001) or most commonly only one (Vorster pers. comm.1). Leaflet margins are usually entire (Figure 1.9), or by comparison less dentate than forest plants, and tips seemingly rounded (Figure 1.9), with leaflets inserted perpendicular to the sun (McLellan & Ndamase 1995). Stems are usually multi-headed clumps possibly due to the effect of fire and the increased ones production (Vorster & Vorster 1974; Proche! & Johnson 2009), which terminates the apical meristem (Jones 2002). Cones are often borne on shorter peduncles, speculated to be light induced (Vorster & Vorster 1974).

The forest form, growing in deep shade, has much longer (up to 2.4m) wider, dark green leaves that are thinner and softer in texture (Figure 1.8). Leaflets are borne on a lax, arching rachis and are vertical in terms of the sun (McLellan & Ndamase 1995). They bear 2–3 leaves, 6 according to Jones (2002), 8–10 according to Whitelock (2002), that are replaced through the production of 1–2 leaves, or 3–4 according to Whitelock (2002), much less frequently than the grassland form. Leaflets are often heavily serrated to dentate or even pectinate and tips acute (Figures 1.8; 1.10;

Vorster & Vorster 1974). A single growth point i.e. an unbranched stem is usually found, possibly as this form produces cones less frequently and fire is excluded from its habitat (Jones 2002). A third form that occurs on the forest edge has been recognised by Grobbelaar (2002) characterised by stems with up to 8 leaves measuring 500–800mm long, which may explain Whitelock and Jones' discordance with other authors in terms of their forest forms.

Reproductive structures between all forms are almost identical and although a probable beetle pollinator (Coleoptera: Nitidulidae) has been found by Proche! & Johnson (2009) the various forms have not been investigated for pollinator specificity. This may well be taxonomically informative as both Bowenia species have different species-specific pollinators (Wilson 2002). Although these variants are environmentally correlated it does not seem to be environmentally induced as they keep their identity when transplanted to very different conditions in cultivation (McLellan & Ndamase 1995; Grobbelaar 2002; Whitelock 2002). The variation is most probably genetic in nature, with differing results in terms of in virtro growth (Resslar 2002 cited in Whitelock 2002) and isozyme allele analysis (McLellan & Ndamase 1995) being found between forms. This could indicate a divergence due to isolation within the species, which might warrant formal taxonomic recognition, the real challenge as yet has been delimitation due to intermediate forms.

1.5.2 Encephalartos The 65 species and two subspecies (Table 1.1) currently recognised in Encephalartos, as with most other cycad genera, are classified based on a morpho-geographic species concept (Vorster 2004a). This places emphasis on the morphological discontinuities (assuming they are indicative of reproductive isolation) coupled to geographical distribution (and isolation) as main criteria for species recognition (Vorster 2004a). Though most species have been delimited by vegetative differences a weighting towards reproductive structures and activities has been implemented of late (Vorster 1993). For example the vegetatively similar E. lehmannii and E. princeps were split based on their morphologically different cones (Dyer 1965) as have most recently described species (Vorster 1996a–e). Also the pollination studies on Encephalartos' two sister genera have yielded taxonomic evidence in terms of the host specificity of pollinators (Forster et al. 1994; Mound & Terry 2001; Terry 2001; Hall et al. 2004; Terry et al. 2004; Terry et al. 2005; Terry et al. 2008). There is some degree of species specificity of pollinators of Encephalartos (Suinyuy et al. 2009) with at least 11 species having specific relationship with beetles in the genus Porthetes (John Donaldson pers. comm.2). Partly on this basis, E. aplanatus was split from E. villosus because apparently different, yet closely related, Porthetes beetles were present on their cones (Vorster 1996a, Vorster & Oberprieler 1999). However, using molecular data, Downie et al. (2008) were unable to differentiate these two and concluded that the same species of pollinator occurs on both

2 Prof. John Donaldson IUCN/SSC Cycad Specialist Group Chair and Red List Subgroup Leader. South African National Biodiversity Institute (SANBI), Kirstenbosch Research Centre. Private Bag X7 Claremont 7735, SA 10

cycads. Downie et al. (2008) also concluded that there did not seem to be evidence for co- speciation between species of Encephalartos and species of Porthetes although the lack of a well resolved phylogeny for Encephalartos made such an analysis difficult. Unpublished results indicate that some Porthetes pollinators have several host plants (John Donaldson pers. comm.2) and recent studies of cone odours in Encephalartos indicate that there is convergence in cone odours that attract Porthetes between unrelated species of Encephalartos (Terence Suinyuy unpublished data). As a result, pollinator associations seems not to corroborate or assist in the current classification.

Taxonomic uncertainty however still abounds with some recognised species speculated to be conspecific, as differences are often small and some forms cannot equivocally be distinguished between species (Vorster 1993, 1995; Grobbelaar 2002)—e.g. the E. delucanus, E. poggei, E. marunguensis, E. schaijesii, E. schmitzii group, which may all be varieties of a single species (Goode 1989; Vorster 2004a); E. natalensis, E. lebomboensis, E. msinganus, E. aemulans, E. senticosus, and E. altensteinii where female cones are essential for confident identification (Goode 2001; Grobbelaar 2002); and E. brevifoliolatus and some forms of E. laevifolius (Grobbelaar 2002). Natural hybridisation further complicates matters (Vorster 1986, Giddy 1980, Goode 2001, Grobbelaar 2002) with hybridisation and the relative ease thereof often indicative of the level of relationship between taxa (Vorster 1986).

Morphological similarities have been reflected in the low genetic variation found in numerous studies, investigating from a few (1-2) to numerous species (78% of the genus in Treutlein et al. 2005). Studies have included analyses of allozyme data (Van der Bank et al. 1998, 2001), DNA fingerprinting (Coetzer 2000), random amplified polymorphic DNA (Coetzer 2000; Chaiprasongsuk et. al. 2007; Viljoen & Van Staden 2006; Ekué et al. 2008; Prakash & Van Staden 2008; Prakash et al. 2008), DNA sequence data from the chloroplast genome (Van der Bank et al. 1998, 2001; Mabunda 2007) and nucleur genome (Van der Bank et al. 1998, 2001; Treutlein et al. 2005), inter- specific sequence repeats (Treutlein et al. 2005; Prakash et al. 2008) and amplified fragment length polymorphism (Mabunda 2007). All these investigations suffer from either a lack of resolution or incomplete sampling and as such relationships between most of the species based on molecular data are uncertain, with no molecular phylogenetic framework available. Low levels of genetic differentiation however seems to be typical of many cycad genera (Sharma et al. 1998, 1999; González & Vovides 2002; Sharma et al. 2004; Caputo et al. 2004; Xiao et al. 2004; González- Astorga et al. 2005; González et al. 2008; Pinares et al. 2009; Sangin et al. 2010). Conversely large amounts of intra-specific morphological variation abounds with several (usually locality bound) forms recognised informally (Appendix 1) by collectors (Grobbelaar 2002). Many imperfectly known species and those occurring over a presumed large geographical range (e.g. E. septentrionalis; E. manikensis, E. natalensis) are thought to be more than a single species, due to

the clear morphological discontinuities within the species (Keppel et al. 2002; González-Astorga et al. 2003, Vorster 2004a). This has been recognised to some extent as E. aemulans was separated from E. natalensis (Vorster 1990), E. senticosus from E. lebomboensis (Vorster 1996e), E. eugene- maraisii (Lavranos & Goode 1988; Robbertse et al. 1989) and E. manikensis were both split into four species (Gilliland 1939; Dyer & Verdoorn 1969). Thus within Encephalartos a situation is currently found where there is little genetic difference between species but large amounts of morphological variation within currently assigned species with some natural hybridisation occurring even between morphologically clearly distinct species.

Taxonomy, in especially the “tropical” African species, where the original descriptions are some of the only in situ work done thereon (Melville 1957; Malaisse 1969; Heenan 1977; Malaisse et al. 1993), is difficult to interpret since there is a lack of field observations specially comparative ones (Vorster 1999; Ekué et al. 2008). Vorster (2004a) unequivocally states that “it seems doubtful if any botanist has seen E. septentrionalis since Schiewnfurth discovered it ! ; consequently the identity of the species is uncertain” and “E. macrostrobilus is also so poorly described it is impossible to visualise”, a trend that has continued in the recent description of E. flavistrobilus (Turner et al. 2006) which resulted in it being immediately subjugated to synonymous rank to E. schaijesi though its exact affinity is uncertian (Osborne et al. in press). Possible reasons for these conditions are the difficulty in making permanent herbarium records, poor decriptions and many species occur in places which are difficult to visit and revisit in order to verify characters (Melville 1957; Goode 2001; Golding & Hurter 2003; Vorster 2004a). Many of the non-South African cycads are thus in serious need of taxonomic attention with unconfirmed reports, photographic evidence and educated speculation indicating simultaneously that there are more species than is currently recognised and that some species may in fact be synonymous. Therefore, despite the fact that their taxonomy has been rather stable for the last decade or so (Haynes 2011; see species numbers Table 1.4), the genus is in dire need of a taxonomic revision. The current molecular phylogenetic study and DNA barcoding of the entire genus subsequently represents the first step towards such action.

1.6 Aims and hypothesis Cycads in general and the genera Encephalartos and Stangeria in particular are considered extremely important for several reasons: their evolutionary history and taxonomic position; their unusual life histories, traits and ecological services; they are considered ‘flagship’ species for conservation; and lastly due to their horticultural appeal and ethnobotanical significance. Taxonomic uncertainty in Encephalartos and Stangeria in terms of intra-generic and intra-specific concepts and relationships can be addressed using molecular techniques, specifically DNA sequencing. We therefore performed phylogenetic analyses of several molecular datasets, including nuclear and plastid regions. The choice of markers was based on previous studies on cycads (Hill et al. 2003; Bogler & Francisco-Ortega 2004; Sangin et al. 2008) and

recommendations by the CBOL Plant Working Group (Chase et al. 2007; Sass et al. 2007; CBOL Plant Working Group 2009). The main objective of this study is to investigate the phylogenetic relations within Encephalartos and Stangeria eriopus based on DNA sequences and determine a gene region, which will be specific for each species, i.e. a DNA barcode. Thus multiple accessions of each species was collected, which also served to address some of the uncertainty surrounding the variation found within species.

Specifically the project aims to: 1. Produce a complete molecular phylogeny for all currently recognised Encephalartos species using nuclear internal transcribed spacer (nrITS), subunit ‘a’ of ribulose 1,5-bisphosphate carboxylase/oxygenase (rbcLa), psbA-trnH intergenic spacer (psbA-trnH), and a portion of maturase K (matK). Hypothesis: The molecular phylogeny will confirm morpho-geographically recognised groups as proposed by Vorster (2004a; Table 1.5). 2. Determine whether morphological forms of Stangeria display DNA sequence differences for the regions nrITS, rbcLa, psbA-trnH, and matK. Hypothesis: Forms of Stangeria will have sequence differences. 3. DNA barcode all individuals sampled to complete the BOLD reference database and determine whether proposed DNA barcoding regions (rbcLa, matK, psbA-trnH, and nrITS) adhere to the guidelines as outlined by the Consortium for the Barcode of Life (CBOL Plant Working Group 2009). Hypothesis: The currently proposed barcode regions will fail to discriminate species. 4. Explore additional gene regions as DNA barcodes for Encephalartos including: 5s, Agamous, !- tubulin, atpFH, CC0702, CC0822, CC1147, CC1606, CC1798, CC2920, ef1-!, ETS, psbKI, SAHH, and trnL. Hypothesis: The additional regions investigated will provide a species specific barcode either individually or in combination.

Table 1.1 World list of Encephalartos species with outgroup species, synonyms, and IUCN Red List Status. Abbreviations: EW=Extinct in the Wild, CR=Critically Endangered, EN=Endangered, VU=Vulnerable, NT=Near Threatened, LC=Least Concern and DD=Data Deficient. KZN=KwaZulu- Natal, EC=Eastern Cape, L=Limpopo, M=Mpumalanga. * indicates genus type. (Osborne et al. in press; IUCN 2010).

Species Synonyms Distribution (province) Red List Encephalartos aemulans Vorster South Africa (KZN) CR E. altensteinii Lehm. E. marumii De Vreise South Africa (EC) VU E. regalis W.Bull E. aplanatus Vorster Swaziland VU E. arenarius R.A. Dyer South Africa (EC) EN E. barteri subsp. barteri Carruth. ex Benin (Bergu), Ghana, Nigeria VU Miq. E. barteri subsp. allochrous L.E. Nigeria (Plateau) EN Newton E. brevifoliolatus Vorster South Africa (L) EW E. bubalinus Melville Kenya, Tanzania (Masai) LC *E. caffer (Thunb.) Lehm. Cycas caffra Thunb. South Africa (EC) NT E. brachyphyllus Lehm. & De Vriese E. cycadis Sweet Zamia caffra (Thunb.) Thunb. Z. cycadis L.f. E. cerinus Lavranos & D.L. Goode South Africa (KZN) CR E. chimanimaniensis R.A. Dyer & I. Mozambique, Zimbabwe EN Verd. E. concinnus R.A. Dyer & I. Verd. Zimbabwe (Mberwenga, Runde) EN E. cupidus R.A. Dyer South Africa (M) CR E. cycadifolius (Jacq.) Lehm. E. elongatus Miq. South Africa (EC) LC E. eximius I. Verd. E. verchaffeltii Regel Zamia cycadifolia Jacq. Z. elongata (Miq.) Heynh. Z. villosa Gaertn. E. delucanus Malaisse, Sclavo & Tanzania (Mpanda) VU Crosiers

E. dolomiticus Lavranos & D.L. E. verrucosus Vorster South Africa (L) CR Goode E. dyerianus Lavranos & D.L. Goode E. graniticolus Vorster South Africa (L) CR E. equatorialis P.J.H. Hurter E. imbricans Vorster Uganda CR E. eugene-maraisii I. Verd. South Africa (L) EN E. ferox Bertol. f. E. kosiensis Hutch. Mozambique, South Africa (KZN) LC E. friderici-guilielmi Lehm. Zamia friderici-guilielmi (Lehm.) Heynh. South Africa (EC, KZN) NT E. ghellinckii Lem. South Africa (EC, KZN) VU E. gratus Prain Malawi, Mozambique VU E. heenanii R.A. Dyer South Africa (M), Swaziland CR E. hildebrandtii A. Braun & C.D. E. hildebrandtii var. dentatus Melville Kenya (Kilifi, Lamu), Tanzania (Lushoto, NT Bouché Tanga), Zanzibar E. hirsutus P.J.H. Hurter South Africa (L) CR E. horridus (Jacq.) Lehm. E. nanus Lehm. South Africa (EC) EN E. van-hallii de Vriese Zamia horrida Jacq. Z. tricuspidata Hort. E. humilis I. Verd. South Africa (M) VU E. inopinus R.A. Dyer South Africa (L) CR E. ituriensis Bamps & Lisowski Dem. Rep. of Congo, Uganda? NT E. kisambo Faden & Beentje E. voiensis A. Moretti, D.W. Stev. & Sclavo Kenya (Taita-Taveta) EN E. kanga Pócs & Q. Luke E. laevifolius Stapf & Burtt Davy South Africa (EC, KZN, L, M), Swaziland CR E. lanatus Stapf & Burtt Davy South Africa (M) NT E. latifrons Lehm. South Africa (EC) CR E. laurentianus De Wild. Angola, Dem. Rep. of Congo NT E. lebomboensis I. Verd. Mozambique, South Africa (KZN), Swaziland EN E. lehmannii Lehm. E. lehmanniana (Ecklon & Zeyher ex Ecklon) South Africa (EC) NT Lehm. E. mauritianus Miq. E. spinulosus Lehm. Zamia spinulosa (Lehm.) Heyhn. E. longifolius (Jacq.) Lehm. E. lanuginosus (Jacq.) Lehm. South Africa (EC) NT E. mackenziei L.E. Newton South Sudan NT E. macrostrobilus S. Jones & J. Northern Uganda VU Wynants E. manikensis (Gilliland) Gilliland Mozambique (Manica), Zimbabwe VU

E. marunguensis Devred Democratic Republic of Congo (Tanganyika) NT E. middelburgensis Vorster E. eugene-maraisii subsp. middelburgensis South Africa (Gauteng, M) CR Lavranos & D.L. Goode E. msinganus Vorster South Africa (KZN) CR E. munchii R.A. Dyer & I. Verd. Mozambique (Manica) CR E. natalensis R.A. Dyer & I. Verd. South Africa (KZN) NT E. ngoyanus I. Verd. South Africa (KZN), Swaziland VU E. nubimontanus P.J.H. Hurter E. venetus Vorster South Africa (L) EW E. paucidentatus Stapf & Burtt Davy South Africa (M), Swaziland VU E. poggei Asch. E. lemarinellianus De Wild. & Th. Dur. Democratic Republic of Congo (Lulua, Lomami, LC Lualaba) E. princeps R.A. Dyer South Africa (EC) VU E. pterogonus R.A. Dyer & I. Verd. Mozambique (Manica) CR E. relictus P.J.H. Hurter Swaziland EW E. schaijesii Malaisse, Sclavo & E. flavistrobilus I. Turner & Sclavo Democratic Republic of Congo (Lualaba) VU Crosiers E. schmitzii Malaisse Democratic Republic of Congo (Haut-Katanga), NT Zambia E. sclavoi A. Moretti, D.W. Stev. & De Tanzania (Tanga) VU Luca E. senticosus Vorster South Africa (KZN), Swaziland VU E. septentrionalis Schweinf. South Sudan, Uganda NT E. tegulaneus subsp. tegulaneus Kenya LC Melville E. tegulaneus subsp. powysii Kenya CR Mirungu & Beentje E. transvenosus Stapf & Burtt Davy South Africa (L) LC E. trispinosus (Hook.) R.A. Dyer E. horridus var. trispinosus Hook. South Africa (EC) VU E. turneri Lavranos & D.L. Goode Mozambique (Nampula) LC E. umbeluziensis R.A. Dyer E. villosus var. umbeluziensis (R.A. Dyer) J.Lewis Mozambique, Swaziland VU E. villosus Lem. South Africa (EC, KZN), Swaziland LC E. whitelockii P.J.H. Hurter E. successibus Vorster Uganda VU E. woodii Sander South Africa (KZN) EW *Stangeria eriopus (Kunze) Baill. Lomaria eriopus Kunze South Africa (KZN) NT Stangeria katzeri Regel S. paradoxa T. Moore S. schizodon Bull

*Lepidozamia hopei (W. Hill) Regel Catakidozamia hopei W. Hill Australia (Queensland) LC Macrozamia hopei (Regel) W. Hill ex C. Moore L. peroffskyana Regel Catakidozamia macleayi Miq. Australia (Queensland, New South Wales) LC E. denisonii (C. Moore & F. Muell) Regel Lepidozamia denisonii (C. Moore & F. Muell) Regel L. minor Miq. Macrozamia denisonii C. Moore & F. Muell M. gigas Miq. M. peroffskyana Regel (Miq.) Macrozamia communis L.A.S. Australia (New South Wales) LC Johnson M. macdonnellii (F. Muell. ex Miq.) E. macdonnellii F. Muell. ex. Miq. Australia (Northern Territory) LC A. DC. M. pauli-guilielmi W. Hill & F. Muell. E. pauli-guilielmi (W. Hill & F. Muell.) F. Muell. Australia (Queensland) EN Macrozamia plumosa A.Mohr M. tenuifolia Miq. Zamia mackenii Miq. M. plurinervia (L.A.S. Johnson) D.L. M. machinii P.I. Forst. & D.L. Jones Australia (New South Wales) NT Jones

Table 1.2 Historical classification (Hill & Stevenson 1998–2005) with species numbers and distribution (Osborne et al. in press).

Suborder Family Sub-Family Tribe Subtribe Genus Species Distribution Cycadinae Pacific Islands, Japan, China, Cycadaceae Cycas L. 105 South-east Asia, India, Sri Lanka, Madagascar Boweniaceae Bowenia Hook. ex Hook. f 2 Australian east coast Stangeriaceae Stangeria T. Moore 1 South African east coast Diooeae Dioon Lindl. 14 Central America mainly Mexico Encephalartinae Encephalartos Lehm. 65 Tropical & Zamianae southern Africa Encephalartoideae Encephalarteae Lepidozamia Regel 2 Australian east Zamiaceae Macrozamiinae coast Macrozamia Miquel 41 Mainly Australian east coast Ceratozamieae Ceratozamia Brongn. 27 Central America mainly Mexico Zamoideae Microcycadinae Microcycas (Miq.) A. DC. 1 Cuba Zamieae Central & South Zaminae Zamia L. 71 America, Gulf of Mexico Islands Chigua D.W. Stevenson 2 Colombia

Table 1.3 Encephalartos distribution per country (Golding & Hurter 2003).

Country Endemics Non-endemics Total number of taxa South Africa 33 8 41 Tanzania 4 4 8 Kenya 3 3 6 Uganda 3 1 4 Mozambique 3 11 14 Democratic Republic of Congo 2 5 7 Nigeria 1 1 2 Swaziland 1 8 9 Zimbabwe 1 2 3 Malawi 1 1 2 Angola - 2 2 Benin - 1 1 Central African Republic - 1 1 Ghana - 1 1 Sudan - 1 1 Zambia - 1 1

Table 1.4 Comparison of World Lists of Cycads (WL) 1–11 with species number (intra-specific taxa/sp. nov.; Modified from Haynes 2011).

Genus WL 1 WL 2 WL 3 WL 4 WL 5 WL 6 WL 7 WL 8 WL 9 WL 10 WL 11 Bowenia 2 2 2 2 2 2 2 2 2 2 2 Ceratozamia 9 (8) 10 11 11 12 12 21 21 21 20 26 Chigua - 2 2 2 2 2 2 2 2 2 - Cycas 15 (29) 17 (41) 33 (2) 39 (2) 52 (2) 64 (13) 98 (8) 99 (8) 99 (8) 100 (8) 105 (7) Dioon 10 (6) 10 (4) 10 (2) 10 (4) 10 (4) 10 (4) 13 13 13 13 14 Encephalartos 43 (4) 51 (5) 51 (2) 54 (2) 54 (2) 61 (2) 65 (2) 65 (2) 65 (2) 65 (2) 65 (2) Lepidozamia 2 2 2 2 2 2 2 2 2 2 2 Macrozamia 14 (3) 13 (3) 16 (3) 21 30 39 40 40 40 41 41 Microcycas 1 1 1 1 1 1 1 1 1 1 1 Stangeria 1 1 1 1 1 1 1 1 1 1 1 Zamia 33 (11) 46 54 44 44 49 58 59 57 57 69 Total 130 (61) 155 (53) 183 (9) 187 (8) 210 (8) 243 (19) 303 (10) 305 (10) 303 (10) 304 (10) 326 (9)

Table 1.5 Species groupings based on morphological and geographic similarities (Vorster 2004a).

Group Number Encephalartos species included 1 E. brevifoliolatus, E. cycadifolius, E. friderici-guilielmi, E. ghellinckii, E. humilis, E. laevifolius, E. lanatus 2 E. arenarius, E. horridus, E. latifrons, E. lehmannii, E. longifolius, E. trispinosus 3 E. princeps 4 E. aplanatus, E. caffer, E. cerinus, E. ngoyanus, E. umbeluziensis, E. villosus 5 E. altensteinii, E. natalensis, E. transvenosus 6 E. woodii 7 E. aemulans, E. lebomboensis, E. msinganus, E. senticosus 8 E. heenanii, E. paucidentatus, E. relictus 9 E. cupidus, E. dolomiticus, E. dyerianus, E. eugene-maraisii, E. middelburgensis, E. nubimontanus and possibly E. hirsutus 10 E. inopinus 11 E. chimanimaniensis, E. concinnus, E. manikensis, E. munchii, E. pterogonus and possibly E. turneri 12 E. gratus 13 E. bubalinus, E. equatorialus, E. hildebrandtii, E. ituriensis, E. kisambo, E. sclavoi, E. tegulaneus (both subspecies), E. whitelockii 14 E. macrostrobilus, E. mackenziei, E. septentrionalis 15 E. laurentianus 16 E. barteri (both subspecies) 17 E. delucanus, E. marunguensis, E. poggei, E. schaijesii, E. schmitzii 18 E. ferox

Figure 1.1 Typical arborescent cycad habit Figure 1.2 Acaulescent habit of (Taken from Giddy 1980). E. villosus in habitat.

Figure 1.3 Cycas thouarsii in habitat, Kaya Figure 1.4 Cycas thouarsii leaf with forest Kinondo, Kenya (Photo: M. Furr). prominent mid-vein (mid-rib).

Figure 1.5 Cycas thouarsii Figure 1.6 Seeds of S. eriopus, note female cone (Photo: A. Vogel). attachment below sporophyll stalk (Taken from Grobbelaar 2002).

Figure 1.7 Bowenia spectabilis with seeds with Figure 1.8 Forest form leaf detail of S. eriopus, white sarcotesta attached below sporophylls Oribi Gorge Nature Reserve. stalks (Photo: K. Elske).

Figure 1.9 Grassland form of S. eriopus in Figure 1.10 Subterranean stem of S. eriopus habitat, Oribi Gorge Nature Reserve. forest form, in habitat Oribi Gorge Nature Reserve.

Figure 1.11 Bowenia spectabilis habit Figure 1.12 Bowenia stem lacking persistent and fern like foliage (Photo: K. Elske). leaf bases (Photo: CSIRO).

Figure 1.13 Terminal S. eriopus male Figure 1.14 Stangeria eriopus female plant cone, note lack of persistent leaf bases. with terminal cone. Note removed male cone.

Figure 1.15 Cones on side branches of Bowenia spectabilis (Photo: K. Elske).

Figure 1.16 Previous morphological classifications (Taken from Hill et al. 2003).

Figure 1.17 (a) Phylogenetic consensus tree between Hill et al. (2003); Chaw et al. (2005) and Hermsen et al. (2006). Symbols for Bowenia, Dioon and Stangeria indicate alternate placements. In all cases the topology is in agreement with Hermsen et al. (2006) except for Encephalartos, Lepidozamia and Macrozamia (indicated with double-headed arrow), and (b) Shimodaira– Hasegawa tests indicating alternate placements for Bowenia, Dioon and Stangeria. (Modified from Zgurski et al. 2008).

Figure 1.18 Worldwide distribution of extant cycads: (1) Ceratozamia, Zamia, Dioon, and, Microcycas, (2) Encephalartos and Stangeria, (3) Cycas, and (4) Macrozamia, Lepidozamia, and Bowenia (Taken from Hill et al. 2003).

Figure 1.19 Approximate distribution of Encephalartos in Africa.

Figure 1.20 Approximate distribution of Encephalartos in South Africa, Swaziland and southern Mozambique.

Figure 1.21 Diversity and status of the world’s cycad flora in the four main regions of diversity. Abbreviations are: EW=Extinct in the Wild, CR=Critically Endangered, EN=Endangered, VU=Vulnerable, NT=Near Threatened, LC=Least Concern, and DD=Data Deficient. (Modified from Donaldson 2003).

Figure 1.22 Encephalartos bark sold at Figure 1.23 Stangeria eriopus stems sold at Faraday muthi market in Johannesburg. Faraday muthi market in Johannesburg.

CHAPTER 2. MATERIAL AND METHODS

2.1 Taxon sampling All 65 species of Encephalartos Lehm. (Osborne et al. in press) and numerous Stangeria eriopus “forms” were collected along with several samples of unknown identity usually with a common name based on possible origin (Appendix 1). Only one of the two subspecies of E. tegulaneus could be sampled as no individuals of E. tegulaneus subspecies powysii could be sourced (Table 2.1). Between one and ten individuals per species were sampled to incorporate species level variation into the sampling. Samples were taken from cultivated material as the plants are extremely rare and narrowly endemic, though in isolated cases material of wild origin was received. In most cases material in cultivation was of wild origin (Appendix 1). Additional information and photographs of individuals was collected (Figure 2.1–2.4) mainly in line with the requirements for submission to the Barcode of Life Datasystems (BOLD; ww.boldsystems.org; Ratnasingham & Herbert 2007) and also to capture characters lost in the pressing and drying process. A few grams of leaf tissue was collected for genetic analysis, cut into small pieces and placed into plastic bags filled with silica gel. A complete leaf was also collected as a voucher, all of which are housed at the herbarium (JRAU) of University of Johannesburg (UJ) (Appendix 1). Collection was undertaken at the Pretoria National Botanical Garden (PBG), University of Pretoria (UP), UJ, Lowveld National Botanical Garden (LBG), Cycad World of Innovation (a private nursery; CWOFI), and two main private collections, with various other samples received as donations from additional sources (Appendix 1). Additional DNA bank samples, housed at the New York Botanical Garden (NYBG) from material collected by Dr. Damon Little, was also incorporated in the study (Appendix 1). Voucher information and GenBank accession numbers of all samples used in the phylogenetic and DNA barcoding analyses are listed in Table 2.1.

2.2 DNA extraction Total DNA was extracted at the African Centre for DNA Barcoding (ACDB; UJ) from on average 0.5g (ranging from 0.3–3g) of fresh or silica dried leaf material. The extraction method employed was the 10X CTAB method (Doyle & Doyle 1987), altered by the addition of 2% PVP (Polyvinylpyrolidone) to aid in phenolic compound removal. All samples were then cleaned with 70% ethanol, dried for one day at room temperature, and re-suspended in 1–1.5ml double distilled water. All DNA extracts are banked at ACDB at -80°C. DNA aliquots were cleaned using QIAquick purification columns (QIAgen, Inc. Hilden, Germany) according to the manufacturer's protocols. Cleaned samples were run on 1% Agarose gel (1g agarose/100ml 1XTBE) with 7!l of Ethidium Bromide at ~110V for between 15–30min and viewed under ultraviolet light.

2.3 Polymerase chain reaction (PCR) All genetic work of the rbcLa gene was done off-site at the Canadian Centre for DNA Barcoding

(CCDB) at the University of Guelph. Total unpurified DNA was prepared in plates by the author according to BOLD handbook (http://www.boldsystems.org/docs/handbook.php) and shipped to Canada for sequencing. Further work at both NYBG and ACBD was done by P. Rousseau.

Primers used for the various proposed barcoding genes are found in Table 2.2 using either premix (Applied Biosystems, Inc., ABI, Warrington, Cheshire, UK) or recipes as described in Table 2.3. In all cases BSA was added to assist in the stabilisation of enzymatic reactions (Savolainen et al. 1995), while in some cases DMSO or Q-solution were added to aid in DNA strand separation due to GC rich regions and complex secondary structures (Winship 1989; Alvarez & Wendel 2003). For the nrITS region the program followed was: 94°C for 1 minute (min); 26 cycles—94°C for 1min, 48°C for 1min, and 72°C for 3min; final extension 72°C for 7min. For failed reactions the program was amended to: 95°C for 2.5min; 35 cycles—95°C for 30sec, 58°C for 30sec, and 72°C for 60sec; final extension 72°C for 10min. For matK the program followed: 95°C for 2.5min; 35 cycles—95°C for 30sec, 52°C for 30sec, and 72°C for 60sec; final extension 72°C for 10min. For psbA-trnH the program followed: 95°C for 2.5min; 35 cycles—95°C for 30sec, 55°C for 30sec, and 72°C for 60sec; final extension 72°C for 10min. For rbcLa PCR conditions were: 94°C for 4min; 35 cycles— 94°C for 30sec, 55°C for 30sec, and 72°C for 1min; final extension 72°C for 10min.

Additional samples tested are listed in Table 2.4, while Table 2.5 lists regions tested. PCR amplification was performed using the following standard program: 95°C for 2.5min, 35 cycles: 95°C for 30sec; (50–60)°C for 30sec; 72°C for (30–120)sec; 72°C for 10min. If reactions were unsuccessful, annealing temperature and extension times were varied as indicated in an attempt to achieve successful amplification. All PCR recipes follow those from the NYBG (Table 2.3) with DNA amounts varying between 0.25–1µl.

2.4 DNA sequencing At NYBG samples were sent for sequencing at the High-Throughput Genomics Unit, Department of Genome Sciences, University of Washington in Washington (United States of America) using ExoSAP-IT (Affymetrix, Santa Clara, CA), while sequencing at ACBD was done by P. Rousseau. In both cases sequencing was done using the ABI Prism BigDyeC V3.1 Terminator Mic (Apllied Biosystems, Inc., ABI, Warrington, Cheshire, UK) kits and methodologies. The ACDB cycle sequencing program followed: 26 cycles—96°C for 10sec, 50°C for 5sec, and 60°C for 4min. After cycle sequencing, products were cleaned using EtOH-NaCl method as prescribed by ABI to remove excess terminators before sequencing on the ABI 3130xl genetic analyser. At the CCDB the EdgeBio® AutoDTR™ 96™ clean-up procedure was followed with protocols as per Appendix 2.

2.5 Sequence alignment and analysis of molecular data Complementary strands were assembled and edited using Sequencher 4.6 (Gene Codes Corp.,

Ann Arbor, Michigan, United States of America). All sequences were aligned manually in PAUP (version 4.0b10, Swofford 2002), with nrITS and psbA-trnH sequences aligned with MUSCLE prior to insertion into PAUP. Lastly sequences used for analyses were uploaded to GenBank (Table 2.1) with all rbcLa sequences (Figure 2.3) and trace files (Figure 2.4) also deposited on BOLD (Appendix 1). Aligned matrices are available from the author ([email protected]).

2.5.1 Phylogenetic analyses The dataset included rbcLa and matK for all 10 recognised Cycadophyta genera, nrITS for Encephalartos, Macrozamia and Lepidozamia, and psbA-trnH for Encephalartos and Lepidozamia. Sequences for nrITS and psbA-trnH were not included for the other genera as alignment and availability was problematic. Numerous molecular studies have also confirmed Encephalartos, Marcozamia and Lepidozamia as a monophyletic group, with Lepidozamia sister to Encephalartos (Treutlein & Wink 2002; Hill et al. 2003; Bogler & Francisco-Ortega 2004 ; Chaw et al. 2005; Sangin et al. 2008; Zgurski et al. 2008). Cycas was used as outgroup in the analyses as its phylogenetic placement has consistently been resolved as basal to all other cycads (Figure 1.18). Selected specimens were not sequenced for all genetic regions, interpreted as missing data in combined matrices. This however did not change the topology of any analysed trees (from visual comparison) and were thus included in the analysis.

Analyses were conducted for (1) each gene separately, (2) combined plastid, (3) nuclear data, and (4) combined plastid and nuclear data sets. All matrices were analysed in PAUP based on Parsimony (MP) in a heuristic search with 1000 replicates of random taxa addition, 10 trees retained at each step and tree-bisection-reconnection (TBR) branch swapping engaged and multrees not in affect. Delayed transformation character optimisation (DELTRAN) was used to calculated branch lengths, due to the reported errors (http://paup.csit.fsu/problems.html) with accelerated transformation optimisation (ACCTRAN) in PAUP v4.0b1. Internal support was calculated using bootstrap analysis (Felsenstein 1985) with 1000 bootstrap replicates, using simple sequence addition, no tree swapping and saving 10 trees per replicate. Only clades with a frequency greater than 50BP (Bootstrap support) are reported. Bootstrap support was classified as: high (85–100BP); moderate (75–84BP); or low (50–74BP).

Bayesian analysis (BI; Huelsenbeck & Ronquist 2001; Ronquist & Huelsenbeck 2003) was performed on only the combined plastid and nuclear dataset using MRBAYES v.3.1.2. For each genetic region the most appropriate model was selected using MODELTEST v. 3.06 (Posada & Crandall 1998). Models selected were: nrITS=K81uf+G, matK=K81uf+G, rbcLa=TrN+I, and psbA- trnH=TrN+I. For models K81uf+G rates were set as gamma and nts=6, for TrN+I rates were set as equal and nts=6. Markov Chain Monte Carlo estimates were run for 2 000 000 generations with trees sampled every 500 generations with log-likelihood scores used to determine the stationary

31 point. All trees prior to the stationary point were discarded as burn-in (first 1000 trees) and removed. Posterior probabilities (PP) were mapped onto a 50% majority-rule consensus tree using MACDRAW PRO with scores classed as high above 0.95PP or low below 0.95PP.

2.5.2 DNA barcoding analyses For DNA barcoding all samples were sequenced for all four barcoding regions with the exception of nrITS2, which was excluded from the analysis. This was done because of the large size and associated difficulties of using the entire region for barcoding. The first fragment nrITS1 was selected above the previously proposed nrITS2 as nrITS2 was found to be much less variable in other taxa (Gao et al. 2010; Liu et al. 2010; Yao et al. 2010). Encephalartos, Lepidozamia, Macrozamia and Stangeria were inlcuded with Stangeria selected as outgroup due to the closer affinity of the other three genera (Zgurski et al. 2008). Settings and methods using PAUP were as described above. Analyses was performed for all genes separately, core barcodes (matK + rbcLa) and core barcode with proposed additional regions (e.g. matK + rbcLa + psbA-trnH and/or nrITS). Species resolution was calculated by the amount of monophyletic species groups each gene and combinations thereof could recover in MP analyses. However in cases where anomalies were retrieved these were considered to be monophyletic should such an event be explainable (e.g. known taxonomic uncertainty, probable identification difficulties).

Pairwise distance was calculated in PAUP using the Kimura-2-parameter setting (Lahaye et al. 2008a, b) for various gene combinations including test genes. These were done for Encephalartos and Stangeria separately with no outgroups included in any matrices. These were then exported to Microsoft excel where mean total genetic, intra- and inter-specific genetic variation and number of genetically identical species was calculated. Significance test were conducted using R Project for Statistical Computing version 2.13.1 (R Development Core Team 2011), calculating p-values by the Kolmogorov-Smirnov test. Success rates were calculated from the amount of taxa successfully sequenced.

Table 2.1 List of samples present in analyses. ‘-‘ Indicates samples without GenBank accession numbers. Collection sources abbreviations: Private Collection 1 (PC1), Private collection 2 (PC2), Private collection 3 (PC3), University of Johannesburg (UJ), University of Pretoria Experimental farm (EF), University of Pretoria Manie van der Schijff Botanical Garden (MBG), Cycad World of Innovation (CWOFI), Lowveld National Botanical Garden (LBG), New York Botanical Garden DNA bank (NYBG), Pretoria National Botanical Garden (PBG), Umtamvuna Nature Reserve (UNR), Rhenosterpoort Nature Reserve (RNR), Xander de Kock donation (XDK), Middelburg Cycad Trail (MCT), Fairchild Tropical Garden (FTG).

