UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Genetic structure and mating patterns of diploid and polyploid Easter daisies (Townsendia hookeri, asteraceae) Thompson, Stacey Lee 2006

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
831-ubc_2006-201107.pdf [ 6.86MB ]
Metadata
JSON: 831-1.0092998.json
JSON-LD: 831-1.0092998-ld.json
RDF/XML (Pretty): 831-1.0092998-rdf.xml
RDF/JSON: 831-1.0092998-rdf.json
Turtle: 831-1.0092998-turtle.txt
N-Triples: 831-1.0092998-rdf-ntriples.txt
Original Record: 831-1.0092998-source.json
Full Text
831-1.0092998-fulltext.txt
Citation
831-1.0092998.ris

Full Text

GENETIC STRUCTURE A N D M A T I N G PATTERNS OF DIPLOID A N D POLYPLOID EASTER DAISIES (TOWNSENDIA HOOKERI, A S T E R A C E A E ) by S T A C E Y L E E THOMPSON B.Sc , The University of Guelph, 1999 A THESIS SUBMITTED IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY in THE F A C U L T Y OF G R A D U A T E STUDIES (Botany) THE UNIVERSITY OF BRITISH C O L U M B I A August 2006 © Stacey Lee Thompson, 2006 Abstract Reproduction in the Rocky Mountain genus Townsendia involves a complex interplay of polyploidy and apomixis. Molecular markers were used to assess genetic structure in relation to patterns of mating mode and ploidy at four hierarchical levels: within the genus, within T. hookeri, within populations of this species, and among progeny from these populations. Phylogenetic analyses of rDNA repeats indicated that polyploid apomixis has evolved multiple times within the genus. Pollen studies and analyses of cpDNA demonstrated that sexual diploids of T. hookeri are found in both northern and southern unglaciated regions, polyploid apomicts have evolved at least once in the north and thrice in the south, and these polyploid apomicts have colonized the post-glacial landscape from two refugia, suggesting that glaciation and not latitude influenced the distribution of apomicts. Pollen studies, flow cytometry, and multilocus tests on A F L P marker genotypes from four Yukon standing populations of T. hookeri indicated sexuality in one male-fertile diploid population, clonality in two male-sterile tetraploid populations, and a combination of sexual and clonal reproduction in one male-sterile polyploid population. This latter population of mixed mating mode included triploids and tetraploids and showed that evidence of cryptic sex may linger in the genomes within a morphologically asexual population. Finally, a new method for mating system analysis, which jointly estimates the rates of outcrossing, selfing, automixis and apomixis, was developed and applied to dominant A F L P marker genotypes from progeny whose mothers arose from three Yukon populations of T. hookeri, one consisting of male-fertile diploids, the other two of male-sterile tetraploids. Despite indications of sexuality in some standing populations, progeny analyses revealed that apomixis is the predominant mating mode in all populations. Levels of outcrossing were moderate in the diploid population and very low in the tetraploids. Selfing/automixis was absent in the diploids and moderate in tetraploids. These findings suggest that the correlation between ploidy and apomixis is not strict when observed on a fine scale, that polyploidy alone does not induce apomixis, and perhaps it is asexuality that selects for polyploidy within this system. Table of contents Abstract ii Table of contents iii List of Tables iv List of Figures v Acknowledgements vi Dedication vii Co-authorship statement viii 1 Intoduction 1 1.1 References 8 2 Recurrent evolution of polyploid apomixis, endemism, and acaulescence in the Rocky mountain agamic complex Townsendia (Astereae, Asteraceae) 11 2.1 Introduction 11 2.2 Material and methods 14 2.3 Results 16 2.4 Discussion 19 2.5 References 24 3 Patterns of recurrent evolution and geographic parthenogenesis within apomictic polyploid Easter daises (Townsendia hookeri) 35 3.1 Introduction 35 3.2 Material and methods 36 3.3 Results 39 3.4 Discussion 42 3.5 References 48 4 Detection of clonality and sexuality in diploid and polyploid populations of the Easter daisy, Townsendia hookeri 60 4.1 Introduction 60 4.2 Material and methods 61 4.3 Results 66 4.4 Discussion 68 4.5 References 73 5 A novel mating system analysis for modes of self-oriented mating applied to diploid and polyploid arctic Easter daisies (Townsendia hookeri) 80 5.1 Introduction 80 5.2 Material and methods 81 5.3 Results 85 5.4 Discussion 87 5.5 References 91 6 Conclusions 97 IV List of Tables Table 2-1 List of accessions and DNA regions sequenced 27 Table 2-2 Summary of pairwise likelihood distances 28 Table 3-1 Designation, locality information, and pollen data for populations of Townsendia hookeri included in this chapter 53 Table 3-2 Haplotype designations based on chloroplast DNA sequences from Townsendia hookeri, showing the constituent polymorphisms at respective alignment positions 55 Table 4-1 Overrepresentation statistics for four sub-arctic populations of the Easter daisy (Townsendia hookeri) 76 Table 4-2 Association