1 SYSTEMATICS OF TRIBE TRICHOCEREEAE AND POPULATION GENETICS OF Haageocereus (CACTACEAE) By MNICA ARAKAKI MAKISHI A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLOR IDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2008
2 2008 Mnica Arakaki Makishi
3 To my parents, Bunzo and Cristina, and to my sisters and brother.
4 ACKNOWLEDGMENTS I want to ex press my deepest appreciation to my advisors, Douglas Soltis and Pamela Soltis, for their consistent support, encouragement and generosity of time. I would also like to thank Norris Williams and Michael Miyamoto, me mbers of my committee, for their guidance, good disposition and positive feedback. Special thanks go to Carlos Ostolaza and Ftima Cceres, for sharing their knowledge on Peruvian Cactaceae, and for providing essential plant material, confirma tion of identifications, and their detailed observations of cacti in the field. I am indebted to the many individuals that have directly or indirectly supported me during the fieldwork: Carlos Ostolaza, Fti ma Cceres, Asuncin Cano, Blanca Len, Jos Roque, Mara La Torre, Richard Aguilar, Nestor Cieza, Olivier Klopfenstein, Martha Vargas, Natalia Caldern, Freddy Pelez, Yammil Ramrez, Eric Rodrguez, Percy Sandoval, and Kenneth Young (Peru); Stephan Beck, Noem Quispe, Lo rena Rey, Rosa Meneses, Alejandro Apaza, Esther Valenzuela, Mnica Zeballos, Freddy Centeno, Alfredo Fuentes, and Ramiro Lopez (Bolivia); Mara E. Ramrez, Mlic a Muoz, and Raquel Pinto (Chile). I thank the curators and staff of the herbaria B, F, FLAS, LPB, MO, USM, U, TEX, UNSA and ZSS, who kindly loaned specimens or made in formation available through electronic means. Thanks to Carlos Ostolaza for providing seeds of Haageocereus tenuis to Graham Charles for seeds of Blossfeldia sucrensis and Acanthocalycium spiniflorum to Donald Henne for specimens of Haageocereus lanugispinus ; and to Bernard Hauser and Kent Vliet for aid with microscopy. I would like to express my gratitude to cu rrent and former members of the Soltis Lab, FLAS Herbarium, and Florida Museum of Natural History, pa rticularly Matt Gitzendanner, Heather Loring, Pablo Speranza, Jennifer Tate Vaughan Symonds, Luiz de Oliveira, Mi-Jeong Yoo, Sam Brockington, Evgeny Ma vrodiev, Summer Scobell, Chris tine Edwards, Josh Clayton,
5 Lucas Majure, Claudia Segovia, Lorena Endara, Mark Whitten, Kent Perkins, Andres Lopez and Tuuli Makinen. My deepest tha nks to my old and new friends, for their patience and encouragement at all times. This work was supported by grants from the American Society of Plant Taxonomy (ASPT), the Botanical Society of America (BSA), the Cactus and Succulent Society of America (CSSA), the International Association for Plant Taxonomy (IAP T), and the National Science Foundation (NSF DEB-0608273).
6 TABLE OF CONTENTS page ACKNOWLEDGMENTS...............................................................................................................4 LIST OF TABLES................................................................................................................. ..........8 LIST OF FIGURES.........................................................................................................................9 ABSTRACT...................................................................................................................................11 CHAP TER 1 INTRODUCTION: SYSTEMATIC STATUS OF THE CACTACEAE, TRIBE TRICHOCE REEAE, AND GENUS Haageocereus ..............................................................13 Cactaceae................................................................................................................................13 Tribe Trichocereeae (Subfamily Cactoideae).........................................................................16 Haageocereus Backeb. ...........................................................................................................18 2 SYSTEMATICS OF TRIBE TRICHOCEREEAE AND GENUS Haageocereus ................24 Introduction................................................................................................................... ..........24 Materials and Methods...........................................................................................................30 Taxon Sampling...............................................................................................................30 DNA Isolation, Amplification, and Sequencing.............................................................. 31 Alignment ans Phylogenetic Analysis............................................................................. 34 Results.....................................................................................................................................36 Descriptive Data.............................................................................................................. 36 Parsimony Analysis......................................................................................................... 36 Maximum Likelihood Analysis....................................................................................... 38 Bayesian Analysis........................................................................................................... 39 Discussion...............................................................................................................................40 Phylogenetic Inference....................................................................................................40 Phylogenetic Relationships in Trichocereeae..................................................................41 Phylogenetic Relationships in Haageocereus .................................................................48 Radiation of Trichocereeae.............................................................................................. 49 Conclusions.............................................................................................................................50 3 DEVELOPMENT OF POLYMORPHIC MICROSATELLITE LOCI FOR THE EXAMINATION OF GENETIC DI VERSITY IN Haageocereus ........................................65 Introduction................................................................................................................... ..........65 Materials and Methods...........................................................................................................66 Results and Discussion......................................................................................................... ..67
7 4 CHROMOSOME COUNTS AND EVIDENCE OF POLYPLOIDY IN Haageocereus AND OTHE R TAXA OF TRIBE TRICHOCEREEAE......................................................... 72 Introduction................................................................................................................... ..........72 Materials and Methods...........................................................................................................73 Results and Discussion......................................................................................................... ..74 5 CLONAL REPRODUCTION AND APOMIXIS IN Haageocereus .....................................82 Materials and Methods...........................................................................................................84 Examination of Apomixis................................................................................................ 85 Microsatellite Amplification and Analysis...................................................................... 85 Results and Discussion......................................................................................................... ..86 Agamospermy..................................................................................................................86 Genetic Diversity.............................................................................................................86 Implications for Conservation......................................................................................... 88 6 EXAMINATION OF RETICULATE EVOLUTION BETWEEN Haage ocereus AND Espostoa ..................................................................................................................................91 Introduction................................................................................................................... ..........91 Materials and Methods...........................................................................................................96 Plant Sampling................................................................................................................. 96 Nuclear Microsatellite Data............................................................................................. 96 Chloroplast DNA Data.................................................................................................... 97 Results and Discussion......................................................................................................... ..98 7 CONCLUDING REMARKS................................................................................................ 108 REFERENCES............................................................................................................................111 BIOGRAPHICAL SKETCH.......................................................................................................125
8 LIST OF TABLES Table page 1-1 Classification histor y at the generic level for tribe Trichocereeae .................................... 22 2-1 List of species in the Trichocere eae and m embers of other tribes (outgroups) included in the present study.............................................................................................. 53 2-2 DNA markers and primer sequences used for amplification and sequencing in the present study. ................................................................................................................. ....57 2-3 Sequence information for individual partitions................................................................. 57 2-4 Sequences of degenerate pr im ers designed for amplification of PI and AP3, and LFY ....58 2-5 Results of parsimony analyses of indi vidual and com bined data sets (including and excluding hybrids)............................................................................................................. 59 3-1 Characterization of 19 polym orphic m icrosatellite loci in Haageocereus tenuis ( Ht ) and Haageocereus pseudomelanostele ( Hp )...................................................................... 70 3-2 AMOVA for three populations of H. pseudomelanostele and five m icrosatellite loci...... 71 3-3 Summary of allele counts for four species of Haageocereus and five m icrosatellite loci......................................................................................................................................71 4-1 Somatic chromosome numbers for taxa in th e Trichocereeae and members of other tribes...................................................................................................................................78 5-1 Multilocus microsatellite genotype of H. tenuis and variability found in three populations of a related widespread sexual species using the sam e set of primers........... 89
9 LIST OF FIGURES Figure page 2-1 Schematic map of LEA FY ..................................................................................................59 2-2 Majority-rule consensus of 10,577 most parsim onious trees from the combined analysis of chloroplast ( rpoB, rpl16, 23S) and nuclear (ITS) sequence data, for 107 taxa representing the Tric hocereeae and 11 outgroups...................................................... 60 2-3 Majority-rule consensus of 5000 most parsim onious trees obtained from the combined analysis of chloroplast ( rpoB rpl16 23S) and nuclear (ITS) sequence data, excluding hybrid individuals..................................................................................... 61 2-4 Maximum likelihood tree from the com bined analysis of chloroplast ( rpoB, rpl16, 23S) and nuclear (ITS) sequence data............................................................................... 62 2-5 Maximum likelihood phylogram obtained from the combined analysis of chloroplast ( rpoB rpl16 23S) and nuclear (ITS) sequence da ta, excluding hybrid individuals. ........ 63 2-6 Majority-rule consensus tree from 1800 post-burn-in trees sampled by Bayesian m cmc analysis of the combined data set of chloroplast ( rpoB, rpl16, 23S) and nuclear (ITS) sequence data, excluding hybrids................................................................ 64 4-1 Somatic chromosomes and photographs of actual plants from diploid and polyploid species representing the Trichocereeae.............................................................................. 81 5-1 Haageocereus tenuis .........................................................................................................90 6-1 Naturally occurring bigeneric hybrids in Cactaceae centered in Peru and Bolivia ......... 101 6-2 Names attributed to Ha ageocereus and Peruvocereus that are possibly referring to xHaagespostoa .................................................................................................................102 6-3 Espostoa x Haageocereus hybrid combinations examin ed in the present study............. 103 6-4 Multilocus microsatellite genotypes for putative parents and hybrid individuals xHaagespostoa climaxantha and putative parents: Espostoa lanata and Haageocereus pseudomelanostele ...................................................................................104 6-5 Multilocus microsatellite genotypes for putative parents and hybrid individuals xHaagespostoa sp., and putative parents: Espostoa melanostele and Haa geocereus pseudomelanostele ...........................................................................................................105 6-6 Multilocus m icrosatellite genotypes for putative parents and hybrid individuals xHaagespostoa sp., and putative parents: Espostoa melanostele and Haageocereus pseudomelanostele ...........................................................................................................106
10 6-7 Amplifications of chloroplast region psbA-trnH, psbE-petL, and 23S from three hybrid com binations involving Espostoa and Haageocereus..........................................107
11 Abstract of Dissertation Pres ented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy SYSTEMATICS OF TRIBE TRICHOCEREEAE AND POPULATION GENETICS OF Haageocereus (CACTACEAE) By Mnica Arakaki Makishi December 2008 Chair: Douglas E. Soltis Cochair: Pamela S. Soltis Major: Botany The Cactaceae comprises approximately 1800 species, found mostly in the tropical deserts of North and South America. The family has been the object of in tensive study but it is not until recently that molecular tools have been used to elucidate its phylogenetic relationships. Most of the Cactaceae found in the Central Andes of Peru and Bolivia belong to tribe Trichocereeae. Phylogenetic relationships within the tribe have been historically controversial, therefore, a molecular study using chloroplast and nuclear mark ers was designed to elucidate phylogenetic relationships among species of Haageocereus and the relationship of this genus to the rest of taxa in the Trichocereeae. We sequenced cpDNA ( rpoB rpl16 and 23S) and nrDNA (ITS1 and ITS2) of 107 ingroup taxa comprisi ng the Trichocereeae and 11 outgroup taxa from different tribes. The analyses support a monophyletic Trichocereeae (including Praecereus euchlorus ), the exclusion of seven genera found to be more closely related to Brazilian Cereeae, and the maintenance of several genera hist orically placed in Trichocereeae, including Haageocereus and Espostoa. Based on our results, the taxono my of the Trichocereeae does not reflect its phylogeny and is in need of re vision. We describe 19 polymorphic dinucleotide microsatellite loci isolated from two species of Haageocereus ( H. tenuis and H. pseudomelanostele ) that have probed their utility in the analysis of populati on differentiation and genetic diversity throughout Haageocereus, as well as in the anal ysis of clonal and hybrid
12 individuals. Chromoso me numbers for a total of 54 indivi duals representing 14 genera and 39 species of Cactaceae, mostly in tribe Trichoceree ae, are reported, demonstrating the presence of polyploidy in the Trichocereeae and its dominance in certain genera. Cl onal propagation and apomixis in the triploid Haageocereus tenuis has significant implications for the evolutionary biology and ecology of Haageocereus and other clonal Cactaceae.
13 CHAPTER 1 INTRODUCTION: SYSTEMATIC STATUS OF THE CACTACEAE, TRIBE TRICHOCE REEAE, AND GENUS Haageocereus Cactaceae The Cactaceae are on e of the largest and mo st morphologically diverse families in the order Caryophyllales. They are prominent among succulent plants and a significant component of the American flora. The family comprises approximately 120 genera and 1800 species (Anderson 2001, Hunt et al 2006) distributed from southern Canada to Patagonia in Argentina and from sea level to elevations of 5 200 m (17060 feet) in the Andes. One species, Rhipsalis baccifera is found in Africa and Sri Lanka. Synapomorphies for Cactaceae include: stems that are usually highly succulent and that replace leaves in their photosyntheti c function, long-lived thick epidermis, and either CAM or C3 phot osynthesis (Gibson and Nobel 1986). Stems may have ridges, tubercles or be flattened. The Cactaceae are highly tolerant to extreme environmental conditions due to the development of important adaptations (e.g., stem succulence for water storage, CAM photosynthesis, waxy cuticular thickenings, hairs and spines). Cactaceae comprise four subfamilies. The subfamily Pereskioideae, sister to the rest of the family (Gibson and Nobel 1986, Mauseth a nd Landrum 1997, Butterworth and Wallace 2005, Edwards et al. 2005), has leaves typical of other e udicots. In contrast, members of the subfamilies Opuntioideae and Maihuenioideae may ha ve small fleshy leaves that are later lost. All cacti have long shoots and shor t shoots (areoles) that are ex tremely congested. Leaves are reduced to spines in most members of subfam ily Cactoideae, and stems of Opuntioideae are covered with irritant hairs (glochids). The flower s are solitary, usually te rminal, and very often the ovary is sunken into surrounding stem tissue calle d a pericarpel (except in Pereskioideae), so that the outer wall of the ovary has areoles. In all Cactaceae the flow er has a fleshy nectar chamber and many petaloid internal tepals intergrading with sepaloid or bract-like tepals toward
14 the outside; the stamens are also numerous and usually arranged in two cycles (Buxbaum 1950a, 1950b, 1953, Anderson 2001). Although most species of cacti live in deserts and are specialized for coping with extreme drought and irradiati on, a large group of cacti occupies tropical wet forests as epiphytes. These epiphytes usually have flattened stems and lose their spines. Examples include Lepismium Hylocereus, and Selenicereus According to Hershkovitz (1991), Hershkovi tz and Zimmer (1997), Applequist and Wallace (2001), Nyffeler (2007), and Stevens (2001-2008), the Cactaceae may be included in a polyphyletic Portulacac eae (showing a close relationship with Anacampseros Portulaca and Talinum ) from which Cactaceae, Basellaceae, Didier eaceae and some other small families of problematic placement arise. However, sampling within Caryophyllales has been insufficient to date to address these relationships. Thus, no strong conclusions can be made in terms of interfamilial relationships. Phylogenetic relationships within Cactaceae are controversial and the family has been the object of intensive study, but only recently have molecular tools been used to elucidate phylogenetic relationships. Studies by Nyffel er (2002), Crozier (2004), Griffith (2004a), Butterworth and Wallace (2005), and Edwards et al. (2005) on basal cactus phylogeny, using morphology and gene or spacer sequences representative of the three plant genomes, agree only partially, suggesting different major clades for the Cactaceae. Edwards et al. (2005) suggest a basal split in Cactaceae betw een Caribbean species of Pereskia and all other cacti. Butterworth and Wallace (2005) report an additional basal clade containing species of Pereskia found primarily in Brazil, Uruguay, Paraguay, Argentin a and Bolivia. Both studies, however, agree that Pereskia might not be monophyletic. Griffith ( 2004a) provides an alternative to the Pereskia -asbasal hypothesis, suggesting that Opuntioideae may represent the basal lineage in the Cactaceae.
15 Supporting evidence includes: the deep lineages of all subfamilies (excluding Pereskioideae) exhibit geophytic storage tissues and a simple architecture; the suggest ed nearest lineage of Cactaceae are the Portulacaceae, which also co ntain small geophytic, not woody, succulents. In studies by Crozier (2004), Blossfeldia was found to be sister to all other Cactoideae. Based on the morphological distinctiveness of the genus, plus a well-supported molecular phylogeny, she proposed the creation of a new s ubfamily: Blossfeldioi deae. The same study proposed the resurrection of subfamily Rhipsalidoideae, which would contain all former members of the Cactoideae (t ribes Browningieae, Calymmant heae, Cereae, Hylocereae, Notocacteae, Pachycereae, Rhipsalideae, Trichocer eae) with the exception of North American globular cacti (tribe Cacteae), which, due to proper application of nomencla tural rules, would be the only tribe remaining in the Cactoideae. Ho wever, proposals by Griffith (2004) and Crozier (2004) have not been widely accepted. The work of Franz Buxbaum (1958), who worked to develop a natural classification of the family based on morphological featur es, continues to be the basis for current cactus classification systems (Anderson 2001). Nevertheless, the many examples of convergent morphologies found in the Cactaceae indicate that morphology-based cl assifications will continue to be problematic, unless they are supported by evolutionary stud ies with a molecular phylogenetic foundation. Morphological characters, though, are very valu able, and several morphology-based groupings will probably be confirmed by current and future studies. Subfamilies Pereskioideae, Opuntioideae, and Cactoideae Schumann ( 1903) and Maihuenioideae Fearn (1996) (see Anderson 2001) have been widely accepted with some minor modifications in the number of tribes and genera they in clude (Buxbaum 1958, Barthlott an d Hunt 1993, Hunt 1999, Anderson 2001, Hunt et al 2006). Since 1984, the International Or ganization for Succulent Plant Study
16 (IOS) through its International Cactaceae Syst ematics Group has been publishing periodical synopses of the classification based on the cons ensus opinion of many cacti specialists (Hunt 1993, 1999, Hunt et al 2006). Although this approach might be controversial, supporters believe that a classification by committee rather than individual efforts is a necessity in the Cactaceae owing to its complex taxonomic history. A useful outcome has been the production of checklists, with information on distribution and taxonomic status, that serve as references for CITES (Convention on International Trade in Endangered Sp ecies) authorities in th e control of species trade. Tribe Trichocereeae (Subfamily Cactoideae) Morphological synapom orphies for tribe Tric hocereeae are a treelike or shrubby habit; columnar or globose, normally unsegmented, tube rculate or ribbed stems; flowers that are usually tubular, radially or bila terally symmetrical, nocturnal or diurnal; pericarpel with scales and/or hairs; fruits fleshy, sometimes dehiscen t; and numerous small to medium-size seeds with hilum and micropyle conjunct or fused (Anderson 2001). The tribe, with the exclusion of Harrisia (of doubtful placement in Trichocereeae), is entirely South American, distributed south of the Equator from the Galapagos Islands to Patagonia. Peru and Bolivia form the second mo st diverse area for Cactaceae, after Mexico (Taylor 1997), and the Trichoceree ae is the dominant group in the region, with about 80 percent of the species being endemic. This area contai ns the greatest concentration of taxonomically doubtful species of Cactaceae and is where most study is needed to resolve questions of taxonomic and conservation status (Taylor 1997). The number of species of Trichocereeae has b een fluctuating throughout the years and as currently recognized the trib e comprises 26 genera and about 400 species (Hunt 1993, 1999, Hunt and Taylor 1990, Hunt et al. 2006). The largest genera are Echinopsis (128 species),
17 Gymnocalycium (71), Cleistocactus (48), and Rebutia (41) Four genera contain 17-20 species ( Espostoa, Haageocereus Harrisia and Matucana), and the remaining 18 genera contain fewer than 10 species each ( Acanthocalycium Arthrocereus, Brachycereus Cephalocleistocactus, Denmoza Discocactus Espostoopsis Facheiroa Lasiocereus, Leocereus Mila Oreocereus, Oroya, Pygmaeocereus, Rauhocereus, Samaipaticereus Weberbauerocereus, and Yungasocereus ). Phylogenetic relationships within the tribe have been historically c ontroversial (Buxbaum 1958, Barthlott and Hunt 1993, Hunt 1999, A nderson 2001). Buxbaum (1958) included 21 species in the Trichocereeae, placed under subtribes Trichocereinae, Rebutinae and Borzicactinae (Table 1-1). The classification of the tribe underwent a nu mber of modifications by Barthlott and Hunt (1993), who merged several genera and added six more. Hunt (1999) and Anderson (2001) for the most part maintained ge nera proposed by Barthlott and Hunt (1993) and added six more. Hunt (1999) and Anderson (2001) is the classification followed widely today. However, recent molecular data have indicate d that important taxonomic rearrangements are needed (Nyffeler 2002, Arakaki 2002, Crozier 2002, Arakaki et al 2003, Lendel et al. 2006, 2007). Phylogenetic studies by Nyffel er (2002), which included repres entatives of all tribes of Cactaceae, did not obtain resolution at the tr ibal level, placing the Trichocereeae in an unresolved clade with four other tribes. The clade has been called RNBCT for RhipsalideaeNotocacteae-Browningieae-Cereeae-Trichocereeae. Studies by Lendel et al. (2006, 2007) support the non-monophyly of Tr ichocereeae and provide evidences for a geographically coherent subclade, comprising Acanthocalycium Cleistocactus, Echinopsis and 17 other genera, but excluding Gymnocalycium and R ebutia In addition, several members of the Trichocereeae (i.e. Discocactus Espostoopsis Facheiroa and Leocereus ) were found to be more closely
18 related to Brazilian memb ers of tribe Cereeae; and Brachycereus was found to be sister to Jasminocereus (both in the Galpagos Isla nds), this clade sister to Armatocereus and nested in the HLP clade of Nyffeler (2002). Molecular work by Arakaki (2002) provided better resolution within some clades in Trichocereeae. For ex ample, their study supports the monophyly of Weberbauerocereus and Cleistocactus, the recognition of Haageocereus as sister to Espostoa (and not a close relationship to Weberbauerocereus as previously suggest ed), the exclusion of Borzicactus and Loxanthocereus from Cleistocactus and the recognition of Lasiocereus as a well-defined genus, placed outside the Trichocer eeae. These studies indicate that Trichocereeae (minus Lasiocereus) is monophyletic. However, only 14 gene ra and 35 species of Trichocereeae were included (mainly from the central Andes). Several genera were repr esented by only a single species. Furthermore, the overall br anch support obtained was low. Haageocereus Backeb. The genus Haageocereus one of the most taxonom ically complex genera in Trichocereeae, is a shrubby or tree-like columnar cactus largely restricted to the western slopes of the Peruvian Andes with one species extend ing into northern Chile. Although some species of Haageocereus cover wide geographic ranges, others ar e restricted to small areas (only one square km) in the vicinity of the most populat ed cities in Peru. The genus is considered vulnerable, and seven species as well as five s ubspecies have been classified according to the categories and criteria of th e IUCN (The World Conservation Union) as in danger of disappearance throughout all or a significant portion of their ranges (Ostolaza 1996, Anderson 2001). As is true of most genera in Trichocereeae, Haageocereus is poorly understood and systematic studies are badly needed, as evidence d by the proliferation of names and descriptions.
19 There are approximately 120 species plus subspecies names for Haageocereus, due in large part to horticultural interests. The taxonomic difficulty in Haageocereus is largely the result of extensive morphological variability. As in most genera in Cactaceae, this variability has been related to environmental gradients (Gibson and Nobel 1986). However, in se veral species groups, variability is related to changes associated with hybridization and polyploidy. Studies in Opuntia (Grant 1971, 1980, Grant and Grant 1979, Rebman and Pinkava 2001) and Mammillaria (Katagiri 1953, Remski 1954) show that polyploidy and hybridization ar e major, but previous ly underappreciated, evolutionary forces in these genera. Patterns of relationship in both genera were found to be complex due to hybrid swarms, vegetative propagation and polyploidy. These processes may be frequent in cacti (Anderson 2001), but the relative importance of hybridization and polyploidy in the family remains uncertain because so few cacti have been examined in detail. In Cactaceae, chromosome counts are few, a nd polyploidy has been re ported sporadically: Opuntia (Pinkava and McLeod 1971, Pinkava et al 1973, 1977, Ross 1981, Pinkava and Parfitt 1982), Mammillaria (Katagiri 1953, Remski 1954), Echinocereus (Cota and Philbrick 1994), and Selenicereus (Lichtenzveig et al 2000). Only 2% of Trichocereeae have published chromosome counts and there are no chromo some counts for the genus Haageocereus It is suggested that both hybridization and polyploidy have played prominent roles in Haageocereus. Hybrids have been reported not only within the genus, but also between Haageocereus and and Espostoa (Rowley 1994). Backeberg (1935) described Haageocereus, whose type is H. pseudomelanostele and many species were described subsequently by W. Rauh, K. Backeberg and F. Ritter (Anderson 2001). Although in 1999 the IOS reduced the number of recognized taxa from 72 species (plus
20 48 infraspecific taxa) to 21 species (plus six infraspecific taxa), disagreement among students of cacti persist, mainly because decisions were ma de without additional studies (Anderson 2001, C. Ostolaza pers. comm.). A recently published taxonomic treatment of Haageocereus (Caldern et al. 2007) has reduced the number of taxa to nine species ( H. acranthus, H. chilensis, H. decumbens, H. lanugispinus, H. platinospinus, H. pseudomelanostele, H. repens, H. tenuis, H. versicolor ), three of which have additional subspecies ( H. acranthus subsp. backebergii, and subsp. zonatus ; H. pseudomelanostele subsp. aureispinus subsp. carminiflorus, and subsp. turbidus ; and H. versicolor subsp. pseudoversicolor). The study is based on the examination of bibliographic records and detailed analysis of herbarium specimens and all species in the field and in cultivation. Some new results are the description of a new subspecies ( H. acranthus subsp. backebergii) and a new combination ( H. versicolor subsp. pseudoversicolor). Systematic and population studies are needed to determine the limits of the species currently recognized in Haageocereus, and to evaluate whether species that are no longer recognized may be, in fact, valid. The contributi on of this dissertation is presented in five chapters, one of them dealing with molecular phylogenetics, one dealing with cytology, and the other three dealing with studi es at the population level. Chapter 2 will clarify generic relationships in the Tricho cereeae and the placement of Haageocereus, and will elucidate interspecific relationships within Haageocereus and Espostoa. The improved phylogeny of Trichocereeae together with the work of Caldern et al. (2007) will create a framework for further research on character evolution, bi ogeography, ecology, and reproductive biology of Haageocereus and other Trichocereeae, and contribut e to conservation efforts of South American Cactaceae. Chapter 3 presents the characterization of microsatellite loci designed for Haageocereus. These markers will be used to determine the amount and distribution of genetic
21 variability in endangered species of Haageocereus compared to widespread close relatives. The same markers will be applied in chapters 5 and 6. Chapter 4 reviews the status of cytogenetic studies in the Trichocereeae and evaluates the prevalence and nature of polyploidy in Haageocereus, the Trichocereeae and othe r tribes of Cactaceae. Chapter 5 evaluates the presence of clonal propagation and apomixis, using H. tenuis one of the most remarkable examples of clonal propagation in the Cactaceae, and Chapter 6 will examine reticulate evolution via hybridization with the genus Espostoa. Chapter 7 is a summary of this research contribution and concluding remarks.