Species Collection Number Source Origin Genebank accession numbers rbcLa matK psbA-trnH nrITS Samples used in phylogenetic analyses E. aemulans PR861 (JRAU) MBG Wild JQ025439 JQ046261 JQ045891 JQ046123 E. aemulans d1018 (NY) NYBG Wild - - - - E. altensteinii PR650 (JRAU) UJ Cultivated - - - - E. altensteinii PR668 (JRAU) PBG Wild JQ025442 JQ046260 JQ045897 JQ046122 E. aplanatus PR922 (JRAU) LBG Wild - - - - E. aplanatus d1126 (NY) NYBG Wild - - - - E. arenarius PR758 (JRAU) CWOFI Cultivated JQ025453 JQ046258 JQ045978 JQ046120 E. arenarius d1020 (NY) NYBG Wild - - - - E. barteri subsp. allochrous PR892 (JRAU) EF Cultivated JQ025458 JQ046255 - JQ046117 E. barteri subsp. barteri d1095 (NY) NYBG Wild - - - - E. brevifoliolatus Xdk1 (JRAU) XDK Wild JQ025460 JQ046254 JQ045976 JQ046116 E. brevifoliolatus Xdk2 (JRAU) XDK Wild JQ025459 JQ046253 JQ045975 JQ046115 E. bubalinus PR885 (JRAU) EF Cultivated JQ025466 JQ046252 JQ045974 JQ046114 E. bubalinus d1021 (NY) NYBG Wild - - - - E. caffer PR729 (JRAU) PC 2 Cultivated JQ025468 JQ046250 JQ045972 JQ046112 E. caffer d1101 (NY) NYBG Wild - - - - E. cerinus PR859 (JRAU) MBG Wild JQ025475 JQ046249 JQ045971 JQ046111 E. cerinus PR744 (JRAU) PC 2 Cultivated - - - - E. chimanimaniensis PR883 (JRAU) LBG Cultivated JQ025477 JQ046248 JQ045970 JQ046110 E. chimanimaniensis PR888 (JRAU) LBG Cultivated JQ025476 JQ046247 JQ046109 E. concinnus PR890 (JRAU) EF Cultivated JQ025479 JQ046246 JQ045968 JQ046108 E. concinnus d1023 (NY) NYBG Wild - - - - E. cupidus PR691 (JRAU) PC 1 Cultivated JQ025481 JQ046245 JQ045967 JQ046107 E. cupidus PR767 (JRAU) CWOFI Cultivated JQ025482 JQ046244 JQ045966 JQ046106

E. cycadifolius PR683 (JRAU) PBG Wild JQ025483 JQ046243 JQ045965 JQ046105 E. cycadifolius d1128 (NY) NYBG Wild - - - - E. delucanus d1129 (NY) NYBG Wild - - - - E. dolomiticus PR865 (JRAU) MBG Wild JQ025489 JQ046242 JQ045964 JQ046104 E. dolomiticus CC343 (NY) NYBG Wild - - - - E. dyerianus PR731 (JRAU) PC 2 Cultivated JQ025491 JQ046241 JQ045963 JQ046103 E. dyerianus CC332 (NY) NYBG Wild - - - - E. equatorialis PR899 (JRAU) LBG Wild JQ025497 JQ046240 JQ045962 JQ046102 E. equatorialis PR900 (JRAU) EF Cultivated JQ025494 JQ046239 JQ045961 JQ046101 E. eugene-maraisii PR872 (JRAU) MBG Wild JQ025502 JQ046238 JQ045960 JQ046100 E. eugene-maraisii d1024 (NY) NYBG Wild - - - - E. ferox PR841 (JRAU) MBG Wild JQ025507 JQ046237 JQ045959 JQ046099 E. ferox d1025 (NY) NYBG Wild - - - - E. friderici-guilielmi PR853 (JRAU) MBG Wild JQ025512 JQ046234 JQ045956 JQ046096 E. friderici-guilielmi CC338 (NY) NYBG Wild - - - - E. ghellinckii Abbott9220 (JRAU) UNR Wild JQ025519 JQ046233 JQ045955 JQ046095 E. ghellinckii PR773 (JRAU) CWOFI Cultivated JQ025518 JQ046232 JQ045954 JQ046094 E. gratus PR774 (JRAU) CWOFI Cultivated JQ025520 JQ046231 JQ045953 JQ046093 E. gratus PR891 (JRAU) LBG Wild JQ025521 JQ046230 JQ045952 JQ046092 E. heenanii PR775 (JRAU) CWOFI Cultivated JQ025528 JQ046229 JQ045951 JQ046091 E. heenanii PR776 (JRAU) CWOFI Cultivated JQ025524 JQ046228 JQ045950 JQ046090 E. hildebrandtii PR824 (JRAU) CWOFI Cultivated JQ025529 JQ046227 JQ045949 JQ046089 E. hildebrandtii d1027 (NY) NYBG Wild - - - - E. hirsutus PR718 (JRAU) PC 2 Cultivated JQ025534 JQ046226 JQ045948 JQ046088 E. hirsutus CC336 (NY) NYBG Wild - - - - E. horridus PR846 (JRAU) MBG Wild JQ025538 JQ046225 JQ045947 JQ046087 E. horridus PR847 (JRAU) MBG Wild JQ025537 JQ046224 JQ045946 JQ046086 E. humilis PR712 (JRAU) PC 2 Cultivated JQ025541 E. inopinus PR778 (JRAU) CWOFI Cultivated JQ025546 JQ046222 JQ045944 JQ046084 E. inopinus PR864 (JRAU) MBG Wild JQ025547 JQ046221 JQ045943 JQ046083 E. ituriensis d1131 (NY) NYBG Wild - - - - E. kisambo PR745 (JRAU) PC 2 Cultivated JQ025551 JQ046220 JQ045942 JQ046082 E. kisambo d1132 (NY) NYBG Wild - - - - E. laevifolius PR803 (JRAU) CWOFI Cultivated JQ025552 JQ046216 JQ045938 JQ046078 E. laevifolius PR845 (JRAU) MBG Wild JQ025555 JQ046215 JQ045937 JQ046077 34

E. lanatus PR828 (JRAU) MBG Wild JQ025562 JQ046213 JQ045935 JQ046075 E. lanatus 587 PRU RNR Wild JQ025560 JQ046214 JQ045936 JQ046076 E. latifrons PR678 (JRAU) PBG Wild JQ025568 JQ046212 JQ045934 JQ046074 E. latifrons PR811 (JRAU) CWOFI Cultivated JQ025566 JQ046211 JQ045933 JQ046073 E. laurentianus PR789 (JRAU) CWOFI Cultivated JQ025572 JQ046210 JQ045932 JQ046072 E. laurentianus d1029 (NY) NYBG Wild - - - - E. lebomboensis PR657 (JRAU) UJ Cultivated JQ025576 JQ046209 JQ045931 JQ046071 E. lebomboensis PR698 (JRAU) PC 1 Cultivated JQ025578 JQ046208 JQ045930 JQ046070 E. lebomboensis PR831 (JRAU) MBG Wild JQ025580 JQ046207 JQ045929 JQ046069 E. lehmannii PR661 (JRAU) PBG Wild JQ025585 JQ046206 JQ045928 JQ046068 E. lehmannii PR780 (JRAU) CWOFI Cultivated JQ025583 JQ046205 JQ045927 JQ046067 E. longifolius PR809 (JRAU) CWOFI Cultivated JQ025591 JQ046204 JQ045926 JQ046066 E. longifolius PR873 (JRAU) MBG Wild JQ025592 JQ046203 JQ045925 JQ046065 E. macrostrobilus CC333 (NY) NYBG Wild - - - - E. manikensis PR903 (JRAU) EF Cultivated JQ025597 JQ046201 JQ045923 JQ046063 E. manikensis PR697 (JRAU) PC 1 Cultivated JQ025600 JQ046202 JQ045924 JQ046064 E. marunguensis PR912 (JRAU) EF Cultivated JQ025602 JQ046200 JQ045922 JQ046062 E. marunguensis CC446 (NY) NYBG Wild - - - - E. middelburgensis PR726 (JRAU) PC 2 Cultivated JQ025608 JQ046199 JQ045920 JQ046060 E. middelburgensis 586 PRU RNR Wild JQ025604 JQ046262 JQ045921 JQ046061 E. munchii PR737 (JRAU) PC 2 Cultivated JQ025614 JQ046196 JQ045917 JQ046057 E. munchii PR855 (JRAU) MBG Cultivated JQ025613 JQ046195 JQ045916 JQ046056 E. natalensis PR802 (JRAU) CWOFI Cultivated JQ025619 JQ046194 JQ045915 JQ046055 E. natalensis d1035 (NY) NYBG Wild - - - - E. ngoyanus PR717 (JRAU) PC 2 Cultivated JQ025626 JQ046193 JQ045914 JQ046054 E. ngoyanus PR799 (JRAU) CWOFI Cultivated JQ025624 JQ046192 JQ045913 JQ046053 E. nubimontanus PR655 (JRAU) UJ Cultivated JQ025631 JQ046191 JQ045912 JQ046052 E. nubimontanus PR704 (JRAU) PC 1 Cultivated JQ025629 JQ046190 JQ045911 JQ046051 E. paucidentatus PR710 (JRAU) PC 2 Cultivated JQ025632 JQ046188 JQ045909 JQ046049 E. paucidentatus PR849 (JRAU) MBG Wild JQ025636 JQ046283 JQ045999 JQ046139 E. poggei PR813 (JRAU) CWOFI Wild JQ025638 JQ046187 JQ045908 JQ046048 E. princeps PR810 (JRAU) CWOFI Cultivated JQ025641 JQ046186 JQ045907 JQ046047 E. princeps PR871 (JRAU) MBG Wild JQ025639 JQ046185 JQ045906 JQ046046 E. pterogonus PR876 (JRAU) PC 1 Cultivated JQ025642 JQ046184 JQ045905 JQ046045 E. pterogonus d1136 (NY) NYBG Wild - - - - E. schaijesii CC447 (NY) NYBG Wild - - - - E. schmitzii PR819 (JRAU) CWOFI Cultivated JQ025644 JQ046183 JQ045904 JQ046044 35

E. schmitzii d1137 (NY) NYBG Wild - - - - E. sclavoi d1038 (NY) NYBG Wild - - - - E. senticosus PR663 (JRAU) PBG Wild JQ025650 JQ046182 JQ045903 JQ046043 E. senticosus PR833 (JRAU) MBG Wild JQ025652 JQ046181 JQ045902 JQ046042 E. septentrionalis d1138 (NY) NYBG Wild - - - - E. tegulaneus subsp. tegulaneus PR825 (JRAU) CWOFI Cultivated JQ025665 JQ046029 JQ046168 E. tegulaneus subsp. tegulaneus PR877 (JRAU) EF Cultivated JQ025664 JQ046180 JQ045901 JQ046041 E. transvenosus PR727 (JRAU) PC 2 Cultivated JQ025671 JQ046179 JQ045900 JQ046040 E. transvenosus PR832 (JRAU) MBG Wild JQ025667 JQ046178 JQ045899 JQ046039 E. trispinosus PR868 (JRAU) MBG Wild JQ025674 JQ046177 JQ045898 JQ046038 E. trispinosus d1043 (NY) NYBG Wild - - - - E. turneri PR886 (JRAU) EF Cultivated JQ025680 JQ046176 JQ046037 E. turneri d1044 (NY) NYBG Wild - - - - E. umbeluziensis PR815 (JRAU) CWOFI Cultivated JQ025687 JQ046175 JQ045896 JQ046036 E. umbeluziensis d1046 (NY) NYBG Wild - - - - E. villosus PR671 (JRAU) PBG Wild JQ025694 JQ046173 JQ045894 JQ046034 E. villosus PR838 (JRAU) MBG Wild JQ025594 JQ046172 JQ045893 JQ046033 E. whitelockii d1048 (NY) NYBG Wild - - - - E. woodii PR675 (JRAU) PBG Wild JQ025702 JQ046170 JQ046031 E. woodii PR875 (JRAU) PC 3 Wild JQ025701 JQ046169 JQ045890 JQ046030 Lepidozamia peroffskyana d1050 (NY) NYBG Wild - - - - Lepidozamia hopei d1049 (NY) NYBG Wild - - - - Macrozamia macdonnellii d1057 (NY) NYBG Wild - - - - Macrozamia pauliguilielmi d1096 (NY) NYBG Wild - - - - Macrozamia plurinervia d1060 (NY) NYBG Wild - - - - Macrozamia communis d1051 (NY) NYBG Wild - - - - Little & Stevenson Wild Cycas thouarsii R. Br. ex Gaudich. 1001 (NY) FTG AF394336 AB116589 - - Little & Stevenson Wild Ceratozamia mexicana Brongn. 1009 (NY) FTG AF394345 AF279794 - - Little & Stevenson Wild Dioon edule Lindl. 1092 (NY) FTG AF531203 AB076193 - - Little & Stevenson Wild Microcycas calocoma (Miq.) A. DC. 1063 (NY) FTG AF531214 AB076194 - - Little & Stevenson Wild Zamia furfuracea L. f. 1098 (NY) FTG AF202959 AF410170 - - Stangeria eriopus PR860 (JRAU) EF Cultivated JQ025704 JQ046263 JQ045979 - 36

Additional samples used in DNA barcoding E. altensteinii PR756 (JRAU) CWOFI Cultivated JQ025441 JQ046312 JQ046028 JQ046167 E. altensteinii PR856 (JRAU) MBG Wild JQ025445 JQ046311 JQ046027 JQ046166 E. altensteinii d1019 (NY) NYBG Wild - - - - E. aplanatus PR757 (JRAU) CWOFI Cultivated JQ025447 JQ046310 JQ046026 JQ046165 E. aplanatus PR682 (JRAU) PBG Wild JQ025446 JQ046259 JQ045969 JQ046121 E. arenarius PR854 (JRAU) MBG Wild JQ025455 JQ046257 JQ045977 JQ046119 E. barteri subsp. barteri PR878 (JRAU) EF Cultivated JQ025457 JQ046256 - JQ046118 E. bubalinus PR910 (JRAU) EF Cultivated JQ025465 JQ046251 JQ045973 JQ046113 E. concinnus PR817 (JRAU) UP Cultivated JQ025478 JQ046309 JQ046025 JQ046164 E. cerinus d1022 (NY) NYBG Wild - - - - E. cupidus PR734 (JRAU) PC 2 Cultivated JQ025480 JQ046308 JQ046024 JQ046163 E. cupidus d1127 (NY) NYBG Wild - - - - E. dyerianus PR769 (JRAU) CWOFI Cultivated JQ025493 JQ046307 JQ046023 JQ046162 E. dyerianus PR820 (JRAU) CWOFI Cultivated JQ025659 JQ046306 JQ046022 JQ046161 E. dyerianus PR821 (JRAU) CWOFI Cultivated JQ025657 JQ046305 JQ046021 JQ046160 E. dyerianus PR863 (JRAU) MBG Wild JQ025492 JQ046304 JQ046020 JQ046159 E. ferox PR651 (JRAU) UJ Cultivated JQ025508 JQ046303 JQ046019 JQ046158 E. ferox PR676 (JRAU) PBG Wild JQ025509 JQ046302 JQ046018 JQ046157 E. ferox PR771 (JRAU) CWOFI Cultivated JQ025505 JQ046301 JQ046017 JQ046156 E. ferox PR844 (JRAU) MBG Wild JQ025506 JQ046236 JQ045958 JQ046098 E. friderici-guilielmi PR733 (JRAU) PC 2 Cultivated JQ025510 JQ046300 JQ046016 JQ046155 E. friderici-guilielmi PR772 (JRAU) CWOFI Cultivated JQ025514 JQ046235 JQ045957 JQ046097 E. ghellinckii CC331 (NY) NYBG Wild - - - - E. horridus PR777 (JRAU) CWOFI Cultivated JQ025536 JQ046299 JQ046015 JQ046154 E. horridus d1028 (NY) NYBG Wild - - - - E. humilis CC340 (NY) NYBG Wild - - - - E. kanga PR907 (JRAU) EF Cultivated JQ025658 JQ046298 JQ046014 JQ046153 E. kisambo PR823 (JRAU) CWOFI Cultivated JQ025550 JQ046297 JQ046013 JQ046152 E. laevifolius PR730 (JRAU) PC 2 Cultivated JQ025556 JQ046219 JQ045941 JQ046081 E. laevifolius PR798 (JRAU) CWOFI Cultivated JQ025554 JQ046218 JQ045940 JQ046080 E. laevifolius PR801 (JRAU) CWOFI Cultivated JQ025553 JQ046217 JQ045939 JQ046079 E. laevifolius CC335 (NY) NYBG Wild - - - - E. lanatus d1133 (NY) NYBG Wild - - - - E. latifrons PR806 (JRAU) CWOFI Cultivated JQ025565 JQ046296 JQ046012 JQ046151 E. lebomboensis PR805 (JRAU) CWOFI Cultivated JQ025663 JQ046293 JQ046009 JQ046148 E. lebomboensis PR796 (JRAU) CWOFI Cultivated JQ025575 JQ046295 JQ046011 JQ046150 37

E. lebomboensis PR800 (JRAU) CWOFI Cultivated JQ025581 JQ046294 JQ046010 JQ046149 E. lebomboensis PR874 (JRAU) MBG Wild JQ025574 JQ046292 JQ046008 JQ046147 E. lebomboensis d1030 (NY) NYBG Wild - - - - E. lehmannii PR835 (JRAU) MBG Wild JQ025584 JQ046291 JQ046007 JQ046146 E. lehmannii d1031 (NY) NYBG Wild - - - - E. longifolius PR673 (JRAU) PBG Wild JQ025587 JQ046290 JQ046006 JQ046145 E. longifolius PR808 (JRAU) CWOFI Cultivated JQ025588 JQ046289 JQ046005 JQ046144 E. manikensis PR795 (JRAU) CWOFI Cultivated JQ025595 JQ046288 JQ046004 E. manikensis d1033 (NY) NYBG Wild - - - - E. middelburgensis PR827 (JRAU) MCT Wild JQ025605 JQ046287 JQ046003 JQ046143 E. middelburgensis CC337 (NY) NYBG Wild - - - - E. msinganus PR701 (JRAU) PC 1 Cultivated JQ025610 JQ046198 JQ045919 JQ046059 E. msinganus PR751 (JRAU) PC 2 Cultivated JQ025611 JQ046197 JQ045918 JQ046058 E. msinganus d1034 (NY) NYBG Wild - - - - E. munchii d1134 (NY) NYBG Wild - - - - E. natalensis x E. woodii PR842A (JRAU) MBG Cultivated JQ025620 JQ046286 JQ046002 JQ046142 E. ngoyanus PR703 (JRAU) PC 1 Cultivated JQ025625 JQ046285 JQ046001 JQ046141 E. ngoyanus d1135 (NY) NYBG Wild - - - - E. nubimontanus PR792 (JRAU) CWOFI Cultivated JQ025630 JQ046284 JQ046000 JQ046140 E. nubimontanus CC328 (NY) NYBG Wild - - - - E. paucidentatus d1036 (NY) NYBG Wild - - - - E. poggei PR911 (JRAU) EF Cultivated JQ025637 JQ046282 JQ045998 JQ046138 E. princeps PR836 (JRAU) MBG Wild JQ025586 JQ046281 JQ045997 JQ046137 E. princeps d1037 (NY) NYBG Wild - - - - E. sclavoi PR738 (JRAU) PC 2 Cultivated JQ025647 JQ046280 JQ045996 JQ046136 E. sclavoi PR790 (JRAU) CWOFI Cultivated JQ025646 JQ046279 JQ045995 JQ046135 E. senticosus PR719 (JRAU) PC 2 Cultivated JQ025654 JQ046278 JQ045994 JQ046134 E. senticosus PR812 (JRAU) CWOFI Cultivated JQ025648 JQ046277 JQ045993 JQ046133 E. senticosus PR830 (JRAU) MBG Wild JQ025651 JQ046276 JQ045992 JQ046132 E. senticosus d1039 (NY) NYBG Wild - - - - E. sp PR913 (JRAU) EF Cultivated JQ025655 JQ046275 JQ045991 JQ046131 E. tegulaneus subsp. tegulaneus PR747 (JRAU) PC 2 Cultivated JQ025666 JQ046274 JQ045990 JQ046130 E. tegulaneus subsp. tegulaneus d1040 (NY) NYBG Wild - - - - E. transvenosus PR665 (JRAU) PBG Wild JQ025672 JQ046273 JQ045989 JQ046129 E. transvenosus PR797 (JRAU) CWOFI Cultivated JQ025669 JQ046272 JQ045988 JQ046128 E. transvenosus PR829 (JRAU) MBG Wild JQ025668 JQ046271 JQ045987 JQ046127 E. transvenosus d1041 (NY) NYBG Wild - - - - 38

E. trispinosus PR680 (JRAU) PBG Wild JQ025675 JQ046270 JQ045986 JQ046126 E. umbeluziensis PR858 (JRAU) MBG Wild JQ025685 JQ046174 JQ045895 JQ046035 E. villosus PR816 (JRAU) CWOFI Cultivated JQ025693 JQ046269 JQ045985 JQ046125 E. villosus PR837 (JRAU) MBG Wild JQ025692 JQ046268 JQ045984 JQ046124 E. whitelockii PR818 (JRAU) CWOFI Cultivated JQ025696 JQ046171 JQ045892 JQ046032 Stangeria eriopus PR706 (JRAU) PC 1 Cultivated JQ025707 JQ046267 JQ045983 - Stangeria eriopus PR753 (JRAU) PC 2 Cultivated JQ025705 JQ046266 JQ045982 - Stangeria eriopus PR842B (JRAU) EF Cultivated JQ025706 JQ046265 JQ045981 - Stangeria eriopus PR843 (JRAU) EF Cultivated JQ025708 JQ046264 JQ045980 -

Table 2.2 Primers used for proposed barcoding regions.

Gene Primers Sequence (3'–5') Reference rbcLa rbcLa-F ATGTCACCACAAACAGAGACTAAAGC Levin 2003 rbcLa-R GTAAAATCAAGTCCACCRCG Kress & Erickson 2007 rbcLajf634R GAAACGGTCTCTCCAACGCAT Frazekas et al. 2008 matK NY1505 ATYGYRCTTTTATGTTTACARGC Li et al. 2011 NY1506 TCAYCCGGARATTTTGGTTCG Li et al. 2011 nrITS AB101 AACACTTCACCGGACCATGAATTCGT Douzery et al. 1999 AB102 AACGGCGAGCGAACCGGGGAATTCTA Douzery et al. 1999 ITS3F GCTGCGTTCTTCATCGATGC Sun et al. 1994 ITS2R GCATCGATGAAGAACGCAGC Sun et al. 1994 NY728 GCCACGATGAAGAACGTAGC Hill et al. 2003 NY514 CTTTTCCTCCGCTTATTGATATG Adams et al. 2002 NY90 GGAAGTAAAAGTCGTAACAAGG White et al. 1990 NY729 GCTACGTTCTTCATCGTGGC Hill et al. 2003 psbA-trnH NY1494 CCGACGACGAACTAACATTTG Sass et al. 2007 NY1493 CGAGCCTGTTTCTGGTTCTC Sass et al. 2007

Table 2.3 Polymerase chain reaction recipes.

University of Guelph reagents 96 well plate 10–20% trehalose 625µl ddH20 200µl 10X buffer 125µl 50mM MgCl2 62.5µl 10µM primers 12.5µl 10mM dNTPs 6.25µl Polymerase (5U/µl) 6µl Total 1050µl DNA template (diluted 10x) 1–2µl

New York Botanical Garden reagents per reaction 10X buffer 1.5µl Q-solution (all CC genes, ef1-!, nrITS, matK, 5S, ETS, SAAH) 3µl BSA 1.5µl 10µM primers 0.75–1.5µl 2.5mM dNTPs 1.2µl Polymerase (41.67U/µl) 0.12µl Total 6.5µl DNA template 0.25–1.5µl

Table 2.4 Encephalartos and Stangeria samples used to test various genes for DNA barcoding.

Sample Species Genes investigated PRU586 E. middelburgensis atpFH, ef1-!, trnL, ETS, 5S, SAHH, psbKI, Agamous, !- tubulin PRU587 E. lanatus atpFH, ef1-!, trnL, ETS, 5S, SAHH, psbKI, Agamous, !- tubulin PR651 E. ferox psbKI PR652 E. villosus SAHH PR653 E. friderici-guilielmi SAHH PR659 E. hildebrandtii psbKI PR662 E. lebomboensis atpFH, ef1-!, trnL, ETS, 5S, SAHH, psbKI, Agamous, !- tubulin PR663 E. senticosus psbKI, CC0822, Agamous, SAHH, CC1606, CC0702, CC1147, CC2920, CC1798 PR675 E. woodii psbKI, CC0822, Agamous, SAHH, CC1606, CC0702, CC1147, CC2920, CC1798 PR676 E. ferox psbKI, CC0822, Agamous, SAHH, CC1606, CC0702, CC1147, CC2920, CC1798 PR677 E. lanatus psbKI, CC0822, Agamous, SAHH, CC1606, CC0702, CC1147, CC2920, CC1798 PR678 E. latifrons psbKI, CC0822, Agamous, SAHH, CC1606, CC0702, CC1147, CC2920, CC1798 PR679 E. arenarius psbKI, CC0822, Agamous, SAHH, CC1606, CC0702, CC1147, CC2920, CC1798 PR681 E. paucidentatus SAHH PR682 E. aplanatus psbKI, CC0822, Agamous, SAHH, CC1606, CC0702, CC1147, CC2920, CC1798 PR683 E. cycadifolius atpFH, ef1-!, trnL, ETS, 5S, SAHH, psbKI, Agamous PR684 E. eugene-maraisii psbKI, CC0822, Agamous, SAHH, CC1606, CC0702, CC1147, CC2920, CC1798 PR687 E. dyerianus psbKI, CC0822, Agamous, SAHH, CC1606, CC0702, CC1147, CC2920, CC1798 PR688 E. aemulans psbKI, CC0822, Agamous, SAHH, CC1606, CC0702, CC1147, CC2920, CC1798 PR691 E. cupidus atpFH, ef1-!, trnL, ETS, 5S, SAHH, psbKI, Agamous, !- tubulin PR692 E. ghellinckii psbKI, CC0822, Agamous, SAHH, CC1606, CC0702, CC1147, CC2920, CC1798 PR693 E. gratus psbKI, CC0822, Agamous, SAHH, CC1606, CC0702, CC1147, CC2920, CC1798 PR697 E. manikensis atpFH, ef1-!, trnL, ETS, 5S, SAHH, psbKI, Agamous, !- tubulin PR703 E. ngoyanus psbKI, SAHH PR706 S. eriopus psbKI PR710 E. paucidentatus atpFH, ef1-!, trnL, ETS, 5S, SAHH, psbKI, Agamous, !- tubulin PR712 E. humilis atpFH, ef1-!, trnL, ETS, 5S, SAHH, psbKI, Agamous, !- tubulin PR714 E. natalensis SAHH PR718 E. hirsutus psbKI, SAHH PR726 E. middelburgensis SAHH PR728 E. eugene-maraisii psbKI PR729 E. caffer psbKI, SAHH PR731 E. dyerianus atpFH, ef1-!, trnL, ETS, 5S, SAHH, psbKI, Agamous, !- tubulin

PR737 E. munchii psbKI, CC0822, Agamous, SAHH, CC1606, CC0702, CC1147, CC2920, CC1798 PR738 E. sclavoi psbKI PR742 E. altensteinii SAHH PR745 E. kisambo psbKI, SAHH PR747 E. tegulaneus subsp. SAHH tegulaneus PR749 E. horridus SAHH PR767 E. cupidus psbKI PR772 E. friderici-guilielmi psbKI, SAHH PR773 E. ghellinckii psbKI PR774 E. gratus SAHH PR775 E. heenanii atpFH, ef1-!, trnL, ETS, 5S, SAHH, psbKI, Agamous, !- tubulin PR776 E. heenanii psbKI, SAHH PR777 E. horridus atpFH, ef1-!, trnL, ETS, 5S, SAHH, psbKI, Agamous, !- tubulin PR788 E. ituriensis SAHH PR789 E. laurentianus SAHH PR790 E. sclavoi SAHH PR793 E. manikensis SAHH PR805 E. lebomboensis SAHH PR817 E. concinnus SAHH PR819 E. schmitzii SAHH PR822 E. nubimontanus psbKI, SAHH PR823 E. kisambo psbKI PR824 E. hildebrandtii SAHH PR832 E. transvenosus psbKI, SAHH PR834 E. senticosus psbKI, SAHH PR836 E. princeps psbKI, CC0822, Agamous, SAHH, CC1606, CC0702, CC1147, CC2920, CC1798 PR842B S. eriopus psbKI, SAHH PR843 S. eriopus SAHH PR844 E. ferox psbKI, SAHH PR848 E. woodii SAHH PR851 E. dolomiticus SAHH PR855 E. munchii atpFH, ef1-!, trnL, ETS, 5S, SAHH, psbKI, Agamous, !- tubulin PR871 E. dolomiticus psbKI, SAHH PR881 E. hildebrandtii psbKI, CC0822, Agamous, SAHH, CC1606, CC0702, CC1147, CC2920, CC1798 PR885 E. bubalinus psbKI, CC0822, Agamous, SAHH, CC1606, CC0702, CC1147, CC2920, CC1798 PR892 E. barteri subsp. allochrous SAHH PR900 E. equatorialis psbKI, SAHH PR903 E. manikensis SAHH PR910 E. bubalinus SAHH PR911 E. poggei SAHH PR912 E. marunguensis psbKI, SAHH Xdk1 E. brevifoliolatus atpFH, ef1-!, trnL, ETS, 5S, SAHH, psbKI, Agamous, !- tubulin Xdk2 E. brevifoliolatus SAHH

Table 2.5 Additional genes tested as possible DNA barcodes.

Gene Primers Sequence (3'-5') Reference 5s NY130 KTMGYGCTGGTATGATCGCA Cox et al. 1992 NY129 TGGGAAGTCCTYGTGTTGCA Cox et al. 1992 Agamous NY48 TGTGTTGTCGACGGCAGTGGGAAGATTGAGATAAA Deyue Yu (Helsinki) unpublished NY868 CCCATRGCTTGYTTRCCCAT Javier Francisco- Ortega unpublished !-tubulin NY975 GAGATACGCATGATGAAGGA D.P. Little unpublished NY976 AGGAAGGCCACCCAATGAT D.P. Little unpublished atpFH NY822 ACTCGCACACACTCCCTTTCC Kim-Joong Kim unpublished NY823 GCTTTTATGGAAGCTTTAACAAT Kim-Joong Kim unpublished CC0702 NY1159 TTTACCAACAATTACAACCCA Kado et al. 2008 NY1160 GTTCTGCAGAGTTTGCCTT Kado et al. 2008 CC0822 NY1161 TGTCTGCCCATTGAGAAGT Kado et al. 2008 NY1162 TCAGATGCCATGTTGATAAGA Kado et al. 2008 CC1147 NY1163 GCACCCATCTCACACTTGTC Kado et al. 2008 NY1164 CCACCCTGTTCATGTGATTCT Kado et al. 2008 CC1606 NY1167 TAACCAGCTTTGCCCTCAG Kado et al. 2008 NY1168 ATACAATTCGCGGCTACCATA Kado et al. 2008 CC1798 NY1169 GGCGGCGGAGATTACTGTA Kado et al. 2008 NY1170 TAGAAGACGCGCATTTGAGAA Kado et al. 2008 CC2920 NY1175 CCGCCACATTCACGCCCTCT Kado et al. 2008 NY1176 CGCAGTTCCAGTAGTTTCTC Kado et al. 2008 ef1- ! NY981 GGAGGATTGTTCTTCCTTGC D.P. Little unpublished NY982 CCCTTGTACCAATCAAGGTT D.P. Little unpublished ETS NY97 ATCTCAGTGGATCGTGGCAG Bult & Zimmer 1993 NY151 CGACTTCTCCTTCCTCTC Soltis et al. 2008 NY153 TCCTATTGTGTTGGCCTT Soltis et al. 2008 psbKI NY824 TTAGCCTTTGTTTGGCAAG Kim-Joong Kim pers. comm. NY825 AGAGTTTGAGAGTAAGCAT Kim-Joong Kim pers. comm. SAHH NY832 TGGTGYTCNTGCAACATCTTCTC Moynihan et al. in press NY835 ACKTTKCCRGCRATCATNACATCAG Moynihan et al. in press trnL NY55 CGAAATCGGTAGACGCTACG Taberlet et al. 1991 NY57 ATTTGAACTGGTGACACGAG Taberlet et al. 1991

Figure 2.1 Information displayed on BOLD under each accession.

Figure 2.2 Photographs as displayed under each specimen profile on BOLD.

Figure 2.3 Sequence and auxillary information of rbcLa for each specimen as displayed on BOLD.

Figure 2.4 Trace file as displayed for each specimen on BOLD.

CHAPTER 3. PHYLOGENETIC RELATIONSHIPS IN ENCEPHALARTOS: RESULTS AND DISCUSSION

These analyses provide the first DNA sequence evidence for the phylogenetic placement of E. friderici-guilielmi, E. brevifoliolatus, E. hirsutus, E. equatorialis, E. ituriensis, E. macrostrobilus, E. poggei, E. marunguensis, E. delucanus, E. schaijesii, and E. schmitzii. For this study E. relictus, E. mackenziei and E. tegulaneus subsp. powysii are the only taxa missing from the analysis.

3.1 Molecular evolution Of the three plastid regions used, matK had the highest number of variable sites (24.3%) compared to rbcLa (15.2%) and psbA-trnH (18.6%). The number of potentially parsimony informative characters for psbA-trnH (1.9%) is much lower than the other two plastid regions 4.2% and 9.6% for rbcLa and matK respectively, though this is most probably due to the lack of outgroup comparative data in psbA-trnH. Variable sites evolved at a similar rate (Table 3.1) and matK and psbA-trnH performed equally well (as measured by the retention index). The nrITS region has variable sites (21.21%) comparable to matK but substantially more potentially informative characters (28.9%) compared to plastid regions (Table 3.1). Evolution rates of all region are very mostly similar being 1.3 (Table 3.1).

Results from three analyses are presented: combined plastid regions (rbcLa + matK + psbA-trnH; Figure 3.1), nrITS (Figure 3.2) and combined plastid and nrITS (Figure 3.3). Results are compared both to each other and previous studies with mention made to significant (defined as supported by bootstrap) and notable (defined as novel but unsupported) results. The combined analysis is discussed in full while the combined plastid and nuclear trees are discussed based on only significant and novel results not found in the combined tree.

3.2 Combined plastid analysis Visual inspection of all three plastid regions (not shown) confirmed no strongly supported incongruence allowing direct combination of the data. As such only the results from the combined plastid dataset are presented and discussed here. The parsimony analysis yielded 1005 most parsimonious trees of 628 steps with a Consistency Index (CI) of 0.89 and Rentention Index (RI) of 0.79 (Table 3.1). The bootstrap consensus tree is illustrated in Figure 3.1. Encephalartos is moderately (78 BP) supported as monophyletic, however relationships within the genus is poorly resolved resulting in a large polytomy. Species groupings are in accordance with the combined analysis except for the significant but poorly supported (63BP) association of E. hirsutus with E. eugene-maraisii. This is noteworthy as it is partially in accordance with Vorster’s (2004a) hypothesis and these species are also geographically well isolated from other possibly related species. They are respectively the most western (E. eugene-maraisii) and northern (E. hirsutus)

46 distributed species in South Africa (above the Eastern Cape) and share many morphological traits. Also striking is the failure of E. heenanii “long leaf form” to group with E. heenanii “short leaf form”. The latter form groups with the presumably related E. paucidentatus (Vorster 2004a) with poor (62BP) support. The “long leaf form” has often been considered a hybrid between E. heenanii and E. paucidentatus but is found here simply to be divergent. The failure of several other species replicates to group together can be ascribed to low sequence divergence as well as missing data resulting in support below 50BP.

3.3 Nuclear ITS analysis Analysis resulted in 1081 equally most parsimonious trees with a tree length of 488 steps, CI of 0.85 and RI of 0.96 (Table 3.1). The bootstrap consensus tree is illustrated in Figure 3.2. Encephalartos is strongly supported (100BP) as monophyletic. Within the genus three major clades are retrieved. Clades A and B are highly supported (99BP and 92BP respectively) and successively sister to Clade C which is moderately supported (79BP). The nrITS analysis is congruent with the combined analysis although markedly less resolved.

3.4 Combined molecular analysis The produced trees are congruent with previous effort on cycad generic systematics (see Zgurski et al. 2008) as Lepidozamia is resolved as sister to Encephalartos (87BP/0.97PP) which, together with Macrozamia forms a supported grouping (81BP/0.53PP). Other genera are reduced to a polytomy (86BP/0.97PP) most probably due to a lack of data, while Stangeria and Cycas are the early diverging lineages to the rest of the phylum. As mentioned previously species and essentially species groups within Encephalartos are currently recognised based on a combination of morphological and geographical data. Here the various lineages found from the combined phylogenetic analysis will be systematically compared to the hypothesis (Section 1.6, hypothesis 1) and other phylogenetic studies with mention made to derived, shared and taxonomically important (Vorster 1993, 1999) characters.

Since the bootstrap consensus trees (Figure 3.1, 3.2) did not conflict with each other, although many taxa were reduced to polytomies in the plastid analysis, the plastid and nrITS datasets were directly combined. Parsimony analysis resulted in 926 equally parsimonious trees (tree length 1151, CI of 0.84 and RI of 0.92. Of the 3735 included characters 2852 were constant, 883 (23.64%) were variable and 440 (11.8%) were potentially parsimoniously informative. Because the maximum parsimony (MP) and bayesian analyses produced similar topologies, the posterior probability values (BI) are mapped onto the maximum parsimony tree (Figure 3.3). Encephalartos has high support (90BP/0.97PP) for being monophyletic consisting of three major clades, with Lepidozamia as sister group.

3.4.1 Clade A: lineage 1 47

E. brevifoliolatus E. cycadifolius E. friderici-guilielmi E. ghellinckii E. humilis E. laevifolius E. lanatus

Clade A: lineage 1 has moderate support (84BP) to be monophyletic and sister to the rest of the genus in the MP analysis but very poor support in the bayesian (0.55PP). It is clearly divided into two sub-lineages. Lineage 1: sub-lineage A is weakly (73BP) supported in the MP and highly supported (1.0PP) in the bayesian. Lineage 1: sub-lineage B has no support though a subset including all but E. ghellinckii has 81BP and 0.56PP support. Lineage 1: sub-lineage A includes two strongly supported sister species E. friderici-guilielmi (96BP/1.0PP) and E. cycadifolius (91BP/1.0PP). Lineage 1: sub-lineage B groups the morphologically distinct E. ghellinckii sister to four closely related taxa (Vorster 1986; 81BP/0.56PP). Notably E. leavifolius is paraphyletic, with one set sister to E. humulis (77BP/0.56PP) while the rest are unresolved with the morphologically almost identical E. brevifoliolatus. This lineage is supported in all taxonomic analyses of the genus: confirming the hypothesis of Vorster (2004a); in an 86 character (morphological, anatomical, biochemical, and other observations) phenetic study (including all except the unanalysed E. brevifoliolatus) (Osborne et al. 1993); in terms of certain anatomical features (all except the unanalysed E. brevifoliolatus) (Koeleman 1978; Koeleman et al. 1981); through a 55 character matrix of morphological and biogeographical characters with 96BP (all except the unanalysed E. brevifoliolatus and E. friderici-guilielmi) (Treutlein et al. 2005); through maximum parsimony and maximum likelihood analyses of the nrITS gene with 89BP (all except the unanalysed E. brevifoliolatus and E. friderici-guilielmi) (Treutlein et al. 2005); and Chaw et al. (2005) from 5.8S and nrITS2 sequences with both species analysed sister to E. longifolius (75BP).

Species included in clade A: lineage 1 occur over a large area along the eastern mountains of South Africa from the Winterberg (Eastern Cape) in the south east, to the Wolkberg (Limpopo) in the north (IUCN 2010; Figure 3.4). They occur at high elevations (700–2000m above sea level) often exposed to frost and even snow, an exceptional condition within the genus (Dyer 1965). Taxa represented in this clade are easily distinguishable vegetatively by their narrow to linear entire leaflets (Figure 3.5), being 8–12mm (Vorster 2004a), but mostly less than 10mm wide (Vorster 2005). Reproductively they are easily identifiable by their extremely pubescent cones (Figure 3.6, 3.7; Vorster 2004a), a trait found to some degree in unrelated members with similar habitat preferences (e.g. E. heenanii). Entomological differentiation also occurs with the cone-associated beetle genus Platymerus (Coleoptera: Curclionidae) occurring on at least two species of the group

48 but on no other Encephalartos species, while the ovule parasitic genus Antliarhinus (Coleoptera: Curclionidae), prevalent on the rest of the genus, is absent from the group (Oberprieler 1995a, b). The split between Platymerus and Antliarhinus, which dates back to the early tertiary (Oberprieler 1995b), mirrors that between clade A: lineage 1 and the rest of the genus. It seems Antliarhinus spp. do not recognise species of clade A as hosts, as under experimental conditions transferred insects make no attempt to oviposit and transferred larvae do not survive (Donaldson 1993a). This situation might have arisen due to a shift in coning phenology (Donaldson 1993a) as the group reproduces six months later than other Encephalartos species and cones disintegrate at least three months earlier than any other species (Vorster 2004a). Various insect species are specific to clade A including: the butterflies Callioratis mayeri and C. curlei (Lepidoptera: Geometridae), beeltes Porthetes spp., Phacecormynes spp., and Amorphocerus spp. (Coleoptera: Curculionidae), Apinotropis spp. (Coleoptera: Anthribidae), and an undescribed Erotylidae beetles species (Coleoptera: Cucujoidea) (Donaldson et al. 1995; Oberprieler 1995a; Staude 2001; Downie et al. 2008; Suinyuy et al. 2009). Analysis of the weevil tribe Amorphocerini suggests that the Amorphocerus species associated with E. cycadifolius is ancestral and basal to all others in the tribe, while Porthetes species found on this lineage are recently divergent (Downie et al. 2008).