statistics, measures of identity, and estimates of effective rate of long-term sexuality and mutation within four populations of the Easter daisy, Townsendia hookeri 76 Table 5-1 Probabilities of gametes from autotetraploid parents assuming double reduction, under co-dominance and dominance, respectively 93 Table 5-2 Probabilities of tetraploid offspring genotypes with dominance and assuming no double reduction, under outcrossing, selfing, automixis and apomixis, respectively) 93 Table 5-3 Theoretical variances and correlations estimates per individual sampled when selfing rate, automixis and outcrossing rate are simultaneously estimated, for two frequencies of the recessive marker q, and three levels of sample size 93 Table 5-4 Estimates of outcrossing t, selfing s, automixis u, and apomixis a, for each of the three populations, under various hypotheses 94 V List of Figures Figure 2-1 Beamans' hypothesis of phylogenetic relationships within Townsendia with indicated characteristics for species 29 Figure 2-2 Strict consensus of 440 equally-parsimonious trees, based on the external transcribed spacer (ETS) of Townsendia 30 Figure 2-3 Strict consensus of 133 equally-parsimonious trees of 171 steps, based on the internal transcribed spacer (ITS) of Townsendia 31 Figure 2-4 Maximum likelihood tree (-InL = 1886.22) based on the external transcribed spacer (ETS) of Townsendia with character states indicated 32 Figure 2-5 Maximum likelihood tree (-InL = 1835.73) based on the internal transcribed spacer (ITS) of Townsendia 33 Figure 2-6 Maximum likelihood tree (-InL = 3780.71) based on the combined analysis of the external and internal transcribed spacer (ETS and ITS) of Townsendia 34 Figure 3-1 Localities of sexual diploid and apomictic polyploid Easter daisy (Townsendia hookeri) populations collected from western North America 56 Figure 3-2 Plot of mean pollen diameter by mean pollen stainability for populations of the Easter daisy (Townsendia hookeri) 57 Figure 3-3 Chloroplast haplotype variation within 3 geographically proximal populations of Easter daisies (Townsendia hookeri) 58 Figure 3-4 Intraspecific chloroplast phylogeny of sexual diploid and apomictic polyploid Easter daisies (Townsendia hookeri) 59 Figure 4-1 D N A content per 2C nucleus as determined through flow cytometry for Yukon populations of the Easter daisy, Townsendia hookeri 77 Figure 4-2 Most parsimonious trees from Yukon populations of the Easter daisy, Townsendia hookeri, based on AFLPs 78 Figure 4-3 Incompatibility distributions for four Yukon populations of the Easter daisy, Townsendia hookeri 79 Figure 5-1 Gene frequency distribution for A F L P markers from one diploid (Tantalus Butte) and two tetraploid (Mile Thirteen and Tachal Dhal) sub-arctic populations of the Easter daisy, Townsendia hookeri 95 Figure 5-2 Log-likelihoods across a range of outcrossing rate t, for other parameters (s, u, a) jointly estimated 96 VI Acknowledgements Thanks to my committee members: • Mary Berbee, Sally Otto, Kermit Ritland and Jeannette Whitton Helpful colleagues and research facilities: • Institut de recherche en biologie vegetale: Anne Bruneau, Simon Joly, Madoka Misumone • Universite de Montreal: Francois-Joseph Lapointe, and Labo L E M E E • University of Colorado at Boulder: Richard D. Noyes, Nan Lederer • University of Guelph: Paul Kron • University of California at Santa Cruz: Katrina Dlugosh • University of Lethbridge: Joanne Golden • University of British Columbia: Linda Jennings, Gina Choe, NAPS Unit, FACS Facility Supportive friends: • Doreen Haven Thompson and the Deline Clan, Jesse Dylan Thompson, Ryan Marshall Driver, Dylan Leblanc, Teika Newton and Enzo Michel Lhermitte Pirelli Collecting permits: • Jasper National Park • Banff National Park • Waterton Lakes National Park • Kluane National Park • Writing-on-Stone Provincial Park • City of Boulder Open Space • Boulder County Open Space • United States Forestry Service (Rocky Mountain Region) • Territorial Government of the Yukon Financial support: • Post-Graduate Scholarship A/B, The Natural Science and Engineering Research Council of Canada • University Graduate Fellowship, The University of British Columbia • Challenge Grants in Biodiversity • The Alberta Conservation Association • Northern Student Training Program • Botanical Society of America • Department of Botany, The University of British Columbia • Several anonymous backers vii D e d i c a t i o n T h i s thesis is dedicated to the endur ing m e m o r y o f L a i k a , the Sovie t spacedog. She was the first l i v i n g o rgan i sm to enter orbit , launched into space inside Sputnik II on 3 N o v e m b e r , 1957. A stray f rom the mean streets o f M o s c o w , and dubbed "Muttnik" by the A m e r i c a n s , L a i k a ' s genetic ancestry l i k e l y c o m p r i s e d a N o r d i c breed and part terrier. Dear L a i k a d ied a f ew hours after launch from stress and overheat ing . H e r cof f in c i r c l ed the earth 2,570 t imes, then incinerated upon reentering the Ear th ' s atmosphere on 4 A p r i l , 