22 Table 1-1. Classification history at the generic level for tribe Trichocereeae; based on Buxbaum (1958), Barthlott and Hunt (19 93), Hunt (1999) and Anderson (2001). Buxbaum (1958) Barthlott and Hunt (1993) Hunt (1999), Anderson (2001) Trichocereus [incl. Heliocereus, Leucostele, ----Roseocereus, Weberbauerocereus ] Echinopsis [incl. Pseudolobivia ] Echinopsis [incl. Trichocereus, Chamaecereus, Acanthocalycium, Echinopsis Soehrensia, Setiechinopsis, Lobivia, Pseudolobivia, Helianthocereus ] Haageocereus [incl Neobinghamia, Peruvocereus ] Haageocereus [incl. Loxanthocereus (in part), Weberbauerocereus] Haageocereus ----Weberbauerocereus Arthrocereus [incl. Setiechinopsis ] Arthrocereus Arthrocereus Espostoa [incl. Pseudospostoa ] Espostoa [incl. Binghamia, Vatricaria, Pseudoespostoa,Thrixanthocereus ] Espostoa Soehrensia ----Acanthocalycium --Acanthocalycium Lobivia [incl. Acantholobivia ] ----Rebutia [incl. Aylostera, Cylindrorebutia, Rebutia [incl. Weingartia, Mediolobivia, Aylostera, Rebutia, Rebutia Digitorebutia, Mediolobivia, Pygmaeolobivia ] Sulcorebutia ] Chamaecereus ----Mila Mila Mila Loxanthocereus [incl. Maritimocereus ] ----Borzicactus [incl. Bolivicereus, Clistanthocereus ] ----Denmoza Denmoza Denmoza Cleistocactus Cleistocactus [incl. Borzicactus, Seticereus, Loxanthocereus Cleistocactus Bolivicereus, Cephalocleistocactus, Akersia, Seticleistocactus, Winteria, Winterocereus, Hildewintera, Borzicactella ]
23 Table 1-1. (cont.) Buxbaum (1958) Barthlott and Hunt (1993) Hunt (1999), Anderson (2001) ----Cephalocleistocactus Oroya Oroya Oroya Matucana Matucana [incl. Submatucana, Eomatucana, Arequipiopsis ] Matucana Oreocereus Oreocereus [incl. Arequipa, Morawetzia ] Oreocereus Morawetzia ----Arequipa ----Seticereus ------Espostoopsis [incl. Gerocephalus, Austrocephalocereus ] Espostoopsis --Brachycereus Brachycereus --Facheiroa [incl. Zehntnerella ] Facheiroa --Samaipaticereus Samaipaticereus --Gymnocalycium [incl. Brachycalycium ] Gymnocalycium --Discocactus Discocactus ----Leocereus ----Pygmaeocereus ----Rauhocereus ----Yungasocereus ----Harrisia ----Lasiocereus
24 CHAPTER 2 SYSTEMATICS OF TRIBE TRICHOCEREEAE AND GENUS Haageocereus Introduction The Cactaceae are on e of the largest and mo st morphologically diverse families in the order Caryophyllales. They are prominent among succulent plants and a significant component of the American flora. The family comprises approximately 120 genera and 1800 species (Anderson 2001, Hunt et al 2006) distributed from southern Canada to Patagonia in Argentina and from sea level to elevations of 5200 m in the Andes. One species, Rhipsalis baccifera is found in Africa and Sri Lanka; Opuntia species have been introduced in Australia, Africa and Asia, where they become invasive (USDA 2008). A lthough most species of cacti live in deserts and are specialized for coping with extreme drought and irradiation, a large group of cacti occupies tropical wet forests as epiphytes. Ce nters of diversity are the southwestern United States and Mexico, the central Andes in South America, and eastern Br azil (Taylor 1997, Nyffeler 2002). Although there are no known fossils of Cactaceae, it has been proposed that the family could not have originated prior to the Late Cretaceous, about 90-65 million years ago (mya), following the break-up of Gondwana (Axelr od 1979, Mauseth 1990, Hershkovitz and Zimmer 1997). However, considering the re latively recent formation of de sert habitats for Cactaceae in North America (about 13 mya) and South America (associated with formation of the Humboldt Current as recently as 5 mya, and the rise of the Andes between 65-17 mya), it is most likely that the explosive diversification of the family o ccurred between 5 and 17 mya (Van der Hammen 1974, Simpson 1975, van der Hammen and Cleff 1986, Mauseth 1990, Hershkovitz and Zimmer 1997, Ritz et al 2007). Other angiosperm families or gene ra proposed to have radiated rapidly after colonization of the Andes include Gentianella (Gentianaceae; von Hagen and Kadereit
25 2001), Halenia (Gentianaceae; von Hagen and Kadere it 2003), Valerianaceae (Bell and Donoghue 2005), and Lupinus (Fabaceae; Hughes and Eastwood 2006), among others. Bell and Donoghue (2005) estimated that the radiation of Valerianaceae in the paramos of the northern Andes occurred at a rate of ca. 0.80.34 speci es/my, and Hughes and Eastwood (2006) found the average-per-lineage species dive rsification rate for the Andean Lupinus to be 2.5-3.72 species/my. Hughes and Eastwood (2 006) suggested that these rapid Andean radiations are comparable to those observed in the coloniza tion of newly formed is lands. Prado and Gibbs (1993) and Prado (2000) proposed th e Pleistocenic Arc hypothesis fo r the range fluctuations in seasonally dry forests of tropical South America. The analysis of distribution patte rns of 80 plant genera or species, including one Cactaceae (Pereskia ), indicates that seasonally dry forests in South America are remnants of once-extensive forest s, with their nucleus in southeastern Brazil, Paraguay, northern Argentina, and Bolivia. Thes e are hypothesized to have reached their dry climatic maxima during the Pleistocene, and at that time extended into the Amazonian region and percolated into the Andean region perhaps vi a the Maran gap (in northern Peru) or from south to north Peru. From there, they expande d in different directions from low to high elevations. Cactaceae comprise four subfamilies: Pere skioideae, Opuntioideae, Cactoideae, and Maihuenioideae (Anderson 2001). Of these subfam ilies, the Cactoideae are the most speciesrich, containing more than 80% of all species in the family. Cactoideae comprise nine tribes, one of which is the Trichocereeae (Anderson 2001, Nyffeler 2002, Hunt et al. 2006). Most of the Cactaceae found in the central Andes of Peru and Bolivia (the second richest area for Cactaceae after Mexico, Taylor 1997) belong to tribe Trichocereeae. This ge ographic area also contains the
26 greatest concentration of taxonomically doubtful sp ecies of Cactaceae and is where most study is needed to resolve questions of taxonomic and co nservation status in the family (Taylor 1997). Members of Trichocereeae are pa rticularly diverse in life form, ranging from globular to columnar, and from small and caespitose to shrubs and tree-like plants. The morphologies associated with their particular pollinators and dispersal agents are equally diverse. Morphological synapomorphies for tribe Trichocereeae include normally unsegmented, tuberculate or ribbed stems; flower s that are usually tubular, radi ally or bilaterally symmetrical, nocturnal or diurnal; pericarpel w ith scales and/or hairs; fleshy fruits, sometimes dehiscent; and numerous small to medium-sized seeds with h ilum and micropyle conjunct or fused (Gibson and Nobel 1986, Anderson 2001). Trichocereeae is entirely South American, with the exception of Harrisia (of doubtful placement in Trichocereeae), which is also found in Florida and the Caribbean. The tribe is distributed south of the Equator from the Ga lpagos Islands to Patagonia. The number of recognized species of Trichocereeae has fluctuate d, and the tribe currently comprises 26 genera and about 400 species (Hunt 1993, 1999, Hunt et al. 2006). The largest genera are Echinopsis (128 species), Gymnocalycium (71), Cleistocactus (48), and Rebutia (41) Four genera contain 17-20 species ( Espostoa, Haageocereus, Harrisia and Matucana), and the remaining 18 genera contain fewer than 10 species each ( Acanthocalycium Arthrocereus, Brachycereus Cephalocleistocactus, Denmoza Discocactus Espostoopsis Facheiroa Lasiocereus, Leocereus, Mila Oreocereus Oroya Pygmaeocereus, Rauhocereus, Samaipaticereus, Weberbauerocereus, and Yungasocereus ). Phylogenetic relationships within Cactaceae in general, and Trichoc ereeae in particular, have been highly controversial (Buxbaum 1958, Barthlott and Hunt 1993, Hunt 1999, Anderson
27 2001), and the family has been the object of in tensive study, but only recently have molecular tools been used to elucidate phylogenetic re lationships (Nyffeler 2002, Crozier 2004, Griffith 2004, Butterworth and Wallace 2005, Edwards et al. 2005). Buxbaum (1958) included 21 species in th e Trichocereeae (Table 1-1), placed under subtribes Trichocereinae, Rebutin ae, and Borzicactinae. The classi fication of the tribe underwent a number of modifications by Barthlott and Hunt (1993), who merged several genera and added six more. Hunt (1999) and Anderson (2001) genera lly retained genera proposed by Barthlott and Hunt (1993) and added six more. Hunt (1999) and Anderson (2001) are the classifications followed widely today. However, recent molecular data have indicated th at important taxonomic rearrangements are needed (Nyffele r 2002, Arakaki 2002, Crozier 2002, Arakaki et al 2003, Lendel et al. 2006, 2007). Phylogenetic studies by Nyffeler (2002), which included representatives of all tribes of Cactaceae, did not obtain resolution at the tribal level, placing the Trichocereeae in an unresolved clade with four other tribes. The clad e has been called RNBCT for Rhipsalideae-Notocacteae-Browningiea e-Cereeae-Trichocereeae. Studies by Lendel et al. (2006, 2007) support the non-monophyly of Trichocereeae and provide evidence for a geographically coherent subclade, comprising Acanthocalycium Cleistocactus Echinopsis and 17 other genera, but excluding Gymnocalycium and Rebutia In addition, several members of the Trichocereeae (i.e., Discocactus, Espostoopsis Facheiroa and Leocereus ) were found to be more closely related to Brazilian members of tribe Cereeae, and Brachycereus was found to be sister to Jasminocereus, this clade sister to Armatocereus and nested in the HLP clade of Nyffeler (2002). Molecular work by Arakaki (2 002) provided better re solution within some clades in Trichocereeae. For exampl e, the study supports the monophyly of Weberbauerocereus and Cleistocactus, the recognition of Haageocereus as sister to Espostoa (rather than closely
28 related to Weberbauerocereus as previously suggested), the exclusion of Borzicactus and Loxanthocereus from Cleistocactus and the recognition of Lasiocereus as a well-defined genus, placed outside the Trichocereeae. This study also indicates that Trichocereeae (minus Lasiocereus ) is monophyletic. However, only 14 genera and 35 species of Trichocereeae were included (mainly from the central Andes). Seve ral genera were represented by only a single species. Furthermore, the overall branch suppor t obtained in parsimony analyses was low. The genus Haageocereus one of the most taxonomically complex genera in Trichocereeae, is a shrubby or tree-like columnar cactus largely restricted to the western slopes of the Peruvian Andes with one species extend ing into northern Chile. Although some species of Haageocereus cover wide geographic ranges, others ar e restricted to small areas (only one square km) in the vicinity of the most populated ci ties in Peru. As is true of most genera in Trichocereeae, Haageocereus is poorly understood, and systematic studies are badly needed, as evidenced by the proliferation of names and de scriptions. There are approximately 120 species plus subspecies names for Haageocereus, due in large part to horticultural interests. The characters used to circumscribe Haageocereus are variable and found in some degree and/or combination in species of closely related genera. This variability has been related to environmental gradients (Gibson and Nobel 1986), as well as to biological phenomena such as polyploidy and hybridization (Rowley 1982, 1994, Anderson 2001). This morphological variability has resulted in a co mplicated taxonomy in which some authors have questioned the validity of most species described for the genus. Backeberg (1935) described Haageocereus, whose type is H. pseudomelanostele and many species were described subsequently by W. Rauh, K. Backeberg, and F. Ritter (Anderson 2001). Although in 1999 the Intern ational Organization for the study of Succulents (IOS)
29 reduced the number of recognized taxa from 72 species (plus 48 in fraspecific taxa) to 21 species (plus six infraspecific taxa), disagreement am ong students of cacti pers ist, mainly because taxonomic decisions were made without additio nal studies (Anderson 2001, C. Ostolaza pers. comm.). A recently published taxonomic treatment of Haageocereus (Caldern et al. 2007) has reduced the number of ta xa to nine species ( H. acranthus, H. chilensis, H. decumbens, H. lanugispinus, H. platinospinus, H. pseudomelanostele, H. repens, H. tenuis, H. versicolor ), three of which have additional subspecies ( H. acranthus subsp. backebergii and subsp. zonatus; H. pseudomelanostele subsp. aureispinus subsp. carminiflorus and subsp. turbidus ; and H. versicolor subsp. pseudoversicolor ). The study is based on the examination of bibliographic records, detailed analysis of herbarium specimens, and examination of all species in the field and in cultivation. Systematic and population studies are needed to determine the limits of the species currently recognized in Haageocereus, and to evaluate whether species that are no longer recognized may be, in fact, vali d. In view of these needs, a molecular study using chloroplast ( rpoB rpl16 23S spacer) and nuclear (ITS1, ITS2) markers was designed to elucidate phylogenetic relationships in the Trichocereeae w ith the primary aim of testing the monophyly of the tribe and examining the relationship of taxa traditionally included in the Trichocereeae with suggested close relatives from tribes Rhipsa lideae, Notocacteae, Browningieae, and Cereeae (Nyffeler 2002). A second objective of the study wa s to elucidate phylogenetic relationships in the genus Haageocereus and supplement these studies with observations of morphology, cytology, reproductive biology, and patterns of distribution.
30 Materials and Methods Taxon Sampling The treatm ent of tribe Trichocer eeae used in this study follo ws that of Buxbaum (1958), with modifications by Barthlott and Hunt (1993) Hunt (1999), and Anderson (2001) (Table 1-1). Twenty-three genera of Trichocereeae (Hunt 1999, Anderson 2001), almost all represented by the type species, were sampled. Genera missing from the analysis are Brachycereus endemic to the Galpagos Islands, and suggested to be more closely related to Armatocereus in the Browningieae (R. Wallace pers comm., see Anderson 2001); Cephalocleistocactus, endemic to Bolivia, suggested to be a hybrid between Cleistocactus and Espostoa (R. Mottram pers. comm., see Anderson 2001) and recently transferred to Cleistocactus (Hunt et al. 2006); and Leocereus endemic to Brazil and found to be more closely related to the Cereeae (Lendel et al. 2007). More extensive sampling was done for larg e genera and for genera with well-known taxonomic difficulties (e.g., Borzicactus, Cleistocactus, Echinopsis, Oreocereus, Weberbauerocereus). Fifteen species of Espostoa (Hunt 1999) and 33 species of Haageocereus (Hunt et al 2006, Caldern et al. 2007) were included. Choice of outgroups [ Eriosyce islayensis (Tribe Notocacteae) Browningia candelaris (Browningieae) Corryocactus melaleucus, Corryocactus meyennii, Cephalocereus senilis (Pachycereeae) Calymmanthium substerile (Calymmantheae), Praecereus euchlorus, Pilosocereus pachycladus (Cereeae), Rhipsalis baccifera (Rhipsalideae), Blossfeldia sucrensis (subfamily Blossfeldioideae),Pereskia sacharosa (subfamily Pereskioideae)] was ba sed on the studies of Crozier et al (2002) and Nyffeler (2002), who identified the major clades of the Cactaceae. A total of 118 species in 39 genera was used to infer intergeneric relationships in the Trichocereeae. Plant material for molecular studies was obtained from living material collected in the field and from specimens in cultivation in the nursery trade (Grigsby Cactus Garden, CA; Arizona
31 Cactus Garden, AZ; Mesa Garden, NM; and the Cactus and Succulent Society of America) (Table 2-1). Voucher specimens were deposit ed at the following herbaria: San Marcos University Herbarium, Lima, Peru (USM), San Agustin University Herbarium, Arequipa, Peru (UNSA), La Paz Herbarium-San Andres University, La Paz, Bolivia (LPB), University of Texas Herbarium, Austin, TX (TEX), and University of Florida Herbarium, Gainesville, FL (FLAS). Comparative studies of morphological features are based on the examination of the few herbarium specimens found deposited at: B, F, MO, TEX, U, USM, and ZSS, as well as observations of all species collect ed in the field and cultivated plants. This work will hopefully complement the detailed morphological study of Haageocereus by Caldern et al (2007), creating a framework for further research on character evolution, biogeography, ecology, and reproductive biology of Haageocereus and other Trichocereeae. DNA Isolation, Amplification, and Sequencing Genom ic DNA was isolated from silica-gel dr ied or fresh specimens using one of three different methods: a modified CTAB-based procedure (Doyle and Doyle 1987) originally intended for DNA extractions in rain forest pl ant species (Scott and Pl ayford 1996), the Soltis Lab CTAB protocol (Soltis et al. 1991), and the DNeasy Mini Kit (Qiagen), which was used in cases when better-quality DNA wa s needed, only a small amout of tissue was available, or samples were highly mucilaginous. Markers were chosen based on proven utility in previous studies and included coding ( rpoB encodes the RNA polymerase subunit, and rpl16 encodes the ribosomal protein L16) and non-coding (23S ribosomal RNA gene spacer region) regions of the chloroplast genome, plus th e nuclear Internal Transcribed Spacer regions (ITS) of the 18S26S rDNA. The chloroplast regions rpoB partial rpl16, and 23S (Table 2-2) were amplified using primers BF/2704R and 2791F/1583R (developed by B. Crozier et al ., pers. comm.) for rpoB ; F71 and R1661 (Jordan et al ., 1996) for rpl16 ; and 23SF and 23S1000R (developed by B.
32 Crozier et al ., pers. comm.) for the amplification of 23S. ITS was amplified using the primers ITS5 and ITS4 (developed by White et al. 1990, modified by Downie and Katz-Downie 1996). Base composition of most of the primers utilized is given in Table 2-3. Efforts were made to develop primers for the amplification of singleor low-c opy nuclear genes (e.g., LEAFY Frohlich and Meyerowitz 1997, Howarth and Baum 2005; PISTILLATA Kim et al. 2004; APETALA3, S. Brockington pers. comm.), but without success. PISTILLATA ( PI ), APETALA3 ( AP3), and LEAFY ( LFY) were chosen because of their prov en utility in phylogenetic studies at lower taxonomic levels (Frohlich and Meyero witz, Bailey and Doyle 1999, Oh and Porter 2003, Montieri et al. 2004). They are known to be singleor low-copy genes and have sizable noncoding regions flanked by conser ved regions where primers can be designed. They are also a source of nuclear sequences, unlinked to the widely used ITS, and useful in providing evidence of reticulate evolution (Bailey and Doyle 1999) Based on phylogenetic analyses, Frohlich and Parker (2000) have suggested that LFY was duplicated in the lineage leading to seed plants but that a copy was lost in the angios perms. Frohlich and Meyerowitz (1997), in a survey of 12 taxa, reported that the size of the second intron of LFY ranges from 88 bp in Peperomia to 7946 bp in Platanus. Sequences of putative orthologues of PI AP3, and LFY from the most closely related taxa to Cactaceae availa ble in GenBank (e.g., Silene, Chenopodium, Fagopyrum, Spinacea ), or provided by S. Brockington (for PI and AP3 in Aizoaceae), were aligned, and the most conserved regions across species were identified. Degenerate pr imers were designed to amplify regions of approximately 1 kb. Amplifications produ ced a number of bands that were either gel purified, cloned, and sequenced, or cleaned and us ed in subsequent nested PCRs with more internal primers.
33 For PI and AP3 total DNA obtained from fresh material was used for amplification with the degenerate primers given in Table 2-4. For LFY, RNA was extracted fro m fresh floral tissue of three species representing three subfamilies of Cactaceae (Pereskioideae, Opuntioideae, and Cactoideae). Initial amplificati ons were performed using degenera te primers (Table 2-4, Fig. 2-1) on cDNA. Sequences obtained from c DNA for the second and third exons of LFY were aligned and used to design specific primers for the amplification of the second intron of LFY from genomic DNA. The resulting bands ranged from 700 bp to 4000 bp. Subsequent isolation of individual bands by gel purific ation or cloning, and sequenci ng revealed only non-specific amplifications (Table 2-4, Fig. 2-1). Polymerase chain reaction (PCR) amplifications were performed in 25 L total volumes, containing 0.01 unit of GoTaq Flexi DNA Polymerase (Promega Corporation, Madison, WI, USA), 0.8X GoTaq buffer, 2.5 M MgCl2, 0.32 M of the forward and reverse primer, 0.2 mM of each dNTP, and 0.4 M Betaine. Betaine was re placed by 5% DMSO for the amplification of the ITS region. For rpoB and rpl16 the PCR protocol consisted of a first cycle with 5 min denaturation at 94C, 45 sec annealing at 48C, 1 min primer extension at 72C, and 32 subsequent cycles with 1 min de naturation at 95C, 45 sec annea ling, where 2 sec were added to each following cycle. The reaction was completed with a 7-min final extension at 72C. The PCR protocol for 23S consisted of a first cycle with 5 min denaturation at 94C, followed by 35 cycles of amplification (1 min denaturation at 94C, 1 mi n annealing at 50C, 45 sec primer extension at 72C, where 3 sec were added to each followi ng cycle). The reaction was completed with a 7min final extension at 72C. For ITS, a stepdow n PCR protocol was used, consisting of a first cycle with 5 min denaturation at 95C, followe d by 2 cycles of amplification with 1 min denaturation at 94C, 1 min annealing at 60C, 1: 30 min primer extension at 72C, two cycles of
34 amplification decreasing the annealing temperatur e to 56C, and 24 cycles of amplification with an annealing temperature of 52C. The reaction was completed with a 7-min final extension at 72C. PCR products were cleaned either with ExoSAP -IT or using Centri-Sep Columns (Princeton Separations, Inc., Adelphia, NJ, USA) The primers used for sequencing were the same as those used for amplification. Sequences were run on an ABI 3730x capillary electrophoresis sequencer, using BigDye Termin ator chemistry (Applied Biosystems Inc., Foster City, CA, USA), at th e DNA sequencing facility of the University of Florida Interdisciplinary Center fo r Biotechnology Research (ICBR), or on a CEQ 8000 capillary sequencer, using the Dye Terminator Cycle Sequencing (DTCS) Quick Start Kit (BeckmanCoulter, Fullerton, CA, USA). Alignment ans Phylogenetic Analysis In m ost cases sequences from both strands of DNA were used to generate a consensus sequence using Sequencher version 4.1 (GeneC odes Corp., Ann Arbor, MI, USA). Sequences were initially aligned using MUSCLE on the web (Edgar 2004) and subsequently refined manually using Se-Al (Rambaut 2007). All phylogene tic analyses were performed on the Florida Museum of Natural History (FLMNH) Phyloinformatics Cluster. Maximum Parsimony (MP) analyses were performed using PAUP* version 4.0b10 (Swofford 2002). All characters were weighted equally, and character-state transitions were treated as unordered. Because inse rtion/deletions (indels) can occur in parallel in two or more lineages (Golenberg et al. 1993, Kechner and Wendel 1996), gaps were treated as missing data. The entire data set was arranged into individual gene partitio ns, as well as chloroplast and nuclear partitions. The Incongruen ce Length Difference (ILD) (Farris et al. 1994) or partition homogeneity test (Swofford 2002) was performed in PAUP* to test for incongruence between
35 assigned partitions. The lack of resolution a nd clade support in the independent plastid and nuclear analyses prompted us to explore merging the two data matrices. Minimal-length trees were obtained by using the heuristic option with 100 random addition sequences and TBR (tree bisecti on reconnection) swapping to sear ch for alternative islands of most parsimonious trees (Maddison et al. 1984), and saving all trees or a maximum of 5000 trees. Characters were optimized by accelerated transformation (ACCTRAN). The consistency index (CI) and the retention i ndex (RI) were calculated auto matically using PAUP*. Mostparsimonious trees were generated for all data sets and data sets combined in a global analysis. A non-parametric bootstrap test (F elsenstein 1985) with heuristic search settings as above and 1000 replicates was used to assess clade support. The selection of a best-fit evolutionary model of nucleotide substitution was performed in Modeltest 3.7 (Posada and Cranda ll 1998) for individual as well as the entire (combined) data sets using the Akaike Information Criterion (AIC, Akaike 1974). The general Time Reversible (GTR) + + I model (Lavane et al. 1984, Tavare 1986, Rodriguez et al. 1990, Yang 1993, Wadell and Penny 1996) was chosen for three of the four data sets, and for the combined data set. The Jukes-Cantor model (JC69, Jukes and Cantor 1969) with and I was chosen for 23S. Heuristic Maximum Likelihood (ML) searches were performed using Garli v. 0.951 (Zwickl 2006), and a non-parametric bootstrap test was applie d to the analysis in four independent runs, with 25 replicates per run, totaling 100 replicates. A Bayesian analysis was performed usi ng MrBayes v. 3.0b4 (Huelsenbeck and Ronquist 2001). Bayesian inference was obtained using a Markov chain Monte Carlo estimation (mcmc), running four chains (one cold and three heat ed), for 2,000,000 generations and saving one tree every 1000 generations. Stationarity levels for th e two independent replic ates conducted for each
36 analysis were obtained (by plotting Mr. Bayes likelihood values over generation numbers), compared, and trees obtained before stationar ity was reached were discarded as burn-in. Additionally, MP, ML, and Bayesian analyses using the same parameters described above, were performed on the combined data set excluding 12 taxa ( Haageocereus platinospinus H. chrysacanthus, H. pseudomelanostele xHaagespostoa climaxantha Cleistocactus acanthurus Echinopsis atacamensis subsp. pasacana, E. cephalomacrostibas Espostoa melanostele 1554a and 1634, E. senilis, Oreocereus tacnaensis and Rauhocereus riosaniensis) for which reticulation has been previously reported or is suspected base d on observations in this study. Results Descriptive Data The total evidence data s et ( rpoB partial = 2552 bp, rpl16 = 1197 bp, 23S = 1313 bp, and ITS = 658 bp) consisted of 5727 characters for 11 8 taxa, of which 11 taxa were outgroups. Characteristics of the sequences are given in Tabl e 2-3. In the combined da ta set, 374 characters (6.4% of the total) were parsimony-informativ e, 585 characters were variable but parsimonyuninformative, and 4768 characters were c onstant. Eighteen seque nces are missing for rpoB five for rpl16 eight for 23S, and five for ITS. Parsimony Analysis MP analyses for individual data sets each produced poorly resolved strict consensus trees from 10,577 trees (including all taxa) and 5000 tr ees (excluding putative hybrids). Descriptive information of the six (four individual data sets, and the combined data set including and excluding putative hybrids) different MP analyses is given in Table 2-5. Congruence among phylogenetic rec onstructions from chloroplast markers was expected because chloroplast genes are physi cally linked (Doyle 1992, de Queiroz et al. 1995); however, the ILD test found the different plastid partitions incongruent (p-value = 0.01). The same result
37 was obtained when comparing the chloroplast and nuclear partitions. The data sets were combined, however, because no major topologic al differences were observed (see below). The combined analysis including hybrids produced 10,577 most-parsimonious trees of length 1652. The 50% majority rule consensus re solved 101 clades (Fig 2-1), and the strict consensus tree resolved 79 clades. The combined analysis excluding hybrids produced 5000 trees of length 1524; the strict consensus resolved 99 clades. Initial analyses revealed a num ber of putative hybrids (arrows in Fig. 2-2) that, together with other reported hybrids (Rowley 1982, 1994), were subsequently removed from the data set ( Echinopsis cephalomacrostibas E. atacamensis subsp. pasacana, Cleistocactus acanthurus, xHaagespostoa climaxantha Haageocereus chrysacanthus, H. platinospinus H. pseudomelanostele Espostoa melanostele E. senilis two individuals of E. lanata Oreocereus tacnaensis, and Rauhocereus riosaniensis ). They were located together or in close proximity to one or both putative parents in the initial analyses. I will focus on the total evidence tree without putative hybrids (Fig. 2-3). With MP, a clade composed of taxa belonging to tribes Pachy cereeae, Notocateae, Calymmantheae, Rhipsalidae, and subfamily Pereskioideae is sister to all the re st of the taxa analyzed. All remaining taxa fall into three clades, one containing Lasiocereus and Rebutia together with Browningia (in the Browningieae). The next clade to diverge contains Espostoopsis, Facheiroa, Discocactus, Gymnocalycium, and Samaipaticereus (Trichocereeae), together with Pilosocereus (member of Brazilian Cereeae). Trichoceree ae is recovered as monophyletic but with the addition of Praecereus euchlorus Several major clades are recove red, and most gene ra are largely monophyletic, although few relationships re ceive bootstrap support (BS) above 50%. Echinopsis forms a clade (with the inclusion of Denmoza ) and is sister to Acanthocalycium This clade is
38 sister to Espostoa + a clade of Cleistocactus and Weberbauerocereus. This ED + [E + CW] clade is in turn sister to the remaining members of the Trichocereeae. Of the remaining taxa, a clade consisting of members of Cleistocactus, Matucana, Moravetzia, Oroya, and Oreocereus [CMMOO] is sister to a monophyletic Haageocereus (which also includes Pygmaeocereus and Mila ). Harrisia is sister to the rest of the Trichocereeae. Maximum Likelihood Analysis The ML tree (-log L = 18425.5837) is shown in Fig. 2-4. A phylogram is also provided to illustrate branch lengths (Fig. 2-5). The MP and ML trees exhibit s lightly different topologies, with some entire clades, rather than individual species, switching positions. Howe ver, there is also little or no BS for these relationships with ML. The outgroup is composed of two main clades, one containing taxa in the Pachycereeae, Notocacteae, Calymmantheae, Rhipsalidae, and Pereskioidea. A second clade contains Espostoopsis, Facheiroa, Discocactus, Gymnocalycium and Samaipaticereus (Trichocereeae), together with Pilosocereus (Cereeae). Lasiocereus (Trichocereeae) and Rebutia are sisters to the rest of the taxa included in th e analyses. ML analysis of the combined data set again revealed a monophyletic Trichoceree ae, but again with the inclusion of Praecereus euchlorus. Harrisia + Acanthocalycium spiniflorum form a clade sister to the rest of the Trichocereeae. A clade of Echinopsis is recovered (again including Denmoza ). This clade is sister to all Trichocereeae after Harrisia A clade comprising part of Haageocereus (and including Pygmaeocereus ) is sister to Mila and this clade is sister to Acanthocalycium This [HP + M] + A clade is in turn sister to a clade containing part of Espostoa + a clade of Cleistocactus and Weberbauerocereus (the same clade that was sister to Echinopsis in the MP tree). The [[HP + M] + A] + [E + [C + W]] clade is sister to a clade that contains the remaining members of
39 Haageocereus + the [CMMOO] clade, and this clade is in turn sister to a clade containing the remaining Espostoa species. Bayesian Analysis The Bayesian analys is of the combined data set (Fig. 2-6) produced a majority-rule consensus tree of 1800 post-burn-in trees that showed less resolution but overall similar topology, and support for clades, as obtained from the ML analyses (Fig. 2-4). As in the MP and ML analyses, the exclusion of taxa suspected to be hybrids increased resolution and support considerably. The Bayesian tree is similar to the ML tree, as expected. The outgroup is composed of three clades. One clade contains Rebutia sister to Browningia, and both sister to the rest of the taxa analyzed. Th e second clade within the outgr oup (containing taxa in the Pachycereeae, Notocateae, Calymmantheae, Rhipsalidae, and Pereskioidea) is sister to Lasiocereus (traditionally of Trichocereea e). The third clade contains Espostoopsis, Facheiroa, Discocactus, Gymnocalycium and Samaipaticereus (Trichocereeae), together with Pilosocereus (Cereeae). Again, Trichocereeae is m onophyletic with the inclusion of Praecereus euchlorus Relationships among five major clades recovere d for Trichocereeae are not resolved. One clade (PP=0.51) contains part of Haageocereus (PP=0.86), including Pygmaeocereus, and is sister to Mila The clade is in an unresolve d position with respect to an Espostoa clade (PP=0.53) and a Cleistocactus + Weberbauerocereus clade (PP=1.00). Acanthocereus and Yungasocereus are also found in this clade in unresolved pos itions. The remaining species of Haageocereus form a clade (PP=0.86) sister to the [CMMOO] clade (PP=0.76). A third clade contains Echinopsis including Denmoza (PP=0.81). A fourth clade contai ns the remaining species of Espostoa (PP=0.52). Harrisia and Acanthocalycium spiniflorum are also found in unresolved positions.