Other shared vegetative traits of Clade A species include leaflets reduced only slightly in size towards the base always with a clear petiole (60–250mm) (Figure 3.7) and yellowish rachis (Figure 3.7; Grobbelaar 2002). Female cones are clearly pedunculate (Grobbelaar 2002) and mostly cylindrical in shape (Goode 1989) with rounded apices (Figure 3.6, 3.7) due to a low sterile sporophyll percentage (>8%; Grobbelaar 2002). Sarcotesta is yellow to amber (Figure 3.8), more or less the same colour as cones (Osborne et al. 1993; Treutlein et al. 2005) which may be indicative of some ecological specialisation of the group (Tang 1983). Other seeds traits shared by the group are a thin sarcotesta (sarcotesta index 7–13%; Grobbelaar 2002) and sclerotesta grooved to various degrees (Grobbelaar 2002; Treutlein et al. 2005). Members of the group also display the full range of Encephalartos growth forms and life histories from arborescent (E. laevifolius; Figure 3.9), intermediate sized ~1–2m (E. lanatus; Figure 3.10) and dwarf subterranean (E. humilis; Figure 3.11) including masting species (Grobbelaar 2002; Raimondo & Donaldson 2003) with desiccating and spontaneously disintegrating cones (Grobbelaar 1989, 1993, 2002). No natural hybrids between the group and other Encephalartos occur even when growing sympatrically, most probably due to the asynchronous cone phenology. Hybridisation experiments ex situ have also suggested some barrier, possibly cytological (Vorster 2004a), casting doubt on the reported partial success between E. friderici-guilielmi ! E. altensteinii (Myburgh 1991). The paraphyly of E. leavifolius is worth investigating as the taxon is one of the few Encephalartos species that has a very large distribution range with marked gaps between populations in the north and those along the eastern coast of South Africa with a large amount of variation found within the species.

3.4.2 Clade B Clade B is supported (82BP/1.0PP) as monophyletic and sister (77BP/0.53PP) to clade C. It contains 17 species in four lineages. Resolution within lineages is generally poor with relationships between the different lineages uncertian due to a lack of resolution on the backbone of the clade. This clade has seen partial recognition by Osborne et al. (1993) where 13 of the 15 species analysed grouped together; Treutlein et al. (2005) found all 17 species of the group analysed to cluster together with 95BP.

Respresentatives of clade B are endemic to South Africa and Swaziland (Figure 3.12) occurring in a more or less continuous range of species pockets from the southernmost distribution point of the genus in the Suuranys mountains near Kareedouw in the Eastern Cape province (E. longifolius) up to the Barberton district in the Mpumalanga province (E. paucidentatus) with the distribution of E. transvenosus increasing the clade's range to the Soutpansberg in the Limpopo province (IUCN 2010). This clade includes all, and only, the arborescent green leaved species endemic to South Africa and Swaziland, except for some species from clade B: lineage 5. These (E. horridus, E. lehmannii, E. princeps, E. trispinosus) are confined to the xeric Eastern Cape habitats and have glaucous blue foliage and/or diminutive statures (Figure 3.13). Insects specific to the clade include various Antliarhinus spp., Porthetes spp., and Amorphocerus spp. (Coleoptera: Curculionidae) (Oberprieler 1995a; Downie et al. 2008). Hybridisation has been successful within and between lineages and with members from both other clades (Table 3.2).

Clade B: lineage 2 E. arenarius E. horridus E. latifrons E. lehmannii E. longifolius E. princeps E. trispinosus

The lineage has no support in the MP analysis but high support in the bayesian analysis (0.97PP). Of the 7 included species only E. princeps (94BP/1.0PP), E. horridus (59BP/0.98PP) and E. longifolius (57BP/1.0PP) are resolved with the relationships between these and the other taxa unresolved. The lack of resolution between the mostly morphologically distinct taxa (Figure 3.13– 3.16) is noteworthy but probably reflective of their evolutionary history in terms of the ease and prevalence of hybridisation between these taxa (Table 3.2) along with low sequence divergence (Table 3.1). This lineage is contrary to the hypothesis but has seen recognition in Vorster (2004a) except for the exclusion of E. princeps; Osborne et al. (1993) encompassing all species; Treutlein

50 et al. (2005) including all but E. princeps in their molecular (66BP) and morphological analyses (77BP); Chaiprasongsuk et al. (2007) found all three of the analysed species to group together based on 9 morphological traits and Random Amplified Polymorphic DNA data analysis; and Van der Bank et al.'s (1998, 2001) allozyme and DNA sequence analyses including all species, with E. princeps sister in some analyses.

Species of this group are the most southerly (E. longifolius) distributed occurring relatively widespread across the Eastern Cape province (Figure 3.17; IUCN 2010). Most have pronounced lobes on their leaflets (Figure 3.13, 3.15; Van der Bank et al. 1998, 2001) and/or glaucous leaves (Figure 3.14, 3.16; Vorster 2005). All species except E. princeps (Figure 3.14) also habitually produce solitary cones (Figure 3.13), a condition uncommon elsewhere in the genus. However this condition is sporadically reversed in cultivation even amongst female individuals (where reproductive effort is significantly higher), possibly indicating the trait as derived with E. princeps being intermediate (Rousseau 2011). Various Porthetes species are restricted to members of the lineage (Oberprieler 1995a; Downie et al. 2008). Other shared characters include a substantial clear petiole (100–350mm; Figure 3.14); a relatively high sterile sporophyll percentage (9–39%; Figure 3.13–3.14) and seed sarcotesta index (32–48%; Grobbelaar 2002), cones that are essentially green but mostly with a russet indumentum; and red sarcotesta (Goode 1989). This is probably the lineage in which the most intra-lineage hybridisation has been observed with several inter-lineage natural hybrids (Table 3.2).

Clade B: lineage 3 E. transvenosus

Clade B: lineage 3 has moderate support in the MP (77BP) and high support in bayesian analysis (1.0PP). Encephalartos transvenosus has always been associated with members of clade B: Lineage 4 previously and is thus not in support of the hypothesis (Vorster 2004a). The species has been group by Vorster (2004a) with clade B: lineage 4; Osborne et al. (1993) with clade B: lineage 4 and E. paucidentatus; and Treutlein et al. (2005) concluded that E. transvenosus is allied with clade B: lineage 4 species in their morphological analysis (85BP).

Encephalartos transvenosus occurs in the Limpopo province from just south of Tzaneen to Thohoyandou in the north (Figure 3.18; IUCN 2010). It is vegetatively umistakable with its very neat incubously overlapping leaves that are not arching (Figure 3.19). Its cones (Figure 3.20-21) however closely resemble those of members of clade B: lineage 4 with which it also shares a large stature, green leaves and a fast growth rate. The species has been hybridised with species in clade B as wel as clade C species (Table 3.2).

Clade B: lineage 4 E. altensteinii E. natalensis E. msinganus E. woodii

Clade B: lineage 4 is not supported in either the MP or bayesian analysis. Encephalartos altensteinii (73BP/0.99PP) although not supported is sister to E. natalensis, E. msinganus and E. woodii which are grouped together with high support (0.95PP). The inability of E. msinganus and E. natalensis replicates to resolve as monophyletic may indicate the conspecificity of the species and the need to reevaluate the species delimitation of E. natalensis. The clear grouping of E. woodii (0.98PP) with a split between material from Kranskloof (0.96PP) and Ngoye (65BP/1.0PP) material is noteworthy. The material from Kranskloof was however found during sequencing of the nrITS gene to be of hybrid origin due to the clear double peaks in the raw data, though when edited overlapping fragments clear of any double peaks could be assembled and used for the analysis.

This group is not in contrast with the hypothesis of Vorster (2004a) but has partially recognised E. altensteinii and E. natalensis are grouped (with E. transvenosus) while E. woodii is considered to be closely related. Osborne et al. (1993) includes E. natalensis, E. woodii and E. altensteinii with other arborescent species (e.g. E. transvenosus). Treutlein et al.'s (2005) morphological analysis found E. transvenosus to be included with E. altensteinii, E. natalensis, and E. woodii (86BP). The relationship of E. woodii with E. natalensis has been the subject of numerous studies including Osborne & Paschke (1993), Osborne & Baijnath (1995), Viljoen & Van Staden (2006) and Prakash & Van Staden (2008).

Encephalartos natalensis and E. altensteinii occur in a more or less continuous range from the Bushman River in the south (E. altensteinii) through to Port Shepstone on the Eastern Cape-KZN border (Figure 3.22) where they occur almost sympatrically, with E. natalensis extending as far north as the Umfolozi River near Vryheid (IUCN 2011). E. msinganus is only known from the type locality in the Msinga district of KZN in the drainage area of the Buffels River (IUCN 2010). Only one male specimen of E. woodii was discovered at the Ngoye forest in KZN (Figure 3.22; IUCN 2010) though a very similar-looking group of individuals are in cultivation and presumable from the Kranskloof area ~160km away (Osborne & Baijnath 1995; Prakash et al. 2008).

Encephalartos natalensis and E. altensteinii are considered by many as conspecific (Grobbelaar 2002; Vorster 2004a) with only trivial differences in their large stature, long green leaves and reproductive structures (Figure 3.23, 3.24) and seemingly more intra-specific than inter-specific variation (compare Figures 3.25-3.27; Grobbelaar 2002). The range of variation of especially E.

52 natalensis is immense (see Grobbelaar 2002) with many other species previously lumped within it (e.g. E. msinganus, E. aemulans). Encephalartos woodii is very distinct morphologically with the crowding of multiple (3–4) teeth at the base of the leaflets upper margin in juveniles and lower leaflets of adults (a trait only found in some members of clade C: lineage 6), its lack of further spinescence including leaflet tips, and its soft, green, often glossy foliage (Figure 3.28, 3.29). Insects restricted include: Antliarhinus sp. n. 1 (Oberprieler 1995a) and Porthetes sp. n. 14 on both E. altensteinii and E. natalensis; Porthetes sp. n. 9 and P. dissimilis is restricted to E. altensteinii (Downie et al. 2008); and a novel beetle of the Erotylidae was found exclusively on E. natalensis in the Pietermaritzburg (KZN) area (Suinyuy et al. 2010) while Ascotis reciprocaria (Lepidoptera: Geomitridae) feeds on no other cycad (Cooper & Goode 2004). Hybrids are common with E. altensteinii, which is the most promiscuous natural hybridiser and has been hybridised artificially with species from all three clades. The hybridisation of E. altensteinii with E. friderici-guilielmi should however be considered suspect (Table 3.2). A notable E. natalensis hybrid is the E. natalensis ! E. woodii (Cafasso et al. 2001) specimens that have matured and were rehybridised with E. woodii pollen (Hurter 2008). As a result of its status (Extinct in the Wild) and only male plants of E. woodii known, it has been a prime subject for artificial hybridisation (Table 3.2).

Clade B: lineage 5 Clade B: lineage 5 received low support in the MP (<50BP) and high support in the bayesian analysis (0.94PP) and includes two sister clades weakly supported in the MP and strongly supported in the bayesian analysis, which comprise E. senticosus, E. aemulans and E. lebomboensis (50BP/0.96PP), and E. paucidentatus and E. heenanii (61BP/1.0PP) respectively. This is the first time these two sub-lineages have been considered as sister groups. Species occur in a more or less continuous range (Figure 3.30) from northern KZN through Swaziland along the Lebombo Mountains through Piggs Peak to Barberton in Mpumalanga (IUCN 2010). These species all share a large stature (2.5–4m; Figure 3.31) and have green leaves with nodules on the adaxial leaf surface (Figure 3.32) to varying degrees (Grobbelaar 2002). Cones are more or less yellow, with a red sarcotesta (Figure 3.33–4.35; Goode 1989; Grobbelaar 2002). The two sub-lineages are however clearly distinguishable from each other in that lineage 5: sub-lineage A has veins (vascular bundle tissue) raised on the adaxial surface (Figure 3.36), conspicuous to the touch. Also all members of lineage 5: sub-lineage A have substantial (150–450mm) clear petioles while in lineage 5: sub-lineage B most have a clear petiole less than 100mm with some leaflets reduced to prickles towards the leaf base (Goode 2001; Grobbelaar 2002).

Lineage 5: sub-lineage A E. heenanii E. paucidentatus

In sub-lineage A the “long leaf form” of E. heenanii is sister to (61BP/1.0PP) the “short leaf form” of E. heenanii plus E. paucidentatus (62BP/1.0PP). This grouping supports the hypothesis of Vorster (2004a) except for the unanalysed E. relictus. This finding was also recognised by Treutlein et al. (2005) in their molecular analyses (65BP) and in their morphological analyses (96BP).

Taxa represented in sub-lineage A occur on mountains in and around Swaziland from Piggs peak to Barberton in the Mpumalanga province from 750–1750m above sea level (IUCN 2010). They share conspicuously raised veins on the abaxial leaf surface (Figure 3.36), found elsewhere to a lesser degree only in the unrelated E. latifrons (clade B: lineage 2). Additionally they share a relatively large stature (Figure 3.31, 3.37), green leaves with nodules on the adaxial leaf surface and deflexed pinnas (Figure 3.37–3.38). Natural hybrids between the two taxa are known (Table 3.2; Grobbelaar 2002). The fact that E. heenanii “long leaf form” and “short leaf form” did not group together is not surprising since it is reflected in their reproductive structures. The description of E. heenanii (Dyer 1972) notes superficially identical male and female reproductive structures while the long leaf form has markedly dissimilar reproductive structures between the sexes (Vorster pers. comm.1). This may warrant formal taxonomic recognition.

Lineage 5: sub-lineage B E. aemulans E. lebomboensis E. senticosus

Lineage 5: sub-lineage B is poorly resolved with only E. lebomboensis resolved and relationships between taxa unclear. This grouping does not support the hypothesis of Vorster (2004a) where E. msinganus is included. Osborne et al. (1993) grouped both species (E. lebomboensis, E. aemulans) recognised at the time along with E. ituriensis (clade C: lineage 10) and Treutlein et al. (2005) in their morphological analysis included E. msinganus with this sub-lineage (64BP).

The group occurs in northern KZN from the Vryheid districts in a narrow band northwards through Swaziland to the Pongola river in Mpumalanga (IUCN 2010). Species share many traits with clade B: lineage 3 and 4, being large arborescent species (stems ~3–4m) with large bright green leaves and less nodulation than lineage 5: sub-lineage A (Grobbelaar 2002). Entomological evidence supports Porthetes species restricted to the group (Oberprieler 1995a; Downie et al. 2008), though Porthetes “vorsteri” was found on E. senticosus and E. lebomboensis from Mananga but not from Pongola populations (Vorster & Oberprieler 1999). Artificial hybrids occur between this group and members of clade B and C, with a notable inter-clade natural hybrid between E. lebomboensis and

1 Dr. Piet Vorster, Encephalartos taxonomist. University of Stellenbosch, Department of Botany and Zoology.

E. villosus (Table 3.2).

3.4.3 Clade C Clade C's 39 species are poorly supported in both analyses (56BP/0.52PP) as monophyletic with E. hirsutus resolved as early divergent without support. Within clade C, three major sub-clades are retrieved: sub-clade I comprises species along the African east coast between northern Malawi and Mombasa (Kenya) (68BP/0.62PP), sub-clade II includes inland species on the equator and southern Democratic Republic of the Congo (DRC) border (61BP/0.95PP) and clade III comprises all southern African members south of the Manica province of Mozambique (0.94PP). The clade has been partially recognised in Vorster (2004a) who recognised tropical and some Mozambican members as related. In Osborne et al. (1993) 29 of the 31 species analysed group together. Treutlein et al. (2005) found all 29 species analysed as monophyletic in their molecular (63BP) analysis. Chaiprasongsuk et al. (2007) found 11 of the 16 species analysed to group.

Representative of clade C occur as isolated species and species groups (Figure 3.39), from south- eastern South Africa in East London up along the coast to Kipini in northern Kenya with three major incursion inland: (1) into the Mpumalanga and Limpopo provinces of South Africa, southern Zimbabwe and from the Zimbabwe-Mozambique border into Mozambique; (2) from south-eastern Zambia along the southern DRC border ending along the Angola-DRC border ~ 400km from the Atlantic Ocean, and (3) along the Kenya-Tanzania border through Kenya and Uganda, the South Sudan-DRC border, probably through the Central African Republic and its neighbours into Nigeria ending in Ghana in the west (Norstog & Nicholls 1997; Jones 2002; IUCN 2010). This clade includes the most northerly and westerly distributed species of the genus. It is by far the most diverse clade in terms of species and lineages and as such few characters are shared by all its members. It includes arborescent, non-emergent, and medium sized species with glaucous/glabrous/entire and/or dentate leaflets. Ecologically the clade is the least well known, though numerous insect taxa are restricted to the clade's members including: Lygaeinae (Hemiptera), Anthribidae, Curculionidae (Coleoptera), and Geometridae (Lepidoptera) (Oberprieler 1995a, b; Staude 2001; Downie et al. 2008; Bayliss et al. 2010).

Clade C: sub-clade I: lineage 6 This lineage is moderately supported (68BP/0.62PP) and contains two sub-lineages A (59BP/1.0PP) and B (98BP/0.63PP). All species, except E. sclavoi, are resolved though relationships between species are unresolved in lineage 6: sub-lineage B. Neither the larger grouping or the sub-lineages are in support of the hypothesis (Vorster 2004a). Partial support for this lineage can be seen in Vorster (2004a) indicating a possible relationship of E. gratus with some members of lineage 6: sub-lineage B. In Osborne et al. (1993) three of the five species analysed group together, while in Treutlein et al. (2005) the species form a monophyletic lineage in their

55 molecular analysis (83BP) and their morphological analysis resolved three of the five species as grouped (66BP) though including members of Clade C: sub-clade II: lineage 7.

Notably two isolated geographic groupings (Figure 3.40) are reflected in the phylogenetic branching (Figure 3.3) in lineage 6. Lineage 6: sub-lineage A occurs in northern Mozambique from Morrumbala and Tuchila and the Ruo in Malawi up to the Niassa province in the north (Figure 3.41). Lineage 6: sub-lineage B is found !1000 km further north from around Dar es Salaam (Tanzania) in the south up north to just past Mombasa (Kenya; Figure 3.42).

Lineage 6: Sub-lineage A E. gratus E. turneri

Lineage 6: sub-lineage A includes two supported sister species (59BP/1.00PP) in E. gratus (85BP/1.0PP) and E. turneri (69BP/1.0PP). No support for this lineage was found in literature.

Recent work by Capela et al. (2003), has resulted in a clear need for a re-evaluation of species delimitation and the possibility of natural hybrids of E. turneri. However it seems the two species share a medium stature (1–3 m); green, glabrous leaves (Figure 3.43–3.44); yellow/orange/red cones in E. gratus (Figure 3.45) with a pinkish bloom in some E. turneri colonies (Figure 3.46); and cylindrical females cones with red sarcotesta (Goode 1989, 2001; Whitelock 2002; Jones 2002; Vorster 2004a). Callioratis grandis (Lepidoptera: Geometridae) is restricted to E. gratus (Bayliss et al. 2010).

Lineage 6: Sub-lineage B E. hildebrandtii E. kisambo E. sclavoi

This lineage is strongly supported (98BP) as monophyletic in the MP and only weakly in the bayesian (0.63PP). Relationships between the species are unresolved and E. sclavoi is not resolved. The lineage was only supported in the molecular study of Treutlein et al. (2005) where all species grouped together with 95BP.

This group shares many characters with lineage 7: sub-lineage A, such as a large stature (1–6m; Figures 3.47), tough leaflets with no clear petiole and male cones produced in succession (Vorster 2004a; Figures 3.48–3.50). The speculation that a concentration of teeth occur on the upper margin of leaflet bases (Vorster 2004a) seems to be an inconsistent character between species (Figures

3.51–3.54). They are characterised by their glabrous yellow cones (Figures 3.48–3.50, 3.55–3.58) with a yellow sarcotesta (Figures 3.56–3.57; Goode 1898, 2001; Jones 2002; Whitelock 2002). Ecologically the beetle Peltostethus gedyei (Coleoptera: Curculionidae; Oberprieler 1995a), butterfly Paraptychodes costimaculata (Lepidoptera: Geomitridae), P. tenuis, and P. kedar (Staude 2001) are restricted to E. hildebrandtii.

Clade C: sub-clade II: lineage 7

This lineage forms a supported (61BP/0.95PP) group of ten species and includes two supported sister clades (A 90BP/0.98PP and B 58BP/0.97PP respectively). Encephalartos mackenzei and E. tugelaneus subsp. powysii are suspected to be associated with these species based on morphological and geographic grounds (Vorster 2004a). The lineage conforms partly to the hypothesis of Vorster (2004a). Osborne et al. (1993) indicated some of the 10 species analysed as related but as a whole the group was dispersed throughout their phylogeny. Chaiprasongsuk et al. (2007) indicated that some of the seven species analysed were related but associated with phylogenetically unrelated taxa. In Treutlein et al.'s (2005) molecular analyses all members of lineage 7: sub-lineage A formed a moderately supported (71BP) group with E. laurentianus as sister (as no other members of lineage 7: sub-lineage B were analysed), while their morphological analysis resolved all members of the lineage (54BP) with species from clade C: sub-clade II: lineage 6 are included.

This group occurs (Figure 3.59) in a horisontal U shaped area skirting the Angola-DRC border in the south-west at Kwango (Angola), moving along the DRC-Zambian/Tanzanian border into its easternmost distribution in central Kenya, through northern Uganda along the DRC-South Sudanese-Central African Republic border, with disjunct distributions on the Jos Plateau and finally south-eastern Nigeria to south-western Ghana (IUCN 2010). The two sub-lineages are clearly distinguishable as lineage 7: sub-lineage A are species with well exposed stems (Figures 3.60– 3.61), dentate leaflets (Figure 3.61) and glabrous (Figures 3.62–3.64) or slightly hairy cones (Figures 3.66–3.67), while lineage 7: sub-lineage B are non-emergent species (Figure 3.68–4.71), with mostly entire leaflets (Figure 3.72) and slightly glaucous cones (Figure 3.73–3.76; Vorster 2004a). They do share a red sarcotesta (Figure 3.77–3.79; Goode 1989) and more or less green cones (Figure 3. 61–3.67, 3.73–3.79), male cones produced in succession (Figure 3.61–3.63, 3.70), and no clear petiole (Figure 3.60–3.62, 3.68–3.70).

Lineage 7: sub-lineage A E. delucanus E. laurentianus E. marunguensis E. poggei 57

E. schaijesii E. schmitzii

Within the strongly supported (90BP/0.98PP) lineage 7: sub-lineage A, three clades are retrieved. Encephalartos laurentianus forms a monophyletic assemblage (64BP/0.99PP). The association of E. schajesii with E. schmitzii is moderately supported (84BP) in the MP analysis and strongly supported in the bayesian analysis (1.0PP). The last clade (53BP/0.99PP) includes E. marunguensis (0.71.0PP) with E. schmitzii (0.86PP) and E. poggei successively sister to it.

Representatives of this group occur along the southern DRC Border from around Tshikapa, Kananga (DRC) and Lunda Sul (northern Angola) in the west to Kundelungu (DRC), Solwezi (Zambia) in the south-east and Marungu (DRC) and Mpanda (Tanzania) in the north-east (Figure 3.82; IUCN 2010).

They are characterised by dwarf or subterranean stems, are considered to be deciduous (Figure 3.71), leaves are entire or with minute spines, and cones glaucous blue-green (Vorster 2004a) to yellow (Figure 3.70; Turner et al. 2006). The poor circumscription of these taxa (Vorster 2004a) might be reflected in the lack of specific resolution as these associations are not geographically supported (UICN 2010). Encephalartos laurentianus is a notable inclusion here as it shares few morphological traits with other species except for a close geographic distribution. It remains an enigma due to it favouring river ravine habitats, its enormous stature (stems up to 15m), soft heavily dentate (Figure 3.83) leaves 4–7m long (Vorster 2004a) and brown indumentum on cones (Figure 3.84–3.85). This group is extremely poorly known ecologically but there are unconfirmed reports of African Elephants eating whole cones of E. poggei (Jones 2002).

Lineage 7: sub-lineage B E. barteri subsp. barteri E. barteri subsp. allochrous E. bubalinus E. equatorialis E. ituriensis E. macrostrobilus E. septentrionalis E. tegulaneus subsp. tegulaneus E. whitelockii

Encephalartos tegulaneus subsp. tegulaneus (99BP; 1.0PP) is sister to the rest of the members in lineage 7: sub-lineage B albiet without support. Within this sub-lineage E. barteri subspecies are

58 unresolved while a clade comprising E. septentrionalis, E. macrostrobilus, E. ituriensis, and E. whitelockii (71BP/0.97PP) and another comprising E. equatorialis and E. bubalinus (62BP/1.0PP) receive weak to strong support.

These species occur over a large area in two belts on and well below the equator (Figure 3.80). They range from east of the Tanzania-Kenya border and the Matthews Range (Kenya) through northern Uganda along the DRC-South Sudanese border, as well as probably through the Central African Republic and its neighbouring countries. Lastly E. barteri subsp. allochrous, which occurs on the Jos plateau (Nigeria) and E. barteri subsp. barteri distributed from Ilorin and Jebba (Nigeria) through to the Volta river in Ghana are disjunct from other species (Norstog & Nicholls 1997; Jones 2002; IUCN 2010).

Species of this group share leaflets reduced proximally to prickles leaving no clear petiole, numerous (up to 15 recorded; Goode 2001) male cones maturing in succession (Figures 3.61– 3.63), green glabrous cones (Figure 3.61–3.63) though somewhat brown with hairs in some members of the group (Figures 3.66–3.67), and a red sarcotesta (Figure 3.77–3.78). Antliarhinus vercourtii (Coleoptera: Curcliunidae; Oberprieler 1995a) and Callioratis apicisecta (Lepidoptera: Geometridae) is restricted to E. tegulaneus subsp. tegulaneus (Staude 2001).

Clade C: sub-clade III: lineage 8 E. aplanatus E. cerinus

The grouping received poor support in the MP analysis (55BP) but strong support in the bayesian analysis (0.99PP) possibly due to missing data for E. aplanatus. This lineage shows a marked departure from the hypothesis (Vorster 2004a) and has seen partial recognition in Vorster (2004a) where it is include with members of clade C: sub-clade III's polytomy and Treutlein et al.'s (2005) morphological analysis as a lineage containing species from clade C: sub-clade III’s polytomy (88BP).

Encephalartos cerinus occurs in only a few colonies in KZN near the coast while E. aplantus is known only from eastern Swaziland and possibly adjascent Mozambique (Figure 3.86). The exclusion of E. villosus from which E. aplantus was split (Vorster 1996a) in the lineage is noteworthy as it has been considered conspecific by various authors (Grobbelaar 1996; Oberprieler 1996; Vorster 1996f) due to the morphological (Figures 3.88–3.93) and habitat similarity (Figure 3.87). One reason for the division is due to a different species specific Porthetes species, P. obscurus on E. aplanatus and P. pearsonii on E. villosus (Vorster & Oberprieler 1999). However Porthetes sp n. 12 was found on both E. aplanatus and E. umbeluziensis, which are the closest 59 species geographically (Downie et al. 2008).

Clade C: sub-clade III polytomy

E. caffer E. ngoyanus E. umbeluziensis E. villosus

The following species are discussed here both as a group and directly after clade C: sub-clade III: lineage 8 as they share various traits with each other and form a group as per hypothesis 1 (Vorster 2004a; Figure 1.5). Only E. umbeluziensis is not resolved as monophyletic with E. ngoyanus (61BP/0.99PP), E. caffer (0.99) and E. villosus (96BP/1.0PP) all resolved but unassociated on the backbone of the sub-clade. These species have seen some recognition in previous analyses: Osborne et al. (1993) groups all species (except the yet undescribed E. aplanatus). Chaiprasongsuk et al. (2007) included all three species analysed with E. aplanatus as sister thereto. In their morphological analysis Treutlein et al. (2005) found all species to group together with 99BP including the previous lineage.

All of these species share a geographic region throughout which clade C: sub-clade III: lineage 8 species occur (Figure 3.94). Species occur along the eastern coast of South Africa from Humansdorp, Eastern Cape in the south (E. caffer) up northwards through KZN (E. villosus, E. ngoyanus) with the northernmost populations near Mlawula (Swaziland) and adjacent Mozambique (E. umbeluziensis). Morphologically they share amongst themselves and with clade C: sub-clade III: lineage 8: non- to slightly emergent stems (Figure 3.95–96); few leaves initially held erect (Figure 4.96) (Vorster 2004a), cones with drooping and/or fringed sporophylls and poor expression of sporophyll facets, and external similarity between male and female cones (Figures 3.97–3.99); Vorster 2005). Encephalartos caffer has two insect species specifically associated with it, the beetle Porthetes bulboscapus (Vorster 1999) and Porthetes sp. n. 11 (Downie et al. 2008). As with the previous groupings Antliarhinus beetles species are absent from E. caffer. The fact that Porthetes sp. n. 12 (Downie et al. 2008) has been found on both E. umbeluziensis and E. aplanatus and the molecular conspecificity of Porthetes sp. n. 12, 11 (E. caffer) and 6 (E. villlosus) (77BP; Downie et al. 2008) supports a phylogenetic link with the previous grouping. Additionally a novel Veniliodes species was found on E. ngoyanus and successfully bred on E. villosus (Staude 2001). Notably Antliarhinus saginatus is absent from E. villosus, though this is thought to be due to the insect being unable to penetrate the thick seed integument (Donaldson 1993a). Additionally several other insect species have been identified solely from E. villosus including: the beetles Xenoscelinae sp. 1, Xenoscelinae sp. 2 (Donaldson 1997), and butterflies Veniliodes pantheraria,

V. inflammata (Lepidoptera: Geomitridae; Staude 2001). Successful hybridisation with members of Clade B and C (Table 3.2) has been achieved.

Encephalartos ferox Encephalartos ferox, though a clear morpholoigcal anomaly in the genus, also shares some traits with the previous species, though such a relationship was not resolved here. In the current analysis it forms a highly (93BP/1.0PP) supported single species group on the backbone of sub-clade III. Encephalartos ferox has seen great movement in placement in phylogenetic analyses but is found in support of the hypothesis (Vorster 2004a). In Osborne et al. (1993) E. ferox was early divergent to all but one other Encephalartos species, whereas Van der Bank et al. (1998, 2001) found it early divergent to clade B: lineage 2 and E. altensteinii. In Treutlein et al.’s (2005) molecular analysis the position of two E. ferox replicates was unresolved (but monophyletic) while in their morphological analysis E. gratus grouped with E. ferox with weak support (59BP).

The species occurs from Kosi Bay in northern KZN up the coast to the Save River 100km north of Vilanculos (Mozambique; IUCN 2010), growing on coastal forest edges (Figure 3.100) often within a hundred meters of the ocean (Goode 1989, 2001; Grobbelaar 2002). It is vegetatively variable, characterised by wide, dark green and heavily dentated leaflets (Figures 3.101–3.104) with a shortly emergent to subterranean stem that is soft in texture (Figure 3.101; Vorster 2004a). It bears glabrous cones in various shades of red (Figure 3.102–3.103), though some are yellow (Figure 3.104). Male cones are produced and mature in succession (Figure 3.102), a trait specific to members of clade C: sub-clade I and II (i.e. lineage 6, 7; Jones 2002), and E. inopinus (Grobbelaar 2002). Its pollen was found to be anomalous when compared to members from clade C: lineage 6 and 8 (Dehgan & Dehgan 1988). Ecologically Porthetes sp. n. 7 (Oberprieler 1995a; Downie et al. 2008), Paraptychodes sp. nov. (Lepidoptera: Geomitridae) and Diptychis meraca (Lepidoptera: Geomitridae; Staude 2001) are restricted to E. ferox. Encephalartos ferox is difficult to hybridise artificially though a few reputed examples are found, amongst others with E. caffer (Table 3.2) possibly indicating some phylogenetic linkage.

Clade C: sub-clade III: lineage 9 E. chimanimaniensis E. concinnus E. manikensis E. munchii E. pterogonus

This lineage forms a well supported (96BP/0.99PP) monophyletic lineage but relationships between

61 the five taxa are unresolved and only E. concinnus is resolved (94BP/1.0PP). This is in accordance with field observations (Capela et al. 2003), which indicate that E. manikensis might be conspecific with all other members of the group despite the clear macromorphological difference in e.g. E. munchii. This clade is not in support of the hypothesis (Vorster 2004a) but has been recognised by: Vorster (2004a) with the exception of the tentative inclusion of E. turneri; Osborne et al. (1993) with the exception of the inclusion of E. barteri subsp. barteri; Chaiprasongsuk et al. (2007) who resolved four species in two sister groups (only E. manikensis unanalysed) although two unrelated species are included; Treutlein et al. (2005) who resolved all species to group in their molecular analyses (91BP) and morphological analysis (90BP), though sister to E. turneri with 98BP.

Species in the group occur in a narrow region (Figure 3.105) from south-central Zimbabwe (E. concinnus) around Gwanda and Runde but is mostly centered around the Zimbabwe-Mozambique border (Capela et al. 2003; IUCN 2010). Species are characterised by soft green leaves, glaucous- blue in E. munchii (Figure 3.106); persistent heavy white indumentum on emerging leaves (Figure 3.107); glabrous green female cones (Figures 3.108–3.109); and a red sarcotesta (Vorster 2004a, 2005). Ecologically insects restricted to the group include Antliarhinus sp.n. 2 (E. concinnus; Oberprieler 1995a).

Clade C: sub-clade III: lineage 10 E. cupidus E. dolomiticus E. dyerianus E. eugene-maraisii E. middelburgensis E. nubimontanus

This lineage receives support (74BP/1.0PP) as monophyletic. Relationships within the group are however poorly supported with only one clade receiving support, namely the grouping of E. dyerianus, E. dolomiticus, and E. cupidus (89BP/1.0PP). This lineage has been supported in previous analyses but is not in support of the hypothesis as Vorster (2004a) tentatively included E. hirsutus in the group. Osborne et al. (1993) grouped all members (except the unanalysed E. nubimontanus), and Treutlein et al. (2005) included all taxa except for the unanalysed E. eugene- maraisii (98BP) in both molecular and morphological analyses.

Members of the group occur along the northern escarpment (Figure 3.110) south from the Gauteng-Mpumalanga border near Bronkhorstspruit up north to Mica in the Limpopo province with E. eugene-mariasii growing disjunctly in the Waterberg to the west. Morphologically species are very similar with three of the species (E. dyerianus, E. dolomiticus and E. middelburgensis) once

62 considered ecotypes of E. eugene-maraisii. They all share glaucous blue leaflets that are isobilateral (Figures 3.111–3.117; Osborne et al. 1993), with cones that are very similar in morphology (Figures 3.112–3.118) and a yellow sarcotesta (Figure 3.119). Very few insect taxa have been collected from the group with only Apinotropis verdoornae (Coleoptera: Anthribidae) restricted to E. eugene-mariasii and Antliarhinus absent from the geographic region (Donaldson 1993a). Hybridisation within the group is probably underestimated due to the previous lumping (Table 3.2).

Encephalartos hirsutus and E. inopinus These species, though enigmatic in the genus, are discussed here together and directly after clade C: sub-clade III: lineage 10 as they may share some relation, though this was not resolved in the current study. Encephalartos hirsutus is unsupported as sister to clade C and E. inopinus’ position is unresolved on the backbone of sub-clade III. Traits shared with clade C: sub-clade III: lineage 10 are their geographic location (Figure 3.120), glaucous blue, unarmed pinnas (Figures 3.121–3.122) (except for E. cupidus and some forms of E. nubimontanus); cone sporophyll morphology (Figure 3.123–125), and orange-yellow sarcotesta (Grobbelaar 2002). Ecologically Antliarhinus spp. have not to date been found in the distribution range of these species (Donaldson 1993a). The placement of E. hirsutus is contrary to the hypothesis (Vorster 2004a) while that of E. inopinus is positive in terms of the hypothesis.

Encephalartos hirsutus has not been included into any phylogenetic analyses to date, except for the hypothesis presented by Vorster (2004a) placing it tentatively with clade C: sub-clade III: lineage 10. It is easily distinguished from all other Encephalartos species by the persistent rough indumentum on its leaves, in copious amounts when young (Figure 3.126; Hurter & Glen 1996). Its cones are somewhat similar to E. inopinus in being glabrous and glaucous blue-green (Figure 3.123). To the knowledge of the author no hybridisation has ever been attempted due to the rarity of material.

Encephalartos inopinus has been associated with various members of the genus in previous analyses. Vorster (2004a) places it in a group of its own; Osborne et al. (1993) resolved it sister to three subterranean species (E. ngoyanus, E. cerinus and E. poggei). Treutlein et al. (2005) finds it separate but unassociated in both their analyses. It is clearly distinguishable due to its soapy green (Vorster 2005) falcate, drooping leaflets (Figure 3.121; Vorster 2004a). Its cones are glaucous (Figure 3.124–3.125) with males produced in succession (Figure 3.125) a trait only shared with E. ferox and members of clade C: subclade I and II (i.e. lineage 6 and 7). No hybrids have been found, most probably due to some genetic barrier (Vorster 2005).

3.5 Conclusions

Three main supported clades were retrieved in Encephalartos. These may serve as a stable sectional classification for the genus much as with the sister genus Macrozamia (Johnson 1959). However before any formal decision is made, the characterisation of the two larger Clades (B and C) needs additional investigation to determine shared and derived characters between the relatively large number of often poorly known taxa. The phylogeny sees several departures from the hypotheses of Vorster (2004a) but often these were either tentative inclusions or mentioned to be of close relation to the taxa found to be phylogenetically related in this study. However, major departures among tropical species, including some of the more poorly known species are significant. A lack of resolution on the backbone of clades B and C as well as most species groups is troublesome and keeps evolutionary understanding of characters out of reach. Additionally a clear need for field observation, at the population level and especially investigating ecological traits, is needed to produce the necessary understanding of the genus. Also, the clear linkage of geographical distribution, even overriding morphological associations, needs further investigation and interpretation within the African continent's meteorological, geomorphological and phytogeographical history.

Table 3.1 PAUP statistics of analyses obtained from separate and combined datasets.

Statistic rbcLa matK psbA-trnH Combined nrITS Combined plastids + plastids nrITS No. of Encephalartos taxa excluded 3 3 6 3 3 3 No. of included characters 546 804 1296 2646 1089 3735 No. of constant characters 463 609 1055 2124 728 2852 No. of variable sites 83 (15.2%) 195 (24.3%) 241 (18.6%) 522 (19.7%) 361 (21.21%) 883 (23.64%) No. of parsimoniously informative 23 (4.2%) 77 (9.6%) 25 (1.9%) 125 (4.7%) 315 (28.9%) 440 (11.8%) characters No. of trees (Fitch) 76 127 178 628 488 1151 No. of steps (Tree Length) 112 240 254 628 488 1151 Consistency Index (CI) 0.84 0.87 0.99 0.89 0.89 0.84 Retention Index (RI) 0.60 0.85 0.96 0.79 0.96 0.92 Average number of changes per 1.3 1.2 1.05 1.2 1.3 1.3 variable sites (number of steps/number of variable sites)

Table 3.2 Hybrids found between members of the genus Encephalartos Species Hybrid partners (Source) Confirmed Natural Confirmed Artificial Unconfirmed Artificial E. aemulans E. woodii (cycadforum.co.za) E. altensteinii E. arenarius (Vorster E. lehmannii (Vorster E. woodii (cycadforum.co.za) 1986) 1986) E. trispinosus E. lebomboensis (lotusland.org) (Vorster 1986) E. latifrons (Vorster E. natalensis (lotusland.org) 1986) E. friderici-guilimi (Myburgh 1991) E. arenarius E. trispinosus E. lebomboensis E. woodii (cycadforum.co.za) (Vorster 2003) (Vorster 1986) E. altensteinii E. paucidentatus E. princeps (cycadforum.co.za) (Vorster 1986) (Holzman 2005) E. longifolius (Vorster E. horridus (cycadforum.co.za) 1986) E. transvenosus E. laurentianus (ebay.co.uk) (Vorster pers. comm. 1) E. natalensis E. lehmannii (lotusland.org) E. latifrons (Vorster pers. comm. 1) E. bubalinus E. hildebrandtii (plantswap.net) [E. natalensis x E. woodii] (plantswap.net) E. caffer E. ferox (Holzman 2005) E. horridus (Vorster pers. comm.1) E. cerinus [E. trispinosus x E. E. eugene-maraisii lehmannii] (Vorster (plantaplam.com) pers comm. 1) E. chimanimaniensis E. munchii (cycadforum.co.za) E. concinnus E. eugene-maraisii (Whitelock 2002) E. delucanus E. schaijesii (http://encephalartos.de) E. nubimontanus E. dolomiticus (cycadforum.co.za) E. eugene-maraisii (cycadforum.co.za) E. dyerianus E. middelburgensis (cycadforum.co.za) E. eugene-maraisii E. concinnus E. cerinus (plantapalm.com) (Whitelock 2002) E. dolomiticus (cycadforum.co.za) E. ferox E. trispinosus (Vorster 1986) E. caffer (Holzman 2005) E. friderici-guilimi E. altensteinii (Myburgh 1991) E. gratus E. manikensis E. woodii (lotusland.org) (Holzman 2005) 66

E. hildebrandtii E. turneri E. bubalinus (plantswap.net (cycadforum.co.za) 2011) E. heenanii E. paucidentatus E. longifolius (cycadforum.co.za) (Grobbelaar 2002) E. horridus E. longifolius (Vorster E. trispinosus E. woodii (cycadforum.co.za) 1986) (Vorster 1986) E. latifrons (Vorster E. senticosus (cycadforum.co.za) pers. comm.1) E. caffer (Vorster E. natalensis (cycadforum.co.za) pers. comm.1) E. transvenosus E. trispinosus (cycadforum.co.za) (Vorster pers. comm.) E. latifrons E. horridus (Vorster E. arenarius (Vorster E. woodii (cycadforum.co.za) pers. comm.1) pers. comm.1) E. altensteinii E. transvenosus (Vorster 1986) (cycadforum.co.za) E. lehmannii (cycadforum.co.za) E. longifolius (cycadforum.co.za) E. laurentianus E. arenarius (ebay.co.uk) E. lebomboensis E. villosus (Vorster E. ngoyanus (Vorster E. lehmannii (cycadforum.co.za) 1986) 1986) E. trispinosus E. woodii (cycadforum.co.za) (Vorster 1986) E. arenarius (Vorster E. longifolius (Stainer 2000) 1986) E. altensteinii (tropicalmigration.com) E. lehmannii E. umbuleziensis E. lebomboensis (Vorster 1986) (cycadforum.co.za) E. arenarius (lotusland.org) E. transvenosus (cycads-n- palms.com) E. trispinosus (cycadforum.co.za) E. latifrons (cycadforum.co.za) E. longifolius E. horridus (Vorster E. transvenosus E. trispinosus (cycadforum.co.za) 1986) (Vorster 1986) E. arenarius (Vorster E. latifrons (cycadforum.co.za) 1986) E. lehmannii (Stainer E. heenanii (cycadforum.co.za) 2000) E. horridus (Vorster E. princeps (cycadforum.co.za) 1986) E. woodii (Vorster pers. comm.1) E. manikensis E. gratus (Holzman E. pterogonus (junglemusic.net) 2005) E. middelburgensis E. dyerianus (cycadforum.co.za) E. munchii E. chimanimaniensis (cycadforum.co.za) E. trispinosus (cycadforum.co.za) E. ngoyanus E. lebomboensis (Vorster 1986) E. nubimontanus E. dolomiticus (cycadforum.co.za)

E. natalensis E. woodii (Cafasso E. horridus (cycadforum.co.za) et al. 2001, Vorster 1986) E. arenarius (Vorster E. bubalinus (plantswap.net) pers. comm.1) E. transvenosus (cycadforum.co.za) E. altensteinii (lotusland.org) E. paucidentatus E. arenarius (Holzman 2005) E. heenanii (Grobbelaar 2002) E. princeps E. arenarius (cycadforum.co.za) E. transvenosus (cycadforum.co.za) E. longifolius (cycadforum.co.za) E. pterogonus E. manikensis (junglemusic.net) E. schaijesii E. delucanus (encephalartos.de) E. senticosus E. woodii (cycadforum.co.za) E. horridus (cycadforum.co.za) E. trispinosus (cycadforum.co.za) E. transvenosus E. woodii (Vorster E. latifrons (cycadforum.co.za) pers. comm.1) E. arenarius (Vorster E. lehmannii (cycads-n- pers. comm.1) palms.com) E. horridus (Vorster E. natalensis (cycadforum.co.za) pers. comm.1) E. princeps (cycadforum.co.za) E. trispinosus E. altensteinii E. latifrons (cycadforum.co.za) (Vorster 1986) E. horridus (Vorster E. woodii (cycadforum.co.za) 1986) E. ferox (Vorster E. senticosus (cycadforum.co.za) 1986) E. arenarius (Vorster E. lehmannii (cycadforum.co.za) 2003) E. villosus (Vorster E. munchii (cycadforum.co.za) 1986) E. turneri E. gratus (cycadforum.co.za E. umbeluziensis E. villosus (Vorster 1986) E. lehmannii (Vorster 1986) E. woodii (Vorster pers. comm.1) E. villosus E. lebomboensis E. trispinosus (Vorster 1986) (Vorster 1986) E. woodii E. natlensis (Cafasso E. altensteinii (cycadforum.co.za) et al. 2001, Vorster 1986) E. transvenosus E. arenarius (cycadforum.co.za) (Vorster pers. comm.1)

E. umbeluziensis E. horridus (cycadforum.co.za) (Vorster pers. comm.1) E. longifolius (Vorster E. lebomboensis pers. comm.1) (cycadforum.co.za) E. trispinosus E. senticosus (cycadforum.co.za) (Vorster 1986) E. trispinosus (cycadforum.co.za) E. aemulans (cycadforum.co.za) E. gratus (lotusland.org) E. bubalinus (plantswap.net) E. longifolius (willowbrooknursery.com)

E dyerianus PR731 E dyerianus CC332 71 E dolomiticus PR865 E dolomiticus CC343 52 E cupidus PR767 E cupidus PR691 E aplanatus d1126 E laevifolius PR845 E laevifolius PR803 50 E laevifolius PR730 E humilis PR791 E humilis PR695 59 E caffer PR729 E barteri subsp bateri d1095 97 E princeps PR871 E princeps PR810 E gratus PR891 54 E ghellinckii PR773 E friderici guilielmi PR853 E friderici guilielmi CC338 E paucidentatus PR710 62 E paucidentatus PR681 E heenanii PR775 short leaf 61 E hirsutus PR718 63 E hirsutus CC336 E eugene maraisii PR872 61 E sclavoi PR747 E lehmannii PR780 62 E ngoyanus PR717 E ngoyanus d1135 65 E longifolius PR873 E longifolius d1032 61 E lehmannii PR661 E ferox d1025 63 E kisambo PR745 E kisambo d1132 95 E inopinus PR778 E inopinus d1130 64 E cycadifolius PR683 E cycadifolius d1128 62 E concinnus PR890 E concinnus d1023 E woodii PR763 kranskloof E woodii PR884 kranskloof E woodii PR675 ngoye E woodii d1139 ngoye E whitelockii PR818 E whitelockii d1048 E villosus PR838 E villosus PR671 E umbeluziensis PR815 E umbeluziensis d1046 E turneri PR886 E turneri d1044 E trispinosus PR680 E trispinosus d1043 E transvenosus PR832 E transvenosus PR727 E tegulaneus subsp tegulaneus PR825 E tegulaneus subsp tegulaneus d1040 E septentrionalis d1138 E senticosus PR834 E senticosus d1039 E sclavoi d1038

Figure 3.1 (b)

70 Figure 3.1 (a)

E schmitzii PR819 78 E schmitzii d1137 E schaijesii CC329 E pterogonus PR876 E pterogonus d1136 E poggei PR813 E nubimontanus PR704 E nubimontanus PR655 E natalensis PR802 kranskloof E natalensis d1035 E munchii PR855 E munchii d1134 E msinganus PR751 E msinganus PR701 E middelburgensis PR726 E middelburgensis CC337 E marunguensis PR912 E marunguensis CC446 E manikensis PR697 E manikensis d1033 E macrostrobilus CC333 E lebomboensis PR698 E lebomboensis PR657 E lebomboensis d1030 E laurentianus PR789 E laurentianus d1029 E latifrons PR811 E latifrons PR678 E lanatus PR828 E lanatus PR677 E laevifolius PR801 E laevifolius PR798 88 E laevifolius CC335 E ituriensis d1131 E horridus PR846 E horridus d1028 E hildebrandtii PR824 E hildebrandtii d1027 E heenanii PR776 long leaf E gratus PR774 E ghellinckii Abbott9220 lowland E ferox PR841 E eugene maraisii d1024 E equatorialis PR900 E equatorialus PR899 E delucanus d1129 E chimanimaniensis PR888 E chimanimaniensis PR883 E cerinus PR859 78 E cerinus PR744 E caffer d1101 E bubalinus PR885 E bubalinus d1021 E brevifoliolatus Xdk2 E brevifoliolatus Xdk1 E barteri subsp allochrous PR892 E arenarius d1020 E arenarius PR758 E aplanatus PR922 E altensteinii PR668 53 E altensteinii PR650 E aemulans PR861 E aemulans d1018 99 Lepidozamia peroffskyana d1050 Lepidozamia hopei d1049 Macrozamia plurinervia d1060 89 99 Macrozamia pauli guilielmi d1096 Macrozamia macdonnellii d1057 Macrozamia communis d1051 100 Zamia furfuracea Microcycas calocoma Bowenia serrulata Dioon spinulosum Ceratozamia mexicana Stangeria eriopus PR860 Cycas thouarsii

Figure 3.1 Bootstrap consensus tree based on the combined plastid regions (matK, rbcLa, and psbA-trnH) with boostrap (> 50BP) values indicated above branches. 71

E whitelockii PR818 E whitelockii d1048 69 E septentrionalis d1138 E macrostrobilus CC333 E ituriensis d1131 E equatorialis PR900 61 62 E equatorialus PR899 E bubalinus PR885 E bubalinus d1021 80 E barteri subsp allochrous PR892 E barteri subsp barteri d1095 100 E tegulaneus subsp tegulaneus PR825 E tegulaneus subsp tegulaneus d1040 87 E schmitzii PR819 62 E schmitzii d1137 E delucanus d1129 E poggei PR813 93 65 E marunguensis PR912 E marunguensis CC446 63 E laurentianus PR789 E laurentianus d1029 E schaijesii CC329 77 E eugene maraisii PR872 E eugene maraisii d1024 E nubimontanus PR704 E nubimontanus PR655 E middelburgensis PR726 E middelburgensis CC337 78 E dyerianus PR731 E dyerianus CC332 E dolomiticus PR865 E dolomiticus CC343 Clade C E cupidus PR767 E cupidus PR691 54 E hildebrandtii PR824 E hildebrandtii d1027 100 E sclavoi PR747 E sclavoi d1038 E kisambo PR745 74 E kisambo d1132 77 E turneri PR886 56 66 E turneri d1044 85 E gratus PR891 E gratus PR774 87 E concinnus PR890 E concinnus d1023 E pterogonus PR876 E pterogonus d1136 97 E munchii PR855 E munchii d1134 E manikensis PR697 E manikensis d1033 E chimanimaniensis PR888 E chimanimaniensis PR883 E cerinus PR859 61 E cerinus PR744 79 E aplanatus PR922 E aplanatus d1126 95 E villosus PR838 E villosus PR671 100 E inopinus PR778 E inopinus d1130 97 E ferox PR841 E ferox d1025 63 E. caffer PR729 E caffer d1101 E umbeluziensis PR815 E umbeluziensis d1046 E ngoyanus PR717 E ngoyanus d1135 100 E hirsutus PR718 Figure 3.2 (b) E hirsutus CC336 72

Figure 3.2 (a) E senticosus PR834 E senticosus d1039 94 E msinganus PR751 61 E lebomboensis PR698 E lebomboensis PR657 E lebomboensis d1030 E aemulans PR861 E aemulans d1018 E paucidentatus PR710 63 E paucidentatus PR681 E heenanii PR776 long leaf E heenanii PR775 short leaf 62 E natalensis PR802 kranskloof 56 E natalensis d1035 65 E altensteinii PR668 E altensteinii PR650 60 E woodii PR884 kranskloof E woodii PR763 kranskloof 64 E woodii PR675 ngoye E woodii d1139 ngoye 92 64 E transvenosus PR832 Clade B E transvenosus PR727 E trispinosus PR680 E trispinosus d1043 E princeps PR871 E princeps PR810 100 E msinganus PR701 E longifolius PR873 E longifolius d1032 E lehmannii PR780 E lehmannii PR661 E latifrons PR811 E latifrons PR678 E horridus PR846 E horridus d1028 E arenarius PR758 E arenarius d1020 E laevifolius PR845 88 E laevifolius PR803 E laevifolius PR730 100 61 E lanatus PR828 E lanatus PR677 97 E laevifolius PR801 E laevifolius PR798 E laevifolius CC335 59 E humilis PR695 Clade A E brevifoliolatus Xdk2 E brevifoliolatus Xdk1 63 99 88 E ghellinckii PR773 E ghellinckii Abbott9220 lowland 91 E friderici guilielmi PR 853 CC338 89 E friderici guilielmi 100 E cycadifolius PR683 E cycadifolius d1128 100 Lepidozamia peroffskyana d1050 Lepidozamia hopei d1049 Macrozamia macdonnellii d1057 Macrozamia pauli guilielmi Macrozamia communis d105d10961 Macrozamia plurinervia d1060 Figure 3.2 Bootstrap consensus tree based on the nrITS dataset, with bootstrap values (>50BP) indicated above branches.

E dyerianus PR731 E dyerianus CC332 6 4 E dolomiticus PR865 89/1.0 E dolomiticus CC343 2 E cupidus PR767 1 E cupidus PR691 1 E nubimontanus PR655 1 E middelburgensis PR726 3 E middelburgensis CC337 10 1 E nubimontanus PR704 74/1.0 2 2 E eugene maraisii PR872 68/0.73 E eugene maraisii d1024 3 E concinnus PR890 1 94/1.0 E concinnus d1023 E pterogonus PR876 E pterogonus d1136 4 E munchii PR855 Sub-clade 9 96/0.99 E munchii d1134 III E manikensis PR697 E manikensis d1033 1 E chimanimaniensis PR888 E chimanimaniensis PR883 E cerinus PR859 1 1 -/0.66 E cerinus PR744 -/0.94 1 51/0.62 8 2 E aplanatus d1126 2 55/0.99 E aplanatus PR922 4 E villosus PR838 96/1.0 E villosus PR671 1 E ngoyanus PR717 1 61/0.99 E ngoyanus d1135 1 8 E inopinus PR778 100/1.0 E inopinus d1130 4 E ferox PR841 46 93/1.0 E ferox d1025 7 1 E caffer PR729 -/0.99 E caffer d1101 1 E umbeluziensis PR815 2 E umbeluziensis d1046 Clade C 2 1 E septentrionalis d1138 65/0.99 E macrostrobilus CC333 1 2 E ituriensis d1131 1 E whitelockii PR818 71/0.97 13 E whitelockii d1048 1 E equatorialis PR900 1 1 1 -/0.67 E equatorialus PR899 1 -/ 1 -/0.68 E bubalinus PR885 0.94 62/1.0 E bubalinus d1021 2 2 B E barteri subsp allochrous PR892 58/0.97 7 E barteri subsp barteri d1095 10 E tegulaneus subsp tegulaneus PR825 1 99/1.0 E tegulaneus subsp tegulaneus d1040 3 7 1 E marunguensis CC446 2 61/0.95 1 -/0.71 E marunguensis PR912 2 -/0.86 3 E schaijesii CC329 Sub-clade 1 II 53/ E poggei PR813 3 0.99 E schmitzii PR819 2 A 90/0.98 E schmitzii d1137 2 84/1.0 1 E delucanus d1129 56/0.52 1 E laurentianus PR789 64/0.99 E laurentianus d1029 2 E hildebrandtii PR824 82/1.0 E hildebrandtii d1027 6 2 E kisambo PR745 B 98/0.63 87/1.0 E kisambo d1132 E sclavoi d1038 Sub-clade 1 I 6 3 E sclavoi PR747 2 68/0.62 2 E turneri PR886 1 2 69/1.0 E turneri d1044 A 1 59/1.0 2 E gratus PR891 1 85/1.0 E gratus PR774 1 10 E hirsutus PR718 100/1.0 4 E hirsutus CC336

Figure 3.3 (b)

Figure 3.3 (a) E lebomboensis PR698 1 1 E lebomboensis PR657 66/ 1.0 E lebomboensis d1030 B 2 E senticosus PR834 15 E senticosus d1039 50/0.96 1 77/ E aemulans PR861 0.53 1 1 5 E aemulans d1018 -/0.94 E paucidentatus PR710 1 E paucidentatus PR681 1 62/1.0 A E heenanii PR775 short leaf 61/ E heenanii PR776 long leaf 1.0 1 1 1 E woodii PR884 kranskloof 1 -/0.96 E woodii PR763 kranskloof 1 -/0.98 2 E woodii PR675 ngoye 85/1.0 E woodii d1139 ngoye 2 7 E msinganus PR751 -/ 2 2 0.77 E natalensis PR802 kranskloof 3 Clade B -/ E natalensis d1035 1 4 0.95 E msinganus PR701 2 E altensteinii PR668 22 73/0.99 E altensteinii PR650 5 2 E transvenosus PR832 2 3 77/1.0 E transvenosus PR727 5 5 E princeps PR871 82/1.0 94/1.0 E princeps PR810 1 E longifolius PR873 59/0.98 163 E longifolius d1032 1 E horridus PR846 90/ 57/1.0 E horridus d1028 0.97 1 1 E trispinosus PR680 2 E trispinosus d1043 -/0.97 2 E lehmannii PR780 1 E lehmannii PR661 E latifrons PR811 E latifrons PR678 E arenarius PR758 1 E arenarius d1020 1 E laevifolius PR803 3 1 E laevifolius PR730 60/ E laevifolius PR845 2 -/ 0.79 3 E humilis PR695 60 77/ 0.50 1 E humilis PR791 0.56 1 87/ E lanatus PR828 0.97 1 4 60/1.0 E lanatus PR677 E laevifolius PR801 81/ Clade A E laevifolius PR798 0.56 1 E laevifolius CC335 B -/0.53 E brevifoliolatus Xdk2 2 E brevifoliolatus Xdk1 11 2 E ghellinckii PR773 2 1 1 84/0.55 83/1.0 E ghellinckii Abbott9220 lowland 81/ 4 E friderici guilielmi PR853 0.53 96/1.0 A 3 E friderici guilielmi CC338 73/1.0 2 E cycadifolius PR683 2 91/1.0 E cycadifolius d1128 130 75 Lepidozamia peroffskyana d1050 40 100/1.0 1 Lepidozamia hopei d1049 6 Macrozamia plurinervia d1060 21 5 100/1.0 4 Macrozamia communis d1051 86/0.97 52 58/0.77 Macrozamia pauli guilielmi d1096 100/1.0 9 Macrozamia macdonnellii d1057 26 24 Zamia furfuracea 19 4 100/1.0 Microcycas calocoma 20 16 Bowenia serrulata 24 Dioon spinulosum Ceratozamia mexican 21 Stangeria eriopus PR860 51 Cycas thouarsii

Figure 3.3 One of the 926 most parsimonious fitch trees (488 steps, CI = 0.8445, RI = 0.9222) from the combined plastid and nrITS dataset. Numbers above branches are fitch lengths (DELTRAN optimisation) and those below branches are MP BP >50%/bayesian PP>0.5. 75

Figure 3.4 Distribution of species included in clade A: lineage 1.

Figure 3.5 Entire, narrowly linear Figure 3.6 Tomentose female cones leaflets of E. friderici-guilielmi. of E. lanatus, note yellow sporophylls.

Figure 3.7 Pubescent cones of E. Figure 3.8 Seeds with yellow sarcotesta friderici-guilielmi. Note clear petiole of E. ghellinckii in habitat around the with yellowing rachis. Drakensberg (Photo: H. Nieuwmeyer).

Figure 3.9 Arborescent E. laevifolius Figure 3.10 Two meter tall E. lanatus in female in cone (Photo: G. van Deventer). habitat near Middelburg (Photo: R. Rousseau).

Figure 3.11 Non-emergent E. humilis female in cone (Photo: K. Haacke).

Figure 3.12 Approximate distribution of species in clade B, colours/numbers denote lineages. 77

Figure 3.13 Diminutive E. horridus (PR847) Figure 3.14 Multiple female cones of with glaucous blue lobed leaves. E. princeps. Note glaucous entire leaflets.

Figure 3.15 Lobed, green leaflets of E. Figure 3.16 Leaflets of E. longifolius, latifrons (PR708). (PR808), note blue waxy covering removed along lower edges to reveal green colour.

Figure 3.17 Approximate distribution of species in clade B: lineage 2 in the Eastern Cape. 78

Figure 3.18 Approximate distribution of Encephalartos transvenosus (clade B).

Figure 3.19 Characteristic leaflet arrangement of E. transvenosus.

Figure 3.20 Female cones of E. transvenosus. Figure 3.21 Male cones of E. transvenosus.

Figure 3.22 Approximate distribution of clade B: lineage 4. Encephalartos cf. woodii refers to material from Kranskloof.

Figure 3.23 Arborescent habit of Figure 3.24 Habit of E. natalensis in habitat at E. altensteinii in cultivation. Oribi Gorge Nature Reserve. 80

Figure 3.25 Female Figure 3.26 Female cone of Figure 3.27 Female cones of cone of E. altensteinii. E. altensteinii. E. natalensis.

Figure 3.28 Kranskloof E. cf. woodii (PR763) Figure 3.29 Leaf of E. woodii with toothed material showing characteristics of E. woodii. margin base (Photo: A. Fanfoni).

Figure 3.30 Approximate distribution of species in clade B: lineage 5. 81

Figure 3.31 Arborescent habit of E. paucidentatus Figure 3.32 Nodules on leaflets of (PR710), note deflexed leaflets. E. heenanii.

Figure 3.33 Tomentose cones of Figure 3.34 Yellow glabrous cone and E. aemulans (Photo: A. Fanfoni). red sarcotesta of E. lebomboensis (PR698).

Figure 3.35 Yellow female Figure 3.36 Conspicuously raised veins cone of E. senticosus. on the abaxial side of E. paucidentatus.

Figure 3.37 Female E. heenanii in cone, Figure 3.38 Heavily deflexed leaflets of note deflexed leaflets (Photo: A. Fanfoni). E. paucidentatus.

Figure 3.39 Approximate distribution of species represented in clade C, colours and numbers denote lineages. 83

Figure 3.40 Approximate distribution of clade C: sub-clade I: lineage 6. Sub-lineage A in blue and sub-lineage B in green.

Figure 3.41 Approximate distribution of lineage 6: sub-lineage A.

Figure 3.42 Approximate distribution of lineage 6: sub-lineage B in Tanzania and Kenya, note distribution on Zanzibar Island. 84

Figure 3.43 Habit of E. gratus (PR891). Figure 3.44 Habit of E. turneri (PR914).

Figure 3.45 Successively maturing, red male Figure 3.46 Male cone of E. turneri cones of E. gratus (PR891). (PR905) exhibiting pinkish bloom.

Figure 3.47 Emergent stem of Figure 3.48 Male cones of Figure 3.49 Male cones of E. E. kisambo (Photo: A. Cameron). E. hildebrandtii (PR917). kisambo (Photo: A. Cameron).

Figure 3.50 Successively Figure 3.51 Teeth at base Figure 3.52 Teeth folded under- maturing E. kisambo male of E. kisambo leaflet neath E. hildebrandtii leaflets cones (Photo: A. Cameron). (Photo: A. Cameron). less crowded than E. sclavoi.

Figure 3.53 Glossy green newly Figure 3.54 Leaf teeth Figure 3.55 Female cones produced E. sclavoi leaves. of E. kisambo (Photo: of E. hildebrandtii. (Photo: A. A. Cameron) Cameron)

Figure 3.56 Female E. sclavoi Figure 3.57 Female E. hildebrandtii Figure 3.58 Male cone of cone with yellow sarcotesta cone with yellow receptive ovules E. hildebrandtii (Photo: A. (Photo: A. Cameron). exposed (Photo: A. Cameron). Cameron).

Figure 3.59 Distribution of species in clade C: sub-clade II: lineage 7. Sub-lineage A in red, sub- lineage B in orange.

Figure 3.59 Arborescent habit of E. tegulaneus Figure 3.60 Emergent stem of E. bubalinus subsp. tegulaneus in habitat, Kenya (PR 894) male. (Photo: A. Cameron).

Figure 3.61 Numerous successively emerging, green, glabrous male cones of E. eqautorialis (PR899). Note heavily dentate leaflets lacking a clear petiole.

Figure 3.62 Green male Figure 3.63 Numerous male Figure 3.64 Female E. whitelockii cones of E. bubalinus. E. equatorialis cones. cones (Photo: A. Lindström).

Figure 3.65 Orange sarcotesta Figure 3.66 Brown E. barteri Figure 3.67 Numerous male cones of E. barteri subsp. barteri subsp. barteri male cone of E. septentrionalis, note brown (Photo: J. Andersson). (Photo: J. Andersson). colour (Photo: A. Lindström).

Figure 3.68 Subterranean stem Figure 3.69 Male cones of Figure 3.70 Male cones of of E. marunguensis (Photo: A. E. delucanus (Photo: A. E. cf. schaijesii/ (Photo:A. Cameron). Cameron). Lindström).

Figure 3.71 Deciduous nature of Figure 3.72 Leaflets of Figure 3.73 Male cone of E. E. schaijesii (Photo: K. Hill). E. marunguensis (PR814). delucanus (Photo: A. Vogel).

Figure 3.74 Female Figure 3.75 E. poggei Figure 3.76 Glaucous blue-green E. marunguensis cone female cone (Photo: E. schmitzii male cone (Photo: A. (Photo: A. Cameron). A. Vogel). Lindström).

Figure 3.177 Seeds of E. Figure 3.78 Mature cones Figure 3.79 Female E. whitelockii poggei with red sarcotesta of E. tegulaneus (Photo: cone with red sarcotesta seeds (Photo: A. Lindström). A Cameron). (Photo: A. Vogel). 89

Figure 3.80 Approximate distribution of species in clade C: sub-clade II: lineage 7: sub-lineage A.

Figure 3.81 Female cone collage of E. bubalinus in habitat in Tanzania (Photos: A. Cameron).

Figure 3.82 Approximate distribution of species in clade C: sub-clade II: lineage 7: sub-lineage B.

Figure 3.83 Heavily Figure 3.84 Brown female Figure 3.85 E. laurentianus male cones, dentate leaflets of cone of E. laurentianus note yellow sporophylls covered by E. laurentianus (PR918). note orange sarcotesta brown indumetum (Photo: L. de Volder). (Photo: B. Schlumpberger).

Figure 3.86 Distribution of clade C: sub-clade III: lineage 8 members in South Africa and Swaziland.

Figure 3.87 Encephalartos villosus in forest habitat at Oribi Gorge Nature Reserve.

Figure 3.88 Male Figure 3.89 Female cone Figure 3.90 Immature male cones cone of E. aplanatus. of E. aplanatus. of E. aplanatus.

Figure 3.91 Habit of Figure 3.92 Female Figure 3.93 Removed cone of E. aplanatus. E. villosus in cone. E. villosus, note red sarcotesta.

Figure 3.94 Clade C: sub-clade III: lineage 8 and morphologically similar species' distribution. 93

Figure 3.95 Slightly emerging stem and Figure 3.96 Underground stem and erect erect leaves of E. caffer. leaves of E. cerinus (PR690).

Figure 3.97 Female Figure 3.98 Female Figure 3.99 Immature male cones of cone of E. caffer. E. cerinus cone. E. cerinus.

Figure 3.100 Approximate distribution of E. ferox, note incursion into South Africa. 94

Figure 3.101 Shortly emergent stem Figure 3.102 Male cones maturing in of a male E. ferox. succession on E. ferox.

Figure 3.103 Red female cones Figure 3.104 Yellow cone form of E. ferox of E. ferox. (Photo: M. Evans).

Figure 3.105 Approximate distribution of species in clade C: sub-clade III: lineage 9. 95

Figure 3.106 Glaucous leaves of E. munchii Figure 3.107 E. manikensis crown detail, note (PR855). note persistant white indumetum (PR795).

Figure 3.108 Female E. manikensis cone. Figure 3.109 Male cones of E. manikensis.

Figure 3.110 Approximate distribution of clade C: sub-clade III: lineage 10 species in South Africa. 96

Figure 3.111 Glaucous Figure 3.112 Female cone of E. Figure 3.113 Female cones of leaves of E. dyerianus. dyerianus, note orange sarcotesta. E. middelburgensis in habitat.

Figure 3.114 Male cones of Figure 3.115 Male cone of Figure 3.116 Female cone of E. middelburgensis (PRU586). E. cupidus (PR691). E. cupidus (PR767).

Figure 3.117 Female cone of Figure 3.118 Female cone Figure 3.119 Yellow sarcotesta E. eugene-maraisii (PR770) of E. nubimontanus (PR704) of E. nubimontanus (PR704) note red indumetum. note sporophyll morphology. seeds. 97

Figure 3.120 Approximate distribution of clade C: sub-clade III: lineage 10 species, E. inopinus (clade C: sub-clade III) and E. hirsutus (clade C: sub-clade III) in South Africa.

Figure 3.121 Foliage of E. inopinus (PR778). Figure 3.122 Foliage of E. hirsutus (PR785).

Figure 3.123 Female E. hirsutus Figure 3.124 Female cone of Figure 3.125 Male E. inopinus cone (Photo: A. Fanfoni). E. inopinus (Photo: A. Fanfoni). cones (Photo: A. Fanfoni).

Figure 3.126 Heavily pubescent, newly produced leaves of E. hirsutus (PR785).

CHAPTER 4. DNA BARCODING: RESULTS AND DISCUSSION

4.1 General introduction DNA barcoding is a relativity new technique that aims to identify all life on earth up to species level using a short, standardised genetic region (Hebert et al. 2003). Implicit in this is the fact that should material not be attributable to any known species it would be considered novel (Ratnasingham & Hebert 2007). If successful, its range of applications are innumerable including biosecurity, biodiversity inventories, forensics tests, detecting illegal trade, identifying fragmented material, monitoring invasive species, and uncovering misrepresentation in trade (Newmaster et al. 2006; Lahaye et al. 2008a, b; Little & Stevenson 2007; Pryer et al. 2010). One of the main distinguishing features of DNA barcoding is its aim to standardise and openly share its protocols and data that in effect would allow biota-wide comparisons to be made (Ratnasingham & Hebert 2007). Since the concept was first introduced by Hebert et al. (2003) it has been widely criticised yet conversely has seen major positive acceptance. Much of the controversy surrounded the role of DNA barcoding is in taxonomy and specifically describing and delimiting species (DeSalle et al. 2005) based simply on arbitrary genetic distance (Ratnasingham & Hebert 2007). Much of this has been addressed and the concept has been more or less accepted for what it aims to be namely a rapid and independent identification tool that could indicate taxonomic anomalies (Pettengill & Neel 2010). Towards this effort and DNA barcoding's goals of standardisation a body was established viz. the Consortium for the Barcoding of Life (CBOL) along with a repository for DNA barcoding information the Barcode of Life Datasystems (BOLD). For DNA barcoding to succeed three main criteria have been outlined (CBOL Plant Working Group 2009; Hollingsworth et al. 2011) to which a potential barcode must adhere: (I) Universality, i.e. it should be sequencable across all land plants (II) Sequence quality and coverage, i.e. it should sequence easily and produce high quality sequences, and (III) Discrimination, i.e. it should be able to distinguish between a high number of even closely related taxa.

Barcoding has been successul in mammals along with some other tested animal (e.g. Lepidoptera) and fungal groups using CO1 often resolving up to 90% of species (Hajibabaei et al. 2007; Virgilio et al. 2010). However in plants it has been less successful (Fazekas et al. 2009) as the plant mitochondrion shows ~10–30 times less divergence than in animals (Wolfe et al. 1987). Efforts were made to explore additional regions primarily in the chloroplast but also in the nucleus (e.g. Kress et al. 2005; Chase et al. 2007; Ford et al. 2009; Lahaye et al. 2008a, b; Hollingsworth et al. 2009; Seberg & Petersen 2009). Generally, this has been done across the entire plant kingdom's diversity or within geographic or ecological communities, with less initial attention on distinguishing closely related species (Pettengill & Neel 2010). A general trend of around 70% discrimination has 100 been found in these studies with almost no barcoding gap between intra- and inter-specific genetic distances, often irrespective of the amount of genes selected (e.g. see Newmaster et al. 2008; Lahaye et al. 2008a; Hollingsworth et al. 2009; CBOL Plant Working Group 2009). Discrimination rates also vary greatly among different plant lineages and are particularly low in many gymnosperm groups (Sass et al. 2007; Hollingsworth et al. 2009). A limitation in many earlier studies is the lack of multiple accessions per species that may result in an overestimation of the discriminatory power (Burgess et al. 2011). In 2009 the CBOL Plant Working Group comprising 52 scientists (many working independently) standardised the plant barcodes, employing for the first time a multi gene barcoding approach. They selected rbcLa and matK as a core two-marker barcode with psbA-trnH and nrITS as additional markers (CBOL Plant Working Group 2009). Other genes that were considered but disregarded include: rpoC1, rpoB, atpF-atpH, psbK–psbI (Hollingsworth et al. 2011, Figure 4.1). The core barcodes were chosen based on the ease of working with rbcLa coupled to the discriminatory power of matK (Hollingsworth et al. 2011). Once this selection was made it paved the way for focussed efforts to test the viability of these barcodes on groups that would benefit greatly from DNA barcoding applications. However, it became evident that successful discrimination was markedly lower than expected in many cases (e.g. Sass et al. 2007; Edwards et al. 2008; Seberg & Petersen 2009; Spooner 2009; Starr et al. 2009; Yao et al. 2009; Ran et al. 2010; Roy et al. 2010; but also see Newmaster et al. 2008; Liu et al. 2010; Yesson et al. 2011). The suggested barcodes were up for review at the Barcoding conference in November 2011 in Adelaide, Australia (Hollingsworth et al. 2011) and the current study along with others will provide insight into the general performance of possible barcodes and combinations thereof.

4.2 Results 4.2.1 Sequencing success rates In terms of the IBOL proposed barcoding genes 97.42% (341/350) of the collected taxa could be sequenced for at least one gene including 64/65 Encephalartos species (Appendix 1) and five Stangeria replicates while 50.85% (178/350) could be sequence for all four genes. For Stangeria (Table 4.1) all plastid genes showed a 100% PCR success rate, while for nrITS1 only 40% of samples could be successful amplified and none were of satisfactory quality. For Encephalartos samples (Table 4.2) the most successfully sequenced genes were rbcLa (96.23%) and matK (88.69%) with the worst performing being nrITS1 (70.28%) and psbA-trnH (54%). Numbers however might be slightly misleading as both matK and nrITS1 had far less success using standard barcoding primers and new cycad and Encephalartos specific primers were developed. A number of failures due to poor sequence quality in nrITS1 was deemed to be due to hybrid taxa being analysed.

The best performing test regions were atpF-atpH and trnL both with a 93.33% success rate, while

101 the poorest preformer was psbK-psbI (11.74%) (Table 4.2), where two different sized bands were inconsistently produced. In all cases these sequences were compared to the National Centre for Biotechnology Information’s GenBank Database using BLAST, and in the following cases they were found to be of fungal origin (success rates in brackets): 5s (8/15=53.33%); ETS single fragment (NY153; 13/15=86.67%); and SAHH (4/78=5.13%). In these cases, Table 4.2 reflects a 0% success rate as the correct sequence was not generated.

For rbcLa a uniform sequence length was produced ranging from 402 to 555 basepairs (bp). For matK sequence lengths varied between 727–788bp with uncomplicated alignment. The nrITS1 sequence length varied based on the quality of fragments produced and alignment was troublesome only in the outgroups. With psbA-trnH a long sequence (857–1210bp) was consistently produced with primer set NY1494/NY1493, which when compared to GenBank accessions, was found to include a sizeable portion of the psbA and trnH genes along with the intergenetic spacer. Alignment posed difficult only in the outgroups and E. barteri subsp. barteri/E. barteri subsp. allochrous/E. ituriensis. This was due to a large deletion, which resulted in long branch attraction. These samples were removed from the analyses based on this result.

4.2.2 Genetic variation Only regions that were of suitable quality, i.e. unambiguous bases to a minimum, were considered for pairwise distance analyses. Thus even though CC0822, ef1-! and psbK-psbI were successfully sequenced these were not analysed due to poor sequence quality. All pairwise distance matrices and manipulations thereof (i.e. averages and standard deviations) are avaliable from the author ([email protected]).

For Stangeria the highest intra-specific variation is found (Table 4.1) in matK (0.65%), followed by psbA-trnH (0.32%) with the lowest being rbcLa (0% variation between all 5 replicates). For gene combinations all three genes had 0.34% variation while the core barcodes (matK + rbcLa) had 0.35% variation. Notably all the observed genetic variation for Stangeria stems from pairwise differences with voucher PR843, all other samples are genetically identical.

Within Encephalartos the highest mean genetic variation is found in nrITS1 (1.7%), followed by rbcLa (0.18%), matK (0.12%), and the lowest being psbA-trnH (0.11%; Table 4.2). For multi-locus barcodes the highest value is derived from the core barcodes (matK + rbcLa) + nrITS1 (0.76%), followed by all four proposed barcode genes (0.54%), other combinations showed little variation as the core barcodes (matK + rbcLa) has 0.14% variation and the core barcodes + psbA-trnH has 0.13% variation (Table 4.2). Mean intra-specific values were consistently lower than inter-specific values (Table 4.3) with high significance (all p values<0.001; Table 4.2). Intra-specific values deviate from the general trend, though nrITS1 is still the most variable (0.18%), rbcLa the second

102 most (0.9%), while psbA-trnH (0.8%), and matK (0.4%) (Table 4.2) remain low. In terms of the test genes samples (Table 4.2) values of the proposed barcodes were recalculated using only the 16 trail samples to ensure statistical comparability. All genes were overestimated except for rbcLa, which shows 0% variation. For the two test genes that were of suitable quality to be analysed, atpF-atpH had 0.14% variation while trnL had 0.16%, which is comparable to the other plastid regions (Table 4.2).

Inter-specific variation values in Encephalartos are very similar (0.10%-0.13%; Table 4.3) except for nrITS1 where the value increased to 1.72%. When species are investigated separately, in many cases they are found to be genetically identical even when all four barcoding genes are employed (Table 4.5). This is surprising as within all genes the inter-specific variation was greater than the intra-specific variation (Table 4.2) all with high significance (p<0.0001).

For the phylogenetic tree statistics (Table 4.4) nrITS1 showed a markedly higher percentage (28.65%) of parsimoniously informative characters compared to the combined plastid dataset (6.04%). All the plastid regions have very little variation individually: rbcLa 2.74%, matK 6.37%, psbA-trnH 1.15%. Where datasets were combined, the core barcodes plus nrITS1 had the largest number of informative characters (15.51%) compared to the core barcodes alone (4.90%) and all four regions combined (12.62%). In terms of variable characters nrITS with 374 (34.34%) has by far the largest percentage, while rbcLa has 61 (11.17%), psbA-trnH 116 (8.9%) and matK only 59 (7.37%).

4.2.3 Species resolution and species concepts For Stangeria eriopus no regions or combination thereof resolved the different replicates as monophyletic (Figure 4.2–4.4, 4.7–4.9).

The best performing gene in Encephalartos was nrITS1 distinguishing 10/60 taxa (16.66%), followed by matK resolving 4/60 (6.66%) species. The poorest performing regions were psbA-trnH and rbcLa, which could only distinguish 2/60 species. Species discrimination for multi-locus barcodes ranged from 8.3% (rbcLa + matK; rbcLa + matK + psbA-trnH) to 20% (rbcLa + matK + nrITS1; rbcLa + matK + psbA-trnH + nrITS1). A list of regions resolving species individually is given in Table 4.6. Notably E. inopinus is the only species that can be resolved by all regions or combinations thereof. No test regions were tested for their discrimination power due to a lack of multiple accessions.