40 Discussion Phylogenetic Inference Although chloroplast genes are known for be ing highly conserved and for providing few charac ters at lower taxonomic levels (Palmer et al. 1988), the selection of markers in this study was influenced by their utility in previous studies of Cactaceae (e.g., rpl16 Nyffeler 2002, Butterworth et al. 2002, Arias et al 2003; see also Jordan et al 1996). Other chloroplast markers explored by Crozier (2002), such as rpoB and 23S, were shown to provide equal or better resolution in the Cactaceae. Se quences for ITS provided additiona l informative characters and a source of nuclear sequence data. Individual data sets were first analyzed separately. The issue of when, and if, separate data sets should be combined for phylogeneti c analysis has been debated (e.g., Bull et al 1993, de Queiroz et al. 1995, Mason-Gamer and Kellog 1996, Seelanan et al 1997, Johnson and Soltis 1998). Conditional combination (Huels enbeck and Bull 1996, Huelsenbeck et al. 1996) was employed, following Mason-Gamer and Kellogg (1996) and Johnson and Soltis (1998). The incongruence length differe nce (ILD) test (Farris et al. 1995) showed significant differences between the plastid and nuclear data sets (p-v alue = 0.01). Recognizing th at the test is highly sensitive to the number of characters present, and to the homogeneity of substitution rates from site to site (Sullivan 1996, Darlu and Lecointre 2002) it alone should not be used to determine data combinability. Therefore, each data set was analyzed sepa rately and topologies compared. Because the comparison of individual gene topol ogies did not show significant incongruence, all data sets were used in combined analyses with the goal of provid ing better estimates of phylogenetic relationships. Hillis (1987) and Kluge (1989) have suggested that the combined analysis of all relevant data can maximize the explanatory power, whethe r or not the individual results are consistent with the combined results. The individual data se ts used here produced
41 poorly resolved topologies. However, clades with BS of more than 50% did not show incongruences among trees when compared. One c oncern was hybridization as the cause of the low support for many clades. As noted, additional analyses were therefore performed excluding suspected reticulate taxa and this discussion is based on these trees. Phylogenetic Relationships in Trichocereeae As expected based on recent treatm e nts (Nyffeler 2002, Arakaki 2002, Lendel et al. 2006, 2007), Espostoopsis dybowsky, Discocactus zehn tneri, Facheiroa ulei, Gymnocalycium kieslingii, Lasiocereus ( L. fulvus and L. rupicola), Rebutia minuscula and Samaipaticereus corroanus (members of tribe Trichocereeae) are placed among members of the outgroup. This confirms suggestions by Lendel et al. (2006, 2007) and Ritz et al (2007) for Espostoopsis and Discocactus ; by Nyffeler (2002) and Lendel et al. (2006, 2007) for Gymnocalycium ; by Arakaki (2002) for Lasiocereus; and by Lendel et al. (2006, 2007) for Facheiroa and Rebutia Espostoopsis dybowsky from Brazil appears as sister to Pilosocereus pachycladus a Brazilian member of the Cereeae. Both share a cephalium ( accumulation of hairs in the flowering zone), a character also present in Espostoa, motivating the inclusion of Espostoopsis in the Trichocereeae in the assumption of a close relationship with Espostoa. The Cactaceae Systematics Group, part of the IOS, accepted the transfer of Espostoopsis from Cereeae to Trichocereeae by Buxbaum (1968, see Anderson 2001) with hesitati on and after considerable debate. Discocactus Facheiroa Gymnocalycium and Samaipaticereus (Bolivia) are placed sister to the EspostoopsisPilosocereus clade (MP-BS=<50%, PP=1.00). Discocactus is found in Brazil, Bolivia, and Paraguay. It possesses an apical cep halium that resembles that of Melocactus (Cereeae); flowers are nocturnal, white, slender, and elongated. The presence of scales in the pericarpel is a character shared with Trichocereeae. Facheiroa is a genus found in northeastern Brazil; it is shrubby or tree-like, usually bearing a cephalium th at protects flowers with pericarpels covered
42 by scales and hairs. Gymnocalycium species are among the most popular for cultivation; they are distributed in southern Brazil, Argentin a, Bolivia, Paraguay, and Uruguay. Plants of Gymnocalycium are globose to cylindrical, with tuberculate stems, with white, pink, yellow or red bell-shaped flowers, and br oad scales covering the pericarp el but lacking hairs. The two species of Lasiocereus ( L. fulvus and L. rupicola ), endemic to a small region in northern Peru, form a clade (BS=90%, PP=1.00) th at is either sister to members of the Pachycereeae ( Corryocactus), Rhipsalidae ( Rhipsalis ), Notocacteae ( Eriosyce ), Blossfeldioideae ( Blossfeldia ), and Pereskioideae (Pereskia ) in the ML and Bayesian analysis (PP=1.00), or one of the most basal branches between Rebutia (found to be outside the Trichocereeae) and Browningia (Browningieae) in the MP analysis. The IO S has long been unsure of the status of Lasiocereus, which was once placed in Haageocereus, but it ultimately provisionally recognized it as a separate genus (Anderson 2001). Plan ts are shrubby or tree -like and develop a pseudocephalium. The flowers are nocturnal, shor t, and wide, with pericarpels covered with dense wool and bristles. Rebutia is another very popular genus among collect ors, and this has resulted in a large number of described species. It is found in Bolivia and Argentina. Genera placed under synonymy with Rebutia are Aylostera, Spegazzinia, Mediolobi via, Weingartia, Digitorebutia and Sulcorebutia Plants are depressed-spherical, with poorly developed ribs or tubercles, numerous diurnal flowers that are funnelform and brightly colo red (red, deep orange, yellow) with pericarpels with numerous scales and some times hairs or bristles. Seeds are distinctively tuberculate to wrinkled (Barthlott and Hunt 2000).
43 Gymnocalycium (71 species) and Rebutia (41 species) are the second and fourth most species-rich genera in the Trichocereeae. Their exclusion from the tribe will reduce the number of species from 400 to about 288 species. In contrast to the genera exclude d from Trichocereeae, described above, Praecereus euchlorus (tribe Cereeae) emerged among members of Trichocereeae in these analyses. Praecereus is found in Venezuela, Trinidad, Ecuador Peru, and Bolivia. The genus has been placed in different genera, and its boundaries ar e still not clear, even when studies by Zappi (1989) and Taylor (1997) (see Anderson 2001) took into consideration geography and anatomy. The DNA results are not supported by morphological features that define the genus: shrubby or tree-like plants lacking mucila ge, with funnelform flowers and thick-walled floral tubes not covered by scales or hairs. Presence of scales and/or hair s in the pericarpel of flowers is the most important characteristic that unifies the Trichocereeae, and Praecereus lacks them. The possibility of Praecereus having a hybrid constitution involving taxa in the Trichocereeae cannot be excluded. Harrisia is one of the most widespread genera in the Trichocereeae; it contains 20 species of shrubby or scandent cacti, with aerial r oots, unsegmented stems, funnelform nocturnal flowers, and fleshy colorful fru its (Applequist and Wallace 2002). Harrisia has been placed in Hylocereeae by Buxbaum (1958), in Echinocereeae by Barthlot and Hunt (1993) based on seed morphology, and more recently in Trichocereeae by Hunt (1999) and Anderson (2001). Being a genus with a confusing taxonomy and the only genus in the tribe distribute d not only in South America, but also in Florida a nd the Carribbean, its pos ition deserved a closer examination. The two species included in the analysis ( H. aboriginum and H. tetracantha ) represent the two subgenera: subgenus Harrisia from the U. S. and the Caribbean, and subgenus Eriocereus, from
44 South America. They form a clade that is sister to the rest of Trichocereeae, with BS=75%, and Posterior probability (PP) of 1.00. A deletion in the trnT-trnL intergenic spacer is shared with other members of Browningieae, Cereeae, and Tr ichocereeae; however, the possibility that the deletion is homoplasious cannot be ruled out (Applequist and Wallace 2002). The presence of hairs in the pericarpel, characte ristic of Trichocereeae, supports its placement in the tribe. Two of the three species of Acanthocalycium ( A. spiniflorum and A. glaucum ) included in the analysis form a weakly supported clade (B S=<50%) in the MP and ML trees, and occupy distant positions in the Bayesian tree. The genus is from Argen tina and was previously included in Echinopsis but excluded again until further informa tion was available. Plants are globose and apically depressed; flowers are diurnal, funnelfor m to bell-shaped, pink to white, with scales and hairs in the pericarpel (Backeb erg 1976, Anderson 2001). The three species of Weberbauerocereus ( W. cuzcoensis, W. r auhii, W. weberbaueri ) included in this analysis form a clade (BS=<50%, PP=0.56) sister to a Cleistocactus clade, and both in turn are sister to Yungasocereus sp. nov. (BS=<50%, PP=1.00). A close relationship between the monotypic Yungasocereus (from Bolivia) and Weberbauerocereus (from Peru) has never been suggested. Instead, Yungasocereus was described as Samaipaticereus but later placed in its own genus. However, recent observations of Yungasocereus inquisivensis in the field, plus examination of a specimen that will be described as a new species by Quispe et al (in prep.), indicated that Yungasocereus is indeed either closely related to or a member of Weberbauerocereus. In contrast, Samaipaticereus is found to be sister to the clade containing Discocactus Espostoopsis Facheiroa and Gymnocalycium among members of the outgroup (BS=<50%, PP=1.00). Samaipaticereus has few ribs, characteristic of the Cereeae, but differs
45 from the Cereeae in having flowers with scales and short clusters of hairs (Backeberg 1976, Anderson 2001). Two widely separated clades contain members of Cleistocactus, or genera recently placed under Cleistocactus (Hunt 1999, Anderson 2001, Hunt et al 2006). In one of these clades Hildewintera aurispina (= Cleistocactus winteri ) is sister to Bolivicereus samaipatanus (= Cleistocactus plagiostoma ), and both are sister to Cleistocactus baumannii (BS=88%, PP=1.00). A second clade contains additional species of Cleistocactus as well as the former genera Borzicactus and Clistanthocereus ( B. fieldianus, B. purpureus, C. samnensis) (BS=72%, PP=0.99). These results support the inclusion of Hildewintera aurispina and Bolivicereus samaipatanus in Cleistocactus (Anderson 2001, Hunt et al 2006) These two taxa, together with Cleistocactus baumannii (the type species in the genus), belong to a strongly supported clade. Members of the second Cleistocactus clade (Cleistocactus purpureus, Clistanthocereus samnensis, and Cleistocactus fieldianus ) belonged at one time to the genus Borzicactus. Results support the maintenance of Borzicactus and suggest a close relationship between Borzicactus and Matucana (BS=82%, PP=1.00) This relationship was previously suggested by Kimnach (1960) based on morphology; however, the combination of Matucana and Borzicactus by Kimnach was not accepted by the IOS, who placed Borzicactus in Cleistocactus. Even though Matucana has been a subject of many studies, its extreme morphologica l variability and the suggested inclusion of species of hybrid origin make it diffic ult to clarify its st atus at this time. Matucana is a genus of globose or cylindrical cacti, at times found in clumps. Borzicactus in contrast, has erect and slender stems. However, both genera posse ss similarly broad and tuberculate ribs. Oreocereus and Oroya have also been suggested to be closely related to Matucana and Borzicactus (Anderson 2001), and our results support this idea (BS=<50%, PP=0.98). Borzicactu s is found
46 from Central Ecuador to Peru, whereas Oreocereus is distributed at middl e to high elevations in southern Peru, Bolivia, northern Chile, and Argentina. Matucana and Oroya are restricted to the Andes of Peru. Flowers and seeds in all of these genera are very similar; flowers are usually funnelform, slightly zygomorphic or radially symmetrical, with yell ow or orange to red perianth, and are pollinated by hummingbirds (Backeber g 1976, Sahley 1995, Anderson 2001). Seeds are medium-sized, oval, and black-brown, with a border that usually e xpands around the hilum, convex relief, and a striated or folded cuticle (Barthlott and Hunt 2000). Three species of Oreocereus included in the analysis ( O. celsianus, O. hempelianus, O. hendricksenianus ) and Moravetzia doeltzina recently transferred to Oreocereus (Anderson 2001), form a clade sister to Loxanthocereus jajoianus Although Loxanthocereus has been transferred to Cleistocactus, there is still controversy with regard to the inclusion of Loxanthocereus plus 12 other genera ( Akersia, Bolivicereus, Borzicactella, Borzicactus, Binghamia, Clistanthocereus, Hildewintera, Mari timocereus, Seticereus, Seticleistocactus, Winteria, Winterocereus) in Cleistocactus, on the basis of similar ity in floral morphology (narrow tubular flowers that barely open) (Anderson 2001, Hunt et al. 2006). Because only one species of Loxanthocereus was included in the analysis, it is not possible to determine with the data at hand if the ge nus is also part of Cleistocactus, as proposed by Hunt (1999), Anderson (2002), and Hunt et al (2006). The decision of placing Loxanthocereus jajoianus (from Peru) under Cleistocactus sepium (a species found only in Ecuador) is controversial (C. Ostolaza and F. Cceres, pers. comm.), because of some mor phological differences, such as flower shape and color (slender, slightly zygomorphi c, and orange to bright red in L. jajoianus as opposed to straight, tubular, red flowers in C. sepium ), and characteristics of the stem and spines. Recent chromosome counts (Chapter 4) revealed that L. jajoianus is tetraploid; however, counts for
47 Cleistocactus sepium from Ecuador report only diploid numbers (Baker 2002). Loxanthocereus is distributed in the western slope s and coast of Ecuador and Peru. All genera included in Cleistocactus have a floral morphology characteristic of hummingbird pollination. Hybridization, and the fact that questionable descriptions are common due in part to Cleistocactus species popularity among collector s, adds to the controversy regarding boundaries between Cleistocactus and related genera. Furthe r phylogenetic studies and observation of species in the field are needed to determine the boundaries of Cleistocactus Echinopsis is poorly understood and has become the largest genus in tribe Trichocereeae due to the merging of several small genera of columnar and globular cacti that, although distinct in vegetative morphology, share nocturnal or diurnal, funnelform flowers with narrow scales bearing hairs (i.e., Acantholobivia, Chamaecereus, Heliantho cereus, Hymenorebutia, Leucostele, Lobivia, Neolobivia, Setiechinop sis, Soehrensia, Trichocereus ). Representatives of globular and columnar Echinopsis were included in the study and we re found to be monophyletic in all analyses. Results also suggest that Denmoza should be placed in Echinopsis (BS=<50%, PP=0.80). The two clades obtained for Echinopsis in the Bayesian analysis seem to agree with their geographic distribution. One clade is composed of species from southern Bolivia and Argentina (PP=0.89), and the second clade (PP=0.96) contains species that occur farther north, with three Peruvian and one Bolivian species. All analyses place Espostoa species together in one clade (BS=<50%, PP=0.67), except for three specimens in two species suggested to be involved in hybridization events (E. melanostele 1554a, E. melanostele 1634, and E. senilis). This Espostoa clade is sister to a clade containing Borzicactus, Cleistocactus, Matucana, Oreocereus, and Oroya.
48 The hybrid constitution of several taxa included in the analyses is consistent with their placement together or in close proximity to one or both suggested putative parents, for example: x Haagespostoa climaxantha together with Haageocereus pseudomelanostele and H. chrysacanthus, is placed in the Espostoa clade; Cleistocactus acanthurus reported to form hybrid swarms with Haageocereus turbidus (Mottram 2004), is embedded in the Haageocereus clade; Oreocereus tacnaensis also embedded in the Haageocereus clade, is believed to be a hybrid between Oreocereus leucotrichus and Weberbauerocereus sp (Charles 2000). However, species of Weberbauerocereus have not been found in close geographic proximity to O. tacnaensis hence, the possibility of Haageocereus being one of the putative parents should be considered. Phylogenetic Relationships in Haa geocereus All analyses indicate that Haageocereus is monophyletic. However, there is very low resolution within the Haageocereus clade except for a subclade (BS=<50%, PP=0.83) that groups prostrate species of Haageocereus : H. chalaensis ( 2n = 44), H. decumbens ( 2n = 22), H. litoralis, H. pluriflorus, and H. tenuis ( 2n = 33). Caldern et al. (2007) placed H. chalaensis, H. litoralis, and H. australis in synonymy, under H. decumbens The inclusion of H. tenuis which is believed to be of hybrid origin (Cha pter 5), indicates that tetraploid H. chalaensis and diploid H. decumbens could be candidates for put ative parents of triploid H. tenuis Intergeneric and infrageneric relationships in Haageocereus have long been considered problematic because the genus has a te ndency to form hybrid swarms with Espostoa and shares several of its diagnostic features (mainly habi t and vegetative morphology) with several other related Cactaceae (Anderson 2001, Hunt et al. 2006). Because there are numerous instances of hybridization between Haageocereus and Espostoa, we added 15 of the 16 species of Espostoa (18 individuals) to the analys es. Although the parsimony analysis reveals a close relationship
49 between these two genera, morphological characters such as the general habit, the presence of cephalia in Espostoa, the flowers, and characters of the fruits and seeds do not strongly support these findings. Both taxa overlap in a great portion of their geographic distributions. They are found typically in drier slopes and inter-Andean valleys at low and medium elevations (0-2000 m). Results show, however, that there are clearly differentiated clades for Haageocereus and Espostoa, which supports the maintenance of these two genera. In contrast to the MP result which shows both Haageocereus and Espostoa to be monophyletic if hybrids are excluded from the an alysis, the ML and Bayesian trees excluding hybrids show both Espostoa and Haageocereus as paraphyletic genera. Most Haageocereus species are contained in a clade sister to Pygmaeocereus baylesianus (BS=<50%, PP=0.68). A second group forms a clade sister to the Borzicactus-Matucana-Oreocereus-Oroya clade. Pygmaeocereus has been proposed to be part of Haageocereus but has been provisionally accepted as a separate genus by the IOS until furt her research is conducted. The results here support the proposal of including Pygmaeocereus in Haageocereus. The main distinctive features of species of Pygmaeocereus are its small size (10-20 cm tall) and slender, funnelform flowers, compared to Haageocereus One clade of Espostoa is sister to members of current Cleistocactus, and the other clade is sister to the Haageocereus-(Borzicactus-MatucanaOreocereus-Oroya) clade. Radiation of Trichocereeae The low sequence divergence and lack of resolution and support for Haageocereus, com pared to other genera such as Cleistocactus or Echinopsis suggest a very rapid and recent diversification. Hughes and Eastwood (2006), anal yzing evidence from various Andean plant groups, concluded that rapid divers ification in these groups might have been driven by ecological opportunity (in the form of extrin sic circumstances), rather than evolutionary innovations.
50 Once a better understanding of generic a nd species relationshi ps is achieved, the biogeographic history of the Trichocereeae can be i nvestigated. The history of the tribe can be linked to the rise of the Andean Cordillera, a nd the development of dese rt habitats in South America, plus their dramatic fluctuations (cont ractions and expansions ) during the Pleistocene (Prado and Gibbs 1993, Prado 2000). I hypothesize the an cestral area for the tribe to be centralSouth America (Bolivia and southern Peru), becaus e this area houses the largest diversity in the tribe and the largest number of narrowly endemic species. The higher genetic differentiation observed in taxa such as Echinopsis and Cleistocactus (which are mainly Bolivian, and probably older than Haageocereus and Espostoa) can be linked to what is expected for ancestral as opposed to more recent species groups (Ito et al. 2006, Crochet et al. 2007). Geographic distribution and phylogenetic rela tionships among species of Trichocereeae (plus outgroups) seem to indicate that the expansion of the gr oup follows the pattern suggested by Prado (2000), from south (Bolivia) to north (Peru) along the arid coast, and from there, within the Cordillera, from low to high elevations, in several direct ions. Very few species of Trichocereeae reach Ecuador, where desert habitats are scarce. Conclusions Results from this study do not differ substantially from those obtained by Nyffeler (2002) and Lendel et al. (2007), yet they provide better reso lution within some clades in the Trichocereeae. Many of the genera are monophyle tic. The relationships among some genera are now clear, but many deeper-level relationships remain enigmatic. Sequen ces of more rapidly evolving genes are needed, especia lly biparentally inherited nucl ear genes that can give more insights into the reticulate nature of several taxa in the Trichocereeae. Efforts to use several such
51 genes [i.e., the floral organ identity genes PISTILLATA ( PI ) and APETALA3 ( AP3), and the inflorescence meristem identity gene LEAFY ( LFY)] were not successful. The study supports the following conclusions: 1) Discocactus, Espostoopsis, Facheiroa, Gymnocalycium Samaipaticereus, Lasiocereus, and Rebutia are not part of Trichocereeae. The position of the first five taxa is closer to Pilosocereus, a Brazilian member of tribe Cereeae. Lasiocereus and Rebutia form a clade sister to Browningia in tribe Browningieae, among members of the outgroup. 2) Praecereus euchlorus (Cereeae) should be pl aced in Trichocereeae based on sequence data. Morphological characteristics, in contrast, place it with tribe Cereeae; however, the possibility of th e taxa having a hybrid constitu tion involving taxa in the Trichocereeae cannot be excluded. 3) Harrisia should be maintained as a member of Trichocereeae. 4) Echinopsis cephalomacrostibas should be maintained in Weberbauerocereus, and Yungasocereus sp. is either closely related to Weberbauerocereus or a new species of Weberbauerocereus 5) Cleistocactus, as currently recognized, is polyphyletic, and deserves a comprehensive examination along with representatives of former Akersia, Bolivicereus, Borzicactella, Borzicactus, Binghamia, Clistanthocereus, Hildewintera, Loxanthocereus, Maritimocereus, Seticereus, Seticleistocactus, Winteria, and Winterocereus 6) The position of Matucana and Borzicactus as sister groups or congeners is strongly supported in all analyses. Both are, in turn, sister to Oreocereus; however, hybridization events, frequent in these genera, cannot be excluded as the cause of the observed relationships 7) All sampled species of Oreocereus, including Moravetzia doeltziana recently transferred to Oreocereus, form a clade. 8) The former Lobivia and Trichocereus should be maintained in Echinopsis along with Denmoza 9) Haageocereus and Espostoa form well-defined clades and should be maintained as separate genera even though hybridization is widespread am ong their members. 10)
52 Pymaeocereus should be included in Haageocereus. 11) A well-supported clade containing prostrate species of Haageocereus ( H. chalaensis H.decumbens H. litoralis, and H. pluriflorus ) agrees with Caldern et al. (2007), who merged three of the four species into H. decumbens. This clade also includes triploid H. tenuis which is believed to be of hybrid origin, with diploid H. decumbens and tetraploid H. chalaensis being the putative parents.
53Table 2-1. List of species in the Trichocereeae and members of other tribes (outgr oups) included in the present study. Voucher specimens are deposited in one or two of the following herbaria: FLAS (University of Florida Herbarium, Gainesville, FL), LPB (La Paz Herbarium, La Paz, Bolivia), TEX (University of Texas Herbarium, Austin, TX), UNSA (San Agustin University Herbarium), USM (San Marcos University Herbarium, Lima, Peru). Species Voucher Provenance Trichocereeae Acanthocalycium glaucum F. Ritter cult. MG (FLAS) AR: Catamarca Acanthocalycium spiniflorum (K. Schumman) Backeb. cult. GC AR Arthrocereus rondonianus Backeb. & Voll cult. ACG (FLAS) BR: Minas Gerais Bolivicereus = Cleistocactus samaipatanus (Cardenas) Hunt MA 1821 (FLAS, LPB) BO: Santa Cruz Borzicactus decumbens (Vaup.) Britton & Rose MA 1537 (USM) PE: Trujillo Borzicacus fieldianus Britton & Rose MA 1480 (USM) PE: Ancash Borzicactus purpureus = Cleistocactus plagiostoma (Vaupel) Hunt MA s/n (USM) PE: Cajamarca Cleistocactus acanthurus (Vaupel) Hunt MA 1629 (USM) PE: Lima Cleistocactus baumannii (Lem.) Lem. cult. MG (FLAS) AR: Crdova Clistanthocereus samnensis = Cleistocactus fieldianus subsp samnensis MA 1710 (USM) PE: Cajamarca (F. Ritter) Ostolaza MA 1494 (USM) PE: La Libertad Denmoza rhodacantha (SD.) Britton & Rose cult. MG (FLAS) AR: Mendoza Discocactus zehntneri Britton & Rose cult. ACG (FLAS) BR: Bahia Echinopsis atacamensis subsp. pasacana (Weber) G. Navarro MA 1723 (FLAS, LPB) BO: Oruro Echinopsis bridgessii Salm-Dyck MA 1768 (FLAS, LPB) BO: Tarija Echinopsis cephalomacrostibas (Werdermann & Backeb.) Friedrich & Rowley MA 1582 (USM) PE: Arequipa Echinopsis cf. calliantholilacina Crdenas MA 1754 (FLAS, LPB) BO: Tarija Echinopsis comarapana Crdenas MA 1805 (FLAS, LPB) BO: Cochabamba Echinopsis eyriesii (Turp.) Pfeiff. & Otto cult. MG (FLAS) AR: Formosa Echinopsis cf. ferox (Britton & Rose) Backeb. MA 1726 (FLAS, LPB) BO: Potos Echinopsis cf. huottii (Cels) Labouret MA 1833 (FLAS, LPB) BO: Cochabamba Echinopsis lageniformis (Frster) Friedrich & Rowley cult. JBLP (FLAS, LPB) BO: La Paz Echinopsis mistiensis = Echinopsis pampana (Britton & Rose) Hunt MA s/n (UNSA, USM) PE: Arequipa Echinopsis obrepanda subsp. calorubra (Cardenas) G. Navarro MA 1815 (FLAS, LPB) BO: Cochabamba Echinopsis pachanoi (Britton & Rose) Friedrich & Rowley cult. CSSA PE: Lima
54Table 2-1. (Cont.) Echinopsis pentlandii (Hooker) Salm-Dyck ex A. Dietrich cult. JBLP (FLAS) BO: Potos Echinopsis peruviana (Britton & Rose) Friedrich & Rowley MA 1705 (USM) PE: Cajamarca Echinopsis tarijensis subsp. herzogiana (Cardenas) G. Navarro MA 1719 (FLAS, LPB) BO: La Paz Echinopsis tarijensis subsp. totorensis (Cardenas) G. Navarro MA 1834 (FLAS, LPB) BO: Cochabamba Echinopsis terscheckii (Parm. ex Pfeiffer) Friedrich & Rowley MA 1735 (FLAS, LPB) BO: Potos Echinopsis uyupampensis (Backeb.) Fredrich & Rowley MA 1605 (UNSA, USM) PE: Arequipa Espostoa ancashensis cult. ACG (FLAS) PE: Ancash Espostoa baumannii Knize cult. ACG, KK 300 (FLAS) PE: Amazonas Espostoa blosfeldiorum (Werdermann) F. Buxbaum MA 1528 (USM) PE: Cajamarca Espostoa calva F. Ritter MA 1692 (USM) PE: Amazonas Espostoa cornifera cult. ACG (FLAS) PE Espostoa huanucoensis Jonson ex F. Ritter MA 1574 (USM) PE: Hunuco Espostoa hylaea F. Ritter cult. MG, KK 296 (FLAS) PE: Amazonas Espostoa lanata (Kunth) Britton & Rose MA 1536 (USM) PE: La Libertad Espostoa lanianuligera F. Ritter MA 1531 (USM) PE: Cajamarca Espostoa laticornua Rauh & Backeb. cult. MG (FLAS), KK 227 (FLAS) PE: Piura Espostoa melanostele (Vaup.) Borg MA 1634 (USM) PE: Lima Espostoa mirabilis F. Ritter MA 1688 (USM) PE: Amazonas Espostoa pariacotensis cult. ACG (FLAS) PE: Ancash Espostoa ritteri Buining cult. MG (FLAS) PE: Amazonas Espostoa senilis (F. Ritter) N. P. Taylor cult. ACG (FLAS) PE: Ancash Espostoa sericata (Backeb.) Backeb. cult. ACG (FLAS) PE: Cajamarca Espostoa superba F. Ritter MA 1666 (USM) PE: Cajamarca Espostoopsis dybowskii (Roland-Gosselin) Buxbaum cult. CSSA (FLAS) BR: Bahia Facheiroa ulei (Grke) Werdermann cult. CSSA, HU 265 (FLAS) BR: Bahia Gymnocalycium kieslingii O. Ferrari cult. CSSA, DG s/n (FLAS) AR: La Rioja Haageocereus acranthus (Vaup.) Backeb. MA 1551 (USM) PE: Lima Haageocereus zonatus Rauh & Backeb. MA 1568 (USM) PE: Lima Haageocereus albispinus (Akers) Backeb. cult. MG, KK 1871 (FLAS) PE: Lima Haageocereus albispinus subsp. albiflorus CO 82217 PE: Lima Haageocereus aureispinus Rauh & Backeb. cult. MG, KK 315 (FLAS) PE: Lima Haageocereus australis Backeb. cult. MG (FLAS) PE: Arequipa Haageocereus cantaensis cult. MG, KK 560 (FLAS) PE: Lima Haageocereus chalaensis F. Ritter MA 1600 (UNSA, USM) PE: Arequipa
55Table 2-1. (Cont.) Haageocereus centrispinus cult. MG, KK 1100 (FLAS) PE: Lima Haageocereus chosicensis Backeb. cult. CO 82052 (USM) PE: Lima Haageocereus chrysacanthus (Akers) Backeb. cult. MG, KK 324 (FLAS) PE: Lima Haageocereus crassiareolatus Rauh & Backeb. cult. ACG (FLAS) PE: Lima Haageocereus decumbens (Vaup.) Backeb. MA 1616 (USM) PE: Ica Haageocereus divaricatispinus Rauh & Backeb. cult. MG, KK 559 (FLAS) PE: Lima Haageocereus horrens Rauh & Backeb. MA 1654 (USM) PE: Lambayeque Haageocereus icosagonoides Rauh & Backeb. MA 1540 (USM) PE: Piura Haageocereus lachayensis Rauh & Backeb. cult. CO 85309 (USM) PE: Lima Haageocereus litoralis Rauh & Backeb. cult. MG, CO 82113 (FLAS) PE: Arequipa Haageocereus pacalaensis Backeb. MA 1538 (USM) PE: La Libertad Haageocereus pectinatus Knize cult. ACG (FLAS) PE: Lima Haageocereus platinospinus (Werdermann & Backeb.) Backeb. MA 1607 (UNSA, USM) PE: Arequipa Haageocereus pluriflorus Rauh & Backeb. FC s/n (UNSA) PE: Arequipa Haageocereus pseudomelanostele (Werdermann & Backeb.) Backeb. MA 1551 (USM) PE: Lima Haageocereus pseudoversicolor Rauh & Backeb. MA 1535 (USM) PE: Cajamarca Haageocereus repens Rauh & Backeb. MA 1539 (USM) PE: La Libertad Haageocereus salmonoides (Akers) Backeb. cult. MG, KK 444 (FLAS) PE: Lima Haageocereus serpens cult. MG, KK 1675 (FLAS) PE: La Libertad Haageocereus setosus (Akers) Rauh & Backeb. cult. MG, CO s/n (FLAS) PE: Lima Haageocereus tenuis F. Ritter MA 1635 (USM) PE: Lima Haageocereus turbidus Rauh & Backeb. cult. MG, KK 551 (FLAS) PE: Lima Haageocereus versicolor (Werdermann & Backeb.) Backeb. MA 1658 (USM) PE: Lambayeque Haageocereus villigera = Neobinghamia villigera Rauh & Backeberg MA s/n (USM) PE: Lima x Haageocereus albisetata (Akers) Rowley = H. albisetatus (Akers) Backeb. cult. MG, KK 248 PE: Lima x Haagespostoa climaxantha (Werdermann) Rowley MA 1550 (USM) PE: Lima Harrisia tetracantha (Labouret) Hunt cult. JBLP (FLAS) BO: Santa Cruz Harrisia aboriginum Small ex Britton & Rose cult. CSSA (FLAS) US: Florida Hildewintera aurispina = Cleistocactus winteri Hunt MA 1790 (FLAS, LPB) BO: Santa Cruz Lasiocereus fulvus F. Ritter MA 1684 (USM) PE: Amazonas Lasiocereus rupicola F. Ritter MA 1698 (USM) PE: Cajamarca Loxanthocereus jajoianus = Cleistocactus sepium (Kunth) Weber ex Roland-Gosselin MA1606 (UNSA, USM) PE: Arequipa Matucana haynei (Otto ex Salm-Dyck) Britton & Rose cult. MG, KK 1548 (FLAS) PE: Lima
56Table 2-1. (Cont.) Mila nealeana Backeb. MA 1627 (USM) PE: Lima Moravetzia doeltziana = Oreocereus doeltzianus (Backeb.) Borg. cult. CO (USM) PE: Ayacucho Oreocereus celsianus (Salm-Dyck) Riccobono MA 1729 (FLAS, LPB) BO: Potos Oreocereus hempelianus (Guerke) Hunt MA 1594 (UNAS, USM) PE: Arequipa Oreocereus hendricksenianus (Backeb.) Backeb. MA 1514 (USM) PE: Tacna Oreocereus tacnaensis F. Ritter MA 1517 (USM) PE: Tacna Oroya borchersii (Boedeker) Backeb. MA 1646 (USM) PE: Ancash Pygmaeocereus baylesianus Andreae & Backeb. cult. MG, KK 1058 (FLAS) PE: Arequipa Rauhocereus riosaniensis Backeb. MA 1665 (USM) PE: Cajamarca Rebutia minuscula Schumann cult. CSSA (FLAS) AR: Salta Samaipaticereus corroanus Cardemas cult. CAI, JL 2098 BO: Santa Cruz Weberbauerocereus cuzcoensis Knize MA 1453 (USM, TEX) PE: Apurmac Weberbauerocereus rauhii Backeb. MA 1461 (USM, TEX) PE: Arequipa Weberbauerocereus weberbaueri (Schumann ex Vaupel) Backeb. MA 1520 (UNSA, USM, TEX) PE: Arequipa Yungasocereus inquisivensis (Cardenas) F. Ritter cult. JBLP, NQ s/n (FLAS) BO: La Paz Outgroup Blossfeldia sucrensis Knize cult. GC, KK1704 BO: Sucre Browningia candelaris (Meyen) Britton & Rose MA 1612 (UNSA, USM) PE: Arequipa Calymmanthium substerile F. Ritter MA 1681 (USM) PE: Amazonas Cephalocereus senilis (Haworth) Pfeiffer cult. ACG (FLAS) MX Corryocactus melaleucus Ritter MA 1561 (USM) PE: Lima Corryocactus meyenii = C. aureus (Meyen) Hutchison ex Buxbaum CO s/n PE: AR Eriosyce islayensis (Foerster) Katterman MA 1620 (USM) PE: Ica Pereskia sacharosa Grisebach MA 1808 (FLAS, LPB) BO: Cochabamba Pilosocereus pachycladus F. Ritter cult. CSSA (FLAS) BR Praecereus euchlorus (Weber) Taylor MA 1663 (USM) PE: Cajamarca Rhipsalis baccifera (J.S. Mueller) Stearn MA 1818 (FLAS, LPB) BO: Cochabamba Abbreviations. Collectors : CO (Carlos Ostolaza), DG (David Griffiths), DH (Donald Henne), FC (Ftima Cceres), GC (Graham Charles), HU (Leopoldo Horst & Werner Uebelmann), JL (Jol Lode), KK (Kar el Knize), MA (Mnica Arakaki), NQ (Noem Quispe). Nurseries : ACG (Arizona Cactus Garden, Arizona), CSSA (Cact us and Succulent Society of America), JBLP (Jardn Botnico La Paz, Bolivia), MG (M esa Garden, New Mexico), CAI (Cactus-Aventures International, Spain). Countries : AR (Argentina), BO (Bolivia), BR (Brazil), CH (Chile), MX (Mexico), PE (Peru), US (United States).
57 Table 2-2. DNA markers and primer sequences used for amplification and sequencing in the present study. DNA marker Primer Sequence 5-3 Reference rpoB BF unpublished B. Crozier, pers. comm. 2704R unpublished B. Crozier, pers. comm. 2791F unpublished B. Crozier, pers. comm. 1583R unpublished B. Crozier, pers. comm. rpl16 F71 GCT ATG CTT AGT GTG TGA CTC G Jordan et al. 1996 R1661 CGT ACC CAT ATT TTT CCA CCA CGA C Jordan et al. 1996 23S 23SF unpublished B. Crozier, pers. comm. 23S1000R unpublished B. Crozier, pers. comm. ITS1-5.8S-ITS2 ITS4 TCC TCC GCT TAT TGA TAT GC White et al 1990 ITS5 GGA AGG AGA AGT CGT AAC AAG G Downie and Katz-Downie 1996 Table 2-3. Sequence information for individual partitions. Sequence characteristics rpoB rpl16 23S ITS Combined data set Sequence length in bp (including indels) 2552 1197 1313 665 5727 No. of parsimony-informative sites 148 94 75 57 374 % of total sites 5.8% 7.9% 5.7% 8.7% 6.4% No. of parsimony-uninformative sites 267 129 110 79 585 No. of constant sites 2137 974 1128 529 4768
58 Table 2-4. Sequences of degenerate primers designed for amplification of PI and AP3, and sequences of degenerate primers used to amplify the second and third exons of LFY for subsequent design of non-degenerate primers (second set). Primer name Primer sequence (5 3) PISTILLATA (PI ) PI-e1F2 C(CT)G GiA A(AG)A GG(CT) iTG GGA TGC (AT)AA GC PI-e2R3 G(CT)T CiA (CT)C(CT) GCA TT(AG) T(CT)(AG) TCA TTi TC PI-e3R2 AGA TG(CT) (AG)AT GTC (CT)TC (CT)CC (AC)TT iAi GTG C APETALA3 ( AP3 ) AP3-e4F3 C(AG)A AG(AG) ATG GG(AGT) GGi (AGT)AT GG(AT) GGA i AP3-e6R1 G(AG)(AG) (AC)TT CTT C(AGC)A GGT TCT TTA CCT T(CT)T T(CT)C LEAFY (LFY) -first set LFY-7F CAA AG(AG) GAA CA(CT) CC(ACT) TT(CT) AT(AT) G LFY-51F (AGT)(AC)G AGG (AGT)AA (AG)AA GAA TGG (AG) LFY-71F T(GT)G ATT AT(CT) TGT T(CT)C AT(CT) T(GC)T ATG A LFY-116F T(ACT)C AAG TTC A(AG)A A(CT)A TT(GT) CTA AGG A LFY-222R TAG TG(CT) CTC ATT TT(GT) GG(CT) TTG LFY-345R GC(AT) AC(ACT) AG(AG) GG(CT) TTG TAA CA(AG) LFY-381R A(AG)A TGG (AC)(AG)T CAA TAT CCC A(AC)C CT LFY-427R GAG CTT (AG)GT GGG (AT)AC ATA CCA AAT LEAFY -second set LFY-1F TAT GAG CAG TGC CGT GAG TT LFY-4F GAG CAG TGC CGT GAG TTC TT LFY-49F AGG AAC GTG GGG AAA AGT G LFY-42R GTT GCT CCT GCT CTC TTT GC FY-47R TGT ACG TTG CTC CTG CTC TCT
59 Figure 2-1. Schematic map of LEAFY, showing second and third exons (boxes) and second intron (line), and approximate position of primers used for amplification and sequencing. Primer sequences are provided in Table 2-4. Table 2-5. Results of parsimony analyses of individual and combined data sets (including and excluding hybrids); SC refers to strict consensus tree. Sequence data rpoB rpl16 23S ITS combined data set incl. hyb excl. hyb Tree length 589 348 225 244 1652 1524 No. of most parsimonious trees 5000 5000 5000 5000 10577 5000 Consistency Index (CI) 0.796 0.763 0.892 0.675 0.481 0.671 Retention Index (RI) 0.820 0.828 0.906 0.835 0.375 0.714 Rescaled Consistency Index (RC) 0.653 0.632 0.808 0.564 0.181 0.479 No. of resolved clades in the SC 62 34 36 59 79 99 427R 7F 51F 71F 116F 381R 345R 222R 1F 4F 49F 42R 116F 222R EXON 2 EXON 3 First set Second set 47R
60 Figure 2-2. Majority-rule consen sus of 10,577 most parsimonious trees from the combined analysis of chloroplast (rpoB, rpl16, 23S) and nuclear (ITS) sequence data, for 107 taxa representing the Trichocereeae and 11 outgroups. Bootstrap values over 50% are above the branches. CI =0.481, RI=0.375. Genera represented by more than one species are color coded. Genera with only one species (mostly types) are in black. indicates reported or su spected hybrid individuals.
61 Figure 2-3. Majority-rule cons ensus of 5000 most parsimonious trees obtained from the combined analysis of chloroplast ( rpoB rpl16 23S) and nuclear (ITS) sequence data, excluding hybrid individuals. Bootst rap values are above the branches. CI=0.671, RI=0.714. Genera represented by more than one species are color coded. Genera with only one species sampled (mostly types) are in black.
62 Figure 2-4. Maximum likelihood tree from the combined analysis of chloroplast (rpoB, rpl16, 23S) and nuclear (ITS) se quence data, assuming the GTR+ + I model, log L = 18425.5837. Bootstrap values ar e above the branches. Genera represented by more than one species are color coded. Genera with only one species sampled (mostly types) are in black.
63 Figure 2-5. Maximum likelihood phylogram obtai ned from the combined analysis of chloroplast (rpoB, rpl16, 23S) and nucle ar (ITS) sequence data, excluding hybrid individuals. Number of supporting characters for each clade are above the branches.
64 Figure 2-6. Majority-rule consensus tree from 1800 post-burn-in trees sampled by Bayesian mcmc analysis of the combined data set of chloroplast (rpoB, rpl16, 23S) and nuclear (ITS) sequence data, ex cluding hybrids. Post erior probabilities of clades are shown above the branches. Genera represented by more than one species are color coded. Genera with onl y one species sampled (mostly types) are in black.
65 CHAPTER 3 DEVELOPMENT OF POLYMORPHIC MICROSATELLITE LOCI FOR THE EXAMINATION OF GENETIC DI VERSITY IN Haageocereus1 Introduction Haageocereus (Cactaceae) is re stricted to the western s lope s of the Andes and is one of the most taxonomically complex genera in Cactaceae. Haageocereus is considered vulnerable to extinction threats, and seven spec ies have been classified as in danger of disappearance throughout all or a significan t portion of their ranges (Ostolaza 1996). Conservation of these taxa requires an unders tanding of population structure and levels of genetic diversity that can be characterized with the use of polymorphic microsatellite markers, and through the study of widespread congeners in addition to the rare species (Gitzendanner and Soltis 2001). Only few reports on isolation of microsatellites have been published for Cactaceae (Otero-Arnaiz et al. 2004, Helsen et al 2006, Terry et al 2006, Hardesty et al. 2008). Examples of recent studies on genetic diversity within the family include the use of RAPDs (Vite et al 1996, Clark-Tapia and Molina-Freaner 2004); and allozyme studies in Pachycereus (Fleming et al. 1998), Pereskia (Nassar et al 2002), Melocactus (Nassar et al 2001, 2008), Lophocereus (Nason et al. 2002), various columnar cacti (Hamrick et al 2002) and Venezuelan cacti (Nassar et al. 2003). Our goal was to develop microsatellite loci to examine the amount of genetic diversity in two rare ( H. tenuis and H. repens) and two widespread yet vulnerable (H. pseudomelanostele and H. acranthus ) species of Haageocereus 1 Submitted for publication in Molecular Ecology Resources (2008).
66 Haageocereus tenuis (2 n = 3 x = 33) consists of a single population restricted to two square kilometers (a complete descrip tion of the species is in Chapter 5); H. pacalaensis subsp. repens (2 n = 2 x = 22) formerly comprised a singl e natural population, 10 % of all known individuals were transloc ated to surrounding areas and botanical institutions before its entire range was converted to agriculture (Ostolaza 2000); H. acranthus (2 n = 4 x = 44) and H. pseudomelanostele (2 n = 2 x = 22) consist of several large populations. Materials and Methods Forty five and 32 individuals were sam p led for the single existing populations of H. tenuis and H. pacalaensis subsp. repens respectively. For the widespread H. pseudomelanostele and H. acranthus three populations, with 30 to 32 individuals per population, were sampled for each. We used an enrichment procedure based on Ernst et al (2004). Genomic DNA was extracted using the DNeasy Mini Ki t (Qiagen) and digested with Sau3AI. DNA was purified and fractioned using Chroma Spin columns (Clontech Laboratories) which removed fragments smaller than 400 base pairs. Remaining DNA was ligated to Sau 3AI linkers and amplified by PCR. The frag ment library was enriched for (CA)n repeats by hybridizing DNA fragments to a biotinylated nucleic acid pr obe. Probe-target fragments were captured using VECTREX Avidin D (Vector Laboratori es, Inc.), amplified by PCR, and ligated into a TOPO TA pCR vector, transformed into One Shot Escherichia coli competent cells (Invitrogen). Colonies were screened by binding them to nitrocellulose membranes (Osmonics) and hybridizing with chemiluminescent (CA)n probe; positive colonies were detected by Lumi-Phos 480 (Lif ecodes). Colonies were grown overnight in Luria broth, plasmids purified using QIAprep Spin Miniprep Kit (Qiagen) and screened
67 once more by dot-blotting serial dilutions on nylon membranes. Plasmids were hybridized to a (CA)n probe and those containing the repeats were sequenced on a CEQ 8000 capillary sequencer (Beckman-Coulter). Primers were developed using Primer 3 (Rozen and Skaletsky 2000) and Operon (Operon Biotechnologies, Inc.). An M 13 tail (5-CACGACGTTGTAAAAC3) was added to each forward primer and labeled w ith D4 (Beckman-Coulter). PCR amplification procedures were optimized for each primer pair following Speranza and Malosetti (2006). PCR amplifications were performed with 10 L total volumes containing 0.2 units of NEB Taq polymerase (New England Biolabs), 1.5 nM MgCl2, 0.15 M of the reverse primer and labeled M13 primer, 0.01 M of the extended forward pr imer, and 0.1 mM of each dNTP. PCR reactions started with a 5 min denaturation time at 94 C, followed by 40 cycles of 15 sec at 94C and 3 min at 53 C, and a final extension step of 5 min at 72 C. PCR products were run on a CEQ 8000 capillary sequencer. A total of 21 primer pairs showed successful amplificati ons; two of the isolated micr osatellites were found to be redundant. Samples were multiplexed by loading 0.3 L (for three samples). 0.35 L of CEQ TM DNA Size Standard Kit-400 (Beckman-Coulter) was added to 25 L of formamide per well. Resulting electropherogram s were visualized using CEQ TM Genetic Analysis System Software (Beckma n-Coulter) and scored manually. Results and Discussion From the 19 polymorphic microsatellite loci (Table 3-1), five were selected ( Ht.Id, Ht.Ie, Ht.IIa, Ht.ms7, Ht.ms11) for the present study and three more loci were characterized in H. pseudomelanostele ( Hp.IVc, Ht.ms8, Ht.ms16) Cross-species
68 amplification and documentation of high variability has shown that five loci are sufficient for addressing questions of population di fferentiation and genetic diversity. AMOVA analyses were executed using Micr osatellite Analyzer [MSA] (Dieringer and Schltterer 2002) (Table 3-2). We calculated Fst and expected heterozygosity (He). All loci were variable. Heterozygosity was relatively high in widespread H. pseudomelanostele (He=0.82), and low levels of population diffe rentiation were found among the three H. pseudomelanostele populations (Fst 0.04), indicating substantial genetic exchange. These data agree with pollination data (Sahley 1995); these plants ar e presumed outcrossers with bat and hummingbird pollination. Heterozygosity in the rare H. repens was lower than in widespread H. pseudomelanostele (He=0.77). Polymorphic loci were analyzed to assess deviation from Hardy-Weinberg equilibrium and linkage disequilibrium among loci using Fishe rs exact tests in GENEPOP version 3.4 (Raymond and Rousset 1995). No significant linkage disequilibrium was detected among each pair of loci. None of th e loci exhibited heterozygote deficiencies. An important limitation in the use of microsat ellites in polyploid species is that the dosage effect of each allele is difficult to determine. We were not able to compute F statistics in the tetraploid H. acranthus and triploid H. tenuis therefore we employed allele number as an indicator of genetic diversity (Table 3-3). As expected, the widespread tetraploid H. acranthus showed higher allele numbers than the widespread diploid H. pseudomelanostele Both possess higher allele numbers than the endangered H. repens, which is restricted to a si ngle population. In view of the results we suggest the protection and rein troduction of H. repens individuals translocated
69 from their original location, and the evaluation of the conservation stat us of the triploid microspecies H. tenuis Widespread H. acranthus and H. pseudomelanostele populations were found to have both high heteroz ygosity and high outcrossing rates. Microsatellites have proved to be very efficient markers for the determination of genetic diversity and populat ion differentiation in other groups (e.g., Jame and Lagoda 1996, Symonds and Lloyd 2003, Vigouroux 2005). Now that we have obtained 19 variable microsatellite markers that work throughout th e genus, 11 of which have been shown to be highly variable and applied su ccessfully to species of Haageocereus, we can continue applying these markers in the study of genetic diversity in other endangered species of Haageocereus. This information is urgently needed for determining species that deserve priority among the large number of endangere d species waiting to be included in current conservation plans.