4.3 Discussion 4.3.1 Success rates The rbcLa region showed a very high success rate with no alignment difficulty, and very consistent

103 results as reported by numerous other studies (e.g. Kress et al. 2005; Kress & Erickson 2007; Kress et al. 2009; Pettengill & Neel 2010; Bafeel et al. 2011; Li et al. 2011).

The success rates of matK, although reported high in some studies (Lahaye et al. 2008a, b; Pettengill & Neel 2010) has been found lower here and elsewhere particularly amongst gymnosperms (Sass et al. 2007; CBOL Plant Working Group 2009; Fazekas et al. 2008; Ford et al. 2009; Kress et al. 2009; Bafeel et al. 2011; Li et al. 2011; but see Liu et al. 2010). However by employing newly designed universal gymnosperm primers (Li et al. 2011) high sequencing success was achieved.

High rates of sequence recovery have been reported for psbA-trnH (Kress et al. 2005; Kress & Erickson 2007; Kress et al. 2009) but is significantly lower in the current study. The large size of the fragment produced is most probably due to duplicated loci found in most cycads (Sass et al. 2007). Also the region suffers from large sequence length variations between gymnosperm groups (Hao et al. 2010). The large deletions in some Encephalartos samples correspond with the phenomenon of inversion resulting in seemingly alignable palindromes in the gene and incorrect groupings of unrelated taxa (see Whitlock et al. 2010).

Reports for nrITS have suggested numerous issues including sequence length variation (Sass et al. 2007), paralogous copies within individuals (Alvarez & Wendel 2003), fungal contamination (Alvarez & Wendel 2003), amplification difficulties using standard primers (Kress & Erickson 2007), and sequencing issues (Kress et al. 2005; Yesson et al. 2011) speculated to be the result of long poly-G, -C, and -A repeats in cycads (Sass et al. 2007). Though not always present, all of these were encountered during the study, though paralogous copies were only identified in GenBank sequences used in early analyses. Most other issues were remedied (except for sequence length variation between Encephalartos and Stangeria; Sass et al. 2007) with the use of specific and internal cycad primers. An issue that could not be resolved was poor quality sequences due to multiple peaks that indicate possible hybridisation (Yesson et al. 2011), a situation known to occur in the genus (Table 3.2).

Few test genes could be successfully amplified even though in some cases these were successful in other gymnosperms (e.g. CC genes for Coniferophyta, Kado et al. 2008; SAHH in Dioon, Moynihan et al. in press; psbK-psbI in Zamia, Dioon, and Ceratozamia, Nicolalde-Morejo et al. 2010, 2011). The results from this study do however concur with the findings of the CBOL Plant Working Group (2009) as the sequence quality and success rates were lower for psbK-psbI while higher for atpF-atpH and trnL than the current core barcodes (rbcLa and matK).

4.3.2 Genetic variation

104

Variation within Stangeria is comparable to other cycad genera (Bogler & Francisco-Ortega 2004; Treutlein & Wink 2002) but the genetic similarity of the four samples used in the current study may be indicative that those are identical forms as confirmed by morphological analyses. The fact that sample PR843 is equally distant from all other samples might indicate some genetic isolation but it does not meet the 1% genetic distance as proposed by CBOL for specific status (Ratnasingham & Hebert 2007).

In Encephalartos the low amount of plastid (mean=0.11–0.18%) and nrITS1 variation (1.7%) found here coupled to the lack of a barcoding gap between intra- and inter-specific genetic distances (Table 4.4) has also been found in other studies but often to a lesser extent (Kress et al. 2005; Newmaster et al. 2006; Little and Stevenson 2007; Kress & Erickson 2007; Fazekas et al. 2009; Yao et al. 2009; Ran et al. 2010). The general trend of nrITS as more variable (Kress & Erikson 2007; Yesson et al. 2011) followed by psba-trnH, matK and rbcLa has been found in numerous studies (see CBOL Plant Working Group 2009). In the current study nrITS1 is 9.4 times more variable than the most variable plastid marker (Table 4.2). However a good barcode should have a barcoding gap between intra- and inter-specific values, whereas in the current study maximum intra-specific values are often much higher than inter-specific values even for nrITS1 with numerous genetically identical species (Table 4.5). This lack of genetic diversity and barcoding gaps has been found in several taxonomically difficult groups. Examples include: Acacia Mill. and Aspalathus L. (Fabaceae; Edwards et al. 2008), Araceae (Luo et al. 2009), Berberis L. (Berberidaceae; Roy et al. 2010), Caladenia R.Br. (Orchidaceae; Farrington et al. 2009), Carex L. (Cyperaceae; Starr et al. 2009), Crocus L. (Iridaceae; Seberg and Petersen 2009), Lamiaceae (Han et al. 2009), Solanum L. (Solanaceae; Spooner 2009), Myristicaceae (Newmaster et al. 2008), Paeoniaceae (Zhang et al. 2009), Parthenium L. (Asteraceae; Kumar et al. 2009), Picea (Pinaceae; Ran et al. 2010), Taxaceae, Cephalotaxaceae, and Podocarpaceae (Hao et al. 2009), and Cycadales (Sass et al. 2007).

The trnL region, though successful in other groups (Taberlet et al. 2006; Kress & Erickson 2007; CBOL Plant Working Group 2009) was found not to be variable within Encephalartos. This result is comparable to a study by Mabunda (2007) where 49 species were included in the analysis. As stated by the author: “Very few variable characters (most of which were autapomorphic) were found within Encephalartos and therefore this region provided very little information to infer species-level relationships”. Although some variation has been detected in atpF-atpH (Ran et al. 2010; Nicolalde-Morejón et al. 2010) in the current study it is as invariant as the suggested plastid barcodes.

The lack of variation found among plants have many speculative causes including: polyploidy, apomixis, microspeciation from ecotypes (Hollingsworth et al. 2011), exceptionally low rates of

105 sequence evolution vs. speciation (Fazekas et al. 2009; Ford et al. 2009), mutation rate and genome repair efficiency (Ford et al. 2009); seed dispersal range (Hollingsworth 2011), pollination mode, inheritance patterns, long life cycle, frequent inter-specific hybridisation (Ran et al. 2010), recent radiation (Treutlein et al. 2005), size of regions investigated (Fazekas et al. 2009), taxonomic issues (Kress & Erikson 2007), retained ancestral polymorphisms, and reticulate evolution (Fazekas et al. 2009). Most of these traits are found in cycads, as a whole and specifically in Encephalartos (Norstog & Nicholls 1997; but see Gorelick & Olson 2011) and may well account for the lack of genetic variation found in the genus.

4.3.3 Species resolution The most important criterion for a DNA barcode is whether or not it is able to distinguish species irrespective of the amount of genetic distance separating them (Little 2011). Though there are numerous methods for determining species resolution (reviewed in Little 2011) maximum parsimony analysis is the accepted method for smaller datasets (see comparisons by Ran et al. 2010). Species resolution however has been one of the more contentious issues in DNA barcoding with many studies reporting high success rates (Lahaye et al. 2008a, b; CBOL Plant Working Group 2009) while others have poor resolution using the same markers (e.g. Ran et al. 2010).

The ability of matK to provide high resolution, often above 80% or 90% (Kress et al. 2009; Pettengill & Neel 2010; Burgess et al. 2011; Yesson et al. 2011), was not replicated in Encephalartos. This has also been found in other groups across most of the angiosperm diversity at order level and some non-vascular plants (Kress & Erikson 2007). In the current study matK resolves only 4/60 (6.66%) species (Figure 4.2) these include E. ngoyanus (61BP), E. kisambo (61BP), E. inopinus (84BP) and E. cycadifolius (86BP). Notable species groupings include E. cupidus, E. dyerianus, and E. dolomiticus with the unrelated E. horridus (56BP). Also the grouping of all but one E. lebomboensis replicate (61BP) might indicate its succesful resolution, though the inclusion of single replicates of four unrelated species with these, complicates such a conclusion. The inclusion of E. msinganus with E. woodii and E. natalensis (62BP) may be indicative of relations and conspecificity but the inclusion of two E. villosus replicates makes such an assessment impossible.

For rbcLa the relatively high success rates publicised amongst many other plant groups (Burgess et al. 2011; Kress & Erikson 2007; Pettengill & Neel 2010) could not be achieved in Encephalartos and is also not found in comparable groups such as Taxus L. (Liu et al. 2010), Picea Mill. (Ran et al. 2010) and in a study done by Kress et al. (2009) which included a diverse range of taxa. Only 2/60 species of Encephalartos could be discriminated namely E. inopinus (65BP) and E. hirsutus (63BP).

106

In psbA-trnH the high species resolving power found in other studies (Kress & Erikson 2007; Kress et al. 2009; Pettengill & Neel 2010; Burgess et al. 2011) was not recovered in the current study, a condition found to a lesser extent in Asphalatus (Edwards et al. 2008), Taxus (Liu et al. 2010), and Myristicaceae (Newmaster et al. 2008). The region resolved only two taxa (Figure 4.4) namely E. inopinus (62BP) and E. hirsutus (65BP) and had exactly the same topology and species groupings as rbcLa, including the E. villosus and E. tegulaneus subsp. tegulaneus mismatch.

The resolving power of nrITS is known from its numerous uses in phylogenetic studies (Hollingsworth et al. 2011) and was used in many DNA barcoding initiatives (Kress & Erikson 2007 Edwards et al. 2008; Liu et al. 2010; Yesson et al. 2011) as well as in the current study. Here nrITS1 distinguishes 10/60 taxa (16.66%) (Figure 4.5). However E. nubimontanus (CC328) is included with E. eugene-maraisii (78BP) which most probably represents a miss-identification. Encephalartos tegulaneus subsp. tegulaneus (PR747), which does not group with other replicates (100BP), is difficult to distinguish from other tropical African species and most probably represents a faulty identification. A similar situation is found with E. kanga, which should be associated with E. kisambo (69BP) with which it appears to be synonymous (see Chapter 1, Table 1.1). Of note are two E. lehmannii specimens that group together (78BP) apart form the other replicates that may indicate a genetic divergence of these two replicates. The inclusion (90BP) of one E. ngoyanus (PR703) replicate with E. villosus and E. aplanatus and the grouping of E. umbeluziensis, E. aplanatus (d1126) and E. cerinus, is probably reflective of the morphological affinity. The grouping of E. schmitzii and E. delucanus (64BP) might indicate conspecificity between these poorly-known and described taxa (Vorster 2004a). The highly supported division of E. laevifolius samples into two groups (84BP and 97BP), the first of which includes the morphologically similar E. brevifoliolatus (97BP) amongst other species (Grobbelaar 2002) is interesting. This might be a reflection of the wide distribution range of E. laevifolius with the species delimitations within E. laevifolius possibly in need of revision.

Although the addition of genes in combined barcodes practically always increases resolution, this quickly becomes saturated even though genetic diversity still increases (Fazekas et al. 2009; Pettengill & Neel 2010; Ran et al. 2010; Burgess et al. 2011). This was also the case in the current study as the maximum number of additional species resolved with the inclusion of any number of barcodes was just three (Figures 4.6–4.9). The barcode combination of matK, rbcLa and nrITS1 performs jointly with the four gene combination (rbcLa, matK, psbA-trnH and nrITS1) as the best resolving, with 12/60 species (20%). Anomalies are as explained under each gene separately. Notably however E. natalensis replicates no longer group with E. altensteinii as found in nrITS1, supporting it’s specific status. All E. dyerianus samples group together (with other species; 53BP) as opposed to nrITS1 results, and E. sp. PR913 (62BP) seems to resolve with E. schajesii while E. poggei (PR911) probably representing a miss-identification. The three plastid barcodes (matK,

107 rbcLa and psbA-trnH) in combination can only resolve four species. The grouping of two E. transvenosus replicates (88BP) may indicate some genetic divergence of these replicates. All other combinations have anomalies as described under each gene separately. Notably no species is resolved in combination that could not be resolved singly, and E. inopinus is the only species that can be resolved by all regions (Table 4.6).

This non-monophyly of many species seen here has been detected in numerous other studies that have used multiple replicates per species (Fazekas et al. 2008; Lahaye et al. 2008a; Newmaster et al. 2008). Reasons for this include the incomplete spread of mutation through a species’ entire distribution range (Hollingsworth et al. 2011), reintegration with sister species, and imperfect taxonomy (Fazekas et al. 2009), once again traits prevalent in Encephalartos.

4.4 Conclusions Stangeria possibly represents a single species as very little intra-specific variation was found. However, as the analysed material was of cultivated origin and morphological forms could not confidently be identified, the results may not be conclusive. Also, based on the high levels of variation found in nrITS1 in Encephalartos, a similar situation may well be present in Stangeria and be worth investigating.

As a result of Encephalartos' phylogenetic isolation (along with other cycads) the application of standardised primers was difficult in most cases. Thus, should a strict standard be employed, sequencing issues with matK, nrITS and psbA-trnH would exclude them as good barcodes. However, should a compromise be made in that specific primers could be employed for matK (see Sass et al. 2007, Li et al. 2011), nrITS, and possibly psbA-trnH, these regions may be considered successful on sequence quality and ease criteria. This however deducts from the efficiency of the barcodes as utterly unidentifiable material (e.g. homogenised herbal medicines) would need to be sequenced using the standard primers first, and only after subsequent failure would group specific primers be used. Should a tiered barcoding approach however be standardised (Newmaster et al. 2006) this would cease to be a concern as the first level testing would clearly place samples within the Cycadales where after the correct primer combinations could be used for the secondary barcodes. Although alignment between higher taxonomic groups in psbA-trnH and specifically nrITS in cycads (unpublished data) is troublesome this is not applicable to certain search tools (Kress & Erikson 2007; Little 2011) and should not pose a concern for barcode acceptance. However the various highlighted issues with psbA-trnH would exclude it as a reliable barcode for Encephalartos.

Based on the lack of resolution of many species as monophyletic groups it may seem feasible to reduce the number of currently recognised Encephalartos species significantly. In fact taxonomic

108 uncertainty in the genus is rife and in some cases the results may be indicative of species delimitations that are in need of revision as highlighted. However, even in species groups recently and thoroughly monographed species could still not be resolved by DNA barcoding (Kress & Erikson 2007). The results from such studies often show that resolution can be made up to closely related species groups (Hollingsworth et al. 2011) as found here and elaborated on in Chapter 3. The genus Encephalartos however suffers from the general lack of genetic diversity found amongst all cycads, most probably due to specific life history and evolutionary traits of the group. Plants species as a whole have been argued to be less clearly definable than for example animals with a much greater overlap between intra-specific and inter-specific genetic distances (CBOL Plant Working Group 2009). Also as mentioned, gene tree paraphyly is possibly extensive. Thus it must be assumed that the thorough sampling of species and replicates has revealed the lack of differentiation speculated by numerous authors (see Sass et al. 2007). As for test regions, all failed at either amplification, sequencing and/or genetic variation and species resolution stages and none can be considered as barcodes by CBOL's standards (CBOL Plant Working Group 2009). However work is underway with specific primer designs for psbK-psbI, which reveals moderate genetic variability along with some single copy nuclear genes including NEEDLY, which shows variation and resolution comparable to nrITS1 (unpublished data). Not withstanding single copy nuclear gene's prospective difficulties (Hollingsworth et al. 2011), these should be investigated as the next generation barcodes.

Finally, based on the results presented here, I strongly advocate the use of a multi-tiered DNA barcode approach that includes a nuclear marker for vascular plants, as also promoted by Newmaster et al. (2006). A multi-tiered approach would negate issues such as primer and even genetic standardisation at the secondary level. A nuclear marker will allow the detection of hybrids, a condition prevalent in many groups (Ran et al. 2010) and since it is biparentally inherited give a more complete and contrasted picture of the species history and may be the only source of barcode genes in parasitic plants (Hollingsworth et al. 2011). Based on the current study nrITS1 is suggested with standardised specialist primers as second tier barcode with either matK and rbcLa as first tier barcodes. However, neither the first nor the second tier barcodes provide enough resolution, and additional barcodes should be investigated along with a thorough review of the currently recognised species within Encephalartos.

109

Table 4.1 DNA barcoding statistics for Stangeria. Gene Success rate Mean genetic variation rbcLa 5/5=100% 0.00% matK 5/5=100% 0.65% psbA-trnH 5/5=100% 0.32% nrITS1 2/5=40% - matK + rbcLa 5/5=100% 0.35% 3 gene combination (rbcLa, 5/5=100% 0.34% matK, psbA-trnH)

110

Table 4.2 DNA barcoding statistics for Encephalartos. ‘Test’ refers to recalculation using only test samples for statistical comparability between proposed and test regions. Gene Success rate Mean genetic Mean inter- Mean intra- Resolution variation specific specific power variation variation Proposed barcodes rbcLa 341/350=97.42% 0.18% Test: 0% 0.13% 0.09% 2/60 (3.33%) matK 309/350=88.28% 0.12% Test: 0.17% 0.12% 0.04% 4/60 (6.66%) psbA-trnH 186/350=53.14% 0.11% Test: 0.13% 0.10% 0.08% 2/60 (3.33%) nrITS1 246/350=70.28% 1.70% Test: 2.10% 1.72% 0.18% 10/60 (16.66%) Test genes atpFH 14/15=93.33% 0.14% 0.14% - - trnL 14/15=93.33% 0.16% 0.16% - - 5s 0/15=0% - - - - AGAMOUS 0/30=0% - - - - !-tubulin 0/15=0% - - - - CC0702 0/15=0% - - - - CC0822 4/15=26.67% - - - - CC1147 0/15=0% - - - - CC1606 0/15=0% - - - - CC1798 0/15=0% - - - - CC2920 0/15=0% - - - - ef1-! 12/47=25.53% - - - - ETS 0/15=0% - - - - SAHH 0/78=0% - - - - psbK-psbI 14/119=11.74% - - - - Gene combinations matK + 178/350=50.58% 0.14% 0.53% 0.06% 5/60 (8.33%) rbcLa Barcodes + 178/350=50.58% 0.13% 0.11% 0.07% 5/60 (8.33%) trnH-psbA Barcodes + 178/350=50.58% 0.76% 0.74% 0.11% 12/60 (20%) nrITS1 4 barcode 178/350=50.58% 0.54% 0.52% 0.10% 12/60 (20%) genes Test genes 178/350=50.58% 0.15% 0.15% - - combined

111

Table 4.3 Inter- and intra-specific variation percentages of Encephalartos species. Where only one replicate was included a ‘–‘sign is inserted. Species rbcLa matK psbA-trnH nrITS1 Barcodes + Barcodes + Barcodes Combined (matK nrITS1 trnH (matK + rbcLa) + rbcLa + psbA- trnH + nrITS1) Intra Intra Inter Intra Inter Inter Intra Inter Intra Inter Intra Inter Intra Inter Intra Inter E. aemulans 0 0 0.07 0 1.58 0.06 0.08 0.15 0 0.69 0.04 0.11 0 0.07 0.03 0.51 E. altensteinii 0.1 0 0.07 0.26 1.58 0.05 0 0.07 0.12 0.69 0.02 0.07 0.04 0.06 0.08 0.48 E. aplanatus 0.18 0.19 0.18 0.17 1.6 0.22 0.08 0.15 0.18 0.65 0.14 0.17 0.19 0.2 0.14 0.46 E. arenarius 0 0 0.07 0 1.45 0.07 0.04 0.07 0 0.63 0.02 0.07 0 0.07 0.01 0.44 E. brevifoliolatus 0 0 0.08 0 3.71 0.07 0 0.13 0 1.27 0 0.1 0 0.07 0 0.84 E. bubalinus 0 0 0.08 0 1.68 0.07 0 0.08 0 0.59 0 0.07 0 0.07 0 0.4 E. caffer - - 0.08 - 1.15 0.07 - 0.08 - 0.52 - 0.08 - 0.07 - 0.37 E. cerinus 0 0.13 0.08 1.37 1.24 0.07 0.08 0.08 0.64 0.55 0.08 0.08 0.08 0.07 0.45 0.39 E. concinnus 0.37 0.13 0.21 0.35 1.9 0.07 0 0.08 0.26 0.71 0.12 0.12 0.23 0.15 0.16 0.48 E. cupidus 0 0 0.29 0 1.55 0.07 0.03 0.08 0 0.64 0.01 0.14 0 0.2 0.01 0.43 E. cycadifolius 0 0 0.33 0 3.07 0.07 0.08 0.14 0 1.39 0.04 0.18 0 0.22 0.03 0.96 E. delucanus - 0.18 - 1.52 0.07 - 0.2 - 0.71 - 0.16 - 0.13 - 0.56 E. dolomiticus - - 0.29 - 1.39 0.07 - 0.08 - 0.69 - 0.14 - 0.2 - 0.48 E. dyerianus 0.04 0 0.29 0.42 1.41 0.08 0 0.08 0.18 0.7 0.01 0.15 0.02 0.2 0.12 0.49 E. eugene-maraisii 0 0 0.06 0 1.55 0.08 0.17 0.25 0 0.68 0.08 0.15 0 0.07 0.06 0.53 E. ferox 2.09 0 0.07 0.02 1.47 0.21 0 0.09 0.45 0.68 0.4 0.11 0.79 0.13 0.29 0.48 E. friderici-guilielmi 0 0 0.19 0 3.45 0.04 0 0.14 0 1.27 0 0.13 0 0.13 0 0.85 E. ghellinckii 0 0.26 0.32 0.79 3.77 0.04 0 0.14 0.36 1.37 0.08 0.18 0.15 0.21 0.22 0.91 E. gratus 0 0.25 0.19 1.08 1.5 0.05 0 0.09 0.55 0.7 0.08 0.11 0.15 0.13 0.37 0.49 E. heenanii 0 0 0.06 0 1.38 0.05 0.09 0.09 0 0.61 0.04 0.07 0 0.06 0.03 0.43

112

E. hildebrandtii - - 0.17 - 1.82 0.05 - 0.09 - 0.81 - 0.11 - 0.12 - 0.56 E. hirsutus 0 0 0.06 0.1 1.92 0.23 0 0.17 0.04 0.86 0 0.15 0 0.13 0.03 0.62 E. horridus 0 0.04 0.17 0 1.27 0.05 0 0.09 0.01 0.6 0.01 0.11 0.03 0.12 0.01 0.43 E. humilis 0 0 0.06 0.19 3.32 0.05 0 0.15 0.08 1.38 0 0.1 0 0.06 0.06 0.98 E. inopinus 0 0 0.31 0.19 1.67 0.23 0.17 0.35 0.08 0.85 0.08 0.31 0 0.28 0.11 0.68 E. kanga - 0.06 - 1.82 0.04 - 0.41 - 0.62 - 0.22 - 0.05 - 0.54 E. kisambo 0 0 0.2 0 1.43 0.05 0.08 0.09 0 0.68 0.04 0.11 0 0.13 0.03 0.47 E. laevifolius 0 0 0.06 0.14 3.57 0.02 0.02 0.15 0.06 1.52 0.01 0.1 0 0.05 0.05 1.04 E. lanatus 0 0.06 0.06 0.1 3.58 0.05 0.04 0.15 0.06 1.49 0.04 0.1 0.04 0.06 0.06 1.04 E. latifrons 0 0.13 0.06 0 1.17 0.05 0 0.09 0.05 0.49 0.04 0.07 0.08 0.06 0.03 0.35 E. laurentianus 0 0 0.06 0 1.25 0.05 0 0.09 0 0.55 0 0.07 0 0.06 0 0.39 E. lebomboensis 0.05 0.02 0.18 0.02 1.61 0.05 0.01 0.1 0.03 0.6 0.02 0.11 0.03 0.13 0.02 0.41 E. lehmannii 0.19 0 0.06 0.08 1.09 0.23 0.03 0.1 0.08 0.53 0.05 0.12 0.08 0.13 0.06 0.38 E. longifolius 0.05 0.13 0.19 0 1.12 0.05 0.04 0.1 0.06 0.54 0.07 0.12 0.09 0.13 0.05 0.39 E. macrostrobilus - - 0.17 - 1.3 0.05 - 0.1 - 0.61 - 0.11 - 0.12 - 0.44 E. manikensis 0 0 0.06 0.05 1.27 0.05 0 0.11 0.02 0.46 0 0.08 0 0.06 0.01 0.33 E. middelburgensis 0 0 0.07 0.05 1.31 0.05 0 0.11 0.02 0.48 0 0.09 0 0.06 0.01 0.34 E. msinganus 0.09 0.06 0.15 1.32 1.43 0.17 0.12 0.22 0.59 0.67 0.1 0.19 0.08 0.16 0.43 0.51 E. munchii 0 0 0.07 0 1.19 0.05 0.09 0.2 0 0.53 0.04 0.12 0 0.06 0.03 0.42 E. natalensis 0 0 0.16 0 1.22 0.18 0.25 0.23 0 0.61 0.12 0.2 0 0.17 0.08 0.48 E. natalensis X E. - - 0.17 - 1.04 0.19 - 0.24 - 0.54 - 0.21 - 0.18 - 0.43 woodii E. ngoyanus 0 0.04 0.18 0.15 0.87 0.04 0 0.12 0.07 0.38 0.01 0.12 0.03 0.13 0.04 0.28 E. nubimontanus 0 0.25 0.27 0.2 1.42 0.05 0.08 0.12 0.17 0.59 0.12 0.15 0.15 0.18 0.14 0.42 E. paucidentatus 0 0 0.06 0 1.33 0.05 0.25 0.19 0 0.59 0.12 0.12 0 0.05 0.08 0.45

113

E. poggei 0 0.25 0.15 0.1 1.07 0.05 0 0.17 0.13 0.51 0.08 0.14 0.15 0.11 0.09 0.4 E. princeps 0 0 0.06 0 1.28 0.06 0.34 0.51 0 0.56 0.16 0.27 0 0.06 0.11 0.54 E. pterogonus 0 0 0.06 0 1.3 0.04 0 0.12 0 0.57 0 0.08 0 0.05 0 0.42 E. schaijesii - - 0.28 - 1.21 0.06 - 0.22 - 0.61 - 0.2 - 0.19 - 0.51 E. schmitzii - - 0.15 - 1.28 0.07 - 0.13 - 0.6 - 0.12 - 0.12 - 0.44 E. sclavoi 0.18 0 0.05 0.2 1.44 0.24 0.17 0.13 0.13 0.65 0.12 0.13 0.08 0.13 0.14 0.47 E. senticosus 0 0 0.05 0.02 1.5 0.08 0 0.15 0.01 0.65 0 0.1 0 0.06 0.01 0.47 E. septentrionalis - - 0.17 - 1.35 0.09 - 0.25 - 0.63 - 0.19 - 0.13 - 0.5 E. tegulaneus 0.09 0 0.03 1.15 2.01 0.1 0.08 0.15 0.51 0.88 0.06 0.1 0.04 0.06 0.37 0.63 subsp. tegulaneus E. transvenosus 0.74 0.08 0.03 0 1.13 0.84 0.09 0.25 0.2 0.69 0.22 0.31 0.35 0.36 0.16 0.54 E. trispinosus 0.09 0 0.04 0 1.23 0.1 0 0.2 0.02 0.57 0.02 0.12 0.04 0.06 0.01 0.44 E. turneri - - 0.04 - 1.33 0.27 - 0.25 - 0.65 - 0.19 - 0.14 - 0.5 E. umbeluziensis 0 0 0.06 0.19 1.05 0.11 0.25 0.38 0.08 0.49 0.12 0.21 0 0.08 0.14 0.43 E. villosus 0.19 0.06 0.06 0 1.46 0.09 0.17 0.37 0.07 0.67 0.14 0.21 0.11 0.08 0.1 0.57 E. whitelockii 0 0 0.13 0 1.8 0.18 0.99 0.33 0 0.86 0.43 0.24 0 0.15 0.3 0.68 E. woodii - - 0.18 - 1.63 0.19 - 0.28 - 0.76 - 0.23 - 0.19 - 0.59 Total Average 0.09 0.04 0.13 0.18 1.65 0.1 0.08 0.17 0.11 0.72 0.07 0.14 0.06 0.12 0.1 0.52 P-value 2.80e-012 1.87e-011 9.05e-009 < 2.2e-16 < 2.2e-16 9.05e-009 1.87e-011 < 2.2e-16

114

Table 4.4 PAUP statistics of analyses obtained from separate and combined datasets.

Statistic rbcLa matK psbA-trnH nrITS Combined Core Core Combined plastids barcodes barcodes + plastids + nrITS nrITS No. of Encephalartos taxa 5 5 5 5 5 5 5 5 excluded No. of included characters 546 801 1303 1089 2650 1347 2436 3739 No. of. constant characters 485 742 1187 715 2414 1227 1942 3129 No. of variable sites 61 (11.17%) 59 (7.37%) 116 (8.9%) 374 (34.34%) 236 (8.91%) 120 (8.91%) 494 (20.28%) 610 (16.31%) No. of parsimoniously 15 (2.75%) 51 (6.37%) 15 (1.15%) 312 (28.65%) 160 (6.04%) 66 (4.90%) 378 (15.51%) 472 (12.62%) informative characters No. of trees (Fitch) 51 3 1 134 2141 1 1 186 No. of steps (tree length) 83 81 83 514 277 191 1007 1034 Consistency index (CI) 0.77 0.74 0.77 0.83 0.87 0.65 0.55 0.65 Retention index (RI) 0.66 0.91 0.66 0.96 0.95 0.76 0.83 0.88 Average number of changes 1.36 1.37 0.1 1.37 1.17 1.59 2.04 1.7 per variable sites (number of steps/number of variable sites)

115

Table 4.5 Number of genetically identical species per gene. Species rbcLa matK psbA nrITS1 Core Core Core 4 gene -trnH Barcodes Barcodes Barcodes + combination (matK + + ITS trnH-psbA rbcLa) E. woodii 40 27 0 0 19 0 0 0 E. whitelockii 46 26 20 1 20 0 8 0 E. villosus 3 3 0 1 3 0 0 0 E. umbeluziensis 39 25 20 5 18 1 7 0 E. turneri 38 24 5 0 17 0 1 0 E. trispinosus 37 23 19 0 16 0 6 0 E. transvenosus 36 22 18 0 15 0 5 0 E. tegulaneus subsp. 35 22 18 0 15 0 5 0 tegulaneus E. septentrionalis 34 2 17 1 1 1 1 1 E. senticosus 33 0 5 0 0 0 0 0 E. sclavoi 32 4 0 1 4 1 0 0 E. schmitzii 31 1 16 0 0 0 0 0 E. schaijesii 31 0 15 0 0 0 0 0 E. pterogonus 30 21 0 0 14 0 0 0 E. princeps 0 20 14 0 0 0 0 0 E. poggei 29 0 5 0 0 0 0 0 E. paucidentatus 28 0 3 0 0 0 0 0 E. nubimontanus 27 0 13 0 0 0 0 0 E. ngoyanus 26 19 13 1 13 1 5 0 E. natalensis x E. 25 3 12 0 3 0 2 0 woodii E. natalensis 0 18 0 0 0 0 0 0 E. munchii 24 1 11 4 1 0 0 0 E. msinganus 23 17 3 0 12 0 0 0 E. middelburgensis 0 0 0 0 0 0 0 0 E. manikensis 22 16 0 0 11 0 0 0 E. macrostrobilus 21 0 11 0 0 0 0 0 E. longifolius 26 15 2 0 11 0 0 0 E. lehmannii 19 15 2 0 10 0 0 0 E. lebomboensis 20 15 10 3 11 1 5 0 E. laurentianus 17 14 9 0 9 0 4 0 E. latifrons 17 0 9 1 0 0 0 0

116

E. lanatus 0 13 9 0 0 0 0 0 E. laevifolius 17 0 9 2 0 0 0 0 E. kisambo 16 2 8 1 2 1 1 0 E. kanga 15 12 7 2 8 2 3 1 E. inopinus 14 11 6 0 7 0 2 0 E. humilis 3 3 2 0 3 0 1 0 E. horridus 13 10 0 1 6 1 0 0 E. hirsutus 2 2 2 0 2 0 1 0 E. hildebrandtii 1 1 1 1 1 1 0 0 E. heenanii 13 0 5 0 0 0 0 0 E. gratus 11 0 5 0 0 0 0 0 E. ghellinckii 10 9 0 0 5 0 0 0 E. friderici-guilielmi 9 1 2 1 1 0 1 0 E. ferox 8 8 0 1 4 0 0 0 E. eugene-maraisii 9 7 4 0 4 0 2 0 E. dyerianus 6 0 0 0 0 0 0 0 E. dolomiticus 5 1 3 0 1 0 0 0 E. delucanus 0 6 3 0 0 0 0 0 E. cycadifolius 4 5 2 0 2 0 0 0 E. cupidus 3 0 0 0 0 0 0 0 E. concinnus 2 0 0 0 0 0 0 0 E. cerinus 2 4 2 0 2 0 0 0 E. caffer 0 4 0 0 0 0 0 0 E. bubalinus 2 3 1 0 2 0 0 0 E. brevifoliolatus 0 2 0 0 0 0 0 0 E. arenarius 1 1 0 0 1 0 0 0 E. aplanatus 1 1 0 0 1 0 0 0 E. altensteinii 0 0 0 0 0 0 0 0 E. aemulans 0 0 0 0 0 0 0 0

117

Table 4.6 Resolution of Encephalartos species per gene and/or gene combinations. Species Gene providing resolution (Boostrap support) E. aemulans None E. altensteinii None E. aplanatus None E. arenarius None E. barteri subsp. barteri Not analysed E. barteri subsp. allochrous Not analysed E. brevifoliolatus None E. bubalinus None E. caffer None E. cerinus None E. chimanimaniensis Not analysed E. concinnus None E. cupidus None E. cycadifolius matK (86BP) E. delucanus None E. dolomiticus None E. dyerianus None E. equatorialis None E. eugene-maraisii nrITS1 (78BP) E. ferox nrITS1 (94BP) E. friderici-guilielmi nrITS1 (53BP) E. ghellinckii None E. gratus None E. heenanii None E. hildebrandtii None E. hirsutus nrITS1 (100BP), psbA-trnH (65BP), rbcLa (63BP) E. horridus None E. humilis None E. inopinus nrITS1, matK (84BP), psbA-trnH (62BP), rbcLa (65BP) E. ituriensis None E. kisambo matK (61BP) E. laevifolius None E. lanatus nrITS1 (54BP) E. latifrons None E. laurentianus nrITS1 (62BP) E. lebomboensis None

118

E. lehmannii None E. longifolius None E. mackenziei Not analysed E. macrostrobilus None E. manikensis None E. marunguensis None E. middelburgensis None E. msinganus None E. munchii None E. natalensis matK + rbcLa + nrITS1 (71BP) E. ngoyanus matK (61BP) E. nubimontanus None E. paucidentatus None E. poggei None E. princeps None E. pterogonus None E. relictus Not analysed E. schaijesii None E. schmitzii None E. sclavoi None E. senticosus None E. septentrionalis None E. tegulaneus subsp. nrITS1 (100BP) tegulaneus E. tegulaneus subsp. Not analysed powysii E. transvenosus nrITS1 (63BP) E. trispinosus None E. turneri None E. umbeluziensis None E. villosus None E. whitelockii None E. woodii None

119

Figure 4.1 Timeline of the considered plant barcodes. Colours (red; blue) are a subjective measure of efforts and possible acceptance of barcodes by the CBOL Plant Working Group (2009). Dashed lines reflect international barcoding conferences (Taken from Hollingsworth et al. 2011). 120

E transvenosus PR665 E sp PR913 E schaijesii CC329 E poggei PR911 E nubimontanus PR792 E longifolius PR673 E horridus PR847 E horridus PR846 E horridus d1028 56 E gratus d1026 E dyerianus PR863 E dyerianus PR821 Levubu E dyerianus PR820 Levubu E dyerianus PR769 E dyerianus PR731 E dyerianus CC332 E dolomiticus CC343 E cupidus PR767 E cupidus PR734 E cupidus PR691 E cupidus d1127 E poggei PR813 E lebomboensis PR874 Piet Retief E lebomboensis PR831 Pongola E lebomboensis PR800 Piet Retief 61 E lebomboensis PR796 E lebomboensis PR698 Mananga E lebomboensis PR657 Piet Retief E lebomboensis d1030 E latifrons PR806 Green Hills E cerinus d1022 E aplanatus PR682 E woodii PR875 E villosus PR837 E villosus d1047 62 E natalensisX E woodii PR842A E natalensis PR802 Kranskloof E natalensis d1035 E msinganus PR751 E msinganus PR701 E aplanatus PR757 E ngoyanus PR799 61 E ngoyanus PR717 E ngoyanus PR703 E ngoyanus d1135 61 E kisambo PR823 E kisambo PR745 E kisambo d1132 62 E transvenosus PR829 E transvenosus PR727 63 E longifolius PR873 Blunt Tip E longifolius d1032 84 E inopinus PR864 E inopinus PR778 86 E cycadifolius PR683 E cycadifolius d1128 E whitelockii PR818 E whitelockii d1048 E villosus PR838 Broad Leaf E villosus PR816 E villosus PR671 E umbeluziensis PR858 E umbeluziensis d1046 E turneri d1044 E trispinosus PR868 E trispinosus PR680 E trispinosus d1043 E transvenosus PR832 Tsualo E transvenosus PR797 E transvenosus d1041 E tegulaneus PR877 E tegulaneus PR747 E tegulaneus d1040 E septentrionalis d1138 E senticosus PR833 E senticosus PR830 E senticosus PR812 E senticosus PR719 E senticosus PR663 E senticosus d1039 E sclavoi PR790 E sclavoi PR738 E sclavoi d1038 E schmitzii PR819 E pterogonus PR876 E pterogonus d1136 E princeps PR871 E princeps PR836 E princeps PR810 E princeps d1037

F igure 4.2 (b) 121

Figure 4.2 (a)

68 E paucidentatus PR849 E paucidentatus PR710 E paucidentatus d1036 E nubimontanus PR704 E nubimontanus PR655 E nubimontanus CC328 E munchii PR855 E munchii PR737 E munchii d1134 E msinganus d1034 E middelburgensis PR827 Middelburg E middelburgensis PR726 E middelburgensis PR586 Wilge River E middelburgensis CC337 E manikensis PR903 Vumba E manikensis PR795 E manikensis PR697 E manikensis d1033 E macrostrobilus CC333 E longifolius PR809 Blunt Tip E longifolius PR808 E lehmannii d1031 E lehmannii PR835 E lehmannii PR780 E lehmannii PR661 E lebomboensis PR805 E laurentianus PR789 E laurentianus d1029 E latifrons PR811 Trappers valley E lanatus PR828 E lanatus PR587 E lanatus d1133 E laevifolius PR845 Kaapsehoop E laevifolius PR803 Mallalosha E laevifolius PR801 E laevifolius PR798 Havelock E laevifolius PR730 E laevifolius CC335 E kanga PR907 E humilis PR712 56 E humilis CC340 E horridus PR777 E hirsutus PR718 E hirsutus CC336 E hildebrandtii PR824 E heenanii PR776 E heenanii PR775 E gratus PR774 E ghellinckii PR773 E ghellinckii CC331 E friderici-guilielmi PR853 E friderici-guilielmi PR772 E friderici-guilielmi PR733 E friderici-guilielmi CC338 E ferox PR844 Yellow cone E ferox PR841 E ferox PR771 E ferox PR676 E ferox PR651 E ferox d1025 E eugene maraisii PR872 E eugene maraisii d1024 100 E delucanus d1129 E concinnus PR890 E concinnus PR817 E cerinus PR859 E caffer d1101 E bubalinus PR910 E bubalinus PR885 E bubalinus d1021 E brevifoliolatus Xdk2 E brevifoliolatus Xdk1 E arenarius PR854 E arenarius PR758 Alexandria 66 E arenarius d1020 E aplanatus d1126 E altensteinii PR856 E altensteinii PR756 E altensteinii d1019 E aemulans PR861 E aemulans d1018 100 Lepidozamia peroffskyana d1050 Lepidozamia hopei d1049 Macrozamia plurinervia d1060 99 Macrozamia pauli-guilielmi d1096 Macrozamia macdonnellii d1057 Macrozamia communis d1051 Stangeria eriopus PR843 Stangeria eriopus PR842B Stangeria eriopus PR753 Stangeria eriopus PR706 Stangeria eriopus PR860 Figure 4.2 Bootstrap tree of the matk region, bootstrap percentage values above branches. 122