70Table 3-1. Characterization of 19 pol ymorphic microsatellite loci in Haageocereus tenuis ( Ht ) and Haageocereus pseudomelanostele ( Hp ). Locus Primer sequence (53) Repeat array micros atellite length in bp N of allele size HO/HE GenBank (plus flanking sequence) alleles range (bp) acc. number Ht.Ia F: *TTGAGGCCAATATGAGATTTGA (AC)23AT(AC)2AT(AC)14(AT)6(AC)6(AT)4(AC)16GT(AT)5 58 (215) ------EF444509 R: TATGGCCCATTTGAAACCAT Ht.Ib F: *CCAAGAAGCCATCTGAGGAG (AC)15(AG)15 60 (154) ------EF444510 R: CACCCCCTTCCTCTCTCTCT Ht.Ic F: *ATCCCAATTT CTTGCCTCCT (TA)6(CA)23(TA)5(GA)13 94 (117) ------EF444511 R: GCTCACGAATGGTCAGAAGA Ht.Id F: *CCAAGATTGGTCGTCGACAT (GT)17AT(GT)6 48 (185) 16 221-253 0.74/0.84 EF444512 R: CTTCGCCCCACCTTCTCTA Ht.Ie F: *TCACCTATTTGATCCCCTTCC (GT)21 42 (147) 13 197-219 0.80/0.86 EF444513 R: CCTAAGTGCTTGAGCCAAAAA Ht.If F: *TCCACATGCTACGACCACTT (AC)21(AT)6 54 (138) ------EF444514 R: AGGCCTCATCCTACCAGAGTC Ht.IIa F: *CATGAATTGAAAGCCACACG (AC)17AT(AC)2GC(AC)4GT(AC)2CCATACAT(AC)4 72 (150) 12 181-225 0.48/0.56 EF444515 R: CCTTGCACTGCATAGTTGGA Hp.IVa F: *CGACAAACCTTGTCCTCTTG (CA)8CG(CA)26(TA)3TG(TA)3 84 (195) ------EF444516 R: AGGTCCGACTGGTCCTAACC Hp.IVc F: *AACCCTTGCAATAAGCTCCA (TG)21 42 (149) 17 201-249 0.87/0.92 EF444517 R: AATGGCAACGAAAGGAGAGA Ht.Vb F: *TCCAATGACATTTGCTATTCCA (CA)6CG(CA)2(CG)4CAGA(CA)5 40 (115) ------EF444518 R: CCCCATCTTCCCTACAGTCA Ht.ms2 F: *CATGGCTATACCCAAAGTGG (CA)32 64 (193) ------EF444520 R: AAATGCCATTGCACATCTTTT Ht.ms4 F: *GCAAGAAAGGGAAGAGAAATC (CA)18(TA)6(CA)11 70 (190) ------EF444519 R: TCAATGTCGTTCCTAAACATGAA Ht.ms7 F: *TTCCCATGACTGCCCTTAG (AT)6(GT)6AT(GT)12 50 (134) 16 179-203 0.73/0.81 EF444521 R: CTGTCCATTTAGCCCCAGAA Ht.ms8 F: *ACCCTTGCTGTTTGTGTGG (CA)12(CG)2CACGCA(CA)3 40 (140) 26 179-209 0.84/0.85 EF444522 R: CCTCCTGCGAAGAAAGAGAA Ht.ms9 F: *TACGGTTTGGGTTTTGGAC (AC)15 30 (208) ------EF444523 R: ACTGGGATCTCCCAACTGC Ht.ms11 F: *GCCATTACCCCAACGTA (GT)6GCGTCT(GT)8(GC)2(GT)6AT(GT)11 74 (179) 42 289-365 0.79/0.96 EF444524 R: CTCACCTCCTTCATCTCTTCAA Ht.ms12 F: *GGTTATGACCATGGAATGC (AC)20(AT)6AG(AC)7 68 (197) ------EF444525 R: ACTTGCAACCGTGTGTGTGT Ht.ms13 F: *TCAAGACATGGCAGCTCAA (GT)14ATGC(GT)4ATGC(GT)4 52 (148) ------EF444526 R: GGCATGGTGCTGATTAGGAT Ht.ms16 F: *CTGGCCAAACCCTATCTACT (TG)9GG(TG)2(CG)2(TG)4GG(TG)14(AG)8AAGA(G)4(AG)8 110 (187) 20 175-339 0.81/0.95 EF444527 R: GGGTACACCGATCAGATAAAGG HO/HE (observed and expected heterozygosity), *M13 tag (CACGACGTTGTAAAACGAC) added to 5 end of primer for amplificati on with flueorecently labeled M13.
71 Table 3-2. AMOVA for three populations of H. pseudomelanostele and five microsatellite loci. Output from Microsatellit e Analyzer [MSA] (Dieringe r and Schltterer 2002) --------------------------------------------------------------------------------------------------------------------Source of Sum of Variance Percentage variation d.f. Squares components of variation --------------------------------------------------------------------------------------------------------------------Among populations 2 14.773 0.08802 Va 4.9 Within populations 181 364.081 2.01150 Vb 95.81 --------------------------------------------------------------------------------------------------------------------Total 183 378.853 2.09951 --------------------------------------------------------------------------------------------------------------------Fixation Index Fst: 0.04192 Significance test P (rand.Value >= obs.Value) = 0.00000+-0.00000 (1023 permutations) Table 3-3. Summary of allele counts for four species of Haageocereus and five microsatellite loci. Locus designation Species Chromosome Number N o of populations Ht.Id Ht.Ie Ht.IIa Ht.ms7 Ht.ms11 All loci H. acranthus 4x = 44 3 (widespread) 17 26 13 27 44 127 H. pseudomelanostele 2x = 22 3 (widespread) 16 13 12 16 42 99 H. repens 2x = 22 only existing pop 18 15 8 14 37 92 H. tenuis 3x = 33 only existing pop 2 3 3 1 3 12 All species 22 27 20 31 66
72 CHAPTER 4 CHROMOSOME COUNTS AND EVIDENCE OF POLYPLOIDY IN Haageocereus AND OTHER TAXA OF TRI BE TRICHOCEREEAE1 Introduction Polyploidy has been suggested to be a prom inent process during angiosperm evolution (e.g., Stebbins 1950, 1971, Grant 1980, 1981, Soltis and Soltis 1993, 1995, Soltis and Soltis 2000, Leitch and Bennett 1997, Wendel 2000, Levin 2002, Soltis et al. 2004, Tate et al 2005, Wendel and Doyle 2005). In Cactaceae, studies in Opuntia (Grant 1971, 1980, Grant and Grant 1979, Rebman and Pinkava 2001) and Mammillaria (Katagiri 1953, Remski 1954) show that polyploidy, as well as hybridizati on, are major evolutionary forces in the family, but perhaps previously underappreciated. Patterns of relation ships in both genera were found to be complex due to polyploidy, interspecifi c hybridization, and vegetative pr opagation. These processes may also be frequent in other cacti (Pinkava et al. 1985, Anderson 2001). For example, Ross (1981) conducted cytological, morphological and reproduc tive studies on 55 species of Cactaceae. His observations on modes of reproduction showed a correlation between polyploidy and selffertility, vegetative reproduction, adventive em bryony, and profuse branching. Nonetheless, the relative importance of hybridi zation and polyploidy in this fa mily (of about 1800 species) remains uncertain because so few cacti have been examined in detail. In spite of the fundamental importance of chromosome number, counts have been concentrated in only a few genera of Cactaceae, mainly from North America (see Index to Plant Chromosome Numbers by Goldblat t and Johnson 1978-2006), including Opuntia Mill. (Remski 1954, Pinkava and McLeod 1971, Pinkava et al. 1973, 1977, Ross 1981, Pinkava and Parfitt 1982, Pinkava et al. 1985), Mammillaria Haw. (Katagiri 1953), Echinocereus Engelm. (Cota and 1 Published. Arakaki, M., D. E. Soltis and P. Esperanza. 2007. New chromosome counts and evidence of polyplody in Haageocereus and related genera in Tribe Trichocereeae and other tribes of Cactaceae. Brittonia 59(3): 290-297.
73 Philbrick 1994, Cota and Wallace 1995), and Selenicereus Britton & Rose (Lichtenzveig et al. 2000). One exception is the work of Lambrou and Till (1993) that surveyed the entire genus Gymnocalycium Pfeiff. ex Mittler. Polyploidy has been reported to be absent in Pereskioideae, widespread in Opuntioideae and sporadic in Cactoi deae, occurring mostly at the tetraploid level (Pinkava et al. 1985). Significantly, only about 15 % of Trichocereeae (Cactoideae) have published chromosome counts and there are no published chromoso me counts for the large genus Haageocereus (Goldblatt and Johnson 2006). Polyploids in Trichocereeae have been reported in 27 species: Trichocereus spachianus with 2n = 4x = 44 (Katagiri 1953), Gymnocalycium bruchii with 2n = 4x = 44, Rebutia kupperiana with 2n = 4x = 44, Rebutia spegazziana with 2 n = 4x = 44 (Ross 1981), 20 species of Gymnocalycium with 2n = 4x = 44 and two species with 2 n = 6x = 66 (Lambrou and Till 1993); and Weberbauerocereus weberbaueri with 2n = 4x = 44 (Sahley 1996). We therefore undertook cytogenetic stud ies in the Trichocereeae (with a focus on Haageocereus, a large genus for which no counts have been reported), plus other tribes in the Cactoideae, to provide the first chromosome counts for most of these groups and to assess whether polyploids are present. Materials and Methods Stem sections and seeds were collected from natural populations (Table 4-1). Vouchers were deposited in Herbario San Marcos, Lima (USM) and Herbario San Agustin, Arequipa (HUSA), Peru. Somatic chromosomes were count ed using root tips. Root tips were obtained in two different ways. We germinated seeds for a few ta xa (Table 4-1) on moist filter paper and then removed root tips from them. For most taxa, we used stems of plants collected from natural populations (Table 4-1). Stems we re used to generate plants, which were maintained in the University of Florida Botany Department greenhous e and induced to develop adventitious roots,
74 which are several times larger than roots from seedlings and hence much easier to use in chromosome squashes. General cytogenetic methods followed Soltis (1980) and Speranza et al. (2003). Root tips from seedlings or those collected in the greenhouse were collected during the early hours of the morning (between 7:00 and 9:00 am), when cell division has been observed to be most active (Cota and Philbrick 1994), and placed in a solu tion of 2 mM 8-hydroxyquinoline for 4 to 6 hours at room temperature and 4 hours to overnight at 4 C. After this treatment, roots were rinsed in distilled water and fixed in a solution of 3:1 abso lute ethanol and glacial acetic acid, for at least 24 hours at room temperature. If not used immediat ely, roots were stored in 70 % ethanol at 4 C. Fixed roots were rinsed in buffe r (40 mM citric acid, 60 mM sodi um citrate), digested at 37 C with a combination of 3 % (w/v) cellulase (Cal biochem, San Diego, CA), 1 % (w/v) cellulase Onozuka RS (Yakult Pharmaceutical, Japan), and 4 % (v/v) pectinase (Sigma, St. Louis, MO). The digestion time varied and had to be adjusted for each individual. Most roots were digested 30 to 90 min. Root tips were dissected in 60 % ac etic acid, stained with 2 % lacto-propionic (1:1) orcein, squashed, and sealed. Two to ten cells pe r individual were examined in each case and separate counts were made for different individu als of the same species or individuals of the same species obtained from different sources. In itial microscopic observations were made under a Nikon Alphaphot-2 microscope and photographs taken under a Zeiss Axioplan microscope with a Kodak MDS 290 digital camera. Results and Discussion Chrom osome numbers for a total of 54 indivi duals, representing 14 genera and 39 species are reported (Table 4-1). Like other Cactaceae, the basic chromo some number for all the taxa examined was x = 11 (Lewis 1980, Pinkava et al. 1985). The diploid number 2 n = 2x = 22 was found in most species of Haageocereus examined, as well as other Trichocereeae we analyzed
75 (e.g., the genera Cleistocactus Lem. Matucana Britton & Rose Mila Britton & Rose). Importantly, polyploid counts were also obtained for some species (Fig. 4-1), including triploid (2 n = 3x = 33), tetraploid (2 n = 4x = 44), hexaploid (2 n = 6x = 66), and octoploid (2 n = 8 x = 88) numbers. No aneuploids were record ed. Six polyploids were detected for Haageocereus: H. acranthus, H. ollowinskianus, H. chalaensis H. fulvus, H. multicolorispinus (2 n = 4 x = 44); and H. tenuis (2 n = 3 x = 33). Among other Trichocereeae, polyploid s were detected in five species belonging to the genera Cleistocactus Lem. Espostoa Britton & Rose and Weberbauerocereus Backeb.: C. sepium (2 n = 4 x = 44), E. lanata (2 n = 4 x = 44, and 2 n = 6x = 66), W. rauhii (2 n = 4x = 44, and 2 n = 8 x = 88), W. weberbaueri (2 n = 4 x = 44), and W. winterianus (2 n = 8 x = 88). In two cases, a single species c ontained different cytotypes: Espostoa lanata, with 2 n = 22, 44 and 66; and Weberbauerocereus rauhii with 2n = 44 and 88. Chrom osomes were very small in size in all taxa examined (3 to 5 m). They were all observed to be of similar size, metacentric or submetacentric, and were not clearly distinguished morphologically. It has been argue d that speciation has occurred rapidly and relatively recently in the Cactaceae and may have been accompanied by very little (or at least cryptic) chromosomal change (e.g., in Mammillaria Remski 1954, 55 taxa of Cactaceae, Ross 1981, Echinocereus, Cota and Philbrick 1994). Some authors also su ggest that the high sim ilarity in chromosome morphology and number observed would explain in pa rt the ease with which cacti can cross and produce fertile inter generic hyb rids even between morphological ly divergent genera (Remski 1954, Gibson and Nobel 1986). But other than work focused on a relatively small group of Cactaceae, very little is know n about karyotypes in the group. These are the first chromosome counts for Haageocereus and the first reports of polyploidy in Haageocereus, Espostoa, and Cleistocactus. Polyploidy in Cactaceae can occur
76 through premeiotic abnormalities (Ross 1981) or so matic doubling in the meristems, as observed in Mammillaria (Remski 1954). These events can lead to the establishment of polyploids when they occur in conjunction with self-fertility or asexual repr oduction (Ross 1981). Polyploidy has been suggested as an important evolutionary mechanism in plants (e.g., Stebbins 1950, 1971, Grant 1980, 1981, Soltis and Soltis 1993, 1995, Levin 2002, Soltis et al 2004, Tate et al. 2005; Wendel and Doyle 2005), and Cactaceae are not an excep tion. It has been suggested that some of the major changes occurring in this group ar e related to chromosome doubling (Gibson and Nobel 1986). In specific genera like Espostoa and Weberbauerocereus the prevalence of high ploidal levels indicates that polyp loidy has played an important ro le in their diversification. No diploids have been detected in Weberbauerocereus and the genus may have an allopolyploid origin given that only tetraploid (2 n = 4 x = 44) and octoploid (2n = 8x = 88) cytotypes have been detected. In spite of the above, the presence of diploids in almost every species in the rest of the genera indicates that differentiation within th e Trichocereeae has been occurring mostly at the diploid level. Previous polyploid counts in Trichocereeae repo rted only tetraploids and hexaploids. It is shown here that, as in the case of Opuntia (Baker and Pinkava 1987), uneven ploidy levels are not only present (most probably as a result of sexual polyploidization), but also fixed by asexual reproduction. In this case the triploid microspecies H. tenuis propagates by apomixis (Chapter 5). Haageocereus polyploids thrive in extremely arid and severe environments compared to most diploid species in the genus. For ex ample, several populations of the polyploid Haageocereus acranthus are found in disturbed areas, usually in dry steep rocky slopes. They receive water only during the short rainy s eason (December to March). The only existing
77 population of the polyploid H. tenuis is found in a sandy plain far fr om any source of fresh water. However, fogs present during the winter months (June-August) maintain these plants, which are also facing habitat loss due to human pressures. Fog also seems to be the main source of water for the polyploids H. chalaensis and H. multicolorispinus Diploids are not observed in such harsh conditions. The polyploids Weberbauerocereus rauhii and W. weberbaueri occupy very dry areas where not many other plants survive. Si nce they set flower and fruit almost year round, during the dry season they become almost the only source of food for birds and bats occupying the area (Sahley 1995). These examples support the idea that polyploidy confers greater ecological tolerance (Remski 1954, Otto and Whitton 2000, Garcia et al. 2006). Most of the polyploids also show some characteristic morphol ogical features, such as a dark green-bluish stem color and reduced number of st em ribs compared to diploids. Additional species of Trichocereeae are being examined and chromosome counts produced. This information will be valuable for ongoing systematic and population genetic studies. We want to evaluate further the prev alence and evolutionary significance of polyploidy in Haageocereus and other genera in the Trichoc ereeae. We suggest that polyploidy, hybridization and clonal reproductio n have played prominent roles in the evolution of some groups within the Trichocereeae.
78 Table 4-1. Somatic chromosome numbers for ta xa in the Trichocereeae and members of other tribes. Polyploid numbers are in bold. Key to the abbreviations is given below the table* Taxon Provenance and voucher specimens Chromosome number (2 n ) TRIBE TRICHOCEREEAE Borzicactus cajamarcencis F. Ritter PE: Cajamarca, San Marcos, MA 1699 (USM) 22 Cleistocactus acanthurus (Vaupel) D.R. Hunt PE: Lima, Huarochir, MA 1629 (USM) 22 Cleistocactus acanthurus (Vaupel) D.R. Hunt PE: Lima, Huarochir, MA 1630 (USM) 22 Cleistocactus sepium (Kunth) F.A.C. Webber ex Rol.-Goss. PE: Arequipa, Arequipa, MA 1606 (USM, HUSA) 44 Cleistocactus serpens (Kunth) F.A.C. Webber ex Rol.-Goss. PE: La Libertad, Otuzco, MA 1714 (USM) 22 Echinopsis eyriesii (Turpin) Pfeiff. & Otto AR: Formosa, KK 1474 (MG) 22 Espostoa blossfeldiorum (Werderm.) Buxb. PE: Amazonas, Chachapoyas, MA 1691 (USM) 22 Espostoa lanata (Kunth) Britton & Rose PE: La Libertad, Trujillo, MA s/n (USM) 22 Espostoa lanata (Kunth) Britton & Rose PE: Lambayeque, Lambayeque, MA 1659 (USM) 44 Espostoa lanata (Kunth) Britton & Rose PE: Lambayeque, Lambayeque, MA 1656 (USM) 66 Espostoa senilis (F. Ritter) N. P. Taylor PE: Cajamarca, San Marcos, MA 1704 (USM) 22 Haageocereus acranthus (Vaupel) Backeb. PE: Lima, Huarochir, MA 1628 (USM) 44 Haageocereus acranthus (Vaupel) Backeb. PE: Lima, Lima, CO s/n (MG) 44 Haageocereus acranthus subsp. ollowinskianus (Backeb.) Ostolaza PE: Lima, Huaura, MA 1644 (USM) 44 Ha ageocereus australis Backeb. PE: Ica, Nazca, MA 1616 (USM) 22 Haageocereus chalaensis F. Ritter PE: Arequipa, Caravel, MA 1600 (USM, HUSA) 44 Haageocereus decumbens (Vaupel) Backeb. PE: Arequipa, Islay, MA 1578 (USM, HUSA) 22 Haageocereus decumbens (Vaupel) Backeb. PE: Arequipa, Islay, MA 1579 (USM, HUSA) 22 Haageocereus decumbens (Vaupel) Backeb. PE: Arequipa, Caravel, MA 1588 (USM, HUSA) 22 Haageocereus fulvus var. yautanensis PE: Ancash, Huaraz, MA 1650 (USM) 44 Haageocereus horrens Rauh & Backeb. PE: Lambayeque, Chiclayo, MA 1652 (USM) 22 Haageocereus icosagonoides Rauh & Backeb. PE: Cajamarca, San Pablo, MA 1707 (USM) 22 Haageocereus icosagonoides Rauh & Backeb. PE: Cajamarca, San Pablo, MA 1711 (USM) 22
79 Table 4-1. (cont.) Haageocereus icosagonoides Rauh & Backeb. PE: No locality information, KK 1639 (MG) 22 Haageocereus multangularis (Willd.) F. Ritter PE: Ancash, Huarmey, MA 1616 (USM) 22 Haageocereus multicolorispinus Buining PE: Ica, Nazca, MA 1617 (USM) 44 Haageocereus pacalaensis subsp. repens (Rauh & Backeb.) Ostolaza PE: La Libertad, Trujillo, MA 1539 (USM) 22 Haageocereus pacalaensis subsp. repens (Rauh & Backeb.) Ostolaza PE: La Libertad, Trujillo, MA s/n (USM) 22 Haageocereus platinospinus (Werderm. & Backeb.) Backeb. PE: Arequipa, Arequipa, MA s/n (USM) 22 Haageocereus platinospinus (Werderm. & Backeb.) Backeb. PE: Arequipa, Arequipa, MA 1614 (USM, HUSA) 22 Haageocereus pseudomelanostele (Werderm. & Backeb.) Backeb. PE: Ancash, Huaraz, MA 1651 (USM) 22 Haageocereus pseudoversicolor Rauh & Backeb. PE: No locality information, KK 1380 (MG) 22 Haageocereus tenuis F. Ritter PE: Lima, Huaura, MA 1635 (USM) 33 Haageocereus versicolor (Werderm. & Backeb.) Backeb. PE: Lambayeque, Lambayeque, MA 1658 (USM) 22 Haageocereus sp. PE: Arequipa, Caravel, MA 1596 (USM, HUSA) 22 Lasiocereus ful vus F. Ritter PE: Amazonas, Chachapoyas, MA 1684 (USM) 22 Lasiocereus rupicola F. Ritter PE: Cajamarca, San Marcos, MA 1698 (USM) 22 Matucana haynei (Otto ex Salm-Dyck) Britton & Rose PE: Lima, Huarochir, KK 1548 (MG) 22 Matucana yanganucensis Rauh & Backeb. PE: Ancash, Huaraz, MA 1647 (USM) 22 Mila caespitosa Britton & Rose PE: Lima, Huarochir, MA 1627 (USM) 22 Mila caespitosa Britton & Rose PE: Lima, Huaura, MA 1638 (USM) 22 Mila caespitosa Britton & Rose PE: Lima, KK 243 (MG) 22 Rauhocereus riosaniensis Backeb. PE: Amazonas, Utcubamba, MA 1671 (USM) 22 Weberbauerocereus jhonsonii F. Ritter PE: Cajamarca, San Pablo, MA 1708 (USM) 88 Weberbauerocereus rauhii Backeb. PE: Arequipa, Caravel, MA 1592 (USM, HUSA) 44 Weberbauerocereus rauhii Backeb. PE: Ica, Nazca, CO 82173 (MG) 88 Weberbauerocereus weberbaueri (K. Schum. ex Vaupel) Backeb. PE: Arequipa, Arequipa, MA 1613 (USM, HUSA) 44 Weberbauerocereus winterianus F. Ritter PE: La Libertad, Otuzco, MA 1713 (USM) 88 Weberbaureocereus winterianus F. Ritter PE: La Libertad, Otuzco, MA s/n (USM) 88
80 Table 4-1. (cont.) TAXA OUTSIDE TRICHOCEREEAE Austrocylindropuntia pachypus (K. Schum. ) Backeb. [Opuntioideae] PE: Lima, Huarochir, MA 1631 (USM) 22 Browningia microsperma (Werderm. & Backeb.) W.T. Marshall [Browningieae] PE: Lambayeque, Lambayeque, MA 1655 (USM) 22 Corryocactus aureus (Meyen) Hutchison ex Buxbaum [Pachycereeae] PE: Arequipa, Arequipa, MA 1603 (USM, HUSA) 22 Eriosyce islayensis (Frster) Katt. [Notocacteae] PE: Arequipa, Caravel, MA 1591 (USM, HUSA) 22 Praecereus euchlorus (F.A.C. Weber) N. P. Taylor [Cereeae] PE: Cajamarca, Jan, MA 1663 (USM) 22 PE: Peru, AR: Argentina, MA: M. Arakaki, KK: K. Knize, CO: C. Ostolaza, USM: Herbario San Marcos, Lima, Peru; HUSA: Herbario Un iversidad San Agustn, Arequipa, Peru; MG: Mesa Garden, New Mexico.
81 Figure 4-1. Somatic chro mosomes and photographs of actual pl ants from diploid and polyploid species representing the Trichocereeae: A) Haageocereus pseudomelanostele (2n = 22); B) H. tenuis (2n = 33); C) H. fulvus (2n = 44); D) H. multicolorispinus (2n = 44); E) Cleistocactus sepium (2n = 44); F) Weberbauerocereus rauhii (2n = 88). 5 m 5 m A B C D E F 5 m 5 m 5 m 5 m
82 CHAPTER 5 CLONAL REPRODUCTION AND APOMIXIS IN Haageocereus1 Introduction Haageocereus tenuis Ritter is a creeping cactus restrict ed to an area of about two square kilometers in a sandy flat close to the city of Lima, Peru (Ostolaza and Rauh 1990, Ostolaza 1990). This species was found by F. Ritter in 1965 but only described in 1981 (Ritter 1981). Ritters description (lacking characters of both flowers and fruits) was completed by Ostolaza (1990). The prostrate stems are gray to bluish green, partially covere d by wind-blown sand and pieces of shells (Fig. 5-1A). They contain up to 15 ribs and are densely covered by grayish spines with microscopic hairs presumably func tioning in the capture of humidity from fog (Ostolaza and Rauh 1990). Haageocereus tenuis flowers only rarely and never profusely. The flowers are nocturnal, white, funnelform, and up to 10 cm long. Although they are presumed to be pollinated by bats or moths, no observations have been reported on the pollination biology and reproduction of H. tenuis (Ostolaza 1990). Clonal propagation was suspected in H. tenuis due to the prostrate habit and sporadic production of flowers and fruits. As in other cacti that reproduce vegetatively ( Opuntia spp., Austrocylindropuntia pachypus, Stenocereus eruca, Lophocereus schottii ) (Grant and Grant 1971, 1980, Parker and Hamrick 1992, Anderson 2001, Clark-Tapia et al. 2005), H. tenuis develops adventitious roots on stems that are in contact with the ground. These stem segments with roots later detach from the mother pl ant (Gibson and Nobel 1986, Anderson 2001). The succulent habit allows the new plants to survive until they become well established (Anderson 2001). As suggested for S. eruca the low production of flowers a nd fruits may be a consequence of pollinator limitation or lack of favorable condi tions for seed establishment (Clark-Tapia and 1 Accepted for publication in Haseltonia (2008).
83 Molina-Freaner 2004). The only known existing population of H. tenuis is found at the lower limit of a loma formation, an island of vegeta tion in the desert supported by winter fogs. Recent chromosome counts revealed that Haageocereus tenuis is triploid (2n = 3x = 33) (Chapter 4, Fig. 5-1D). In Taraxacum one of the most famous triploid apomicts, apomixis allows the reproduction of plants with unbala nced chromosome numbers, in which meiotic disturbances would cause pronounced sterility (R ichards 1970). Apomixis is often associated with hybridity and a perennial habit (Stebbi ns 1950, 1979, Asker and Jerling 1992). A given genotype that is adaptively s uperior in a certain ecological niche can expand by vegetative reproduction while the production of asexual seed s can increase the plan ts capabilities for dispersal, establishment, and range expansi on (Grant and Grant 1980, Asker and Jerling 1992). Apomixis as defined by Stebbins (1950) and Richards (1997) may include both vegetative reproduction and agamospermy (production of fertile seeds without sexual fusion between gametes). Asker and Jerling (1992) re stricted the term apomixis to agamospermy arguing that all kinds of situations may exist where vege tative reproduction (i.e., by propagules, stem fragmentation, rhizomes) is only a comple ment to sexual reproduction. Two forms of agamospermy are recognized: adventitious embryony when the embryos arise from somatic cells of the nucellus or integument, and gametophytic agamospermy in which embryo sacs are produced from unreduced initial ce lls giving rise to plants of the maternal genotype (Stebbins 1950, Asker and Jerling 1992, Richards 1997). Apomicts have also been classi fied as facultative and obligate, although the latter is considered improbable in nature (Asker and Jerling 1992, Spillane et al. 2001). Apomixis is not rare in angiosperms, including the Cactaceae (Nogler 1984, Asker and Jerling 1992, Naumova 1992); howe ver only adventitious embryony has been
84 documented in the family, in Opuntia (Lakshmanan and Ambegaokar 1984, Asker and Jerling 1992, Negron-Ortiz 1998). Molecular markers have been used in other Cactaceae to examine genetic variation within and between populations and reproductiv e systems. For example, the cactus Stenocereus eruca known as the creeping devil, is restricted to inhospitable sandy flats in the Magdalena region of Baja California and has been considered by Gibson and Nobel (1986) as the most remarkable case of vegetative reproduction in the Cactaceae. However, isozyme studies on the diversity of this species showed moderate le vels of genetic varia tion (higher than those obtained from other clonal cacti) and suggested that both sexual a nd clonal propagation are present (Clark-Tapia et al. 2005). However, not all variabil ity can be attributed to sexu al reproduction. Various types of vegetative mutations also occur in apomictic pl ants, but the importance of such variation is unknown (Asker and Jerling 1992). In analyses of genotypic variation in 21 clonal plant species (Ellstrand and Rose 1987), all but two species revealed multiple clon es within and among populations. The two uniclonal species mentioned ( Gaura triangulata and Taraxacum obliquum) showed very restricted ecological ranges compared to their congeners. The examination of genetic diversity in Haageocereus tenuis using microsatellite markers has the following objectives: 1) to examine gene tic variation within the single known population, 2) to assess whether agamospermy occurs, and 3) to discuss the taxonomic and conservation status of H. tenuis Materials and Methods Study Site and Sampling The site, loc ated in the Province of Huaral (L ima, Peru), close to the town of Chancay, supports the only known population, with about 400 individuals, in an area of approximately two square kilometers. Plants are partially covere d by sand and are difficult to see from great
85 distances. On average, each plan t has 10-15 branches of about 40 cm in length. Old branches are brown, woody, and spineless at the base, with new growths which are green to bluish green and densely covered with gray to whitish soft spin es at the tips (Fig. 5-1B, C). Spines trap sand, small pieces of shells, and feathers/litter carried by the wind from adjacent chicken farms. Small portions of young stems (approximately 20 cm long) were taken from 45 individuals covering the entire population and gr owing at least 3 meters apart from one another. Stems were dried in silica gel and used fo r DNA extraction and subsequent mi crosatellite analysis. Fruits collected from a plant collected in the same population and main tained in a garden (by C. Ostolaza, Lima) were used to obtain seeds and assess the occurrence of agamospermy. Examination of Apomixis Mature seeds were dissected to obse rve polyem bryony, a phenomenon frequently associated with apomixis (Asker and Jerli ng 1992). To examine agamospermy, seeds were germinated and seedlings used to extract DNA and run microsatellites. Ro ot tips were collected and used to count chromosome numbers (Chapter 4). Microsatellite Amplification and Analysis The general m ethods employed for microsatel lite capture are described in Chapter 3. Genomic DNA was extracted using the DNeasy Mi ni Kit (Qiagen Inc., Valencia, CA, USA). PCR amplifications were performed with 10 L total volume containing 0.2 units of NEB Taq polymerase (New England Biolabs Inc., Ipswich, MA, USA), 1.5 nM MgCl2, 0.15 M of the reverse primer and labeled M13 primer, 0.01 M of the extended forward primer, and 0.1 mM of each dNTP. PCR reactions started with a 5-min de naturation time at 94C, followed by 40 cycles of 15 sec at 94C and 3 min at 53 C, and a final extension step of 5 min at 72C. PCR products were run on a CEQ 8000 (Beckman-Coulter, Fullerton, CA, USA) capillary sequencer. We employed microsatellite markers to determine th e individuals genotypes. These markers were
86 designed for H. tenuis and previously applied in the determination of ge netic variation in three other species of Haageocereus (Chapter 3). Five loci were used for this study ( Ht.Id, Ht.Ie, Ht.IIa, Ht.ms7, Ht.ms11). Results and Discussion Agamospermy Seed germ ination took three to seven days and germination success was 60% (18 of 30 seeds). Six out of the 18 seedlings that germin ated showed three cotyledons instead of the expected two (Fig. 5-1E). This phe nomenon has also been reported in Opuntia dillenii (Maheshwari and Chopra 1955, Johri et al. 1992). In Opuntia the zygotic embryo is replaced by several embryos of nucellar origin; however, usually only one or two such embryos reach maturity fusing and forming seedlings with three cotyledons Genetic Diversity All indiv iduals analyzed, incl uding nine individuals obtaine d from seeds, exhibited the same heterozygous genotype. Three alleles were present in all individuals for loci Ht.Ie, Ht.IIa, and Ht.ms11, two for locus Ht.Id and a single alle le for locus Ht.ms7 (Table 5-1). Because H. tenuis is probably of hybrid orig in, the amplification of a single allele at locus Ht.ms7 may be due to either a shared allele between the parental species or the inability of the primers to amplify one parental copy. Th e absence of multiple genotypes across the entire population, even in highly variable microsatellite loci, may indi cate that the genotype is of recent origin and mutation has not generated observable molecular va riability yet. Further interpretation of the genetic constitution of this spec ies requires the identification of its genome donors. Since the population is isolated and growing far away from other congeners, possibl e parental species are difficult to determine. However, given its trip loid nature, it likely arose through hybridization between a tetraploid and a di ploid parent. The tetraploid Haageocereus chalaensis is proposed as
87 one of the possible parental species. Both H. tenuis and H. chalaensis thrive on sandy plains at sea level and share several morphological charac ters, including a prostrate habit. However, H. chalaensis is currently found 600 km south of H. tenuis Two possible diploi d parental species are H. pseudomelanostele which is widespread th roughout central Peru, and H. decumbens also with a prostrate habit and gr owing in close proximity to H. chalaensis Haageocereus pseudomelanostele is an extremely variable, medium-s ize shrub, with nocturnal or diurnal flowers. Although its morphology is quite distinct from that of H. tenuis its current distribution is close to that of H. tenuis The stable propagation of a genotype with an uneven ploidal level is often regarded as evidence of apomixis itself (Bon illa and Quarin 1997). Even if mechanisms may exist for the generation of a certain degree of variability in populations with uneven ploidal levels (Grant 1981), in our sample this possibility is ruled ou t by the presence of the same highly heterozygous genotype in all individuals a nd evidence of nucellar embryony. The genetic variability contained in the singl e clonal genotype of H. tenuis is highly restricted. In related H. pseudomelanostele a total of 99 alleles were scored for the same set of loci in three populations (w ith an average of 65 alleles per population), while in H. tenuis only 12 alleles were amplified (Table 5-1). The production of asexual seeds in a geogr aphically restricted species becomes advantageous by increasing the plants capabilities for dispersion beyond that accomplished by stem fragmentation. Although 60% of the seeds germinated in the lab, no seedlings were observed in the field. Recruitment of seedlings ha s been shown to be low in succulent plants living in arid environments (Mandujano et al. 1996, Turner et al. 1996, Negrn-Ortiz 1998) typically due to ecological stress rather than problems in seed pr oduction. In spite of the lack of
88 direct evidence, windows of opportunity may appear for the establishment of new seedlings of H. tenuis in its native location (i.e., years with El Nio events characterized by rains and formation of ephemeral vegetation in otherwise extremely arid areas). Implications for Conservation Although it is represented by onl y a single uniclonal population, H. tenuis is morphologically and genetically distinct from ot her congeners and should perhaps be considered a microspecies. Microspecies, as defined by Gran t and Grant (1971) and Grant (1981), are plant populations that reproduce mainly but not exclusively by uniparental methods, are morphologically homogeneous, occupy a restricted geographical area, an d are differentiated from related species. Microspecies often have a hybrid constitution that leads to sexual sterility. Other prostrate species of Haageocereus are also suspected to be occasional apomicts. This phenomenon may have significant implications for the evolutionary biology and ecology of Haageocereus and other clonal Cactaceae. Even in the absence of sexual recombinati on, clonal species still have a potential to accumulate genetic variation. In a narrowly defined monoclonal microspecies such as H. tenuis the amount of observable variability is dependent on the resolution of th e technique used to assess variability and the period of time since the establishment of the original hybrid (Ellstrand and Rose 1987, Loxdale and Lushai 2003) The evolutionary potential of Haageocereus tenuis depends on its long-term survival in nature ; however the preservation of a single known population is currently threatened by urban development.
89 Table 5-1. Multilocus micr osatellite genotype of H. tenuis and variability found in three populations of a related widesp read sexual species using the same set of primers. *Allele sizes in bp. Locus Allele 1* Allele 2* Allele 3* Number of alleles found in H. pseudomelanostele Ht.Id 231 257 --16 Ht.Ie 199 211 221 13 Ht.IIa 207 209 211 12 Ht.ms7 195 ----16 Ht.ms11 237 243 255 42
90 Figure 5-1. Haageocereus tenuis : A) habitat, B) stem, C) detail of spines, D) chromosomes ( 2n = 3x = 33), E) seedling show ing three cotyledons. A C E B D
91 CHAPTER 6 EXAMINATION OF RETICULATE EVOLUTION BETWEEN Haage ocereus AND Espostoa Introduction Hybridization is defined as crossing between two individual s belonging to populations, or groups of populations, which are dist inguishable on the basis of one or more heritable characters, regardless of whether the resulting hybrids are sterile or fe rtile (Stebbins 1969, Harrison 1990). Arnold (1997) adopts this definition but lim its it to hybridizations that take place in natural settings and successfully produce progeny, so th at there are potentia l ongoing evolutionary effects from hybrids that pass the F1 generation. Natural hybridization is more frequent in plants than in animals, and is found more frequently in perennial plants than in annuals (Grant 1981, Jones and Luchsinger 1987). However, hybridization has been recorded with certain frequency in some animal groups including insects, fishes and birds (Hubbs 1955, Grant and Grant 1992, Arnold 1997). Hybridization has been suggested as an important phenomenon in angiosperm evolution, with the majority of plant species possibly derived from past hybridization events and subsequent polyploidization (Stebbins 1959, Grant 1981, Masterson 1994, Arnold 1997, Vriesendorp and Bakker 2005). Some evolutionary advantages attributed to hybridization are: (1) hybr idization gives rise to a large increase in the si ze of the gene pool, (2) hybrids sometimes respond better to the environment than their parents, and (3) hybrids may show new gene combinations and characters not present in either parent (Ste bbins 1969, Knobloch 1972, Arnold 1997, Rieseberg et al. 2003). In contrast, hybridization may be a local phe nomenon with only ephemeral, rather than evolutionary effects (Rieseberg et al. 2003). The range of patterns considered characteristic of hybrids (e.g. polymorphic nucleotides, additivity of AFLP bands, incongruence between gene trees, intermediate morphology, etc.) may not apply equally to all hybrid individuals
92 (Vriesendorp and Bakker 2005) First-generation hybrids (F1s) in general ha ve characteristics that are intermediate between the parental speci es, and are also intermediate in their habitat requirement (Jones and Luchsinger 1987). These F1s generally backcross to one of the parental species because far more indivi duals having the parental genot ype are present. If successive backcrossing with one of the parents occurs over many generations, the result would be the reversion of the hybrid offspring to that parental type (Jones and Luchsinger 1987, McDade 1990, Rieseberg 1999). Advanced (later, or deri ved) generation hybrids may more closely resemble one of the parents or they may e xhibit novel characters (Vriesendorp and Bakker 2005). The results of hybridization can be either negligible or profound, depending upon the environment in which it takes place and the environmental opportunities that are offered to the resulting progeny (Stebbins 1969, Rieseberg et al 2003). Introgressive hybridization and positive selection are important factors that can facilitate habitat invasion and range expansion (Arnold 1997). However, an F1 hybrid, although heterozygous at many or all loci, will not breed true for an adaptive phenotype, and the hybrid genotype will undergo exte nsive segregation in later generations. Also, segregants from wide cr osses can be less adaptive than the parental genotypes, especially in the original parental environment. Sterility, partial or complete, will become a barrier to the later influence of hybridization (Stebbins 1969, Jones and Luchsinger 1987). Hybrid swarms are largely ignored by ecologi sts and evolutionary biologists because of their complexity (Hewitt 1988); how ever, they constitute areas of great genetic variation and unique gene combinations, where selecti on should be intense (Fisher 1930, Keim et al. 1989, Arnold 1997). Keim et al. (1989) and Arnold (2007) empha size natural hybr idization as a
93 process of significant evolutionary consequen ces, and that has greater impact in certain taxonomic groups than others. A survey of five floras by Ellstrand et al. (1996) suggests that spontaneous hybridization is not a general feature of all plants, but is co ncentrated in certain families (Scrophulariaceae, Gesneriaceae, Salicaceae, Rosaceae, Onagraceae), and certain genera ( Euphrasia, Cyrtandra, Salix, Rosa, Epilobium ), of mainly perenni al outcossing taxa. A systematic survey of the frequency of hybridizat ion in these groups is a starting point for later exploration of the potential adva ntages or disadvantages of hybr idization (if hybrid individuals are less or more fit than the parental genotype s) and the generation of novel genotypes with novel adaptations (Arnold 1997). In our efforts to produce a classification system that will reflect evolutionary relationships in tribe Trichocereeae of the Cactaceae, we are confronted with many difficulties in understanding species limits, in that barriers for reproductive is olation are poorly developed in Cactaceae (Friedrich 1974, Gibson and Nobel 1986, Rowley 1982, 1994, Anderson 2001). The effectiveness of hybridization in the evolution of Cactaceae has been attributed to the constancy in number, size, and structure of chromosomes (Friedrich 1974). Many examples of interspecific and intergeneric hybrids can be found both in na ture and in cultivation, some of them fairly extreme, involving taxa widely separated systematically (e.g. the natural hybrid Pachycereus pringlei x Bergerocactus emoryi the artificial hybrid Heliocereus speciosus x Pilosocereus palmeri and the four-fold hybrid Heliocereus speciosus x Aporocactus flagelliformis x Epiphyllum crenatum x Selenicereus grandiflorus ) (Friedrich 1974). Cacti flowers are zoophilous, and most species are self-incompatible; hence, there are numerous opportunities for hybrids to arise from chance visits from bird s, bats, bees, moths and other insects.
94 Several supposed new species have now been regarded as hybrids. The best-known cases are in the genus Opuntia (Grant and Grant 1971, 1979, Be nson 1982, Baker and Pinkava 1987, Rowley 1994, Rebman and Pinkava 2001, Griffith 2003, 2004). Although there are many interspecific hybrids in Opuntia there are no reports of intergeneric hybrids in subfamily Opuntioideae. In the wild, hybrids have been found to be more frequent in cereiform (columnar) than in cactiform (globular) genera. However, in cultivation the situati on is reversed (Rowley 1994). Rowley (1994) analyzed hybrid combinations in relation to pollination syndromes in the 20 known bigeneric hybrids found in nature. The number of hybrids documented for Cactaceae would increase considerably if artificial hybrids were included. Considering the ease with which cacti hybridize the attention turns to locating where the barriers do exist that cannot be bridged by hybridization (Rowley 1994). All South American intergeneric hybrids reported by Rowley (1982, 1994) for Cactaceae belong to tribe Trichocereeae, with the exception of x Espostingia which involves Browningia in the Browningieae (Fig. 6-1). South American cact i (including hybrids) are considered the least studied taxa; thus, we can expect more hybrids to be discovered as the number of studies increases. Evidence for hybridi zation in this group has to date been mostly based on morphological data. However, given the morphol ogical variability found within populations of putative parental species, molecular genetic t ools may provide more conclusive evidence. Natural hybrids between Espostoa and either Haageocereus or Cleistocactus have been considered distinct genera and have therefor e resulted in much taxonomic and nomenclatural confusion (Anderson 2001). The name Peruvocereus was given by Akers (1947) to a putative hybrid involving Haageocereus pseudomelanostele and Espostoa melanostele. Binghamia is another generic name, applied by Werd ermann (1937), to a hybrid involving Haageocereus
95 albispinus and Espostoa melanostele Binghamia was later transferred to Neobinghamia by Backeberg (1950). Ritter (1981) reinterpre ted these hybrids, a pplying the names x Haagespostoa (x H. albisetata and xH. climaxantha ) to hybrids involving Espostoa and Haageocereus and xEspostocactus to hybrids involving Espostoa and Cleistocactus (Rowley 1994) (Fig. 6-1). Caldern et al. (2007) recognized the preval ence of hybridization in Haageocereus and suggested additional species that could possibly have a hybrid constitution (Fig. 6-2). To study the role that hybr idization has played in Haageocereus and related genera, three putative intergen eric hybrids ( Haageocereus pseudomelanostele x Espostoa melanostele = xHaagespostoa albisetata, H. pseudomelanostele subsp. carminiflorus x Espostoa lanata and H. pseudomelanostele subsp. acanthocladus x Espostoa melanostele ) (Fig. 6-3) of the seven suggested intergeneric hybrid s (the above-mentioned plus H. fulvus x E. melanostele H. versicolor x E. melanostele H. chryseus x E. nana, and H. albispinus x E. melanostele = x Haagespostoa climaxantha ) were examined (Fig. 6-3). Espostoa lanata individuals are tree-like shrubs, up to 4 m high, with stems 6-10 cm in diameter, 18-25 ribs, lateral cephalia (accumulation of woolly hairs in th e flowering zone) with brown to gray wool, and light purple, funnelform flowers. Espostoa melanostele individuals are shorter, up to 2 m high, with stems 8-10 cm in diameter, 18-25 ribs, lateral cephalia with whitish to brownish wool, and white, bell -shaped nocturnal flowers Haageocereus pseudomelanostele and H. chrysacanthus individuals are m edium-size shrubs, branching profusely at the base, with stems up to 70 cm high and 6-8 cm in diameter, without cephalia. Vegetative charac teristics, such as number, size and color of spines, are extremely variable, and populations have sometim es been described as different species or varieties based on these characters. Spine color ranges from whitish gray to brown or yellowish,
96 with or without accompanying bristles. Flowers ar e tubular to funnelform, white, greenish white, or white and light purple inside, up to 10 cm long, nocturnal, although they may sometimes remain open during the early hours of the morning. Putative hybrid individuals (x Haagespostoa) are tall shrubs, to 1.5 m, branching at the base, with dark, grayish green st ems, and a variable number of ribs, up to 27. Spines are fine and variable in length and color, ra nging from gray to yellowish or brown; lateral cephalia are irregularly formed, with accumulations of hairs in distinct parts of the flowering zone in upper portions of the stems. The flowers ar e funnelform and nocturnal, similar to Espostoa flowers, and range in color from white to pink or light green. Morphological characters and molecular analysis were used to evaluate the hypothesis of hybrid origin of these thr ee putative hybrids involving Haageocereus, thus providing more insight into the reticulate evolution that seems to be important in the genus. Materials and Methods Plant Sampling Suspected intergeneric hybrids, determ ined by intermediacy in morphological characters, were sampled along with individu als suspected to be the putativ e parents. Unrecognized hybrids within Haageocereus may be revealed by this study. Stem tissue from Espostoa, the hybrid individuals, and 3032 individuals of Haageocereus were sampled from each of three populations found in Quebrada Tinajas, Lima (MA1551,1554), and Fortaleza River Valley, Ancash (MA1642), Peru. The collected tissue was dr ied in silica gel and transported to the laboratory for DNA isolation, sequencing and genotyping. Nuclear Microsatellite Data Five to 11 microsatellite loci designed for Haageocereus (Chapter 3) were used for this portion of th e study. PCR conditions were adjusted so that most microsatellites would require a
97 common amplification profile. PCR amp lifications were performed with 10 L total volumes containing 0.2 unit of NEB Taq polymerase (New England Biolabs Inc., Ipswich, MA, USA), 1.5 M MgCl2, 0.15 M of the reverse primer and labeled M13 primer, 0.01 M of the extended forward primer, and 0.1 mM of each dNTP. PCR reactions started with a 5-min denaturation time at 94 C, followed by 40 cycles of 15 sec at 94 C and 3 min at 53 C, and a final extension of 5 min at 72 C. PCR products were run on a CEQ 8000 (B eckman-Coulter, Fullerton, CA, USA) capillary sequencer. The multilocus genotype data were used to char acterize the level of introgression in each individual and to assign indi viduals to one of six categ ories (parent 1, parent 2, F1, F2, backcross with parent 1, backcross with parent 2) using a Bayesian m odel-based method implemented in the program NewHybrids, version 1.1 Beta 3 ( http://ib.berkeley.edu/labs/ slatkin/eriq/ software/ software.htm#NewHybs) by Anderson and Thom pson (2002). The m odel uses a Markov chain Monte Carlo (MCMC) sampling procedure to de termine deviations from Hardy-Weinberg equilibrium among multilocus genotypes to assess th e posterior probabilities that an individual belongs to one of two pure lineages or is a hybrid mixture (Wares et al 2004). The method does not require that allele frequencie s are known for the parental species, nor that parental classes be sampled separately or that the species posse ss unique alleles (Anderson and Thompson 2002). The data sets were analyzed several times, w ith increasing burn-in and total iterations until similar results were obtained, which was at both 10000 burn-in iterations and 50000 total iterations. Because putative parents were not kn own with certainty, no prior information on this was included. Chloroplast DNA Data Inheritance of chloroplast DNA has been shown to be m aternal in six Cactaceae examined by Corriveau and Coleman (1988). However, th e same study reports two epiphytic cacti,
98 members of tribe Rhipsalideae, to have biparental plastid inheri tance. Assuming that members of Trichocereeae exhibit maternal plas tid inheritance, to determine which species is the maternal parent in the hybrid combinations, we amplif ied and sequenced three relatively conserved chloroplast DNA regions: psbA-trnH (Sang et al 1997), psbE-petL (Popp et al 2005), and 23S (B. Crozier, pers. comm.). Amplification and sequ encing procedures are explained in detail in Chapter 2. Since Espostoa individuals were rare in populations dominated by Haageocereus, we expected the less common species ( Espostoa ) to be the maternal parent, and therefore, the resulting hybrids would have inherited its cytoplasm (Soltis and Soltis 1989). Results and Discussion The m icrosatellite products amplified were as expected: one peak for homozygous individuals and two peaks for he terozygous individuals. If an une xpected number of peaks was amplified, which occurred in four individuals, these samples were cloned and compared to the original captured sequences. The extra peaks were found to be unspecific amplifications. Allele sizes varied from 179 bp to 395 bp. In general, longer and uninterrupted microsatellites were more variable, but also presented more di fficulties during amplif ication and scoring. Microsatellite data indi cate that hybrids between Espostoa and Haageocereus are present and that hybridization has gone beyond the F1 generation, with sampled plants being either F2 or backcrosses. Several alleles from each putat ive parent could be assigned to each hybrid; however, in none of the three cases was there comple te additivity of the pa rental alleles (Fig. 64A to 6-6A). Genotype distributions obtained with NewHybrids are shown in Fig. 6-4B to 6-6B. The sampling procedure tried to include individua ls that represent the morphological variability observed in the populations; however, several back cross plants were erroneously assigned to pure Haageocereus due to our inability to distinguish between pure Haageocereus and hybrids using morphological features. All three hybrid populations ex amined showed that several
99 individuals assigned to Haageocereus species have higher probabilit ies of being later-generation hybrids or backcrosses to Haageocereus (Fig. 6-4B to 6-6B). Thes e populations meet the criteria of hybrid swarms (Grant 1981). Hybrids that sh ow completely intermediate morphologies are rare, and most have the mor phological attributes of one of the parents, in this case Haageocereus, the more abundant parent at each s ite. The evidence obtained here for the occurrence of introgression indicat es that hybrid individuals ar e fertile. However, we do not know the proportion of F1 hybrid individuals that succe ssfully reproduce sexually after hybridization. Because individuals of Espostoa are extremely rare at a location dominated by Haageocereus, we hypothesized Haageocereus to be the pollen contribut or (paternal parent) and Espostoa the chloroplast donor (maternal pare nt). The three plastid regions ( psbA-trnH, psbEpetL, and 23S) sequenced for each of the hybrid comb inations support this hypothesis (Fig. 6-7), indicating that introgressi on is mostly due to pollen being transferred from Haageocereus (dominant at each site) to Espostoa. We could infer that if pollen continues moving towards Espostoa and early-generation hybrids, hybrid offspring will revert to the Haageocereus nuclear genotype, while retaining the plastid genome of Espostoa. Contrary to reports for se veral other plant groups ( Populus Keim et al 1989; Turnera ulmifolia complex, Shore et al 1994; Polystichum Mullenniex et al 1999; Helianthus, Rieseberg et al 2003), most cacti that form intergeneric hybrids are sympatric, have similar habitat requirements and share pol linators. It appears that geogra phic barriers are more important than intrinsic reproductive barriers to hybridiza tion. Even genera with very different pollination syndromes (birds-diurnal vs. bats-nocturnal) can hybr idize. Bees that are ge neralists seem to be the most effective pollinators in Cactaceae unde r drought conditions regardless of the floral
100 morphology (Rowley 1994). Several features of cacti (i.e. their pres ence and diversification in relatively young habitats, low genetic differe ntiation) suggest a lack of morphological specialization, with flowers that are not complete ly adapted to one pollinator or another (Rowley 1997). There are several examples of floral dimorphism in the family (Rowley 1994, Sahley 1995), and these dimorphic flow ers attract distinct pollinat ors depending on the relative abundance of these pollinators, times of the da y, and environmental conditions. Sahley (1995) reported an interesting case of floral dimorphism in Weberbauerocereus weberbaueri, where funnelform, white flowers and zygomorphic brownish -purple flowers are maintained in the same population. White flowers open at night and are visited by bats that are present in the area during normal years. Bat densities decrease during year s with an El Nio event (characterized by drought conditions in southwestern Peru), and ba ts are replaced by hummingbirds that visit the brownish-purple flowers that open during the night but remain ope n during the early hours of the morning. In other cases, tubular red or orange flowers are not onl y visited by birds that feed on nectar but also by bees that are presumably attracted by the pollen (Rowley 1994). It would be worth examining more populati ons and hybrid combinations proposed by Rowley (1994) and Caldern et al. (2207) between different species of Espostoa and Haageocereus, and between species of Espostoa and Cleistocactus, Matucana, Weberbauerocereus and Browningia to confirm that they are hybr ids and determine if what has been observed in hybrid combinations of Espostoa and Haageocereus is typical of other taxa in the Trichocereeae. Some genera of Cactaceae, such as the entire genus Weberbauerocereus, where only tetraploid and octoploid counts are re corded (Chapter 4), may have an allopolyploid origin. Also, the triploid apomict H. tenuis most probably has an allopo lyploid origin (Chapter5).