Macrozamia plurinervia d1060 63 Macrozamia pauli-guilielmi d1096 Macrozamia macdonnellii d1057 Macrozamia communis d1051 63 E vilosus PR816 E tegulaneus subsp. tegulaneus d1040 65 E inopinus PR864 E inopinus PR778 63 E hirsutus PR718 E hirsutus CC336 Lepidozamia peroffskyana d1050 Lepidozamia hopei d1049 E woodii PR875 E whitelockii PR818 E whitelockii d1048 E villosus PR838 E villosus PR837 E villosus PR671 E villosus d1047 E umbeluziensis PR858 E umbeluziensis d1046 E turneri d1044 E trispinosus PR868 E trispinosus PR680 E trispinosus d1043 E transvenosus PR832 Tsualo E transvenosus PR829 E transvenosus PR797 E transvenosus PR727 E transvenosus PR665 E transvenosus d1041 E tegulaneus PR877 E tegulaneus PR747 E sp PR913 E septentrionalis d1138 E senticosus PR833 E senticosus PR830 E senticosus PR812 E senticosus PR719 E senticosus PR663 E senticosus d1039 E sclavoi PR790 E sclavoi PR738 E sclavoi d1038 E schmitzii PR819 E schaijesii CC329 E pterogonus PR876 E pterogonus d1136 E princeps PR871 E princeps PR836 E princeps PR810 E princeps d1037 E poggei PR911 E poggei PR813 E paucidentatus PR849 E paucidentatus PR710 E paucidentatus d1036 E nubimontanus PR792 E nubimontanus PR704 E nubimontanus PR655 E nubimontanus CC328 E ngoyanus PR799 E ngoyanus PR717 E ngoyanus PR703 E ngoyanus d1135 E natalensis x E woodi PR842A E natalensis PR802 Kranskloof E natalensis d1035 E munchii PR855 E munchii PR737 E munchii d1134 E msinganus PR751 E msinganus PR701 E msinganus d1034 E middelburgensis PR827 Middelburg E middelburgensis PR726 E middelburgensis PRU586 E middelburgensis CC337 E manikensis PR903 Vumba E manikensis PR795 E manikensis PR697 E manikensis d1033 E macrostrobilus CC333 E longifolius PR873 Blunt Tip E longifolius PR809 Blunt Tip E longifolius PR808 E longifolius PR673 E longifolius d1032

Figure 4.3 (b) 123

Figure 4.3 (a)

77 E lehmannii d1031 E lehmannii PR835 E lehmannii PR780 E lehmannii PR661 E lebomboensis PR874 Piet Retief E lebomboensis PR831 Pongola E lebomboensis PR805 E lebomboensis PR800 Piet Retief E lebomboensis PR796 E lebomboensis PR698 Mananga E lebomboensis PR657 Piet Retief E lebomboensis d1030 E laurentianus PR789 E laurentianus d1029 E latifrons PR811 Trappers valley E latifrons PR806 Green Hills E lanatus PR828 E lanatus PRU587 E lanatus d1133 E laevifolius PR845 Kaapsehoop E laevifolius PR803 Mallalosha E laevifolius PR801 E laevifolius PR798 Havelock E laevifolius PR730 E laevifolius CC335 E kisambo PR823 E kisambo PR745 E kisambo d1132 E kanga PR907 E humilis PR712 E humilis CC340 E horridus PR847 E horridus PR846 E horridus PR777 E horridus d1028 E hildebrandtii PR824 E heenanii PR776 E heenanii PR775 E gratus PR774 E gratus d1026 E ghellinckii PR773 E ghellinckii CC331 E friderici guilielmi PR853 E friderici guilielmi PR772 97 E friderici guilielmi PR733 E friderici guilielmi CC338 E ferox PR844 Yellow cone E ferox PR841 E ferox PR771 E ferox PR676 E ferox PR651 E eugene maraisii PR872 E eugene maraisii d1024 E dyerianus PR863 E dyerianus PR821 Levubu E dyerianus PR820 Levubu E dyerianus PR769 E dyerianus PR731 E dyerianus CC332 E dolomiticus CC343 E delucanus d1129 E cycadifolius PR683 E cycadifolius d1128 E cupidus PR767 E cupidus PR734 E cupidus PR691 E cupidus d1127 E concinnus PR890 E concinnus PR817 E cerinus PR859 E cerinus d1022 E caffer d1101 E bubalinus PR910 E bubalinus PR885 E bubalinus d1021 E brevifoliolatus xdk2 E brevifoliolatus Xdk1 E arenarius PR854 E arenarius PR758 Alexandria E arenarius d1020 E aplanatus PR757 E aplanatus PR682 E aplanatus d1126 E altensteinii PR856 E altensteinii PR756 E altensteinii d1019 E aemulans PR861 E aemulans d1018 E ferox d1025 Stangeria eriopus PR843 Stangeria eriopus PR842B Stangeria eriopus PR753 Stangeria eriopus PR706 Stangeria eriopus PR860 Figure 4.3 Bootstrap tree of the rbcLa region, bootstrap percentage values above branches. 124

Macrozamia plurinervia d1060 63 Macrozamia pauli-guilielmi d1096 Macrozamia macdonnellii d1057 Macrozamia communis d1051 64 E villosus PR816 E tegulaneus subsp. tegulaneus d1040 62 E inopinus PR864 E inopinus PR778 65 E hirsutus PR718 E hirsutus CC336 Lepidozamia peroffskyana d1050 Lepidozamia hopei d1049 E woodii PR875 E whitelockii PR818 E whitelockii d1048 E villosus PR838 E villosus PR837 E villosus PR671 E villosus d1047 E umbeluziensis PR858 E umbeluziensis d1046 E turneri d1044 E trispinosus PR868 E trispinosus PR680 E trispinosus d1043 E transvenosus PR832 Tsualo E transvenosus PR829 E transvenosus PR797 E transvenosus PR727 E transvenosus PR665 E transvenosus d1041 E tegulaneus subsp. tegulaneus PR877 E tegulaneus subsp. tegulaneus PR747 E sp PR913 E septentrionalis d1138 E senticosus PR833 E senticosus PR830 E senticosus PR812 E senticosus PR719 E senticosus PR663 E senticosus d1039 E sclavoi PR790 E sclavoi PR738 E sclavoi d1038 E schmitzii PR819 E schaijesii CC329 E pterogonus PR876 E pterogonus d1136 E princeps PR871 E princeps PR836 E princeps PR810 E princeps d1037 E poggei PR911 E poggei PR813 E paucidentatus PR849 E paucidentatus PR710 E paucidentatus d1036 E nubimontanus PR792 E nubimontanus PR704 E nubimontanus PR655 E nubimontanus CC328 E ngoyanus PR799 E ngoyanus PR717 E ngoyanus PR703 E ngoyanus d1135 E natalensis x E woodi PR842A E natalensis PR802 Kranskloof E natalensis d1035 E munchii PR855 E munchii PR737 E munchii d1134 E msinganus PR751 E msinganus PR701 E msinganus d1034 E middelburgensis PR827 Middelburg E middelburgensis PR726 E middelburgensis PRU586 E middelburgensis CC337 E manikensis PR903 Vumba E manikensis PR795 E manikensis PR697 E manikensis d1033 E macrostrobilus CC333 E longifolius PR873 Blunt Tip E longifolius PR809 Blunt Tip E longifolius PR808 E longifolius PR673 E longifolius d1032

Figure 4.4 (b) 125

Figure 4.4 (a)

79 E lehmannii d1031 E lehmannii PR835 E lehmannii PR780 E lehmannii PR661 E lebomboensis PR874 Piet Retief E lebomboensis PR831 Pongola E lebomboensis PR805 E lebomboensis PR800 Piet Retief E lebomboensis PR796 E lebomboensis PR698 Mananga E lebomboensis PR657 Piet Retief E lebomboensis d1030 E laurentianus PR789 E laurentianus d1029 E latifrons PR811 Trappers valley E latifrons PR806 Green Hills E lanatus PR828 E lanatus PRU587 E lanatus d1133 E laevifolius PR845 Kaapsehoop E laevifolius PR803 Mallalosha E laevifolius PR801 E laevifolius PR798 Havelock E laevifolius PR730 E laevifolius CC335 E kisambo PR823 E kisambo PR745 E kisambo d1132 E kanga PR907 E humilis PR712 E humilis CC340 E horridus PR847 E horridus PR846 E horridus PR777 E horridus d1028 E hildebrandtii PR824 E heenanii PR776 E heenanii PR775 E gratus PR774 E gratus d1026 E ghellinckii PR773 E ghellinckii CC331 E friderici-guilielmi PR853 E friderici-guilielmi PR772 98 E friderici-guilielmi PR733 E friderici guilielmi CC338 E ferox PR844 Yellow cone E ferox PR841 E ferox PR771 E ferox PR676 E ferox PR651 E eugene-maraisii PR872 E eugene-maraisii d1024 E dyerianus PR863 E dyerianus PR821 Levubu E dyerianus PR820 Levubu E dyerianus PR769 E dyerianus PR731 E dyerianus CC332 E dolomiticus CC343 E delucanus d1129 E cycadifolius PR683 E cycadifolius d1128 E cupidus PR767 E cupidus PR734 E cupidus PR691 E cupidus d1127 E concinnus PR890 E concinnus PR817 E cerinus PR859 E cerinus d1022 E caffer d1101 E bubalinus PR910 E bubalinus PR885 E bubalinus d1021 E brevifoliolatus xdk2 E brevifoliolatus Xdk1 E arenarius PR854 E arenarius PR758 Alexandria E arenarius d1020 E aplanatus PR757 E aplanatus PR682 E aplanatus d1126 E altensteinii PR856 E altensteinii PR756 E altensteinii d1019 E aemulans PR861 E aemulans d1018 E ferox d1025 Stangeria eriopus PR843 Stangeria eriopus PR842B Stangeria eriopus PR753 Stangeria eriopus PR706 Stangeria eriopus PR860 Figure 4.4 Bootstrap tree of the psbA-trnH region, bootstrap percentage values above branches. 126

78 E nubimontanus CC328 E eugene-maraisii PR872 E eugene-maraisii d1024 E nubimontanus PR792 E nubimontanus PR704 E nubimontanus PR655 E middelburgensis PR827 Middelburg E middelburgensis PR726 E middelburgensis PR586 Wilge River E middelburgensis CC337 51 E dyerianus PR863 E dyerianus PR821 Levubu E dyerianus PR769 E dyerianus PR731 E dyerianus CC332 E dolomiticus CC343 E cupidus PR767 E cupidus PR734 E cupidus PR691 E cupidus d1127 E whitelockii PR818 70 E whitelockii d1048 E septentrionalis d1138 E macrostrobilus CC333 66 E kanga PR907 E msinganus d1034 E bubalinus PR910 75 E bubalinus PR885 E bubalinus d1021 100 E tegulaneus subsp. tegulaneus PR877 85 E tegulaneus subsp. tegulaneus d1040 64 E schmitzii PR819 E delucanus d1129 62 E laurentianus PR789 93 E laurentianus d1029 E sp PR913 E schaijesii CC329 E poggei PR911 E poggei PR813 53 E hildebrandtii PR824 E gratus d1026 E tegulaneus subsp. tegulaneus PR747 69 E sclavoi PR790 E sclavoi d1038 62 E kisambo PR823 E kisambo PR745 66 E kisambo d1132 E sclavoi PR738 66 84 E turneri d1044 E gratus PR774 E pterogonus PR876 E pterogonus d1136 E munchii PR855 E munchii PR737 84 E munchii d1134 E manikensis PR903 Vumba E manikensis PR795 E manikensis PR697 E manikensis d1033 E concinnus PR890 E concinnus PR817 E villosus PR838 Broad Leaf E villosus PR837 E villosus PR816 90 E villosus PR671 E villosus d1047 E ngoyanus PR703 E aplanatus PR757 E aplanatus PR682 E ferox PR844 Yellow cone E ferox PR841 94 E ferox PR771 E ferox PR676 E ferox PR651 E ferox d1025 61 E umbeluziensis PR858 E cerinus PR859 E aplanatus d1126 98 E inopinus PR864 E inopinus PR778 100 E hirsutus PR718 E hirsutus CC336 E umbeluziensis d1046 E ngoyanus PR799 E ngoyanus PR717 E ngoyanus d1135 E dyerianus PR820 Levubu E caffer d1101 Figure 4.5 (b) 127 Figure 4.5 (a)

E transvenosus PR832 Tsualo E transvenosus PR829 63 E transvenosus PR797 E transvenosus PR727 55 E transvenosus PR665 E transvenosus d1041 E woodii PR875 E paucidentatus PR849 62 E paucidentatus PR710 E paucidentatus d1036 E heenanii PR776 E heenanii PR775 62 E natalensis PR802 Kranskloof E natalensis d1035 E altensteinii PR756 78 E lehmannii d1031 E lehmannii PR835 E trispinosus PR868 E trispinosus PR680 E trispinosus d1043 E senticosus PR833 E senticosus PR830 E senticosus PR812 E senticosus PR719 E senticosus PR663 E senticosus d1039 E princeps PR871 E princeps PR836 E princeps PR810 E princeps d1037 E natalensis x E woodii PR842A E msinganus PR751 52 E msinganus PR701 E longifolius PR873 Blunt Tip E longifolius PR809 Blunt Tip E longifolius PR808 Blue E longifolius PR673 100 E longifolius d1032 E lehmannii PR780 E lehmannii PR661 E lebomboensis PR874 Piet Retief E lebomboensis PR831 Pongola E lebomboensis PR805 E lebomboensis PR800 Piet Retief E lebomboensis PR796 E lebomboensis PR698 Mananga E lebomboensis PR657 Piet Retief E lebomboensis d1030 E latifrons PR811 Trappers valley E latifrons PR806 Green Hills E horridus PR847 E horridus PR846 E horridus PR777 E horridus d1028 E cerinus d1022 E arenarius PR854 E arenarius PR758 Alexandria E arenarius d1020 E altensteinii PR856 E altensteinii d1019 100 E aemulans PR861 E aemulans d1018 E laevifolius PR845 Kaapsehoop 82 E laevifolius PR803 Mallalosha E laevifolius PR730 E humilis CC340 54 E lanatus PR828 E lanatus PR587 97 E lanatus d1133 E laevifolius PR801 E laevifolius PR798 Havelock E laevifolius CC335 59 E humilis PR712 E brevifoliolatus Xdk2 53 E brevifoliolatus Xdk1 E friderici-guilielmi PR853 53 E friderici-guilielmi PR772 E friderici-guilielmi PR733 98 77 E friderici-guilielmi CC338 100 E cycadifolius PR683 E cycadifolius d1128 E ghellinckii PR773 E ghellinckii CC331 100 Lepidozamia peroffskyana d1050 Lepidozamia hopei d1049 Macrozamia macdonnellii d1057 Macrozamia pauli-guilielmi d1096 Macrozamia communis d1051 Macrozamia plurinervia d1060 Figure 4.5 Bootstrap tree of the nrITS1 region, bootstrap percentage values above branches. 128

60 E transvenosus PR829 E transvenosus PR727 66 E transvenosus PR832 Tsualo E transvenosus PR797 E transvenosus PR665 E transvenosus d1041 E paucidentatus PR849 64 E paucidentatus PR710 E paucidentatus d1036 E heenanii PR776 E heenanii PR775 71 E natalensis PR802 Kraanskloof E natalensis d1035 65 E longifolius PR873 Blunt Tip E longifolius d1032 62 E lehmannii d1031 E lehmannii PR835 E woodii PR875 E trispinosus PR868 E trispinosus PR680 E trispinosus d1043 E senticosus PR833 E senticosus PR830 E senticosus PR812 E senticosus PR719 E senticosus PR663 E senticosus d1039 E princeps PR871 E princeps PR836 E princeps PR810 E princeps d1037 52 E natalensis x E woodii PR842A E msinganus PR751 E msinganus PR701 E longifolius PR809 Blunt Tip E longifolius PR808 Blue E longifolius PR673 E lehmannii PR780 E lehmannii PR661 E lebomboensis PR874 Piet Retief E lebomboensis PR831 Pongola E lebomboensis PR805 E lebomboensis PR800 Piet Retief E lebomboensis PR796 E lebomboensis PR698 Mananga E lebomboensis PR657 Piet Retief E lebomboensis d1030 E latifrons PR811 Trappers valley E latifrons PR806 Green Hills E horridus PR847 E horridus PR846 E horridus PR777 E horridus d1028 E cerinus d1022 E arenarius PR854 E arenarius PR758 Alexandria E arenarius d1020 E altensteinii PR856 E altensteinii PR756 E altensteinii d1019 E aemulans PR861 E aemulans d1018 82 E nubimontanus CC328 E eugene-maraisii PR872 E eugene-maraisii d1024 E nubimontanus PR792 E nubimontanus PR704 E nubimontanus PR655 E middelburgensis PR827 Middelburg E middelburgensis PR726 E middelburgensis PR586 Wilge River E middelburgensis CC337 59 E dyerianus PR863 E dyerianus PR821 Levubu E dyerianus PR820 Levubu E dyerianus PR769 E dyerianus PR731 E dyerianus CC332 E dolomiticus CC343 E cupidus PR767 E cupidus PR734 E cupidus PR691 E cupidus d1127 F igure 4.6 (b) 129

Figure 4.6 (a) 58 E septentrionalis d1138 E macrostrobilus CC333 63 E whitelockii PR818 E whitelockii d1048 62 E kanga PR907 55 E msinganus d1034 E bubalinus d1021 52 71 E bubalinus PR910 E bubalinus PR885 100 E tegulaneus subsp. tegulaneus PR877 E tegulaneus subsp. tegulaneus d1040 74 60 E sp PR913 E schaijesii CC329 E poggei PR911 83 87 E schmitzii PR819 E delucanus d1129 60 E laurentianus PR789 E laurentianus d1029 E poggei PR813 66 E kisambo PR823 E kisambo PR745 E kisambo d1132 56 69 E hildebrandtii PR824 E gratus d1026 57 E tegulaneus subsp. tegulaneus PR747 E sclavoi PR790 64 E sclavoi d1038 E sclavoi PR738 86 E turneri d1044 E gratus PR774 E pterogonus PR876 E pterogonus d1136 E munchii PR855 E munchii PR737 81 E munchii d1134 E manikensis PR903 Vumba E manikensis PR795 E manikensis PR697 E manikensis d1033 E concinnus PR890 E concinnus PR817 83 E villosus PR837 E villosus d1047 E aplanatus PR757 82 E villosus PR838 Broad Leaf 56 E villosus PR816 E villosus PR671 E ngoyanus PR703 E aplanatus PR682 E ferox PR844 Yellow cone E ferox PR841 68 E ferox PR771 E ferox PR676 E ferox PR651 E ferox d1025 62 E umbeluziensis PR858 E cerinus PR859 E aplanatus d1126 100 E inopinus PR864 E inopinus PR778 100 E hirsutus PR718 E hirsutus CC336 E umbeluziensis d1046 E ngoyanus PR799 E ngoyanus PR717 E ngoyanus d1135 E caffer d1101 E laevifolius PR845 Kaapsehoop 83 E laevifolius PR803 Mallalosha E laevifolius PR730 E humilis CC340 62 E lanatus PR828 E lanatus PR587 97 E lanatus d1133 E laevifolius PR801 E laevifolius PR798 Havelock E laevifolius CC335 E humilis PR712 E brevifoliolatus Xdk2 54 E brevifoliolatus Xdk1 E friderici-guilielmi PR853 74 E friderici-guilielmi PR772 66 E friderici-guilielmi PR733 100 97 E friderici-guilielmi CC338 87 E cycadifolius PR683 E cycadifolius d1128 E ghellinckii PR773 E ghellinckii CC331 100 Macrozamia plurinervia d1060 57 100 Macrozamia communis d1051 Macrozamia pauli-guilielmi d1096 Macrozamia macdonnellii d1057 100 Lepidozamia peroffskyana d1050 Lepidozamia hopei d1049 Stangeria eriopus PR860 64 Stangeria eriopus PR842B Stangeria eriopus PR753 Stangeria eriopus PR706 Stangeria eriopus PR843 Figure 4.6 Bootstrap tree of the combined regions: matK, rbcLa and nrITS1, bootstrap percentage values shown above branches. 130

65 E sclavoi PR738 E sclavoi d1038 62 E princeps d1037 E tegulaneus subsp. tegulaneus PR747 E lehmannii PR835 E ghellinckii PR773 53 E friderici-guilielmi PR853 E friderici-guilielmi PR772 E friderici-guilielmi PR733 E friderici-guilielmi CC338 E ngoyanus PR799 65 E ngoyanus PR717 E ngoyanus PR703 E ngoyanus d1135 E longifolius PR808 59 E lehmannii d1031 E lehmannii PR661 E ferox d1025 61 E princeps PR871 E princeps PR810 E paucidentatus d1036 63 E paucidentatus PR849 E paucidentatus PR710 E heenanii PR775 88 E transvenosus PR829 E transvenosus PR727 64 E longifolius PR873 Blunt Tip E longifolius d1032 96 E inopinus PR864 E inopinus PR778 62 E hirsutus PR718 E hirsutus CC336 62 E cycadifolius PR683 E cycadifolius d1128 E woodii PR875 E whitelockii PR818 E whitelockii d1048 E villosus PR838 Broad Leaf E villosus PR837 E villosus PR816 E villosus PR671 E villosus d1047 E umbeluziensis PR858 E umbeluziensis d1046 E turneri d1044 E trispinosus PR868 E trispinosus PR680 E trispinosus d1043 E transvenosus PR832 Tsualo E transvenosus PR797 E transvenosus PR665 E transvenosus d1041 E tegulaneus subsp. tegulaneus PR877 E tegulaneus subsp. tegulaneus d1040 E sp PR913 E septentrionalis d1138 E senticosus PR833 E senticosus PR830 E senticosus PR812 E senticosus PR719 E senticosus PR663 E senticosus d1039 E sclavoi PR790 E schmitzii PR819 E schaijesii CC329 E pterogonus PR876 E pterogonus d1136 E princeps PR836 E poggei PR911 E poggei PR813 E nubimontanus PR792 E nubimontanus PR704 E nubimontanus PR655 E nubimontanus CC328 E natalensis x E woodii PR842A E natalensis PR802 Kraanskloof E natalensis d1035 E munchii PR855 E munchii PR737 E munchii d1134 E msinganus PR751 E msinganus PR701 E msinganus d1034 E middelburgensis PR827 Middelburg E middelburgensis PR726 E middelburgensis PR586 Wilge River E middelburgensis CC337

Figure 4.7 (b) 131 Figure 4.7 (a)

82 E manikensis PR903 Vumba E manikensis PR795 E manikensis PR697 E manikensis d1033 E macrostrobilus CC333 E longifolius PR809 Blunt Tip E longifolius PR673 E lehmannii PR780 E lebomboensis PR874 Piet Retief E lebomboensis PR831 Pongola E lebomboensis PR805 E lebomboensis PR800 Piet Retief E lebomboensis PR796 E lebomboensis PR698 Mananga E lebomboensis PR657 Piet Retief E lebomboensis d1030 E laurentianus PR789 E laurentianus d1029 E latifrons PR811 Trappers valley E latifrons PR806 Green Hills E lanatus PR828 E lanatus PR587 E lanatus d1133 E laevifolius PR845 Kaapsehoop E laevifolius PR803 Mallalosha E laevifolius PR801 E laevifolius PR798 Havelock E laevifolius PR730 E laevifolius CC335 E kisambo PR823 E kisambo PR745 E kisambo d1132 E kanga PR907 E humilis PR712 E humilis CC340 E horridus PR847 E horridus PR846 E horridus PR777 E horridus d1028 E hildebrandtii PR824 E heenanii PR776 71 E gratus PR774 E gratus d1026 E ghellinckii CC331 E ferox PR844 Yellow cone E ferox PR841 E ferox PR771 E ferox PR676 E ferox PR651 E eugene-maraisii PR872 E eugene-maraisii d1024 E dyerianus PR863 E dyerianus PR821 Levubu E dyerianus PR820 Levubu E dyerianus PR769 E dyerianus PR731 E dyerianus CC332 E dolomiticus CC343 E delucanus d1129 E cupidus PR767 E cupidus PR734 E cupidus PR691 E cupidus d1127 E concinnus PR890 100 E concinnus PR817 E cerinus PR859 E cerinus d1022 E caffer d1101 E bubalinus PR910 E bubalinus PR885 E bubalinus d1021 E brevifoliolatus Xdk2 E brevifoliolatus Xdk1 E arenarius PR854 E arenarius PR758 Alexandria E arenarius d1020 E aplanatus PR757 E aplanatus PR682 E aplanatus d1126 E altensteinii PR856 E altensteinii PR756 E altensteinii d1019 E aemulans PR861 E aemulans d1018 100 Lepidozamia peroffskyana d1050 Lepidozamia hopei d1049 Macrozamia plurinervia d1060 99 Macrozamia pauli-guilielmi d1096 Macrozamia macdonnellii d1057 Macrozamia communis d1051 Stangeria eriopus PR860 62 Stangeria eriopus PR842B Stangeria eriopus PR753 Stangeria eriopus PR706 Stangeria eriopus PR843 Figure 4.7 Bootstrap tree of the gene combination rbcLa, matK, and psbA-trnH, bootstrap percentage values shown above branches. 132

E transvenosus PR665 E sp PR913 E schaijesii CC329 E poggei PR911 E nubimontanus PR792 E longifolius PR673 E horridus PR847 E horridus PR846 E horridus d1028 52 E gratus d1026 E dyerianus PR863 E dyerianus PR821 Levubu E dyerianus PR820 Levubu E dyerianus PR769 E dyerianus PR731 E dyerianus CC332 E dolomiticus CC343 E cupidus PR767 E cupidus PR734 E cupidus PR691 E cupidus d1127 E woodii PR875 E villosus PR837 E villosus d1047 66 E natalensis x E woodii PR842A E natalensis PR802 Kranskloof E natalensis d1035 E msinganus PR751 E msinganus PR701 E aplanatus PR757 E ngoyanus PR799 63 E ngoyanus PR717 E ngoyanus PR703 E ngoyanus d1135 62 E kisambo PR823 E kisambo PR745 E kisambo d1132 64 E villosus PR816 E tegulaneus subsp. tegulaneus d1040 62 E transvenosus PR829 E transvenosus PR727 63 E longifolius PR873 Blunt Tip E longifolius d1032 95 E inopinus PR864 E inopinus PR778 62 E hirsutus PR718 E hirsutus CC336 85 E cycadifolius PR683 E cycadifolius d1128 E whitelockii PR818 E whitelockii d1048 E villosus PR838 Broad Leaf E villosus PR671 E umbeluziensis PR858 E umbeluziensis d1046 E turneri d1044 E trispinosus PR868 E trispinosus PR680 E trispinosus d1043 E transvenosus PR832 Tsualo E transvenosus PR797 E transvenosus d1041 E tegulaneus subsp. tegulaneus PR877 E tegulaneus subsp. tegulaneus PR747 E septentrionalis d1138 E senticosus PR833 E senticosus PR830 E senticosus PR812 E senticosus PR719 E senticosus PR663 E senticosus d1039 E sclavoi PR790 E sclavoi PR738 E sclavoi d1038 E schmitzii PR819 E pterogonus PR876 E pterogonus d1136 E princeps PR871 E princeps PR836 E princeps PR810 E princeps d1037 E poggei PR813 E paucidentatus PR849 E paucidentatus PR710 E paucidentatus d1036 E nubimontanus PR704 E nubimontanus PR655 E nubimontanus CC328

F igure 4.8 (b) 133

Figure 4.8 (a)

E munchii PR855 50 E munchii PR737 E munchii d1134 E msinganus d1034 E middelburgensis PR827 Middelburg E middelburgensis PR726 E middelburgensis PR586 Wilge River E middelburgensis CC337 E manikensis PR903 Vumba E manikensis PR795 E manikensis PR697 E manikensis d1033 E macrostrobilus CC333 E longifolius PR809 Blunt Tip E longifolius PR808 E lehmannii d1031 E lehmannii PR835 E lehmannii PR780 E lehmannii PR661 E lebomboensis PR874 Piet Retief E lebomboensis PR831 Pongola E lebomboensis PR805 E lebomboensis PR800 Piet Retief E lebomboensis PR796 E lebomboensis PR698 Mananga E lebomboensis PR657 Piet Retief E lebomboensis d1030 E laurentianus PR789 E laurentianus d1029 E latifrons PR811 Trappers valley E latifrons PR806 Green Hills E lanatus PR828 E lanatus PR587 E lanatus d1133 E laevifolius PR845 Kaapsehoop E laevifolius PR803 Mallalosha E laevifolius PR801 E laevifolius PR798 Havelock E laevifolius PR730 E laevifolius CC335 E kanga PR907 E humilis PR712 E humilis CC340 E horridus PR777 E hildebrandtii PR824 100 E heenanii PR776 E heenanii PR775 E gratus PR774 E ghellinckii PR773 E ghellinckii CC331 E friderici-guilielmi PR853 E friderici-guilielmi PR772 E friderici-guilielmi PR733 E friderici-guilielmi CC338 E ferox PR844 Yellow cone E ferox PR841 E ferox PR771 E ferox PR676 E ferox PR651 E ferox d1025 E eugene-maraisii PR872 E eugene-maraisii d1024 E delucanus d1129 E concinnus PR890 E concinnus PR817 E cerinus PR859 E cerinus d1022 E caffer d1101 E bubalinus PR910 E bubalinus PR885 E bubalinus d1021 E brevifoliolatus Xdk2 E brevifoliolatus Xdk1 E arenarius PR854 E arenarius PR758 Alexandria E arenarius d1020 E aplanatus PR682 E aplanatus d1126 E altensteinii PR856 E altensteinii PR756 E altensteinii d1019 E aemulans PR861 E aemulans d1018 Stangeria eriopus PR860 64 Stangeria eriopus PR842B 100 Stangeria eriopus PR753 Stangeria eriopus PR706 Stangeria eriopus PR843 99 Lepidozamia peroffskyana d1050 Lepidozamia hopei d1049 Macrozamia plurinervia d1060 Macrozamia pauli-guilielmi d1096 Macrozamia macdonnellii d1057 Macrozamia communis d1051 Figure 4.8 Bootstrap tree of the core barcodes (matK and rbcLa), bootstrap percentage values shown above branches. 134

E laevifolius PR845 Kaapsehoop 82 E laevifolius PR803 Mallalosha E laevifolius PR730 E humilis CC340 55 E lanatus PR828 E lanatus PR587 95 E lanatus d1133 E laevifolius PR801 E laevifolius PR798 Havelock E laevifolius CC335 E humilis PR712 E brevifoliolatus Xdk2 57 E brevifoliolatus Xdk1 E friderici guilielmi PR853 75 E friderici guilielmi PR772 58 E friderici guilielmi PR733 95 E friderici guilielmi CC338 86 E cycadifolius PR683 E cycadifolius d1128 E ghellinckii PR773 68 E ghellinckii CC331 100 Macrozamia plurinervia d1060 62 100 Macrozamia communis d1051 Macrozamia pauli-guilielmi d1096 76 Macrozamia macdonnellii d1057 100 Lepidozamia peroffskyana d1050 Lepidozamia hopei d1049 E nubimontanus PR792 E dyerianus PR863 E dyerianus PR821 Levubu E dyerianus PR820 Levubu E dyerianus PR769 53 E dyerianus PR731 E dyerianus CC332 E dolomiticus CC343 E cupidus PR767 E cupidus PR734 E cupidus PR691 E cupidus d1127 61 63 E nubimontanus CC328 74 E eugene maraisii PR872 E eugene maraisii d1024 E nubimontanus PR704 E nubimontanus PR655 E middelburgensis PR827 Middelburg E middelburgensis PR726 E middelburgensis PR586 Wilge River E middelburgensis CC337 62 E whitelockii PR818 E kanga PR907 50 E whitelockii d1048 E septentrionalis d1138 E macrostrobilus CC333 100 E tegulaneus subsp. tegulaneus PR877 58 E tegulaneus subsp. tegulaneus d1040 E msinganus d1034 E bubalinus PR910 E bubalinus PR885 62 E bubalinus d1021 63 E sp PR913 62 E schaijesii CC329 E poggei PR911 E poggei PR813 90 87 E schmitzii PR819 E delucanus d1129 60 E laurentianus PR789 E laurentianus d1029 65 E kisambo PR745 59 E kisambo d1132 E kisambo PR823 59 E hildebrandtii PR824 75 E gratus d1026 E tegulaneus subsp. tegulaneus PR747 68 E sclavoi PR790 E sclavoi PR738 E sclavoi d1038 87 E turneri d1044 E gratus PR774 E pterogonus PR876 E pterogonus d1136 E munchii PR855 E munchii PR737 83 E munchii d1134 E manikensis PR903 Vumba E manikensis PR795 E manikensis PR697 E manikensis d1033 E concinnus PR890 E concinnus PR817 58 E villosus PR837 82 E aplanatus PR757 E villosus d1047 81 E villosus PR838 Broad Leaf E villosus PR816 E villosus PR671 E ngoyanus PR703 F igure 4.9 (a) E aplanatus PR682 135

Figure 4.9 (b)

100 60 E transvenosus PR832 Tsualo E transvenosus PR797 72 E transvenosus PR665 86 E transvenosus PR829 E transvenosus PR727 E transvenosus d1041 E ferox PR844 Yellow cone E ferox PR841 87 E ferox PR771 E ferox PR676 E ferox PR651 E ferox d1025 E longifolius PR673 51 E horridus PR847 E horridus PR846 E horridus d1028 62 E umbeluziensis PR858 E cerinus PR859 E aplanatus d1126 64 E paucidentatus PR849 E paucidentatus PR710 E heenanii PR775 65 E longifolius PR873 Blunt Tip E longifolius d1032 56 E lehmannii d1031 E lehmannii PR835 100 E inopinus PR864 E inopinus PR778 100 E hirsutus PR718 E hirsutus CC336 E woodii PR875 E umbeluziensis d1046 E trispinosus PR868 E trispinosus PR680 E trispinosus d1043 E senticosus PR833 E senticosus PR830 E senticosus PR812 E senticosus PR719 E senticosus PR663 65 E senticosus d1039 E princeps PR871 E princeps PR836 E princeps PR810 E princeps d1037 E paucidentatus d1036 E ngoyanus PR799 E ngoyanus PR717 E ngoyanus d1135 E natalensis x E woodii PR842A E natalensis PR802 Kranskloof E natalensis d1035 E msinganus PR751 E msinganus PR701 E longifolius PR809 Blunt Tip E longifolius PR808 Blue E lehmannii PR780 E lehmannii PR661 E lebomboensis PR874 Piet Retief E lebomboensis PR831 Pongola E lebomboensis PR805 E lebomboensis PR800 Piet Retief E lebomboensis PR796 E lebomboensis PR698 Mananga E lebomboensis PR657 Piet Retief E lebomboensis d1030 E latifrons PR811 Trappers valley E latifrons PR806 Green Hills E horridus PR777 E heenanii PR776 E cerinus d1022 E caffer d1101 E arenarius PR854 E arenarius PR758 Alexandria E arenarius d1020 E altensteinii PR856 E altensteinii PR756 E altensteinii d1019 E aemulans PR861 E aemulans d1018 Stangeria eriopus PR843 Stangeria eriopus PR842B Stangeria eriopus PR753 Stangeria eriopus PR706 Stangeria eriopus PR860 Figure 4.9 Bootstrap tree of the combined plastid and nuclear genes, bootstrap percentage values shown above branches. 136

CHAPTER 5. GENERAL CONCLUSIONS

A substantial amount of reference material has been sourced, much of which is from documented wild origin, and all of which is freely available from the ACDB DNA Bank. These were also supplemented with collections done by the New York Botanical Garden on wild accessioned material. In terms of the general aims of the project only a single taxon could not be sourced while a further two could not be sequenced for inclusion in the analysis. Generally the study has seen several obstacles from PCR failures using standard primers to a lack of resolution even when employing multiple genes found to be highly variable in other plant groups (CBOL Plant Working Group 2009).

Hypothesis 1 stating morphological and geographic groups will be supported by the molecular results has been refuted in most cases with only 3/18 groups as proposed by Vorster (2004a) confirmed. Departures were often only due to single species shifts which were tentative inclusions. The current molecular phylogeny provides DNA sequence evidence for the placement of several taxa for the first time and is both the most completely sampled and resolved molecular phylogenetic hypothesis produced thus far. Importantly three major clades are resolved which may serve as a sub-generic or sectional classification. When currently resolved groups were compared to other datasets and taxonomically important characters they were found to be generally well supported and sensible. However, in many cases little other data are available, which hinders robust characterisation of lineages. A major concern is the continued lack of resolution at the species level due to a lack of genetic variation, even though some of the most variable genes currently in use among plants were employed. This lack of species differentiation may alternatively be as a result of over-splitting in the genus, a condition which calls for taxonomic attention.

Hypothesis 2 stating that Stangeria eriopus form will have sequence difference has also been refuted. Very little sequence divergence was retrieved from any Stangeria samples, though the divergence of a single specimen may indicate poor sampling. Wild collected material must be employed for the taxonomic questions to be finally addressed. Also the application of the more promising nuclear genetic regions should be considered before finality can be reached.

Hypothesis 3 stating proposed DNA barcodes will adhere to CBOL’s standards in the Africa’s endemic cycads has been confirmed as the proposed barcodes did not meet the CBOL Plant Working Group's (2009) standards. Sequencing was difficult in the case of nrITS, matK and to some extent psbA-trnH, with specific primers needed in the former two cases. Alignment, though uncomplicated within Encephalartos, was challenging between cycad genera and will no doubt be even more so outside the Cycadales. The region psbA-trnH was found to contain genetic

137 anomalies such as multiple copies and palindromes which would immediately exclude the region as a possible barcode for Encephalartos and probably all other cycad genera. Finally the resolution within the genus is poor with a maximum of 20% of species resolvable even when a multi-locus barcode of all four regions was used. This may be somewhat of an underestimate based on the tree building method employed as opposed to character recognition methods (Little 2011), these should be employed to determine the true genetic resolvability of species using the proposed barcodes. Despite this a barcoding library has been constructed for the genus including 61 of the 67 recognised taxa (including sub-species) applying all four proposed regions. This will provide identification for at least some species, a significant improvement in many respects, for example with completely untrained law enforcement officers or otherwise wholly unidentifiable material. When a species-specific identification cannot be achieved, at least the choice will be narrowed to a few usually closely related and locality bound species whereafter other identification means could be employed (see Cousins et al. 2011) to further narrow down possibilities. Additional regions should be investigated in the genus to try and increase the resolution power of DNA barcoding. The more promising regions in this regard seem to be single copy and other nuclear regions which are being investigated at the time of writing, e.g. NEEDLY has 9.84% parsimoniously informative characters (118/1199 characters; Rousseau, Little, Van der Bank unpublished data).

Hypothesis 4 stating that additional regions will provide the a species specific barcode has been refuted as the failure of previously suggested and other promising regions at either amplification or sequencing stages, and the lack of genetic variation would make these poor selections as barcodes. As some regions could not be successfully sequenced these may still provide higher levels of genetic variation, e.g. ETS which is often comparable to ITS (Van der Bank pers. comm.3). However at the time of writing the region found to be highly promising in other cycads, psbK-psbI (Nicolalde-Morejo et al. 2010, 2011), has been sequenced but revealed low genetic variation with 3.76% bases parsimoniously informative (38/1010 characters; Rousseau, Little, Van der Bank unpublished data).

Finally several issues beyond the scope of the project have been identified ranging from the prevalence of hybrids in cultivation and amongst material of wild origin, the non-monophyly of several species in contrast to the lack of separation of several closely related taxa, and the gaps in knowledge surrounding synapomorphic characters of many lineages. These issues need formal and holistic taxonomic investigation, which will eventually culminate in a monograph and revision of the genus which has been outstanding for several decades now (Osborne et al. 1988).

3 Prof. Michelle van der Bank, CBOL Plant Working Group. African Center for DNA barcoding, Department Botany and Plant Biotechnology, University of Johannesburg. 138

CHAPTER 6. REFERENCES

ADAMS, P.A. PANDEY, N. REZZI, S. & CASANOVA, J. 2002. Geographic variation in the Random Amplified Polymorphic DNAs (RAPDs) of Juniperus phoenicea, J. p. var. canariensis, J. p. subsp. eumediterranea, and J. p. var. turbinata. Biochemical Systematic Ecology 30: 223–229.

ALVAREZ, I. & WENDEL, J.F. 2003. Ribosomal ITS sequences and plant phylogenetic inference. Molecular Phylogenetics and Evolution 29: 417–434. ARTABE, A.E. & STEVENSON, D.W. 1999. Fossil Cycadales of Argentina. The Botanical Review 65: 219–238. BAFEEL, S.O. ARIF, I.A. BAKIR, M.A. KHAN, H.A. AL FARHAN, A.H. AL HOMAIDAN, A.A. AHAMED, A. & THOMAS, J. 2011. Comparative evaluation of PCR success with universal primers of maturase K (matK) and ribulose-1, 5-bisphosphate carboxylase oxygenase large subunit (rbcL) for barcoding of some arid plants. Plantomics Journal 4: 195–198. BAILLON, H.E. 1892. Histoire des plantes 12: 68. L. Hachette & Co. Paris, France. BAYLISS, J. BURROW, C. MARTELL, S. & STAUDE, H. 2010. An ecological study of the relationship between two living fossils in Malawi: the Mulanje Tiger Moth (Callioratis grandis) and the Mulanje Cycad (Encephalartos gratus). African Journal of Ecology 48: 472–480. BOGLER, D.J. & FRANCISCO-ORTEGA, J. 2004. Molecular systematic studies in cycads: Evidence from trnL intron and ITS2 rDNA sequences. The Botanical Review 70: 260–273. BOWE, L.M. COAT, G. & DEPAMPHILIS, C.W. 2000. Phylogeny of seed plants based on all three genomic compartments: Extant gymnosperms are monophyletic and Gnetales' closest relatives are conifers. Proceedings of the National Academy of Sciences of the United States of America 97: 4092–4097. BRENNER, E.D. STEVENSON, D.W. & TWIGG, R.W. 2003. Cycads: evolutionary innovations and the role of plant-derived neurotoxins. Trends in Plant Science 8: 446–452. BRIQUET, J. & RENDELL, A.B. (eds). 1935. International Rules of Botanical Nomenclature, edn 3. Jena. BULT, C.J. & ZIMMER, E.A. 1993. Nuclear ribosomal RNA sequences for inferring tribal relationships in Onagraceae. Systematic Botany 18: 48–63. BURGESS, K.S. FAZEKAS, A.J. KESANAKURTI, P.R. GRAHAM, S.W. HUSBAND, B.C. NEWMASTER, S.G. PERCY, D.M. HAJIBABAEI, M. & BARRETT, S.C.H. 2011. Discriminating plant species in a local temperate flora using the rbcL + matK DNA barcode. Methods in Ecology and Evolution 2: 333–340. CAFASSO, D. COZZOLINO, S. CAPUTO, P. & DE LUCA, P. 2001. Maternal inheritance of plastids in Encephalartos Lehm. (Zamiaceae, Cycadales). Genome 44: 239–241. CAPELLA, C. CHASSOT, C. & OSBORNE, R. 2003. Notes on the cycads of Mozambique. Encephalartos 76: 13–17.

139

CAPUTO, P. COZZOLINO, S. DE LUCA, P. MORETTI, A. & STEVENSON, D. 2004. Molecular phylogeny of Zamia (Zamiaceae). In: WALTERS, T. & OSBORNE, R. (eds.). Cycads classification concepts and recommendations. Wallingford: CABI Publishing. CBOL PLANT WORKING GROUP. 2009. A DNA barcode for land plants. Proceedings of the National Academy of Sciences USA 106: 12794–12797. CHAIPRASONGSUK, M. MINGMUANG, M. THONGPAN, A. & NAMWONGPROM, K. 2007. Molecular identification of Encephalartos (Zamiaceae) species and their relationships to morphological characters. Kasetsart Journal (Natural Science) 41: 43–60. CHAMBERLAIN, C.J. 1915. A phylogenetic study of cycads. Proceedings of the National Academy of Sciences of the United States of America 1: 86–90. CHAMBERLAIN, C.J. 1935. Gymnosperms: structure and evolution. University of Chicago Press, Chicago and Illinois. CHASE, M.W. COWAN, R.S. HOLLINGSWORTH, P.M. VAN DEN BERG, C. MADRIÑÁN, S. PETERSEN, G. SEBERG, O. JØRGSENSEN, T. CAMERON, K.M. CARINE, M. PEDERSEN, N. HEDDERSON, T.A.J. CONRAD, F. SALAZAR, G.A. RICHARDSON, J.E. HOLLINGSWORTH, M.L. BARRACLOUGH, T.G. KELLY, L. & WILKINSON, M. 2007. A proposal for a standardised protocol to barcode all land plants. Taxon 56: 295–299. CHAW, S. ZHARKIKH, A. SUNG, H. LUU, T. & LI–L, W. 1997. Molecular phylogeny of extant gymnosperms and seed plant evolution: analysis of nuclear 18s rRNA sequences. Biology and Evolution 14: 56–68. CHAW, S.M. PARKINSON, C.L. CHENG, Y. VINCENT, T.M. & PALMER, J. 2000. Seed plant phylogeny inferred from all three genomes: monophyly of extant gymnosperms and origin of Gnetales from conifers. Proceedings of the National Academy of Sciences USA 97: 4086– 4091. CHAW, S.M. WALTERS, T.W. CHANG, C.C. HU, S.H. & CHEN, S.H. 2005. A phylogeny of cycads (Cycadales) inferred from chloroplast matK gene, trnK intron and nuclear rDNA ITS region. Molecular Phylogenetics and Evolution 37: 214–234. CHRISTENHUSZ, M.J.M. REVEAL, J.L. FARJON, A. GARDNER, M.F. MILL, R.P. & CHASE, M.W. 2011. A new classification and linear sequence of extant gymnosperms. Phytotaxa 19: 55–70. COETZER, M.E. 2000. A molecular systematic study of the genus Encephalartos. Unpublished Msc. Thesis, University of the Orange Free State. COOPER, M.R. & GOODE, D. 2004. The cycads & cycads moths of KZN. Peroniceras Press, 7 Ridge Road, 3610 New Germany. COUSINS, S. 2009. An ethno-ecological assessment of Encephalartos spp. in the medicinal plant trade on the eastern seaboard of South Africa. Unpublished Msc. Thesis. University of the Witwatersrand. COUSINS, S.R. WILLIAMS, V.L. & WITKOWSKI, E.T.F. 2011. Uncovering the cycad taxa

140

(Encephalartos species) traded for traditional medicine in Johannesburg and Durban. South African Journal of Botany 78: 129–138. COX, A.V. BENNETT, M.D. & DYER, T.A. 1992. Use of the polymerase chain reaction to detect spacer size heterogeneity in plant 5S-rRNA gene clusters and to locate such clusters in wheat (Triticum aestivum L.). Theoretical and Applied Genetics 83: 684–690. CRANE, P.R. 1988. Major clades and relationships in the higher gymnosperms. In: BECK, C.B. (ed.). Origin and evolution of the gymnosperms. Columbia University Press, New York. DEHGAN, B. & DEHGAN, N.B. 1988. Comparative pollen morphology and taxonomic affinities in Cycadales. American Journal of Botany 75: 1501–1516. DE LAUBENFELS, D.J. 1999. The families of the Cycadaceae. Encephalartos 59: 7–9. DESALLE, R. EGAN, M.G. & SIDALL, M. 2005. The unholy trinity: taxonomy, species delimitation and DNA barcoding. Philosophical Transactions of the Royal Society B: Biological Sciences 360: 1905–1916. DONALDSON, J.S. (ed). 2003. Cycads—status survey and conservation action plan. IUCN/SSC Cycad Specialist Group. IUCN, Gland and Cambridge. DONALDSON, J.S. & BÖSENBERG, J.D. 1999. Changes in the abundance of South African cycads during the 20th century: preliminary data from the study of matched photographs. In CHEN, C-J. (ed.). Biology and Conservation of Cycads—Proceedings of the 4th International Conference on Cycad Biology Panzhihua, China. Academic Publishers, Beijing. DONALDSON, J.S. 1993a. Insect predation of ovules in the South African species of Encephalartos (Cycadales: Zamiaceae). In: STEVENSON, D.W. & NORSTOG, K.J. (eds.). Proceedings of Cycad 90, the Second International Conference on Cycad Biology. Palm and Cycad Societies of Australia, Milton, Queensland. DONALDSON, J.S. 1993b. Local extinction of insects associated with cycad strobili: an evaluation of possible causes and their consequences for cycad conservation (Abstract). In: VORSTER, P. (ed.). Proceedings of the Third International Conference on Cycad Biology. Cycad society of South Africa, Stellenbosch. DONALDSON, J.S. 1997. Is there a floral parasite mutualism in cycad pollination? The pollination biology of Encephalartos villosus (Zamiaceae). American Journal of Botany 84: 1398–1406. DONALDSON, J.S. 2008. South African Encephalartos species. NDF Workshop Case Studies: Case Study 4: Encephalartos. Mexico. DONALDSON, J.S. NANNI, I. & BÖSENBERG, J.D. 1995. The role of insects in pollination of the African cycad Encephalartos cycadifolius. In: VORSTER, P. (ed.). Proceedings of the Third International Conference on Cycad Biology. Cycad Society of South Africa, Stellenbosch. DOUZERY, E.J.P. PRIDGEON, A.M. KORES, P. LINDER, H.P. KURZWEIL, H. & CHASE, M.W. 1999. Molecular phylogenetics of Diseae (Orchidaceae): a constribution from nuclear ribosomal ITS sequences. American Journal of Botany 86: 887–899. DOWNIE, D.A. DONALDSON, J.S. & OBERPRIELER, R.G. 2008. Molecular systematics and

141

evolution in an African cycad-weevil interaction: Amorphocerini (Coleoptera: Curculionidae: Molytinae) weevils on Encephalartos. Molecular Phylogenetics and Evolution 47: 102–116. DOYLE, J.A. & DONOGHUE, M.J. 1992. Fossils and seed plant phylogeny reanalyzed. Brittonia 44: 89–106. DOYLE, J.J. & DOYLE, J.L. 1987. A rapid DNA isolation procedure for small quantities of fresh leaf tissue. Phytochemical Bulletin 19: 11–15. DYER, R.A. & VERDOORN, I. 1969. Encephalartos manikensis and its near allies. Kirkia 7: 147– 58. DYER, R.A. 1965. The cycads of southern Africa. Bothalia 8: 405–515. DYER, R.A. 1972. A new species of Encephalartos from Swaziland. Bothalia 10: 539–546. EDWARDS, D. HORN, A. TAYLOR, D. SAVOLAINEN, V. & HAWKINS, J.A. 2008. DNA barcoding of a large genus, Aspalathus L. (Fabaceae). Taxon 57: 1317–1327. EKUÉ, M.R.M. GAILING, O. HÖLSCHER, D. SINSIN, B. & FINKELDEY, R. 2008. Population genetics of the cycad Encephalartos barteri ssp. barteri (Zamiaceae) in Benin with notes on leaflet morphology and implications for conservation. Belgian Journal of Botany 141: 78–94. EZEMVELO KZN WILDLIFE. 2004. A management plan for cycads in KwaZulu–Natal. Available online: http: //www.cycadsg.org/publications/KwaZulu–Natal–Cycad–Management–Plan– 2004.pdf. FARRINGTON, L. MACGILLIVRAY, P. FAAST, R. & AUSTIN, A. 2009. Investigating DNA barcoding options for the identification of Caladenia (Orchidaceae) species. Australian Journal of Botany 57: 276–286. FAZEKAS, A.J. BURGESS, K.S. KESANAKURTI, P.R. GRAHAM, S.W. NEWMASTER, S.G. HUSBAN, B.C. PERCY, D.M. HAJIBABAEI, M. & BARRETT, S.C.H. 2008. Multiple multilocus DNA barcodes from the plastid genome discriminate plant species equally well. PLoS One 7: e2802. doi: 10.1371/journal.pone.0002802. FAZEKAS, A.J. KESANAKURTI, P.R. BURGESS, K.S. PERCY, D.M. GRAHAM, S.W. BARRETT, S.C.H. NEWMASTER, S.G. HAJIBABAEI, M. & HUSBAND, B.C. 2009. Are plant species inherently harder to discriminate than animal species using DNA barcoding markers?. Molecular Ecology Resources 9: 130–139. FELSENSTEIN, J. 1985. Confidence limits on phylogenies: an approach using the bootstrap. Evolution 39: 783–791. FISHER, J.B. & MARLER, T.E. 2006. Eccentric growth but no compression wood in a horizontal stem of Cycas micronesica (Cycadales). IAWA Journal 27: 377–382. FORD, C.S. AYRES, K.L. TOOMEY, N. HAIDER, N. VAN ALPHEN STAHL, J. KELLY, L.J. WIKSTROM, N. HOLLINGSWORTH, P.M. DUFF, R.J. HOOT, S.B. COWAN, R.S. CHASE, M.W. & WILKINSON, M.J. 2009. Selection of candidate coding DNA barcoding regions for use on land plants. Botanical Journal of the Linnean Society 159: 1–11. FORSTER, P.I. MACHIN, P.J. MOUND, L. & WILSON, G.W. 1994. Insects associated with

142

reproductive structures of cycads in Queensland and northeast New South Wales, Australia. Biotropica 26: 217–222. GAO, T. YAO, H. SONG, J.Y. LIU, C. ZHU, Y.J. MA, X.Y. PANG, X.H. XU, H.X. & CHEN, S.L. 2010. Identification of medicinal plants in the family Fabaceae using a potential DNA barcode ITS2. Journal of Ethnopharmacology 130: 116–121. GIDDY, C. 1980. Cycads of South Africa. 2nd edition, 3rd impression. Purnell, Cape Town. GILLILAND, H.B. 1939. A note on the Rhodesian cycad. Proceedings of the Rhodesian Science Association 37: 133–134. GOLDING, J.S. & HURTER, P.J.H. 2003. A Red List account of Africa’s cycads and implications of considering life-history and threats. Biodiversity and Conservation 12: 507–528. GONZÁLEZ-ASTORGA, J. VOVIDES, A.P. CRUZ-ANGON, A. OCTAVIO-AGUILAR, P. & IGLESIAS, C. 2005. Allozyme variation in the three extant populations of the narrowly endemic cycad Dioon angustifolium Miq. (Zamiaceae) from north–eastern Mexico. Annals of Botany 95: 999–1007. GONZÁLEZ-ASTORGA, J. VOVIDES, A.P. FERRER, M.M. & IGLESIAS, C. 2003. Population genetics of Dioon edule Lindl. (Zamiaceae, Cycadales): biogeographical and evolutionary implications. Biological Journal of the Linnean Society 80: 457–467. GONZALEZ, D. & VOVIDES, A.P. 2002. Low intralineage divergence in Ceratozamia (Zamiaceae) detected with nuclear ribosomal DNA ITS and chloroplast DNA trnL–F non–coding region. Systematic Botany 27: 654–661. GONZÁLEZ, D. VOVIDES, A.P. & BÁRCENAS, C. 2008. Phylogenetic relationships of the neotropical genus Dioon (Cycadales, Zamiaceae) based on nuclear and chloroplast DNA sequence data. Systematic Botany 33: 229–236. GOODE, D. 1989. Cycads of Africa. Struik, Cape Town. GOODE, D. 2001. Cycads of Africa Volume I. D.E. Cycads of Africa Publishers, Gallo-manor, South Africa. GORELICK, R. & OLSON, K. 2011. Is lack of cycad (Cycadales) diversity a result of a lack of polyploidy? Botanical Journal of the Linnean Society 165: 156–167. GROBBELAAR, N. 1989. Disintegration of Encephalartos megastrobili. South African Journal of Botany 55: 581–585. GROBBELAAR, N. 1993 Senescence patterns in the female cones of Encephalartos. In: STEVENSON, D.W. & NORSTOG, K.J. (eds.). The Biology, Structure and Systematics of the Cycadales. Proceedings of CYCAD 90, the Second International Conference on Cycad Biology. Milton, Queensland, Palm and Cycad Societies of Australia Ltd. GROBBELAAR, N. 1996. Encephalartos aplanatus. Encephalartos 46: 28. GROBBELAAR, N. 2002. Cycads—with special reference to the southern African species. Published by the author, Pretoria, Gauteng, South Africa. HAJIBABAEI, M. SINGER, G.A.C. HEBERT, P.D.N. & HICKEY, D.A. 2007. DNA barcoding: how it

143

complements taxonomy, molecular phylogenetics and population genetics. Trends in Genetics: 23: 167-172. HALL, J.A. WALTER, G.H. BERGTROM, D.M. & MACHIN, P.J. 2004. Pollination ecology of the Australian cycad Lepidozamia peroffskyana (Zamiaceae). Australian Journal of Botany 52: 333–343. HAN, J.P. SHI, L.C. YAO, H. SONG, J.Y. XU, H.X. SUN, C. XIE, C.X. & CHEN, S.L. 2009. Testing potential DNA barcoding regions in the Labiatae medicinal plants. Planta Medica 75: 407–407. HAO, D.C. CHEN, S.L. & XIAO, P.G. 2010. Sequence characteristics and divergent evolution of the chloroplast psbA-trnH noncoding region in gymnosperms. Journal of Applied Genetics 51: 259–273. HAO, D.C. HUANG, B. CHEN, S.L. & MU, J. 2009. Evolution of the chloroplast trnL–trnF region in the gymnosperm lineages Taxaceae and Cephalotaxaceae. Biochemical Genetics 47: 351– 369. HAYNES, J.L. 2011. World List of Cycads: A historical review. IUCN/SSC Cycad Specialist Group. Available online: http: //www.cycadsg.org/publications/Haynes–Historical–Review–of–World– List–of–Cycads–2011.pdf. HEBERT, P.D.N. CYWINSKA, N.A. BALL, S.L. & DEWAARD, J.R. 2003. Biological identifications through DNA barcodes. Proceedings of the Royal Society B: Biological Sciences 270: 313– 321. HEENAN, D. 1977. Some observations on the cycads of central Africa. Journal of the Linnean Society, Botany 74: 279–288. HERMSEN, E.J. TAYLOR, T.N. TAYLOR, E.L. & STEVENSON, D.W. 2006. Cataphylls of the middle Triassic cycad Antarcticycas schopfii and new insights into cycad evolution. American Journal of Botany 93: 724–738. HILL, K.D. & STEVENSON, D.W. 1998–2005. The Cycad Pages. Available Online. http: //plantnet.rbgsyd.gov.au/PlantNet/cycad/index.html. HILL, K.D. CHASE, M.W. STEVENSON, D.W. HILLS, H.G. & SCHUTZMAN, B. 2003. The families and genera of cycads: a molecular phylogenetic analysis of cycadophyta based on nuclear and plastid DNA sequences. International Journal of Plant Science 164: 933–948. HOLLINGSWORTH, M.L. CLARK, A.A. FORREST, L.L. RICHARDSON, J. PENNINGTON, R.T. LONG, D.G. COWAN, R. CHASE, M.W. GAUDEUL, M. & HOLLINGSWORTH, P.M. 2009. Selecting barcoding loci for plants: evaluation of seven candidate loci with species-level sampling in three divergent groups of land plants. Molecular Ecology Resources 9: 439–457. HOLLINGSWORTH, P.M. GRAHAM, S.W. & LITTLE, D.P. 2011. Choosing and using a plant DNA barcode. PLoS ONE 6: e19254. doi: 10.1371/journal.pone.0019254. HOLZMAN, G. 2005. Designer cycads. Encephalartos 83: 18. HUELSENBECK, J.P. & RONQUIST, F. 2001. MrBayes: Bayesian inference of phylogeny. Biometrics 17: 754–755.

144

HURTER, M. 2008. The different Encephalartos woodii plants. Encephalartos 94: 31–35. HURTER, P.J.H. & GLEN, H.F. 1996. Encephalartos hirsutus (Zamiaceae): a newly described species from South Africa. Bothalia 31: 197–199. IUCN. 2010. IUCN Red List of threatened species. Version 2010.4. . Downloaded on 24 February 2011. IUCN/SSC CYCAD SPECIALIST GROUP. 2003. Review of significant trade: cycads. Available Online: www.cites.org/eng/com/pc/14/E–PC14–09–02–02–A1.pdf. JOHNSON, L.A.S. 1959. The families of cycads and the Zamiaceae of Australia. Proceedings of The Linnean Society of New South Wales 84: 64–117. JONES, D.L. 2002. Cycads of the World, ancient plants in today's landscape. Second edition. New Holland Publishers, Australia. KADO, T. MATSUMOTO, A. UJINO-IHARA, T. & TSUMURA, Y. 2008. Amounts and patterns of nucleotide variation within and between two Japanese conifers, sugi (Cryptomeria japonica) and hinoki (Chamaecyparis obtuse) (Cupressaceae sensu lato). Tree Genetetics Genomes 4: 133–14. KEPPEL, G. LEE, S.W. & HODGSKISS, P.D. 2002. Evidence for long isolation among populations of a Pacific cycad: genetic diversity and differentiation in Cycas seemannii A. Br. (Cycadaceae). Journal of Heredity 93: 133–139. KLAVINS, S.D. KELLOGG, D.W. KRINGS, M. TAYLOR, E.L. & TAYLOR, T.N. 2005. Coprolites in a Middle Triassic cycad pollen cone: evidence for insect pollination in early cycads? Evolutionary Ecology Research 7: 479–488. KLAVINS, S.D. TAYLOR, E.L. KRINGS, M. & TAYLOR, T.N. 2003. Gymnosperms from the middle Triassic of Antarctica: the first structurally preserved cycad pollen cone. International Journal of Plant Sciences 164: 1007–1020. KOELEMAN, A. 1978. 'n Morfologies-taksonomiese studie van die blare van die genus Encephalartos Lehm. in Suid-Afrika. Unpublished Msc. Thesis, University of Pretoria. KOELEMAN, A. ROBBERTSE, P.J. & EICKER, A. 1981. Die anatomie van die pinnas van die Suid-Afrikaanse spesies van Encephalartos Lehm. Journal of South African Botany 47: 247– 271. KRESS, J. & ERICKSON, D.L. 2007. A two-locus global DNA barcode for land plants: the coding rbcL gene complements the non-coding trnH- psbA spacer region. PLoS One 6: 1–10. KRESS, J.W. WURDACK, K.J. ZIMMER, E.A. WEIGT, L.A. & JANZEN, D.H. 2005. Use of DNA barcodes to identify flowering plants. Proceedings of the National Academy of Sciences 102: 8369–8374. KRESS, W.J. ERICKSON, D.L. JONES, F.A. SWENSOND, N.G. PEREZ, R. SANJUR, O. & BERMINGHAM, E. 2009. Plant DNA barcodes and a community phylogeny of a tropical forest dynamics plot in Panama. Proceedings of the National Academy of Sciences 106: 18621– 18626.

145

KUMAR, S. HAHN, F.M. MCMAHAN, C.M. CORNISH, K. & WHALEN, M.C. 2009. Comparative analysis of the complete sequence of the plastid genome of Parthenium argentatum and identification of DNA barcodes to differentiate Parthenium species and lines. BMC Plant Biology 9: 13. KUNZE, G. 1839. Additamentum enumeratiunus fiicum maxicanarum. Linnaea 13: 129–153. LAHAYE, R. SAVOLAINEN, V. DUTHOIT, S. MAURIN, O. & VAN DER BANK, M. 2008a. A test of psbK-psbI and atpF-atpH as potential plant DNA barcodes using the flora of the Kruger National Park as a model system (South Africa). Nature Proceedings: http: //hdl.handle.net/10101/npre.2008.1896.1. LAHAYE, R. VAN DER BANK, M. BOGARIN, D. WARNER, J. PUPULIN, F. GIGOT, G. MAURIN, O. DUTHOIT, S. BARRACLOUGH, T.G. & SAVOLAINEN, V. 2008b. DNA barcoding the floras of biodiversity hotspots. Proceedings of the National Academy of Sciences of the United States of America 105: 2923–2928. LAVRANOS, J.J. & GOODE, D.L. 1988. Notes on southern African Cycadales. Bulletin du Jardin botanique national de Belgique 58: 219–224. LEHMANN, J.G.C. 1834. Novarum et minus cognitarum stirpium pugillus I-X Addita enumeratione plantarum omnium. Pugillus descriptarum 6: 1–14. Johannes Augusi Meissner, Hamburg, Germany. LEVIN, R.A. WAGNER, W.L. HOCH, P.C. NEPOKROEFF, J.C. PIRES, J.C. ZIMMER, E.A. & SYTSMA, K.J. 2003. Family-level relationships of Onagraceae based on chloroplast rbcL and ndhF data. American Journal of Botany 90: 107–115. LI, Y. GAO, L.-M. POUDEL, R.C. LI, D.-Z. & FORREST, A. 2011. High universality of matK primers for barcoding gymnosperms. Journal of Systematics and Evolution 49: 169–175. LINDSTRÖM, A.J. 2009. Typification of some species names in Zamia L. (Zamiaceae), with an assessment of the status of Chigua D. Stev. Taxon 58: 265–270. LINNAEUS, C. 1753. Species plantarum, vol 1. Available online: http: //books.google.co.za/books?id=6A0PAAAAQAAJ&printsec=frontcover&dq=Linnaeus+species+ plantarum&hl=en&ei=Un5oTsSfDvHnmAWAlu2qDA&sa=X&oi=book_result&ct=result&resnum =2&ved=0CC4Q6AEwAQ#v=onepage&q&f=false. LITTLE, D.P. & STEVENSON, D.W. 2007. A comparison of algorithms for identification of specimens using DNA barcodes: examples from gymnosperms. Cladistics 23: 1–21. LITTLE, D.P. 2011. DNA barcode sequence identification incorporating taxonomic hierarchy and within taxon variability. PLoS ONE 6: e20552. doi: 10.1371/journal.pone.0020552. LIU, J. MOLLER, M. GAO, L.M. ZHANG, D.Q. & ZHUKI, D.E. 2010. DNA barcoding for the discrimination of Eurasian yews (Taxus L. Taxaceae) and the discovery of cryptic species. Molecular Ecology Resources 11: 89–100. LUO, K. CHEN, S.L. CHEN, K.L. SONG, J.Y. & YAO, H. 2009. Application of DNA barcoding to the medicinal plants of the Araceae family. Planta medica 75: 416–416.

146

MABUNDA, M.A. 2007. Species-level phylogenetic reconstruction of the African cycad genus Encephalartos (Zamiaceae). Unpublished Msc. thesis, University of the Western Cape. MALAISSE, F. 1969. Encephalartos schmitzii Malaisse, Cycadacée nouvelle du Congo-Kinshasa. Bulletin du Jardin botanique national de Belgique 39: 401–406. MALAISSE, F. SCLAVO, J.P. & CROSIERS, C. 1993. Recherches sur les Encephalartos Lehm. (Zamiaceae) d'Afrique centrale 2. Apport de la morphologie foliaire dans la différenciation spécifique. Bulletin du Jardin botanique national de Belgique 62: 205–219. MANDER, M. 1998. Marketing of indigenous medicinal plants in South Africa: A case study in KwaZulu-Natal. Food and Agricultural organization of the United Nations. Rome. MCLELLAN, T. & NDAMASE, D. 1995. Morphological and genetic comparison of forest and grassland Stangeria eriopus. In: VORSTER, P. (ed.). Proceedings of the Third International Conference on Cycad Biology. Cycad Society of South Africa, Stellenbosch. MCNEILL, J. BARRIE, F.R. BURDET, H.M. DEMOULIN, V. HAWKSWORTH, D.L. MARHOLD, K. NICOLSON, D.H. PRADO, J. SILVA, P.C. SKOG, J.E. WIERSEMA, J.H. & TURLAND, N.J. 2006. International Code of Botanical Nomenclature (Vienna Code). Regnum Vegetabile 146. A.R.G. Gantner Verlag KG. MELVILLE, R. 1957. Encephalartos in Central Africa. Kew Bulletin 12: 237–257. MIRINGU, B.W. 1999. The population and conservation status of the Kenyan cycads, and their potential as a horticultural crop. In: CHEN, C.J. (ed.) Biology and Conservation of Cycads. Proceedings of the Fourth International Conference on Cycad Biology. International Academic Publishers, Beijing. MOORE, T. 1853. List of Mr. Plant's Natal plants. Hooker’s Journal of Botany and Kew Garden Miscellany 5: 225–229. MOUND, L.A. & TERRY, I. 2001. Thrips pollination of the central Australian cycad, Macrozamia macdonnellii (Cycadales). International Journal of Plant Science 162: 147–154. MOYNIHAN, J. STEVENSON, D.W. LEWIS, C.E. VOVIDES, A.P. CAPUTO, P. & FRANCISCO- ORTEGA, J. In press. A phylogenetic study of Dioon Lindl. (Zamiaceae, Cycadales), based on morphology, nuclear ribosomal DNA, a low copy nuclear gene, and plastid RFLP. Proceedings of the 8th International Conference on Cycad Biology (CYCAD 2008). Memoirs of the New York Botanical Garden. MYBURG, L.M 1991. Hybrid of different species of cycads in gardens. Encephalartos 62: 26. NEWMASTER, S.G. FAZEKAS, A.J. & RAGUPATHY, S. 2006. DNA barcoding in land plants: evaluation of rbcL in a multigene tiered approach. Canadian Journal of Botany-Revue Canadienne De Botanique 84: 335–341. NEWMASTER, S.G. FAZEKAS, A.J. STEEVES, R.A.D. & JANOVEC, J. 2008. Testing candidate plant barcode regions in the Myristicaceae. Molecular Ecology Resources 8: 480–490. NICOLALDE-MOREJÓN, F. VERGARA-SILVA, F. GONZALEZ-ASTORGA, J. & STEVENSON, D.W. 2011. Character-based, population-level DNA barcoding in Mexican species of Zamia L.

147

(Zamiaceae: Cycadales). Mitochondrial DNA 21: 51–59. NICOLALDE-MOREJÓN, F. VERGARA-SILVA, F. GONZALEZ-ASTORGA, J. STEVENSON, D.W. VOVIDES, A.P. & SOSA, V. 2010. A character-based approach in the Mexican cycads supports diverse multigene combinations for DNA barcoding. Cladistics 26: 1–15. NORSTOG, K.J. & NICHOLS, T.J. 1997. The biology of the cycads. Ithaca, New York: Cornell University Press. OBERPRIELER, R.G. 1995a. The weevils (Coleoptera: Curculionoidea) associated with the cycads. 1. Classification, relationships, and biology. In: VORSTER, P. (ed.). Proceedings of the Third International Conference on Cycad Biology. Cycad society of South Africa, Stellenbosch. OBERPRIELER, R.G. 1995b. The weevils (Coleoptera: Curculionoidea) associated with the 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. OBERPRIELER, R.G. 1996. Encephalartos aplanatus Vorster: the weevil perspective. Encephalartos 48: 10–12. OSBORNE, R. & BAIJNATH, H. 1995. Morphometric analysis of vegetative characters of Encephalartos woodii, E. natalensis and an apparent intermediate. In: VORSTER, P. (ed.). Proceedings of the Third International Conference on Cycad Biology. Cycad society of South Africa, Stellenbosch. OSBORNE, R. & PASCHKE, R.T. 1993. Morphometric analysis of vegetative characters of Encephalartos woodii, E. natalensis and an apparent intermediate. South African Journal of Botany 59: 455–456. OSBORNE, R. GROBBELAAR, N. & VINCENT, P.L.D. 1993. A numerical phenetic study of the genus Encephalartos Lehm. In: STEVENSON, D.W. & NORSTOG, K.J. (eds.). The Biology, Structure and Systematics of the Cycadales. Proceedings of CYCAD 90, the Second International Conference on Cycad Biology. Milton, Queensland: Palm and Cycad Societies of Australia Ltd. OSBORNE, R. GROBBELAAR, N. & VORSTER, P. 1988. South African cycad Research: progress and prospects. Suid Afrikaanse Tydskrif vir Wetenskap 84: 981–896. OSBORNE, R. GROVE, A. OH, P. MABRY, T.J. NG, J.C. & SEAWRIGHT, A.A. 1994. The magical and medicinal usage of Stangeria eriopus in South Africa. Journal of Ethnopharmacology 43: 67–72. OSBORNE, R. STEVENSON, D.W. HILL, K.D. & STANBERG, L. In press. The World List of Cycads. Proceedings of the 8th International Conference on Cycad Biology (CYCAD 2008). Memoirs of the New York Botanical Garden. PANT, D.D. 1987. The fossil history and phylogeny of the Cycadales. Geophytology 17: 125–162. PETTENGILL, J.B. & NEEL, M.C. 2010. An evaluation of candidate plant DNA barcodes and assignment methods in diagnosing 29 species in the genus Agalinis (Orobanchaceae).

148

American Journal of Botany 97: 1391–1406. PILGER, R. 1926. Cycadaceae. In: ENGLER, A. (ed.). Die naturlichen Pflanzenfamilien edition 2 13: 44–82. PINARES, A. GONZÁLEZ-ASTORGA, J. VOVIDES, A.P. LAZCANO, J. & VENDRAME, W.A. 2009. Genetic diversity of the endangered endemic Microcycas calocoma (Miq.) A.DC. (Zamiaceae, Cycadales): implications for conservation. Biochemical Systematics and Ecology 37: 385–394. POSADA, D. & CRANDALL, K.A. 1998. Modeltest: testing the model of DNA substitution. Bioinformatics 14: 817–818. PRAKASH, S. & VAN STADEN, J. 2008. Genetic variability and species identification within Encephalartos using Random Amplified Polymorphic DNA (RAPD) markers. South African Journal of Botany 74: 735–739. PRAKASH, S. GROBBELAAR, N. & VAN STADEN, J. 2008. Diversity in Encephalartos woodii collections based on Random Amplified DNA markers (RAPD’s) and inter-Specific Sequence Repeats (ISSR’s). South African Journal of Botany 74: 341–344. PROCHE!, !. & JOHNSON, S.D. 2009. Beetle pollination of the fruit-scented cones of the South African cycad Stangeria eriopus. American Journal of Botany 96: 1722–1730. PRYER, K.M. SCHUETTPELZ, E. HUIET, L. GRUSZ, A.L. ROTHFELS, C.J. AVENT, T. SCHWARTZ, D. & WINDHAM, M.D. 2010. DNA barcoding exposes a case of mistaken identity in the fern horticultural trade. Molecular Ecology Resources 10: 979–985. RAI, H.S. O'BRIEN, H.E. REEVES, P.A. OLMSTEAD, R.G. & GRAHAM, S.W. 2003. Inference of higher-order relationships in the cycads from a large chloroplast data set. Molecular Phylogenetics and Evolution 29: 350–359. RAIMONDO, D.C. & DONALDSON, J.S. 2003. Responses of cycads with different life histories to the impact of plant collecting: simulation models to determine important life history stages and population recovery times. Biological Conservation 111: 345–358. RAN, J. GAO, H. & WANG, X. 2010. Fast evolution of the retroprocessed mitochondrial rps3 gene in Conifer II and further evidence for the phylogeny of gymnosperms. Molecular Phylogenetics and Evolution 54: 136–149. RAN, J.H. WANG, P.P. ZHAO, H.J. & WANG, X.Q. 2010. A test of seven candidate barcode regions from the plastome in Picea (Pinaceae). Journal of Integrative Plant Biology 52: 1109– 1126. RATNASINGHAM, S. & HEBERT, P.D.N. 2007. BOLD: The Barcode of Life Data System (http: //barcodinglife.org). Molecular Ecology Notes 7: 355–364. R DEVELOPMENT CORE TEAM. 2010. R: A language and environment for staistical computing. R Foundation for statistical computing, Vienna, Austria. ISBN 3-900051-07-0, URL http://www.r-project.org/ REN, B.Q. XIANG, X.G. & CHEN, Z.D. 2010. Species identification of Alnus (Betulaceae) using

149

nrDNA and cpDNA genetic markers. Molecular Ecology Resources 10: 594–605. RESSLAR, P.M. 2002. Genotypic variation in the in vitro formation of callus on the petioles of Stangeria. In: STEVENSON, D.W. The biology, Structure, and Systematics of the Cycadales: Proceedings of Cycad 99, the fifth international conference on cycad biology. Memoirs of the New York Botanical Garden 87: 35. ROBBERTSE, P.J. VORSTER, P. & VAN DER WESTHUIZEN, S. 1989. Encephalartos middelburgensis (Zamiaceae): a new species from the Transvaal. South African Journal of Botany 55: 122–126. RONQUIST, F. & HUELSENBECK, J.P. 2003. MrBAYES 3: Bayesian phylogenetic inference under mixed models. Bioinformatics 19: 1572–1574. ROUSSEAU, P. 2011. Multiple cones in Eastern Cape species: Evolutionary considerations. Encephalartos 105: 22–23. ROY, S. TYAGI, A. SHUKLA, V. KUMAR, A. SINGH, U.M. CHAUDHARY, L.B. DATT, B. BAG, S.K. SINGH, P.K. NAIR, N.K. HUSAIN, T. & TULI, R. 2010. Universal plant DNA barcode loci may not work in complex groups: A case study with Indian Berberis species. PLoS ONE 5: e13674. doi: 10.1371/journal.pone.0013674. SANGIN, P. FORSTER, P.I. MINGMUANG, M. & KOKUBUGATA, G. 2008. A phylogeny for two cycad families (Stangeriaceae and Zamiaceae) based on Chloroplast DNA Sequences. Bulletin of the National Museum of Nature and Science. Series B, Botany, National Science Museum 34: 75–82. SANGIN, P. LINDSTRÖM, A.J. KOKUBUGATA, G. CHAIPRASONGSUK, M. & MINGMUANG, M. 2010. Phylogenetic relationships within Cycadaceae inferred from non–coding regions of chloroplast DNA. Kasetsart Journal (Natural Science) 44: 544–557. SASS, C. LITTLE, D.P. STEVENSON, D.W. & SPECHT, C.D. 2007. DNA Barcoding in the Cycadales: Testing the Potential of Proposed Barcoding Markers for Species Identification of Cycads. PLoS ONE 2: e1154. doi: 10.1371/journal.pone.0001154. SAVOLAINEN, V. CUÉNOUD, P. SPICHIGER, R. MARTINEZ, M.D.P. CRÉVECOEUR, M. & MANEN, J.-F. 1995. The use of herbarium specimens in DNA phylogenetics: Evaluation and improvements. Plant Systematics and Evolution 197: 87–98. SCHNEIDER, D. WINK, M. SPORER, F. & LOUNIBOS, P. 2002. Cycads: their evolution, toxins, herbivores and insect pollinators. Naturwissenschaften 89: 281–294. SCHUTZMAN, B. & DEHGAN, B. 1993. Computer–assisted systematics in the Cycadales. In: STEVENSON, D.W. & NORSTOG, K.J. (eds.). TheBiology, Structure and Systematics of the Cycadales. Proceedings of CYCAD 90, the Second International Conference on Cycad Biology. Milton, Queensland: Palm and Cycad Societies of Australia Ltd. SEBERG, O. & PETERSEN, G. 2009. How many loci does it take to DNA barcode a Crocus? PLoS ONE 4: e4598. doi: 10.1371/journal.pone.0004598. SHARMA, K. JONES, D.L. & FORSTER, P.I. 2004. Genetic differentiation and phenetic