101 Polyploidy, after a hybridization event, may be an important means of originating new species in Cactaceae, as happens in many other angiosperm groups. Pollination Naturally occurring bigeneric hybrids syndrome Bees Weingartia Diurnal x Weingartiopsis Moths Echinopsis Nocturnal x Oreonopsis Oreocereus x Cleistoreocereus x Oreocana Birds Diurnal Cleistocactus x Cleistocana Matucana x Cleipaticereus x Maturoya Samaipaticereus Oroya x Espostocactus x Espocana Espostoa Bats x Weberbostoa x Haagespostoa x Espostingia Nocturnal Weberbauerocereus Haageocereus Browningia Figure 6-1. Naturally occurring bi generic hybrids in Cactaceae centered in Peru and Bolivia (from Rowley 1994).
102 Haageocereus pseudomelanostele subsp. carminiflorus H. albisetatus H. albispinus H. albispinus var. floribundus H. climaxanthus H. comosus H. divaricatispinus H. seticeps H. seticeps var. robustispinus Peruvocereus albicephalus P. albicephalus var armatus P. albisetatus var robustus Espostoa melanostele H. dichromus var pallidior Haageocereus pseudomelanostele subsp aureispinus H. pseudomelanostele subsp chryseus H. zehnderi Espostoa nana Figure 6-2. Names attributed to Haageocereus and Peruvocereus that are possibly referring to xHaagespostoa, according to Caldern et al. (2007).
103 A B C Figure 6-3 Espostoa x Haageocereus hybrid combinations examined in the present study: A) Espostoa lanata 1550a, xHaagespostoa climaxantha 1550, Haageocereus pseudomelanostele 1551. B) Espostoa melanostele 1641, xHaagespostoa sp. 1642, Haageocereus chrysacanthus 1643. C) Espostoa melanostele 1554a, xHaagespostoa albisetata 1554, Haageocereus pseudomelanostele 1553.
104 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 123456789101112131415161718192021222324252627282930313233IndividualsPosterior probabilities Bx_B Bx_A F2 F1 Pure B Pure A Figure 6-4. Multilocus micros atellite genotypes for putative pa rents and hybrid individuals: A) xHaagespostoa climaxantha and putative parents: Espostoa lanata and Haageocereus pseudomelanostele B) Posterior probabi lities of individuals belonging to each of six categories: pure pa rent A, pure parent B, F1, F2, backcross with parent A (Bx_A), backcross with parent B (Bx_B). H H a a a a g g e e o o c c e e r r e e u u s s p p s s e e u u d d o o m m e e l l a a n n o o s s t t e e l l e e 1551 E E s s p p o o s s t t o o a a l l a a n n a a t t a a 1 1 5 5 5 5 0 0 a a x x H H a a a a g g e e s s p p o o s s t t o o a a c c l l i i m m a a x x a a n n t t h h a a 1 1 5 5 5 5 0 0 170 190 210 230 250 270 290 310 330 350 370 Id Ie Iia Ms7 Ms11 Individuals Allele Size (bp) xH. climaxantha E. lanata H. pseudomelanostele A B
105 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 1234567891011121314151617181920212223242526272829IndividualsPosterior probabilities Bx_B Bx_A F2 F1 Pure B Pure A Figure 6-5. Multilocus microsatellite genotypes for putative pare nts and hybrid individuals: A) xHaagespostoa sp., and putative parents: Espostoa melanostele and Haageocereus pseudomelanostele B) Posterior probabilities of i ndividuals belonging to each of six categories: pure parent A, pure parent B, F1, F2, backcross with parent A (Bx_A), backcross with parent B (Bx_B). Individuals E E s s p p o o s s t t o o a a m m e e l l a a n n o o s s t t e e l l e e 1 1 5 5 5 5 4 4 a a H H a a a a g g e e o o c c e e r r e e u u s s p p s s e e u u d d o o m m e e l l a a n n o o s s t t e e l l e e 1 1 5 5 5 5 3 3 x x H H a a a a g g e e s s p p o o s s t t o o a a a a l l b b i i s s e e t t a a t t a a 1 1 5 5 5 5 4 4 150 200 250 300 350 400 Id Ie Iia Ms7 Ms11 E E s s p p o o s s t t o o a a x x H H a a a a g g e e s s p p o o s s t t o o a a H H a a a a g g e e o o c c e e r r e e u u s s p p s s e e u u d d o o m m e e l l a a n n o o s s t t e e l l e e Allele Size (bp) A B
106 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 123IndividualsPosterior probabilities Bx_B Bx_A F2 F1 Pure B Pure A Figure 6-6. Multilocus microsatellite genotypes for put ative parents and hybrid individuals: A) xHaagespostoa sp., and putative parents: Espostoa melanostele and Haageocereus pseudomelanostele B) Posterior probabil ities of individuals belonging to each of six categories: pure parent A, pure parent B, F1, F2, backcross with parent A (Bx_A), backcross with parent B (Bx_B). E E s s p p o o s s t t o o a a m m e e l l a a n n o o s s t t e e l l e e 1 1 6 6 4 4 1 1 H H a a a a g g e e o o c c e e r r e e u u s s c c h h r r y y s s a a c c a a n n t t h h u u s s 1 1 6 6 4 4 3 3 x x H H a a a a g g e e s s p p o o s s t t o o a a s s p p 1 1 6 6 4 4 2 2 Individuals 15 17 19 21 23 25 27 29 Id Ie Iia Ms7 Ms11 (-50bp) Iv Ms2 Ms6 Ms8 Ms13 Ms16 Allele size (bp) E E s s p p o o s s t t o o a a x x H H a a a a g g e e s s p p o o s s t t o o a a H H a a a a g g e e o o c c e e r r e e u u s s h h A B
107 E.lan 1550a TTAAG TTA G T C TTTTTTTTT A TTA T G T C T AAAAC G T A T AAATTTTTA TTTA G A T A T G AAAT x H.cli 1550 TTAAG TTA G T C TTTTTTTTT A TTA T G T C T AAAAC G T A T AAATTTTTA TTTA G A T A T G AAAT H.pse 1551 TTAAG TTA G--------A T A TTA -GT A T AAAAC G T A T AAATTTA T A T A T A G A T A T G AAAT Figure 6-7. Amplifications of chloroplast regions from thr ee hybrid combinations involving Espostoa and Haageocereus. A) Amplification of psbA-trnH portion of psbA-trnH sequence is shown inside box. B) Intergenic spacer psbE-petL. C) Gene and spacer 23S. Individuals are in the following order: 1. Espostoa lanata 1550, 2. xHaagespostoa climaxantha 1550, 3. Haageocereus pseudomelanostele 1551, 4. Espostoa melanostele 1641, 5. x Haagespostoa sp. 1642, 6. Haageocereus chrysacanthus 1643, 7. Espotoa melanostele 1554, 8. x Haagespostoa albisetata 1554, 9. Haageocereus pseudomelanostele 1553. 1 2 3 4 5 6 7 8 9 psbA-trnH psbE-petL 23S 1 2 3 4 5 6 7 8 9 1 2 3 4 5 6 7 8 9 A B C
108 CHAPTER 7 CONCLUDING REMARKS The Cactaceae com prises approximately 1800 speci es, found mostly in th e tropical deserts of North and South America. The family has been the object of intensive study but it is not until recently that molecular tools have been used to elucidate its p hylogenetic relationships. Most of the Cactaceae found in the Centra l Andes of Peru and Bolivia belong to tribe Trichocereeae. Phylogenetic relationships within the tribe have been historically contro versial, therefore, a molecular study using chloroplas t and nuclear markers was desi gned to elucidate phylogenetic relationships in tribe Trichocereeae and genus Haageocereus. We sequenced cpDNA ( rpoB rpl16 and 23S) and nrDNA (ITS1 and ITS2) of 107 ingroup taxa comprising the Trichocereeae, and 11 outgroup taxa from different tribes. The analyses support a monophyletic Trichocereeae (including Praecereus euchlorus). Seven genera were found to be outside the Trichocereeae and more closely related to Brazilian members of tribe Cereeae. The study also supports the maintenance of several genera histor ically placed in Trichocereeae, including Haageocereus and Espostoa. Based on our results, the taxonomy of th e Trichocereeae does not reflect its phylogeny and is in need of revision. Additional studies at the popul ation level were carried ou t to examine the role of popyploidy and hybridization in the evolution of Haageocereus, Espostoa and other Trichocereeae. Chromosome counts for a total of 54 individuals representing 14 genera and 39 species of Cactaceae, mostly in tribe Trichocereeae, are repo rted. Five additional taxa examined belong to subfamily Opuntioideae and othe r tribes of Cactoideae (Bro wningieae, Pachycereeae, Notocacteae, and Cereeae). Among Trichocereeae, counts for 34 taxa in nine genera are reported, with half of these (17 species) for the genus Haageocereus. These are the first
109 chromosome numbers reported for 38 of the 39 total species examined, as well as the first counts for the genus Haageocereus. Both diploid and polyploid coun ts were obtained. Twenty eight species were diploid with 2 n = 2x = 22. Polyploid counts were obtained from the genera Espostoa, Cleistocactus, Haageocereus, and Weberbauerocereus; we detected one triploid (2 n = 3x = 33), nine tetraploids (2 n = 4x = 44), one hexaploid (2 n = 6 x = 66), and three octoploids (2 n = 8 x = 88). In two cases, different counts were reco rded for different individuals of the same species (Espostoa lanata, with 2n = 22, 44, 66; and Weberbauerocereus rauhii, with 2n = 44, 88). These are the first repor ted polyploid counts for Haageocereus, Cleistocactus, and Espostoa Our counts support the hypothesis th at polyploidy and hybridization have played prominent roles in the evolution of Haageocereus Weberbauerocereus, and other Trichocereeae. Nineteen polymorphic dinucleotide microsatel lite loci isolated from two species of Haageocereus ( H. tenuis and H. pseudomelanostele ) are described. Microsatellites were isolated from a genomic library enriched for CA repeat motifs; 255 individuals an d five microsatellite loci were employed in the preliminary study of genetic diversity and popu lation differentiation in four species of Haageocereus ( H. tenuis, H. repens, H. acranthus, and H. pseudomelanostele ), yielding between one and 44 alleles per locus. Three additional loci were characterized for H. pseudomelanostele These markers will facilitate analys is of population differentiation and genetic diversity throughout Haageocereus. Haageocereus tenuis a triploid cactus restricted to a sm all area of two square kilometerse was examined in detail. Propagation via agam ospermy was documented and adventitious embryony was also inferred as a mechanism. Although seedling recruitment has not been observed in nature, we showed th at asexually produced seeds are viable. Microsatellite analysis confirms that individuals from the only existing population are genetically identical and that the
110 population likely represents a sing le clone. The absence of mutations in any individual, even in highly variable microsatellite loci may indicate that the species is also of recent origin. Other prostrate species of Haageocereus are suspected to be occasional apomicts. This phenomenon has significant implications for th e evolutionary biology and ecology of Haageocereus and other clonal Cactaceae.
111 REFERENCES Akaike, H. 1974. A new look at the statistical m odel identification. IEEE Transactions on Automatic Control 19 (6): 716. Akers, J. F. 1947. Peruvocereus Akers. Cactus and Succulent Journal (LA) 19: 67 Anderson, E. C. and E. A. Thompson. 2002. A model-based method for identifying species hybrids using multilocus genetic data. Genetics 160: 1217-1229. Anderson, E. N. 2001. The Cactus Family. Timber Press, Portland, Oregon. Applequist, W. L. and R. S. Wallace. 2001. Phylogeny of the Portulacaceous cohort based on ndhF sequence data. Systematic Botany 26(2): 406-419. Applequist, W. L. and R. S. Wallace. 2002. Deletions in the plastid trn Ttrn L intergenic spacer define clades within Cactaceae subfamily Cactoideae. Plant Systematics and Evolution 231: 153-162. Arakaki, M. 2002. Systematics of the genus Weberbauerocereus Backeb. (Cactaceae). Thesis to obtain the degree of M. A. Univ ersity of Texas at Austin. Arakaki, M., J. Panero and B. Crozier. 2003. Relaciones taxonmicas en el gnero peruano Weberbauerocereus Backeb. Quepo 17: 62-72. Arias, S., T. Terrazas and K. Cameron. 2003. Phylogenetic analysis of Pachycereus (Cactaceae, Pachycereeae) based on chloropl ast and nuclear DNA sequences. Systematic Botany 28(3): 547-557. Arnold, M. L. 1997. Natural Hybridization and Evolution. Oxford Series in Ecology and Evolution, Oxford University Press, New York. Asker, S. E. and L. Jerling. 1992. Apomixis in Plants. CRC Press, Inc. Boca Raton, Florida. Axelrod, D. I. 1979. Desert vegetation, its age origin. In : J. R. Goodin and D. K. Northington (ed.), Arid Land Plant Resources. Texas Tech University. Lubbock. Bailey, C. D. and J. J. Doyle. 1999. Potential ph ylogenetic utility of the low-copy nuclear gene pistillata in dicotyledonous plants: comparison to nrDNA ITS and trnL intron in Sphaerocardamum and other Brassicaceae. Molecular Phylogenetics and Evolution 13(1): 20-30. Baker, M. A. and D. J. Pinkava. 1987. A cytologi cal and morphometric an alysis of a triploid apomict, Opuntia x kelvinensis (Subgenus Cylindropuntia Cactaceae). Brittonia 39: 387401. Backeberg, C. 1950. Neobinghamia Backeb. Cactus and Succulent Journal (LA) 22: 154. Backeberg, C. 1976. Cactus Lexico n. Translation of German 3rd ed. Blandford Press, Poole.
112 Barthlott, W. and D. R. Hunt. 1993. Cactaceae. In : K. Kubitzki, J. G. Rohwer, and V. Bittrich (eds.), The Families and Genera of Vascular Plants II: Flowering Plants, Dicotyledons, Magnoliid, Hamamelid and Caryophyllid Fa milies. Springer-Verlag. Berlin. Barthlott, W. and D. R. Hunt. 2000. Seed-diver sity in the Cactaceae, subfamily Cactoideae. Succulent Plant Research 5: 1-173. Bell, C. D. and M. J. Donoghue. 2005. Phylogeny and biogeography of Valerianaceae (Dipsacales) with special reference to the South American valerians. Organisms Diversity and Evolution 5: 147-159. Benson, L. 1982. The Cacti of the United States and Canada. Stanford University Press, Stanford, California. Bonilla, J. R. and C. L. Quarin. 1997. Diplosporous and aposporous apomixes in a pentaploid race of Paspalum minus Plant Science 127: 97-104. Bull, J. J., J. P. Huelsenbeck, C. W. Cunningham, D. L. Swofford and P. J. Waddell. 1993. Partitioning and combining data in phylogenetic analysis. Systematic Biology 42: 384397. Butterworth, C. A., J. H. Cota-Sanchez and R. S. Wallace. 2002. Molecular systematics of tribe Cacteae (Cactaceae: Cactoide ae): a phylogeny based on rpl16 intron sequence variation. Systematic Botany 27(2): 257-270. Butterworth, C. A. and R. S. Wallace. 2005. Molecu lar phylogenetics of the leafy cactus genus Pereskia (Cactaceae). Systematic Botany 30(4): 800-808. Buxbaum, F. 1950a. Morphology of cacti. Section I. Roots and stems. Abbey Garden Press, Pasadena. Buxbaum, F. 1950b. Morphology of cacti. Section II. The flower. Abbey Garden Press, Pasadena. Buxbaum, F. 1953. Morphology of cacti. Section III. Fruits and seeds. Abbey Garden Press. Pasadena. Buxbaum, F. 1958. The phylogenetic division of the subfamily Cereoideae, Cactaceae. Madroo 14: 177-206. Caldern, N., D. Zappi, N. Taylor and A. Ceroni. 2007. Taxonomy and conservation of Haageocereus Backeb. (Cactaceae) in Peru. Bradleya 25: 45-124. Charles, G. 2000. Oreocereus tacnaensis Ritter a natural hybrid?. Cactus and Succulent Journal (U.S.) 72(5): 243-246.
113 Clark-Tapia, R., C. AlfonsoCorrado, L. E. Eguiarte and F. Molina-Freaner. 2005. Clonal diversity and distribution in Stenocereus eruca (Cactaceae), a narrow endemic cactus of the Sonoran Desert. American Journal of Botany 92(2): 272-278. Clark-Tapia, R. and F. Molina-Freaner. 2004. Re productive ecology of th e rare clonal cactus Stenocereus eruca in the Sonoran Desert. Plant Systematics and Evolution 247 (3-4): 155-164. Corriveau, J. L. and A. W. Coleman. 1988. Rapid sc reening method to detect potential biparental inheritance of plastid DNA and resu lts for over 200 Angiosperm species. American Journal of Botany 75(10): 1443-1458. Cota, J. H. and C. T. Philbrick. 1994. Chromosome number variation and polyploidy in the genus Echinocereus (Cactaceae). American Journal of Botany 81: 1054-1062. Cota. J. H. and R. S. Wallace. 1995. Karyotypic studies in the genus Echinocereus (Cactaceae) and their taxonomic significance. Caryologia 48: 105-122. Crochet, P. A., J. Z. Chen, J.-M. Pons, J. D. Lebreton, P. D. Hebert and F. Bonhomme. 2007. Genetic differentiation at nucle ar and mitochondrial loci am ong large white-headed gulls: sex-biased interespecific gene flow? Evolution 57(12): 2865-2878. Crozier, B. S. 2002. Systematics of Cact aceae Juss.: Phylogeny, cpDNA evolution, and Classification, with em phasis on the genus Mammillaria Haw. Ph. D. Dissertation. University of Texas at Austin. Crozier, B. S. 2004. Subfamilies of Cactaceae Ju ss., including Blossfeldioideae subfam. nov. Phytologia 86(2): 52-64. Darlu, P. and G. Lecointre. 2002. When does th e Incongruence Length Difference test fail? Molecular Biology and Evolution 19: 432-437. Dieringer, D. and C. Schltterer. 2002. Microsat ellite Analyser (MSA): a platform independent analysis tool for large microsatellite data sets. Molecular Ecology Notes 3(1): 167-169. Downie S. R. and D. S. Katz-Downie. 1996. A molecular phylogeny of Apiaceae subfamily Apioideae: evidence from nuclear ribosomal DNA internal transcribed spacer sequences. American Journal of Botany 83(2): 234-251. Doyle, J. J. 1992. Gene trees and species trees: molecular systematics as one-character taxonomy. Systematic Botany 17: 144-163. Doyle, J. J. and J. L. Doyle. 1987. A rapid DNA is olation procedure for small quantities of fresh leaf tissue. Phytochemical Bulletin 19: 11-15. Edgar, R. C. 2004. MUSCLE : multiple sequence alignment with high accuracy and high throughput. Nucleic Acid Research 32(5): 1792-1797.
114 Edwards, E. J., R. Nyffeler and M. J. D onoghue. 2005. Basal cactus phylogeny: implications of Pereskia (Cactaceaea) paraphyly for the transi tion to the cactus life form. American Journal of Botany 92(7): 1177-1188. Ellstrand, N. C. and M. L. Rose. 1987. Patterns of genotypic diversity in clonal plant species. American Journal of Botany 74(1): 123-131. Ellstrand N. C., R. Whitkus and L. H. Rieseberg. 1996. Distribution of spontaneous plant hybrids. Proceedings of the Nati onal Academy of Sciences 93: 5090-5093. Ernst, J. A., L. C. Branch, A. M. Clark a nd D. G. Hokit. 2004. Polymorphic microsatellite markers for the Florida scrub lizard ( Sceloporus woody ). Molecular Ecology Notes 4: 364-365. Farris, J. S, M. Kllersj, A. G. Kluge and C. Bult. 1994. Testing significance of congruence Cladistics 10: 315-319. Farris, J. S, M. Kllersj, A. G. Kluge and C. Bult. 1995. Constructing a significance test for incongruence. Systematic Biology 44: 570-572. Felsenstein, J. 1985. Confidence limits on phylogenies: An approach using the bootstrap. Evolution 39: 783-791. Fisher, R. A. 1930. The Genetical Theory of Natural Selection. Clarendon Press. Oxford. Fleming, T. H., S. Maurice and J. L. Hamrick. 1998. Geographic variation in the breeding system and the evolutionary stability of trioecy in Pachycereus pringlei (Cactaceae). Evolutionary Ecology 12: 279-289. Friedrich, H. 1974. Hybridization as a fact or in the evoluti on of the Cactaceae. Cactus and Succulent Journal (U.S.) 46(5):213-214. Frohlich, M. W. and E. M. Me yerowitz. 1997. The search for homeotic gene homologs in basal angiosperms and Gnetales: a potential new sour ce of data on the evol utionary origin of the flowers. International Journal of Plant Sciences 158: S131-S142. Garcia, S., T. S. Garnatje, S. Dariimaa, S. Ts ooj and J. Valls. 2006. New or rarely reported chromosome numbers in taxa of subtribe Artemisiinae (Anthemideae, Asteraceae) from Mongolia. Botanical Journal of the Linnaean Society 150: 203-210. Gibson, A. C. and P. S. Nobel. 1986. The Cactus Primer. Harvard University Press, Cambridge. Gitzendanner, M. A. and P. S. Soltis. 2001. Ge netic variation in rare and widespread Lomatium species (Apiaceae): a co mparis on of AFLP and SSCP data. Edinburgh Journal of Botany 58(2): 347.