150

relatedness among seven species of the Macrozamia plurinervia complex (Zamiaceae). Biochemical Systematics and Ecology 32: 313–327. SHARMA, K. JONES, D.L. FORSTER, P.I. & YOUNG, A.G. 1998. The extent and structure of genetic variation in the Macrozamia pauli–guilielmi complex (Zamiaceae). Biochemical Systematics and Ecology 26: 45–54. SHARMA, K. JONES, D.L. FORSTER, P.I. & YOUNG, A.G. 1999. Low isozymic differentiation among five species of the Macrozamia heteromera group (Zamiaceae). Biochemical Systematics and Ecology 27: 67–77. SOLTIS, D.E. MAVRODIEV, E.V. DOYLE, J.J. RAHSCHER J. & SOLTIS, P.S. 2008. ITS and ETS sequence data and phylogeny reconstruction in allopolyploids and hybrids. Systematic Botany 33: 7–20. SPOONER, D.M. 2009. DNA barcoding will frequently fail in complicated groups: an example in wild potatoes. American Journal of Botany 96: 1177–1189. STAINER, O. 2000. The first sexual reproduction of Encephalartos in Europe? Encephalartos 63: 30. STARR, J.R. NACZI, R.F.C. & CHOUINARD, B.N. 2009. Plant DNA barcodes and species resolution in sedges (Carex, Cyperaceae). Molecular Ecology Resources 9: 151–163. STAUDE, H. 2001. African cycads and moths: An intricate relationship of ancient origin. In: GOODE, D. 2001. Cycads of Africa Volume I. D.E. Cycads of Africa Publishers, Gallo-manor, South Africa. STEVENS, P.F. 2001. Angiosperm Phylogeny Website. Version 9, June 2008 [and more or less continuously updated since]. Available online: http: //www.mobot.org/MOBOT/research/APweb. STEVENSON, D.W. 1981. Observations on ptyxis, phenology, and trichomes in the Cycadales and their systematic implications. American Journal of Botany 68: 1104–1114. STEVENSON, D.W. 1992. A formal classification of the extant cycads. Brittonia 44: 220–223. SUINYUY, T.N. DONALDSON, J.S. & JOHNSON, S.D. 2009. Insect pollination in the African cycad Encephalartos friderici-guilielmi Lehm. South African Journal of Botany 75: 682–688. SUINYUY, T.N. DONALDSON, J.S. & JOHNSON, S.D. 2010. Scent chemistry and patterns of thermogenesis in male and female cones of the African cycad Encephalartos natalensis (Zamiaceae). South African Journal of Botany 76: 717–725. SUN, Y. SKINNER, D.Z. LIANG, G.H. & HULBERT, S.H. 1994. Phylogenetic analysis of Sorghum and related taxa using internal transcribed spacers of nuclear ribosomal DNA. Theoretical and Applied Genetics 89: 26–32. SWOFFORD, D.L. 2002. PAUP* Sinauer Associates, Sunderland Massachusetts. TABERLET, P. COISSAC, E. POMPANON, F. GIELLY, L. MIQUEL, C. VALENTINI, A. VERMAT, T. CORTHIER, G. BROCHMANN, C. & WILLERSLEV, E. 2006. Power and limitations of the chloroplast trnL (UAA) intron for plant DNA barcoding. Nucleic Acids Research 35: e14 doi:10.1093/nar/gkl938

151

TABERLET, P. GIELLY, L. PAUTOU, G. & BOUVET, J. 1991. Universal primers for amplification of three non-coding regions of chloroplast DNA. Plant Molecular Biology 17: 1105–1109. TANG, W. 1983. Seed displays in cycads. Cycad Newsletter 6: 7–11. TERRY, I. 2001. Thrips and weevils as dual specialist pollinators of the Australian cycad Macrozamia communis (Zamiaceae). International Journal of Plant Science 162: 1293–1305. TERRY, I. FORSTER, P.I. MOORE, C.J. ROEMER, R.B. & MACHIN, P.J. 2008. Demographics, pollination syndrome and conservation status of Macrozamia platyrhachis (Zamiaceae), a geographically restricted Queensland cycad. Australian Journal of Botany 56: 321–332. TERRY, I. MOORE, C.J. WALTER, G.H. FORSTER, P.I. ROEMER, R.B. DONALDSON, J.D. & MACHIN, P.J. 2004. Association of cone thermogenesis and volatiles with pollinator specificity in Macrozamia cycads. Plant Systematics and Evolution 243: 233–247. TERRY, L.I. WALTER, G.H. DONADLSON, J.S. SNOW, E. FORSTER, P.I. & MACHIN, P.J. 2005. Pollination of Australian Macrozamia cycads (Zamiaceae): effectiveness and behavior of specialist vectors in a dependent mutualism. American Journal of Botany 92: 931–940. THUNBERG, C.P. 1775. Cycas caffra, nova palmae species, decripta. Nova Acta Regiae Societatis Scientiarum Upsaliensis 2: 283–288. TREUTLEIN, J. & WINK, M. 2002. Molecular phylogeny of cycads. Naturwissenschaften 89: 221– 225. TREUTLEIN, J. VORSTER, P. & WINK, M. 2005. Molecular relationships in Encephalartos (Zamiaceae, Cycadales) based on nucleotide sequences of nuclear ITS 1, 2, rbcL, and genomic ISSR fingerprinting. Plant Biology 7: 1–12. TURNER, I. SCLAVO, J.-P. & MALAISSE, F. 2006. Un nouvel Encephalartos Lehm. (Zamiaceae) de Zambie. Biotechnology, Agronomy, Society and Environment 10: 181–183. VAN DER BANK, F.H. VORSTER, P. VAN DER BANK, M. & WINK, M. 1998. Phylogeny of Encephalartos: Some Eastern Cape species. The Botanical Review 70: 250–259. VAN DER BANK, F.H. WINK, M. VORSTER, P. TREUTLEIN, J. BRAND, L. VAN DER BANK, M. & HURTER, J. 2001. Allozyme and DNA sequence comparisons of nine species of Encephalartos (Zamiaceae). Biochemical Systematics and Ecology 29: 241–266. VILJOEN, D.C. & VAN STADEN, J. 2006. The genetic relationship between Encephalartos natalensis and E. woodii determined using RAPD fingerprinting. South African Journal of Botany 72: 642–645. VIRGILIO, M. BACKELJAU, T. NEVADO, B. & DE MEYER, M. 2010. Comparative performances of DNA barcoding across insect orders. BMC Bioinformatics 11: 206. VORSTER, P. & OBERPRIELER, R.G. 1999. Entomological evidence for and against taxonomical decisions in Encephalartos. In: CHEN, C.J. (ed). Biology and Conservation of Cycads. Proceedings of the Fourth International Conference on Cycad Biology, Panzhihua, China. International Academic Publishers, Beijing. VORSTER, P. & VORSTER, E. 1974. Stangeria eriopus. Excelsa 4: 79–88.

152

VORSTER, P. 1986. Hybridization in Encephalartos. Excelsa 12: 101–106. VORSTER, P. 1990. Encephalartos aemulans (Zamiaceae), a new species from northern Natal. South African Journal of Botany 56: 239–243. VORSTER, P. 1993. Taxonomy of Encephalartos (Zamiaceae): taxonomically useful external characteristics. In: STEVENSON, D.W. & NORSTOG, K.J. (eds.). Proceedings of Cycad 90, the Second International Conference on Cycad Biology. Palm and Cycad Societies of Australia, Milton, Queensland. 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. VORSTER, P. 1996a. Encephalartos aplanatus (Zamiaceae): a new species from Swaziland. South African Journal of Botany 62: 57–60. VORSTER, P. 1996b. Encephalartos brevifoliolatus (Zamiaceae): a new species from the Northern province. South African Journal of Botany 62: 61–64. VORSTER, P. 1996c. Encephalartos msinganus (Zamiaceae): a new species from KwaZulu-Natal. South African Journal of Botany 62: 67–70. VORSTER, P. 1996d. Encephalartos venetus (Zamiaceae): a new species from the Northern province. South African Journal of Botany 62: 71–75. VORSTER, P. 1996e. Encephalartos senticosus (Zamiaceae): a new species from northern KwaZulu–Natal and Swaziland. South African Journal of Botany 62: 76–79. VORSTER, P. 1996f. Encephalartos aplanatus: Taxonomic ranking and Encephalartos aplanatus. Encephalartos 47: 25. VORSTER, P. 1999. A review of the arborescent tropical species of Encephalartos. In: CHEN, C.J. (eds). Biology and Conservation of Cycads—Proceedings of the 4th International Conference on Cycad Biology Panzhihua, China. Academic Publishers, Beijing. VORSTER, P. 2003. Growing cycads at 34ºS. Encephalartos 74: 32. VORSTER, P. 2004a. Chapter 6: Classification concepts in Encephalartos (Zamiaceae). In WALTERS, T. & OSBORNE, R. (eds.). Cycad Classification: Concepts and recommendations. CABI Publishing, Wallingford. VORSTER, P. 2004b. Typification of Encephalartos. Bothalia 34: 112–113. VORSTER, P. 2005. Subdivision of Encephalartos. In: VOVIDES, A.P, STEVENSON, D.W. & OSBORNE, R. (eds). Proceedings of Cycad 2005: the 7th International Conference on Cycad Biology. Memoirs of the New York Botanical Garden 97: 597–610. VORSTER, P. & OBERPRIELER, R.G. 1999. Entomological evidence for and against taxonomical decisions in Encephalartos. In: CHEN, C.J. (ed). Biology and Conservation of Cycads. Proceedings of the Fourth International Conference on Cycad Biology, Panzhihua, China. International Academic Publishers, Beijing. WALTERS, T. & OSBORNE, R. (eds.). 2004. Cycad Classification: Concepts and

153

recommendations. CABI Publishing, Wallingford. WHITE, T.J. BRUNS, T. LEE, S. & TAYLOR, J. W. 1990. Amplification and direct sequencing of fungal ribosomal RNA genes for phylogenetics. In: INNIS, M.A. GELFAND, D.H. SNINSKY, J.J. & WHITE, T.J. (eds.). PCR Protocols: A guide to methods and applications. Academic Press, Inc. New York.

WHITELOCK, L.M. 2002. The Cycads. Portland, Oregon Timber Press. WHITLOCK, B.A. HALE, A.M. & GROFF, P.A. 2010. Intraspecific inversions pose a challenge for the trnH–psbA plant DNA barcode. PLoS ONE 5: e11533. doi: 10.1371/journal.pone.0011533. WILLIAMS, V.L. 2003. Hawkers of health: an investigation of the Faraday Street traditional Medicine Market in Johannesburg, Gauteng. Plant Ecology and Conservation Series No 15 (Report to Gauteng Directorate of Nature Conservation, DACEL). University of the Witwatersrand. WILLIAMS, V.L. BALKWILL, K. & WITKOWSKI, E.T.F. 2001. A lexicon of plants traded in the Witwatersrand umuthi shops, South Africa. Bothalia 31: 71–79. WILSON, G.W. 2002. Insect pollination in the cycad genus Bowenia Hook. ex Hook. f. (Stangeriaceae). Biotropica 34: 438–441. WINSHIP, P.R. 1989. An improved method for directly sequencing PCR amplified material using dimethyl sulphoxide. Nucleic Acids Research 17: 1266. WOLFE, K.H. LI, W.H. & SHARP, P.M. 1987. Rates of nucleotide substitution vary greatly among plant mitochondrial, chloroplast, and nuclear DNAs. Proceedings of the National Academy of Sciences of the United States of America 84: 9054–9058. XIAO, L. GE, X. GONG, X. HAO, G. & ZHENG, S. 2004. ISSR Variation in the endemic and endangered plant Cycas guizhouensis (Cycadaceae). Annals of Botany 94: 133–138. YAO, H. SONG, J.Y. LIU, C. LUO, K. HAN, J.P. LI, Y. PANG, X.H. XU, H.X. ZHU, Y.J. XIAO, P.G. & CHEN, S.L. 2010. Use of ITS2 Region as the Universal DNA Barcode for Plants and Animals. PLoS One 5: e13102. doi:10.1371/journal.pone.0013102 YAO, H. SONG, J.Y. MA, X.Y. LIU, C. LI, Y. XU, H.X. HAN, J.P. DUAN, L.S. & CHEN, S.L. 2009. Identification of Dendrobium species by a candidate DNA barcode sequence: the chloroplast psbA-trnH intergenic region. Planta Medica 75: 667–669. YESSON, C. BARCENAS, R.T. HERNANDEZ, H.M. RUIZ-MAQUEDA, M.DLZ. PRADO, A. RODRIGUEZ, V.M. & HAWKINS, J.A. 2011. DNA barcodes for Mexican Cactaceae, plants under pressure from wild collecting. Molecular Ecology Resources 11: 775–783. ZGURSKI, J.M. RAI, H.S. FAI, Q.M. BOGLER, D.J. FRANCISCO-ORTEGA, J. & GRAHAM, S.W. 2008. How well do we understand the overall backbone of cycad phylogeny? New insights from a large, multigene plastid data set. Molecular Phylogenetics and Evolution 47: 1232–1237. ZHANG, J.M. WANG, J.X. XIA, T. & ZHOU, S.L. 2009. DNA barcoding: species delimitation in tree peonies. Science China Life Sciences 52: 568–578.

154

Appendix 1. Complete collection list

List of all samples collected or received by the author including BOLD accession numbers. Collection sources abbreviations: Private collection (PC), University of Johannesburg (UJ), University of Pretoria (UP) experimental farm (EF) and Manie van der Schijff Botanical Garden (MBG), Cycad World of Innovation (CWOFI), Lowveld National Botanical Garden (LBG), New York Botanical Garden (NYBG), Pretoria National Botanical Garden (PBG).

Species Collection ID Source BOLD Form name E. aemulans Vorster PR688 PC 1 CYAF039 E. aemulans Vorster PR748 PC 2 CYAF094 E. aemulans Vorster PR755 CWOFI CYAF101 E. aemulans Vorster PR861 UP, MBG CYPAF309 E. aemulans Vorster d1018 NYBG DNA bank - - E. altensteinii Lehm d1019 NYBG DNA bank - - E. altensteinii Lehm. PR658 UJ CYAF009 E. altensteinii Lehm. PR668 PBG CYAF019 E. altensteinii Lehm. PR742 PC 2 CYAF088 E. altensteinii Lehm. PR756 CWOFI CYAF102 E. altensteinii Lehm. PR856 UP, MBG CYPAF304 E. altensteinii Lehm. PR857 UP, MBG CYPAF305 E. aplanatus Vorster PR682 PBG CYAF033 E. aplanatus Vorster PR757 CWOFI CYAF103 E. aplanatus Vorster PR840 UP, MBG CYAF185 E. aplanatus Vorster PR922 LBG CYPAF370 E. aplanatus Vorster PR923 LBG CYPAF371 E. aplanatus Vorster d1126 NYBG DNA bank - - E. arenarius R.A. Dyer PR667 PBG CYAF018 E. arenarius R.A. Dyer PR679 PBG CYAF030 E. arenarius R.A. Dyer PR741 PC 2 CYAF087 E. arenarius R.A. Dyer PR758 CWOFI CYAF104 Alexandria E. arenarius R.A. Dyer PR759 CWOFI CYAF105 Blue E. arenarius R.A. Dyer PR854 UP, MBG CYPAF302 E. arenarius R.A. Dyer d1020 NYBG DNA bank - - E. barteri subsp. allochrous L.E. Newton PR892 UP, EF CYPAF340 E. barteri subsp. barteri Carruth. ex Miq. PR754 PC 2 CYAF100 E. barteri subsp. barteri Carruth. ex Miq. PR878 UP, EF CYPAF326 E. barteri subsp. barteri Carruth. ex Miq. d1095 NYBG DNA bank - - E. brevifoliolatus Vorster Xdk 1 Xanderk de Kock CYPAF378 Type Locality Donation E. brevifoliolatus Vorster Xdk 2 Xanderk de Kock CYPAF379 Type Locality Donation E. brevifoliolatus Vorster PR752 PC 2 CYAF098 E. brevifoliolatus Vorster PR760 CWOFI CYAF106 Wolkberg E. brevifoliolatus Vorster PR761 CWOFI CYAF107 Down's Wolkberg E. brevifoliolatus Vorster PR762 CWOFI CYAF108 Wolkberg E. bubalinus Melville PR885 UP, EF CYPAF333 E. bubalinus Melville PR893 LBG CYPAF341 E. bubalinus Melville PR894 LBG CYPAF342 E. bubalinus Melville PR910 UP, EF CYPAF358 Blue E. bubalinus Melville d1021 NYBG DNA bank - - E. caffer (Thunb.) Lehm. PR689 PC 1 CYAF040

155

E. caffer (Thunb.) Lehm. PR729 PC 2 CYAF076 E. caffer (Thunb.) Lehm. PR764 CWOFI CYAF109 East London E. caffer (Thunb.) Lehm. d1101 NYBG DNA bank - - E. cerinus Lavranos & D.L. Goode PR690 PC 1 CYAF041 E. cerinus Lavranos & D.L. Goode PR744 PC 2 CYAF090 E. cerinus Lavranos & D.L. Goode PR766 CWOFI CYAF111 E. cerinus Lavranos & D.L. Goode PR859 UP, MBG CYPAF307 E. cerinus Lavranos & D.L. Goode PR862 UP, MBG CYPAF310 E. cerinus Lavranos & D.L. Goode d1022 NYBG DNA bank - - E. cf. macrostrobilus S. Jones & J. Wynants PR919 LBG CYPAF367 E. cf. dyerianus Lavranos & D.L. Goode PR820 CWOFI CYAF165 Levubu E. cf. dyerianus Lavranos & D.L. Goode PR821 CWOFI CYAF166 Levubu E. cf. kisambo Faden & Beentje PR745 PC 2 CYAF091 E. cf. lebomboensis I. Verd. PR805 CWOFI CYAF150 E. cf. lebomboensis I. Verd. PR739 PC 2 CYAF085 E. cf. natalensis R.A. Dyer & I. Verd. PR884 Nat Grobbelaar CYPAF332 Donation E. cf. whitelockii P.J.H. Hurter PR707 PC 1 CYAF058 E. chimanimaniensis R.A. Dyer & I. Verd. PR883 LBG CYPAF331 E. chimanimaniensis R.A. Dyer & I. Verd. PR888 LBG CYPAF336 E. concinnus R.A. Dyer & I. Verd. PR817 CWOFI CYAF162 E. concinnus R.A. Dyer & I. Verd. PR890 UP, EF CYPAF338 E. concinnus R.A. Dyer & I. Verd. d1023 NYBG DNA bank - - E. cupidus R.A. Dyer PR691 PC 1 CYAF042 E. cupidus R.A. Dyer PR734 PC 2 CYAF080 E. cupidus R.A. Dyer PR767 CWOFI CYAF112 E. cupidus R.A. Dyer CC536 NYBG DNA bank - - E. cupidus R.A. Dyer d1127 NYBG DNA bank - - E. cycadifolius (Jacq.) Lehm. PR683 PBG CYAF034 E. cycadifolius (Jacq.) Lehm. PR722 PC 2 CYPAF288 E. cycadifolius (Jacq.) Lehm. PR765 CWOFI CYAF110 E. cycadifolius (Jacq.) Lehm. PR850 UP, MBG CYPAF298 E. cycadifolius (Jacq.) Lehm. d1128 NYBG DNA bank - - E. delucanus Malaisse, Sclavo & Crosiers d1129 NYBG DNA bank - - E. delucanus Malaisse, Sclavo & Crosiers CC448 NYBG DNA bank - - E. dolomiticus Lavranos & D.L. Goode PR725 PC 2 CYAF072 E. dolomiticus Lavranos & D.L. Goode PR768 CWOFI CYAF113 E. dolomiticus Lavranos & D.L. Goode PR851 UP, MBG CYPAF299 Hairy Leaf E. dolomiticus Lavranos & D.L. Goode PR865 UP, MBG CYPAF313 E. dolomiticus Lavranos & D.L. Goode CC343 NYBG DNA bank - - E. dyerianus Lavranos & D.L. Goode PR687 PBG CYAF038 E. dyerianus Lavranos & D.L. Goode PR731 PC 2 CYAF078 E. dyerianus Lavranos & D.L. Goode PR769 CWOFI CYAF114 E. dyerianus Lavranos & D.L. Goode PR863 UP, MBG CYPAF311 E. dyerianus Lavranos & D.L. Goode CC332 NYBG DNA bank - - E. equatorialis P.J.H. Hurter PR898 LBG CYPAF346 E. equatorialis P.J.H. Hurter PR899 LBG CYPAF347 E. equatorialis P.J.H. Hurter PR900 UP, EF CYPAF348 E. equatorialis P.J.H. Hurter PR902 LBG CYPAF350 E. equatorialis P.J.H. Hurter PR904 LBG CYPAF352 E. eugene-maraisii I. Verd d1024 NYBG DNA bank - - E. eugene-maraisii I. Verd. PR684 PBG CYAF035

156

E. eugene-maraisii I. Verd. PR728 PC 2 CYAF075 E. eugene-maraisii I. Verd. PR770 CWOFI CYAF115 E. eugene-maraisii I. Verd. PR866 UP, MBG CYPAF314 E. eugene-maraisii I. Verd. PR870 UP, MBG CYPAF318 Pallala E. eugene-maraisii I. Verd. PR872 UP, MBG CYPAF320 E. ferox Bertol. f. PR651 UJ CYAF002 E. ferox Bertol. f. PR676 PBG CYAF027 E. ferox Bertol. f. PR715 PC 2 CYAF066 E. ferox Bertol. f. PR771 CWOFI CYAF116 E. ferox Bertol. f. PR841 UP, MBG CYAF186 Inhambane E. ferox Bertol. f. PR844 UP, MBG CYAF190 Yellow cone E. ferox Bertol. f. d1025 NYBG DNA bank - - E. friderici-guilielmi Lehm. PR653 UJ CYAF004 E. friderici-guilielmi Lehm. PR674 PBG CYAF025 E. friderici-guilielmi Lehm. PR733 PC 2 CYAF079 E. friderici-guilielmi Lehm. PR772 CWOFI CYAF117 E. friderici-guilielmi Lehm. PR853 UP, MBG CYPAF301 E. friderici-guilielmi Lehm. CC338 NYBG DNA bank - - E. ghellinckii Lem. PR692 PC 1 CYAF043 E. ghellinckii Lem. PR716 PC 2 CYAF067 E. ghellinckii Lem. PR773 CWOFI CYAF118 E. ghellinckii Lem. PR908 UP, EF CYPAF356 Oribi Gorge E. ghellinckii Lem. Abbott9220 Umtamvuna Nature SAFH441 Reserve E. ghellinckii Lem. CC331 NYBG DNA bank - - E. gratus Prain PR693 PC 1 CYAF044 E. gratus Prain PR740 PC 2 CYAF086 E. gratus Prain PR774 CWOFI CYAF119 E. gratus Prain PR891 LBG CYPAF339 E. gratus Prain d1026 NYBG DNA bank - - E. gratus Prain CC522 NYBG DNA bank - - E. heenanii R.A. Dyer PR686 PBG CYAF037 E. heenanii R.A. Dyer PR775 CWOFI CYAF120 E. heenanii R.A. Dyer PR776 CWOFI CYAF121 Long leaf E. heenanii R.A. Dyer PR925 LBG CYPAF373 E. heenanii R.A. Dyer PR926 LBG CYPAF374 E. hildebrandtii A. Braun & C.D. Bouché d1027 NYBG DNA bank - - E. hildebrandtii A. Braun & C.D. Bouché CC516 NYBG DNA bank - - E. hildebrandtii A.Br. & Bouche´ PR659 UJ CYAF010 E. hildebrandtii A.Br. & Bouche´ PR824 CWOFI CYAF169 E. hildebrandtii A.Br. & Bouche´ PR881 UP, EF CYPAF329 Red cone E. hildebrandtii A.Br. & Bouche´ PR917 LBG CYPAF365 E. hirsutus P.J.H. Hurter PR694 PC 1 CYAF045 E. hirsutus P.J.H. Hurter PR718 PC 2 CYAF069 E. hirsutus P.J.H. Hurter PR785 CWOFI CYAF130 E. hirsutus P.J.H. Hurter CC336 NYBG DNA bank - - E. horridus (Jacq.) Lehm. PR666 PBG CYAF017 E. horridus (Jacq.) Lehm. PR749 PC 2 CYAF095 E. horridus (Jacq.) Lehm. PR777 CWOFI CYAF122 E. horridus (Jacq.) Lehm. PR846 UP, MBG CYPAF294 E. horridus (Jacq.) Lehm. PR847 UP, MBG CYPAF295 E. horridus (Jacq.) Lehm. d1028 NYBG DNA bank - -

157

E. humilis I. Verd. PR695 PC 1 CYAF046 E. humilis I. Verd. PR712 PC 2 CYAF063 E. humilis I. Verd. PR791 CWOFI CYAF136 E. humilis I. Verd. CC340 NYBG DNA bank - - E. inopinus R.A. Dyer PR696 PC 1 CYAF047 E. inopinus R.A. Dyer PR724 PC 2 CYPAF290 E. inopinus R.A. Dyer PR778 CWOFI CYAF123 E. inopinus R.A. Dyer PR864 UP, MBG CYPAF312 E. inopinus R.A. Dyer PR939 UJ CYPAF375 E. inopinus R.A. Dyer d1130 NYBG DNA bank - - E. ituriensis Bamps & Lisowski d1131 NYBG DNA bank - - E. ituriensis Bamps & Lisowski PR788 CWOFI CYAF133 E. kanga Pócs & Luke PR907 UP, EF CYPAF355 E. kisambo Faden & Beentje d1132 NYBG DNA bank - - E. kisambo Faden & Beentje PR700 PC 1 CYAF051 E. kisambo Faden & Beentje PR823 CWOFI CYAF168 E. laevifolius Stapf & Burtt Davy CC335 NYBG DNA bank - - E. laevifolius Stapf & Burtt Davy CC530 NYBG DNA bank - - E. laevifolius Stapf & Burtt Davy PR669 PBG CYAF020 E. laevifolius Stapf & Burtt Davy PR730 PC 2 CYAF077 E. laevifolius Stapf & Burtt Davy PR798 CWOFI CYAF143 Havelock E. laevifolius Stapf & Burtt Davy PR801 CWOFI CYAF146 E. laevifolius Stapf & Burtt Davy PR803 CWOFI CYAF148 Mallalosha E. laevifolius Stapf & Burtt Davy PR804 CWOFI CYAF149 Swaziland x Tugella Ferry E. laevifolius Stapf & Burtt Davy PR807 CWOFI CYAF152 Kaapsehoop E. laevifolius Stapf & Burtt Davy PR845 UP, MBG CYPAF293 Kaapsehoop E. lanatus Stapf & Burtt Davy d1133 NYBG DNA bank - - E. lanatus Stapf & Burtt Davy PRU587 Rhenosterpoort Nature CYPAF377 Reserve E. lanatus Stapf & Burtt Davy PR677 PBG CYAF028 E. lanatus Stapf & Burtt Davy PR713 PC 2 CYAF064 E. lanatus Stapf & Burtt Davy PR779 CWOFI CYAF124 E. lanatus Stapf & Burtt Davy PR828 UP, MBG CYAF173 E. latifrons Lehm. PR678 PBG CYAF029 E. latifrons Lehm. PR708 PC 1 CYAF059 E. latifrons Lehm. PR709 PC 1 CYAF060 E. latifrons Lehm. PR806 CWOFI CYAF151 Green Hills E. latifrons Lehm. PR811 CWOFI CYAF156 Trappers valley E. latifrons Lehm. CC539 NYBG DNA bank - - E. laurentianus De Wild. PR789 CWOFI CYAF134 E. laurentianus De Wild. PR895 LBG CYPAF343 E. laurentianus De Wild. PR897 LBG CYPAF345 E. laurentianus De Wild. PR918 LBG CYPAF366 E. laurentianus De Wild. d1029 NYBG DNA bank - - E. lebomboensis I. Verd. PR657 UJ CYAF008 Piet Retief E. lebomboensis I. Verd. PR662 PBG CYAF013 E. lebomboensis I. Verd. PR698 PC 1 CYAF049 Mananga E. lebomboensis I. Verd. PR699 PC 1 CYAF050 Piet Retief E. lebomboensis I. Verd. PR735 PC 2 CYAF081 Piet Retief E. lebomboensis I. Verd. PR796 CWOFI CYAF141 E. lebomboensis I. Verd. PR800 CWOFI CYAF145 Piet Retief

158

E. lebomboensis I. Verd. PR831 UP, MBG CYAF176 Pongola E. lebomboensis I. Verd. PR874 UP, MBG CYPAF322 Piet Retief E. lebomboensis I. Verd. d1030 NYBG DNA bank - - E. lehmannii Lehm. PR660 UJ CYAF011 E. lehmannii Lehm. PR661 PBG CYAF012 E. lehmannii Lehm. PR721 PC 2 CYPAF287 E. lehmannii Lehm. PR780 CWOFI CYAF125 E. lehmannii Lehm. PR835 UP, MBG CYAF180 E. lehmannii Lehm. d1031 NYBG DNA bank - - E. longifolius (Jacq.) Lehm. PR673 PBG CYAF024 E. longifolius (Jacq.) Lehm. PR711 PC 2 CYAF062 E. longifolius (Jacq.) Lehm. PR781 CWOFI CYAF126 E. longifolius (Jacq.) Lehm. PR808 CWOFI CYAF153 Blue E. longifolius (Jacq.) Lehm. PR809 CWOFI CYAF154 Blunt Tip E. longifolius (Jacq.) Lehm. PR873 UP, MBG CYPAF321 Blunt Tip E. longifolius (Jacq.) Lehm. d1032 NYBG DNA bank - - E. mackenziei L.E. Newton CC529 NYBG DNA bank - - E. macrostrobilus S. Jones & J. Wynants CC333 NYBG DNA bank - - E. manikensis (Gilliland) Gilliland PR697 PC 1 CYAF048 E. manikensis (Gilliland) Gilliland PR793 CWOFI CYAF138 E. manikensis (Gilliland) Gilliland PR794 CWOFI CYAF139 E. manikensis (Gilliland) Gilliland PR795 CWOFI CYAF140 E. manikensis (Gilliland) Gilliland PR887 UP, EF CYPAF335 Elizabethville E. manikensis (Gilliland) Gilliland PR889 UP, EF CYPAF337 Chivala or Manica E. manikensis (Gilliland) Gilliland PR896 UP, EF CYPAF344 Bandula E. manikensis (Gilliland) Gilliland PR903 UP, EF CYPAF351 Vumba E. manikensis Gilliland (Gilliland) d1033 NYBG DNA bank - - E. marunguensis Devred PR814 CWOFI CYAF159 E. marunguensis Devred PR912 UP, EF CYPAF360 E. marunguensis Devred CC446 NYBG DNA bank - - E. middelburgensis Vorster PRU586 Rhenosterpoort Nature CYPAF376 Wilge River Reserve E. middelburgensis Vorster PR685 PBG CYAF036 E. middelburgensis Vorster PR726 PC 2 CYAF073 E. middelburgensis Vorster PR782 CWOFI CYAF127 E. middelburgensis Vorster PR827 Middelburg Natural CYAF172 habitat E. middelburgensis Vorster CC337 NYBG DNA bank - - E. msinganus Vorster PR701 PC 1 CYAF052 E. msinganus Vorster PR751 PC 2 CYAF097 E. msinganus Vorster PR921 LBG CYPAF369 E. msinganus Vorster d1034 NYBG DNA bank - - E. munchii R.A. Dyer & I. Verd. d1134 NYBG DNA bank - - E. munchii R.A. Dyer & I. Verd. PR702 PC 1 CYAF053 E. munchii R.A. Dyer & I. Verd. PR737 PC 2 CYAF083 E. munchii R.A. Dyer & I. Verd. PR784 CWOFI CYAF129 E. munchii R.A. Dyer & I. Verd. PR855 UP, MBG CYPAF303 E. munchii R.A. Dyer & I. Verd. PR901 UP, EF CYPAF349 E. natalensis R.A. Dyer & I. Verd. d1035 NYBG DNA bank - - E. natalensis R.A. Dyer & I. Verd. PR650 UJ CYAF001 E. natalensis R.A. Dyer & I. Verd. PR664 PBG CYAF015 E. natalensis R.A. Dyer & I. Verd. PR714 PC 2 CYAF065

159

E. natalensis R.A. Dyer & I. Verd. PR802 CWOFI CYAF147 Kranskloof E. natalensis R.A. Dyer & I. Verd. PR879 Nat Grobbelaar CYPAF327 Donation E. natalensis R.A. Dyer & I. Verd. Abbott9258 Umtamvuna Nature CYPAF286 Umtamvuna Reserve E. natalensis R.A. Dyer & I. Verd. X E. woodii Sander PR842A UP, MBG CYAF187 E. ngoyanus I. Verd. PR703 PC 1 CYAF054 E. ngoyanus I. Verd. PR717 PC 2 CYAF068 E. ngoyanus I. Verd. PR799 CWOFI CYAF144 E. ngoyanus I. Verd. d1135 NYBG DNA bank - - E. nubimontanus P.J.H. Hurter PR655 UJ CYAF006 E. nubimontanus P.J.H. Hurter PR704 PC 1 CYAF055 E. nubimontanus P.J.H. Hurter PR720 PC 2 CYAF071 E. nubimontanus P.J.H. Hurter PR792 CWOFI CYAF137 E. nubimontanus P.J.H. Hurter PR822 CWOFI CYAF167 Robust E. nubimontanus P.J.H. Hurter CC328 NYBG DNA bank - - E. paucidentatus Stapf & Burtt Davy d1036 NYBG DNA bank - - E. paucidentatus Stapf & Burtt Davy Xdk 3 Xanderk de Kock CYPAF380 Strydom Tunnel Donation E. paucidentatus Stapf & Burtt Davy PR681 PBG CYAF032 E. paucidentatus Stapf & Burtt Davy PR710 PC 2 CYAF061 E. paucidentatus Stapf & Burtt Davy PR786 CWOFI CYAF131 E. paucidentatus Stapf & Burtt Davy PR849 UP, MBG CYPAF297 E. poggei Asch. PR813 CWOFI CYAF158 E. poggei Asch. PR911 UP, EF CYPAF359 E. princeps R.A. Dyer PR672 PBG CYAF023 E. princeps R.A. Dyer PR750 PC 2 CYAF096 E. princeps R.A. Dyer PR810 CWOFI CYAF155 E. princeps R.A. Dyer PR836 UP, MBG CYAF181 E. princeps R.A. Dyer PR871 UP, MBG CYPAF319 E. princeps R.A. Dyer d1037 NYBG DNA bank - - E. pterogonus R.A. Dyer & I. Verd. PR876 PC 1 CYPAF324 E. pterogonus R.A. Dyer & I. Verd. d1136 NYBG DNA bank - - E. relictus P.J.H. Hurter PR732 Martin Bruwer CYPAF291 Donation E. schaijesii Malaisse, Sclavo & Crosiers CC329 NYBG DNA bank - - E. schaijesii Malaisse, Sclavo & Crosiers CC447 NYBG DNA bank - - E. schmitzii Malaisse PR819 CWOFI CYAF164 E. schmitzii Malaisse d1137 NYBG DNA bank - - E. sclavoi A. Moretti, D.W. Stev. & De Luca d1038 NYBG DNA bank - - E. sclavoi De Luca, D. Stevenson & Moretti PR738 PC 2 CYAF084 E. sclavoi De Luca, D. Stevenson & Moretti PR790 CWOFI CYAF135 E. sclavoi De Luca, D. Stevenson & Moretti PR920 LBG CYPAF368 E. senticosus Vorster PR654 UJ CYAF005 E. senticosus Vorster PR663 PBG CYAF014 E. senticosus Vorster PR719 PC 2 CYAF070 E. senticosus Vorster PR812 CWOFI CYAF157 E. senticosus Vorster PR830 UP, MBG CYAF175 E. senticosus Vorster PR833 UP, MBG CYAF178 E. senticosus Vorster PR834 UP, MBG CYAF179 E. senticosus Vorster PR852 UP, MBG CYPAF300 Thornless E. senticosus Vorster d1039 NYBG DNA bank - -

160

E. septentrionalis Schweinf. d1138 NYBG DNA bank - - E. sp PR906 UP, EF CYPAF354 Morogora E. sp PR909 UP, EF CYPAF357 Bangoran E. sp PR913 UP, EF CYPAF361 Tsikapa E. sp. PR746 PC 2 CYAF092 E. sp. PR826 CWOFI CYAF171 E. tegulaneus subsp. tegulaneus Melville PR747 PC 2 CYAF093 E. tegulaneus subsp. tegulaneus Melville PR825 CWOFI CYAF170 E. tegulaneus subsp. tegulaneus Melville PR877 UP, EF CYPAF325 E. tegulaneus subsp. tegulaneus Melville d1040 NYBG DNA bank - - E. transvenosus Stapf & Burtt Davy d1041 NYBG DNA bank - - E. transvenosus Stapf & Burtt Davy PR656 UJ CYAF007 E. transvenosus Stapf & Burtt Davy PR665 PBG CYAF016 E. transvenosus Stapf & Burtt Davy PR727 PC 2 CYAF074 E. transvenosus Stapf & Burtt Davy PR797 CWOFI CYAF142 E. transvenosus Stapf & Burtt Davy PR829 UP, MBG CYAF174 E. transvenosus Stapf & Burtt Davy PR832 UP, MBG CYAF177 Tsualo E. transvenosus Stapf & Burtt Davy CS28 Modjadjiskloof KNPA691 E. transvenosus Stapf & Burtt Davy X E. woodii Sander PR783 CWOFI CYAF128 E. trispinosus (Hook.) R.A. Dyer PR680 PBG CYAF031 E. trispinosus (Hook.) R.A. Dyer PR723 PC 2 CYPAF289 E. trispinosus (Hook.) R.A. Dyer PR787 CWOFI CYAF132 E. trispinosus (Hook.) R.A. Dyer PR867 UP, MBG CYPAF315 E. trispinosus (Hook.) R.A. Dyer PR868 UP, MBG CYPAF316 E. trispinosus (Hook.) R.A. Dyer PR869 UP, MBG CYPAF317 E. trispinosus (Hook.) R.A. Dyer d1043 NYBG DNA bank - - E. turneri Lavranos & D.L. Goode d1044 NYBG DNA bank - - E. turneri Lavranos & D.L. Goode PR886 UP, EF CYPAF334 E. turneri Lavranos & D.L. Goode PR905 LBG CYPAF353 E. turneri Lavranos & D.L. Goode PR914 LBG CYPAF362 E. turneri Lavranos & D.L. Goode PR915 LBG CYPAF363 E. turneri Lavranos & D.L. Goode PR916 LBG CYPAF364 E. umbeluziensis R.A. Dyer PR670 PBG CYAF021 E. umbeluziensis R.A. Dyer PR736 PC 2 CYAF082 E. umbeluziensis R.A. Dyer PR815 CWOFI CYAF160 E. umbeluziensis R.A. Dyer PR858 UP, MBG CYPAF306 E. umbeluziensis R.A. Dyer d1046 NYBG DNA bank - - E. villosus Lem. PR652 UJ CYAF003 E. villosus Lem. PR671 PBG CYAF022 E. villosus Lem. PR743 PC 2 CYAF089 E. villosus Lem. PR816 CWOFI CYAF161 E. villosus Lem. PR837 UP, MBG CYAF182 E. villosus Lem. PR838 UP, MBG CYAF183 Broad Leaf E. villosus Lem. PR839 UP, MBG CYAF184 Thin Leaf E. villosus Lem. d1047 NYBG DNA bank - - E. whitelockii P.J.H. Hurter PR705 PC 1 CYAF056 E. whitelockii P.J.H. Hurter PR818 CWOFI CYAF163 E. whitelockii P.J.H. Hurter PR880 UP, EF CYPAF328 E. whitelockii P.J.H. Hurter PR882 LBG CYPAF330 E. whitelockii P.J.H. Hurter d1048 NYBG DNA bank - - E. woodii Sander PR675 PBG CYAF026

161

E. woodii Sander PR763 Nat Grobbelaar CYPAF292 Kranskloof Donation E. woodii Sander PR848 UP, MBG CYPAF296 E. woodii Sander PR875 PC, Palaborwa CYPAF323 E. woodii Sander d1139 NYBG DNA bank - - Stangeria eriopus (Kunze) Baill. PR706 PC 1 CYAF057 Grassland S. eriopus (Kunze) Baill. PR753 PC 2 CYAF099 Forest S. eriopus (Kunze) Baill. PR842B UP, MBG CYAF188 Forest S. eriopus (Kunze) Baill. PR843 UP, MBG CYAF189 Forest S. eriopus (Kunze) Baill. PR860 UP, MBG CYPAF308 S. eriopus (Kunze) Baill. PR924 LBG CYPAF372

162

Appendix 2. Canadian Center for DNA barcoding protocols

163

164

165

166

167