115 Goldblatt, P. and D. E. Johnson. 1978-2006. Index to Plant Chromosome Numbers (IPCN). Missouri Botanical Garden w3TROPICOS. URL: http://mobot.mobot.org/W3T/Search/ ipcn.htm l. Golemberg, E. M., M. T. Clegg, M. L. Durbin, J. Doebley and D. P. Ma. 1993. Evolution of a noncoding region of the chloroplast genome. Molecular Phylogenetics and Evolution 2: 52-64. Grant, V. 1971. Natural hybridization between the cholla cactus species Opuntia spinosior and Opuntia versicolor. Proceedings of the Nati onal Academy of Sciences 68: 1993-1995. Grant, V. 1980. Clonal microspecies of hybrid origin in the Opuntia lindheimeri group. Botanical Gazette 141:101-106. Grant, V. 1981. Plant Speciation. 2nd ed. Columbia University Press, New York. Grant, P. R. and B. R. Grant. 1992. Hybridization of bird species. Science 256:193-197. Grant, V. and K. A. Grant. 1971. Dynamics of clonal microspecies in cholla cactus. Evolution 25: 144-155. Grant, V. and K. A. Grant. 1979. Hybridization and variation in the Opuntia phaeacantha group in Central Texas. Botanical Gazette 140(2): 208-215. Griffith, M. P. 2003. Using molecular data to elucidate retic ulate evolution in Opuntia Madroo 50: 162-169. Griffith, M. P. 2004a. What did the first cact us look like? An attempt to reconcile the morphological and molecular evidence. Taxon 53(2): 493-499. Griffith, M. P. 2004b. The origins of an important cactus crop, Opuntia ficus-indica (Cactaceae): new molecular evidence. American Journal of Botany 91(11): 1915-1921. von Hagen, K. B. and J. W. Kadereit. 2001. The phylogeny of Gentianella (Gentianaceae) and its colonization of the southern hemisphere as revealed by nuclear and chloroplast DNA sequence variation. Organisms Diversity and Evolution 1: 61-79. von Hagen, K. B. and J. W. Kadereit. 2003. The diversification of Halenia (Gentianaceae): ecological opportunity versus key innovation. Evolution 57(11): 2507-2518. van der Hammen, T. 1974. The Pleistocene changes of vegetation and climate in tropical South America. Journal of Biogeography 1: 3. van der Hammen, T. and A. M. Cleef. 1986. Devel opment of the High Andean pramo flora and vegetation. In: F. Vuilleumier and M. Monasterio (eds.), High Tropical Altitude Biogeography. Oxford University Press, Nueva York.
116 Hamrick, J. L., J. D. Nason, T. H. Fleming, J. M. Nassar. 2002. Genetic diversity in columnar cacti. In: T. H. Fleming and A. Valiente-Banue t (eds.), Columnar Cacti and their Mutualists: Evolution, Ecology and Conservati on. University of Arizona Press, Tucson. Hardesty, B. D., S. L. Hughes, V. M. Rodrguez and J. A. Hawkins. 2008. Characterization of microsatellite loci fo r the endangered cactus Echinocactus grusonii and their crossspecies utilization. Molecular Ecology Notes 8(1): 164-167. Harrison, R. G. 1990. Hybrid zones: windows on evolutionary process. Oxford Surveys in Evolutionary Ecology 7: 69-128. Helsen, P., P. Verdyck, A Tye, K. Desender, N. Van Houtte and S. Van Dongen. 2006. Isolation and characterization of polymorphic microsat ellite markers in Galapagos prickly pear ( Opuntia ) cactus species. Molecular Ecology Notes 7(3): 454-456. Hershkovitz, M. A. 1991. Phylogenetic assessm ent and revised circumscription of Cistanthe (Portulacaceae). Annals of the Missouri Botanical Garden 78: 1009-1021. Hershkovitz, M. A. and E. A. Zimmer. 1997. On the evolutionary origins of the cacti. Taxon 46: 217-232. Hewitt, G. M. 1988. Hybrid zones natural laboratories for evolutionary studies. Trends in Ecology and Evolution 3: 158. Hillis, D. M. 1987. Molecular versus morphological approaches to systematics. Annual Review in Ecology and Systematics 18: 23-42. Hubbs, C. L. 1955. Hybridization be tween fish species in nature. Systematic Zoology 4: 1-20. Huelsenbeck, J. P. and J. J. Bull. 1996. A likel ihood ratio test for detection of conflicting phylogenetic signal. Systematic Biology 45: 92-98. Huelsenbeck, J. P., J. J. Bull and C. W. C unningham. 1996. Combining data in phylogenetic analysis. Trends in Ecology and Evolution 11: 152-158. Huelsenbeck, J. P. and F. Ronquist. 2001. MR BAYES: Bayesian inference of phylogeny. Bioinformatics 17: 754-755. Hughes, C. and R. Eastwood. 2006. Island radiation on a continental scale: exceptional rates of plant diversification after uplift of the Andes. Proceedings of the National Academy of Sciences 102(27): 10334-10339. Hunt, D. R. 1993. CITES Cactaceae checklist. Royal Botanic Gardens Kew & International Organization for Succulent Plant Stu dy (IOS). Remous Ltd., Milborne Port Hunt, D. R. and N. P. Taylor (comp.). 1990. The genera of Cactaceae: progress towards consensus. Bradleya 8: 85-107.
117 Hunt, D. R., N. Taylor, G. Charles and IC SG (International Cactaceae Systematics Group) members. 2006. The New Cactus Lexicon. DH Books, England. Ito, M., A. Soejima and M. Wada. 2006. Low genetic differentiation between Adiantum capillusveneris L. and A. ogawarense Tagawa (Adiantaceae), en endemic pteridophyte of the Bonin (Ogasawara) island. Journal of Plant Research 105(1): 161-165. Jame, P. and P. J. L. Lagoda. 1996. Microsatellite s, from molecules to populations and back. Trends in Ecology and Evolution 11: 424-430. Johnson L. A., and D. E Soltis. 1998. Assessing congruence: empirical examples from molecular data. In: D. E. Soltis, P. S. Soltis, and J. J. Doyle (eds.), Molecular Systematics of Plants 2: DNA seque ncing. Kluwer, Boston. Johri, B. M., K. B. Ambegaokar and P. S. Srivastava. 1992. Comparative Embryology of Angiosperms, Vol I. Springer-Verlag, Berlin. Jones, S. B. and A. E. Luchsinger. 1987. Plant Systematics. 2nd ed. Mc Graw-Hill Company, New York. Jordan, W. C., W. M. Courtney and E. J. Ne igel. 1996. Low levels of intraspecific genetic variation at a rapidly evol ving chloroplast DNA locus in North American duckwoods (Lemnaceae). American Journal of Botany 83: 430-439. Jukes, T. C. Cantor. 1969. Evol ution of protein molecules. In : H. Munro (ed), Mammalian Protein Metabolism. Academic Press, New York. Katagiri, S. 1953. Chromosome numbers and polyploidy in certain Cactaceae. Cactus and Succulent Journal (U.S) 25: 141-142. Kechner, S. A. and J. F. Wendel. 1996. Hairpins create minute inversions in non-coding regions of chloroplast DNA. Current Genetics 30: 259-262. Keim, P., K. N. Paige, T. G. Whitham and K. G. Lark. 1989. Genetic analysis of an interspecific hybrid swarm of Populus : occurrence of unidirectional introgression. Genetics 123: 557565. Kimnach, M. 1960. A revision of Borzicactus Cactus and Succulent Journal (U.S.) 32(1): 8-13. Kluge, A. G. 1989. A concern for evidence and a phyloge netic hypothesis of relationships among Epicrates (Boidae, Serpentes). Systematic Zoology 38: 7-25. Knobloch, I. W.1972. Intergeneric hybr idization in flowering plants. Taxon 21: 97. Lakshmanan, K. K. and K. B. Ambegaokar. 1984. Polyembryony. In : Johri B. M. (ed.), Embryology of Angiosperms. Springer-Verlag, Berlin.
118 Lambrou, M. and W. Till. 1993. Zur Karyologie der Gattung Gymnocalycium Arbeitsgruppe Gymnocalycium 6: 85-88. Lanave, C., G. Preparata, C. Saccone a nd G. Serio. 1984. A new method for calculating evolutionary substitution rates. Journal of Molecular Evolution 20: 86-93. Leitch, I. J. and M. D. Bennett. 1997. Polyploidy in angiosperms. Trends in Plant Sciences 2: 470-476 Lendel, A., U. Eggli and R. Nyffeler. 2006. Phyloge netic relationships in the tribe Trichocereeae (Cactaceae) inferred from cpDNA sequence anal ysis. Abstract for oral presentation at Botany 2006, Chico, California. Lendel, A., U. Eggli and R. Nyffeler. 2007. Mole cular Systematics of the tribe Trichocereeae (Cactaceae) with special refere nce to biogeography and grow th-form evolution. Abstract for poster presentation at Biology 2007, Zurich. Levin, D. A. 2002. The Role of Chromosomal Cha nge in Plant Evolution. Oxford University Press, New York. Lewis, W. H. 1980. Polyploi dy in species populations. In : W. H. Lewis (ed.), Polyploidy: Biological Relevance. Plenum Press, New York. Lichtenzveig, J., S. Abbo, A. Nerd, N. TelZur and Y. Mizrahi. 2000. Cytology and mating systems in the climbing Hylocereus and Selenicereus. American Journal of Botany 87: 1058-1065. Loxdale, H. D. and G. Lushai. 2003. Rapid changes in clonal lines: the de ath of a sacred cow. Biological Journal of the Linnaean Society 79:3-16. Maddison, W. P., M. J. Donoghue and D. R. Maddison. 1984. Outgroup analysis and parsimony. Systematic Zoology 33: 83-103. Maheshwari, P. and R. N. Chopra. 1955. The struct ure and development of the ovule and seed of Opuntia dillenii Haw. Phytomorphology 5: 112-122. Mandujano, M. C., C. Montaa and L. E. E guiarte. 1996. Reproductive ecology and inbreeding depresion in Opuntia rastrera (Cactaceae) in the Chihuahua n Desert: why are sexually derived recruitments so rare? American Journal of Botany 83(1): 63-70. Mason-Gamer, R. J. and E. A. Kellogg. 1996. Testing for phylogenetic conflict among molecular data sets in the tribe Triticeae (Gramineae). Systematic Biology 45: 524. Masterson, J. 1994. Stomatal size in fossil plants: evidence for poplyploidy in majority of angiosperms. Science 264:421-424. Mauseth, J. D. 1990. Continental drift, climate and the evolution of cacti. Cactus and Succulent Journal (U. S.) 62: 302-308.
119 Mauseth, J. D. and J. V. Landrum. 1997. Relictu al vegetative anatom ical characters in Cactaceae: the genus Pereskia Journal of Plant Research 110: 55-64. McDade, L. A. 1990. Hybrids and phylogenetic system atics I. Patterns of ch aracter expression in hybrids and their implications for cladistic analysis. Evolution 44: 1685. Montieri, S., L. Gaudio and S. Aceto. 2004. Isolation of the LFY/FLO homologue in Orchis italica and evolutionary analysis in some European orchids. Gene 333: 101-109. Mullenniex, A., T. M. Hardig and M. R. Mesl er. 1999. Molecular confirmation of hybrid swarms in the fern genus Polystichum (Dryopteridaceae). Systematic Botany 23(4): 421-426. Nason, J. D., J. L. Hamrick and T. H. Flemi ng. 2002. Historical vicari ance and postglacial colonization effects on the evolut ion of genetic structure in Lophocereus a Sonoran Desert columnar cactus. Evolution 56: 2214-2226. Nassar, J. M., J. L. Hamrick and T. H. Flem ing. 2001. Genetic variation and population structure of the mixed-mating cactus, Melocactus curvispinus (Cactaceae). Heredity 87: 69-79. Nassar, J. M., J. L. Hamrick and T. H. Fleming. 2002. Allozyme diversity and genetic structure of the leafy cactus (Pereskia guamacho [Cactaceae]). Journal of Heredity 93: 193-200. Nassar, J. M., J. L. Hamrick and T. H. Fleming. 2003. Population genetic structure of Venezuelan quiropterophilous columnar cacti (Cactaceae). American Journal of Botany 90: 1628-1637. Naumova, T. N. 1992. Apomixis in Angiosperms: Nucellar and Integumentary Embryony. CRC Press, Boca Raton. Negrn-Ortiz, V. 1998. Reproductiv e biology of a rare cactus, Opuntia spinosissima (Cactaceae), in the Florida Keys: why is seed set very low? Sexual Plant Reproduction 11: 208-212. Nogler, G. A. 1984. Gametophytic apomixis. In : Johri B. M. (ed.), Embryology of Angiosperms Springer-Verlag, Berlin. Nyffeler, R. 2002. Phylogenetic relationships in the cactus family (Cactaceae) based on evidence from trnK/matK and trnL-trnF sequences. American Journal of Botany 89(2): 312-326. Oh, S-H. and D. Porter. 2003. Phylogene tic utility of the second intron of LEAFY in Neillia and Stephanandra (Rosaceae) and implicati ons for the origin of Stephanandra Molecular Phylogenetics and Evolution 29: 202-215. Ostolaza, C. 1990. Haageocereus tenuis Ritt. emend. Rauh et Ostolaza. QUEPO 4(1-2):76-80. Ostolaza, C. 1996. A closer look at the conservation status of cacti in the vicinity of Lima, Peru. British Cactus and Succulent Journal 14(4):158-156.
120 Ostolaza, C. 2000. Haageocereus tenuis and H. repens rescued Peruvian cacti. British Cactus and Succulent Journal 18(3): 129-131. Ostolaza, C. and W. Rauh. 1990. Bemerkungen zur Kakteenvegetation der peruanischen Kstenwste, insbesondere zu Haageocereus tenuis Ritter, einer wenig bekannten Art. Kakteen und Andere Sukkulenten 41(2): 40-44. Otero-Arnaiz, A., J. CruseSanders, A. Casas and Hamrick, J. L. 2004. Isolation and characterization of microsate llites in the columnar cactus: Polaskia chichipe and crossspecies amplification within th e Tribe Pachycer eeae (Cactaceae). Molecular Ecology Notes 4(2): 265-267. Otto, S. P. and J. Whitton. 2000. Polyploid incidence and evolution. Annual Review of Genetics 34: 401-437. Palmer, J. D., R. K. Jansen, H. J. Michaels, M. W. Chase and J. R. Manhart. 1988. Chloroplast DNA variation and plant phylogeny. Annals of the Missouri Botanical Garden 75: 11801206. Parker, K. C. and J. L. Hamrick. 1992. Genetic diversity and clonal stru cture in a columnar cactus, Lophocereus schottii American Journal of Botany 79(1): 86-96. Pinkava, D. J., M. A. Baker, B. D. Parfitt, M. W. Mohlenbrock and R. D. Worthington. 1985. Chromosome numbers in some cacti of western North America V. Systematic Botany 10: 471-483. Pinkava, D. J., L. A. McGill, T. Reeves and R. C. Brown. 1977. Chromosome numbers in some cacti of western North America III. Bulletin of the Torrey Botanical Club 104: 105-110. Pinkava, D. J. and M. G. McLeod. 1971. Chromoso me numbers in some cacti of western North America. Brittonia 23: 77-94. Pinkava, D. J., M. G. McLeod, L. A. McGill a nd R. C. Brown. 1973. Chromosome numbers in some cacti of western North America II. Brittonia 25: 171-176. Pinkava, D. J. and B. D. Parfitt. 1982. Chromo some numbers in some cacti of western North America IV. Bulletin of the Torrey Botanical Club 109: 121-128. Popp, M., P. Erixon, F. Eggens and B. Oxelma n. 2005. Origin and evolution of a circumpolar polyploid species complex in Silene (Caryophyllaceae). Systematic Botany 30: 302-313. Posada, D. and K. A. Crandall. 1998. Modelt est: testing the model of DNA substitution. Bioinformatics 14: 817-818. Prado, D. E. 2000. Seasonal dry forests of tropical South America: from forgotten ecosystems to a new phytogeographic unit. Edinburgh Journal of Botany 57(3): 437-461.
121 Prado, D. E. and P. E. Gibbs. 1993. Patterns of speci es distributions in the dry seasonal forests of South America. Annals of the Missouri Botanical Garden 80: 902-927. de Queiroz, A., M. J. Donoghue and J. Kim. 1995. Separate versus combined analysis of phylogenetic evidence. Annual Review of Ecology and Systematics 26: 657-681. Rambaut, A. 2007. Sequence alignment editor version 2.0. Available from http://tree.bio.ed. ac. uk/software/seal/. Raymond, M. and F. Rousset. 1995. Genepop (ver sion 1.2), population gene tics software for exact tests and ecumenicism. Journal of Heredity 86: 248-249. Rebm an, J. P. and D. J. Pinkava. 2001. Opuntia cacti of North America an overview. Florida Entomologist 84: 474-483. Remski, M. F. 1954. Cytological investigations in Mammillaria and some associated genera. Botanical Gazette 116: 163-171. Richards, A. J. 1970. Eutriploid facultative agamospermy in Taraxacum New Phytologyst 69: 761-774. Richards, A. J. 1997. Plant Breeding Systems. 2nd ed. Chapman and Hall, Cambridge. Rieseberg L. H., M. A. Archer and R. K. Wayne. 1999. Transgressive segregation, adaptation and speciation. Heredity 83: 363. Rieseberg, L. H., O. Raymond, D. M. Rosenthal, Z. Lai, K. Livingstone, T. Nakazato, J. L. Durphy, A. E. Schwarzbach, L. A. Donovan and C. Lexer. 2003. Major ecological transitions in wild sunflowers facilitated by hybridization. Science 301: 1211-1216. Ritter, F. 1981. Kakteen in Sdamerika, Band IV (Peru). Spangenberg, Germany: Friedrich Ritter. Ritz, C. M, L. Martins, R. Mecklenburg, V. Goremykin and F. H. Hellwig. 2007. The molecular phylogeny of Rebutia (Cactaceae) and its allies de monstrates the influence of Paleogeography on the evolution of South American mountain cacti. American Journal of Botany 94(8): 1321-1332. Rodriguez, F., J. F. Oliver, A. Marin and J. R. Medina. 1990. The genera l stochastic model of nucleotide substitutions. Journal of Theoretical Biology 142: 485-501. Ross, R. 1981. Chromosome counts, cytol ogy, and reproduction in the Cactaceae. American Journal of Botany 68: 463-470. Rowley, G. D. 1982. Intergeneric hybrids in succulents. National Cactus and Succulent Journal 37(2): 45-49. Rowley, G. D. 1994. Spontaneous bigeneric hybrids in Cactaceae. Bradleya 12: 2-7.
122 Sahley, C. T. 1995. Bat and hummingbird pollination of two species of columnar cacti: effects on fruit production and pollen dispersal. Ph. D. Dissertation. University of Miami, Coral Gables, FL. Sahley, C. T. 1996. Bat and hummingbird pollination of an autotetraploid columnar cactus, Weberbauerocereus weberbaueri (Cactaceae). American Journal of Botany 83(10): 13291336. Sang, T., D. J. Crawford and T. F. Stue ssy. 1997. Chloroplast DNA phylogeny, reticulate evolution and biogeography of Paeonia (Paeon iaceae). American Journal of Botany 84: 1120. Scott, K. D. and J. Playford. 1996. DNA extraction technique for PCR in rainforest plant species. Biotechniques 20(6): 974-978. Seelanan, T., A. Schnabel and J. F. Wendel. 1997. Congruence and consensus in the cotton tribe (Malvaceae). Systematic Botany 22: 259-290. Shore, J. S., K. L. McQueen and S. H. L ittle. 1994. Inheritance of plastid DNA in the Turnera ulmifolia complex (Turneraceae). Americal Journal of Botany 81(12): 1636-1639. Simpson, B. B. 1975. Pleistocene changes in the flora of the hi gh tropical Andes, Paleobiology 1: 273. Soltis, D. E. 1980. Karyotypic rela tionships among species of Boykinia, Heuchera, Mitella, Sullivantia, Tiarella, and Tolmiea (Saxifragaceae). Systematic Botany 5: 17-29. Soltis, D. E., M. S. Mayer, P. S. Soltis and M. Edgerton. 1991. Chloroplast DNA variation and intraspecific phylogeography in Tellima grandiflora (Saxifragaceae). American Journal of Botany 78: 1379-1390. Soltis. D. E. and P. S. Soltis. 1989. Allopolyploid speciation in Tragopogon: insights from chloroplast DNA. American Journal of Botany 76: 1119-1124. Soltis, D. E. and P. S. Soltis. 1993 Molecular data and the dynamic nature of polyploidy. Critical Reviews in Plant Science 12: 243-273. Soltis, D. E. and P. S. Soltis. 1995. The dynamic nature of polyploidy genomes. Proceedings of the National Academy of Sciences 92: 8089-8091. Soltis, P. S. and D. E. Soltis. 2000. The role of genetic and genomic attrib utes in the success of polyploids. Proceedings of the National Academy of Sciences 97(13): 7051-7057. Soltis, D. E., P. S. Soltis and J. Tate. 2004. A dvances in the study of polyploidy since Plant Speciation. New Phytologist 161: 173-191. Speranza, P. and M. Malosetti. 2006. Nuclear and cytoplasmic microsatellite markers for the species of the Dilatata group of Paspalum (Poaceae). Plant Genetic Resources 5(1): 1-14.
123 Speranza, P., M. Vaio and C. Mazella. 2003. Karyotypes of two cytotypes of Paspalum quadrifarium Lam. (Poaceae). An alternative techni que for small chromosomes in plants. Genetics and Molecular Biology 26: 499-503. Spillane, C., A. Steimer and U. Grossniklaus. 2001. Apomixis in agriculture: the quest for clonal seeds. Sexual Plant Reproduction 14: 179-187. Stebbins, G. L. 1950. Variation a nd Evolution in Plants. Columbia University Press, New York. Stebbins, G. L. 1959. The role of hybridization in evolution. Proceedings of the American Philosophical Society 103: 231-251. Stebbins, G. L. 1969. The significance of hybr idization for plant taxonomy and evolution. Taxon 18: 26-35. Stebbins, G. L. 1971. Chromosomal evolution in higher plants. Addison-Wesley Co. Mass. Stebbins, G. L. 1979. Polyploidy in plan ts: unsolved problems and prospects. In : Lewis W. H. (ed.), Polyploidy, Biological Releva nce. Plenum Press, New York. Stevens, P. F. 2001 onwards. Angiosperm Phylogeny Website. Version 9, June 2008. Sullivan J. 1996. Combining data with different distributions of amongsite rate variation. Systematic Biology 45: 375-380. Swofford, D. L. 2002. PAUP*: phylogenetic analysis using parsimony (* and other methods), version 4.0.b10. Sinauer, Suderland, Ma ssachusetts, USA. Available at http://paup. csit.fsu.edu/about.htm l Symonds, V. V. and A. M. Lloyd. 2003. An an alysis of microsatellite loci in Arabidopsis thaliana : mutational dynamics and applications. Genetics 165: 1475-1488. Tate, J. A., D. E. Soltis and P. S. Soltis. 2005. Polyploidy in plants. In: T. R. Gregory (ed.), The Evolution of the Genome. Elsevier, San Diego. Tavare, S. 1986. Some probabilistic and statistical prob lems in the analysis of DNA sequences. Lectures in Mathematical Life Science 17: 57-86. Taylor, N. P. Cactaceae. 1997. In : S. Oldfield (comp.), Cactus and Succulent Plants Status Survey and Conservation Action Plan. IU CN/SSC Cactus and Succulent Specialist Group. IUCN, Gland, Switzerland and Cambridge, UK. Terry, M., A. E. Pepper and J. R. Manhart. 2006. Development and characterization of microsatellite loci in endangered Astrophytum asterias (Cactaceae). Molecular Ecology Notes 6(3): 865-866. Turner, R. M., S. M. Alcorn and J. A. Booth. 1966. The influence of shade, soil and water on saguaro seedling establishment. Botanical Gazette 127: 95-102.
124 USDA, ARS, National Genetic Resources Prog ram. 2008. Germplasm Resources Information Network (GRIN) [Online Database]. National Germplasm Resources Laboratory, Beltsville, Maryland. URL: http://www.ars-grin.gov/cgi-bin/npgs/html/taxon.pl?25872. Viewed: 11 Nov. 2008. Vigoroux, Y., S. Mitchell, Y. Matsuoka, M. Hamblin, S. Kresovich, J. S. C. Sm ith, J. Jacqueth, O. S. Smith and J. Doebley. 2005. An analys is of genetic diversity across the maize genome using microsatellites. Genetics 169: 1617-1630. Vite, F., E. Portilla, J. A. Zavala-Hurtado, P. L. Valverde and A. Diaz-Solis. 1996. A natural hybrid population between Neobuxbaumia tetetzo and Cephalocereus columna-trajani (Cactaceae). Journal of Arid Environments 32: 395-405. Vriesendorp, B. and F. T. Bakker. 2005. Recons tructing patterns of re ticulate evolution in angiosperms: what can we do? Taxon 54(3): 593-604. Waddell, P. J. and D. Penny. 1996. Evolutionary tr ees of apes and humans from DNA sequences. In : A. J. Lock and C. R. Peters (eds.), Handbook of Symbolic Evolution. Clarendon Press, Oxford. Wares, J. P., D. Alo, T. F. Turner. 2004. A gene tic perspective on management and recovery of federally endangered trout ( Onchorynchus gilae ) in the American Southwest. Canadian Journal of Fisheries and Aquatic Sciences 61: 1890-1899. Wendel, J. F. 2000. Genome evolution in polyploids. Plant Molecular Biology 42: 225-249. Wendel, J. F. and J. Doyle. 2005. Po lyploidy and evolution in plants. In : R. J. Henry (ed.), Plant Diversity and Evolution, Genotypic and Phe notypic Variation in Higher Plants. CABI Publishing, Wallingford, UK. Werdermann, E. 1937. Binghamia Werder. Repertorium Specierum Novarum Regni Vegetabilis 42: 4. White, T. L., T. Bruns, S. Lee, and J. Taylor. 1991 Amplification and dire ct sequencing of fungal ribosomal RNA genes for phylogenetics. In : M. A. Innis, D. H. Gelfand, J. J. Sninsky, and T. J. White (eds.), PCR protocols. Academic Press, San Diego, CA. Wiens, J. J. 1998. Combining data sets with different phylogenetic histories. Systematic Biology 47: 568. Yang, Z. 1994. Maximum likelihood phylogenetic esti mation from DNA sequences with variable rates over sites: approximate methods. Journal of MolecularEvolution 39: 306. Zwickl, D. J. 2006. Genetic algorithm approaches for the phylogenetic analysis of large biological sequence data sets under the maximum likelihood criterion. Ph.D. dissertation, The University of Texas at Austin.
125 BIOGRAPHICAL SKETCH Mnica Arakaki M., daughter of Bunzo Arakak i a nd Cristina Makishi, was born in Lima, Peru. She attended elementary and high school in Huanuco. In 1994 she received a Bachelors degree in Biological Sciences, and in 1999 a Professi onal Title of Biologist with specialization in Botany from Universidad Nacional Mayor de San Marcos, Lima. During the following years she worked as a researcher at Museum of Natural History and San Marcos Herbarium, Lima. In 1999 she entered The University of Texas at Aus tin to pursue an M.A. degree working on the systematics of a Peruvian genus of Cactaceae. At University of Florida, she continued working with Cactaceae, particularly on molecular phylo genetics and population genetics of the South American tribe Trichocereeae.