1958. H e r true cause o f death was not made pub l i c unt i l more than 40 years after the fl ight, w i t h off ic ia ls stating at the t ime that she was pa in less ly euthanized w i t h po i soned food. Russ i an bureaucrats have since expressed p u b l i c regret for a l l o w i n g L a i k a to d ie . A l t h o u g h L a i k a suffered, the exper iment p roved that a l i v i n g passenger c o u l d surv ive the weight lessness o f orbi t , p a v i n g the way for eventual spacefl ight by human beings. Vll l Co-authorship statement The research presented in this thesis was identified, designed, performed and analyzed the author, Stacey Thompson, with the exceptions listed below. • Jeannette Whitton identified the research problem in chapter 2. The ITS sequencing in this chapter was performed by various members of Jeannette Whitton's lab group, including Katrina Dlugosch. The author designed and performed the study of ETS variation, analyzed and prepared the chapter as presented here. • Jeannette Whitton identified the research problem in chapter 3 and helped with some of the writing. • The AFLP reactions presented in chapter 4 were run by Gina Choe, under the author's supervision, as part of an undergraduate NSERC project • The program used to analyze the data in chapter 5 was written by Kermit Ritland • The simulations presented in chapter 5 were performed by Kermit Ritland This thesis was written and prepared by the author, Stacey Thompson, with guidance and editorial input from all of her committee members. 1 1. Introduction Plants exhibit a wide variety of reproductive modes, impacting levels of recombination, patterns of genetic variation, rate of mutational meltdown, and hence the species' ability to evolve over time. As the adaptive significance of recombination (Maynard Smith 1978) remains one of the key issues in evolutionary biology (reviewed in Otto and Lenormand 2002), plants that deviate from random mating provide a fertile testing ground for hypotheses surrounding these phenomena. In this thesis, some of these issues are explored using species of the genus Townsendia Hook. (Astereae, Asteraceae). Townsendia consists of perennial, biennial, and annual daisy-like herbs, largely distributed throughout western North America. Populations of Townsendia can vary in ploidy-level and reproductive mode across a species' range, thus providing opportunities for insights into the factors promoting the evolution of mating systems and genetic variation, as related to ploidy. Apomixis is a generic word that has historically been applied to a host of reproductive modes that do not involve fertilization. Here, we use the term in its most commonly-practiced sense, as synonymous with gametophytic agamospermy, the formation of asexual seed from an egg-like precursor (Nogler 1984). This excludes vegetative propagation as well as adventitious embryony (the formation of somatic embryos). Gametophytic agamospermy (herein referred to as "apomixis") comprises two distinct phenomena: apomeiosis and parthenogenesis. Apomeiosis is when the embryo sac precursor does not undergo meiosis so that embryo sacs are derived from diploid (2n) cells, instead of the expected haploid (n) meiotic products (Renner 1916). Apomeiosis can occur through a myriad of mechanisms and embryo sac development is often dichotomously-categorized as diplosporous (arising from derivatives of the megaspore) or aposporous (having non-megasporic origins). Parthenogenesis is the ability to produce an embryo without fertilization of the egg. In apomicts, the endosperm can be formed in the presence or the absence of fertilization of the central cell of the female gametophyte (i.e. apomixis can be pseudogamous or non-pseudogamous). Apomictic plant species are typically categorized as obligate or facultative apomicts. While obligate apomicts are characterized as undergoing little or no recombination, asexual populations, in general, can harbor much genetic variation, often on par with sexuals (Parker 1979; Ellstrand and Roose 1987; Hamrick and Godt 1989; Widen et al. 1994; Loxdale and Lushai 2003). This genetic diversity can have many sources, including multiple origins of apomicts from sexuals and immigration, in addition to somatic mutation and sporadic episodes of meiotic and mitotic recombination. Typically, apomictic plant populations have been characterized with descriptive metrics such as the proportion of distinguishable genets, diversity indices and clonal evenness (e.g. Ellstrand and Roose 1987). However, as some researchers question the existence of obligate 2 apomixis (Asker and Jerling 1992), putatively obligate apomicts should be explicitly tested for recombination. ' '-• . Polyploidy, the condition of possessing more than two chromosome sets, results from a pervasive and frequently occurring class of mutation that is both prevalent and important in the evolution of angiosperms (reviewed in Stebbins 1950; Grant 1981; Soltis and Soltis 1993; Soltis and Soltis 2000; Otto and Whitton 2000). A well-established aspect of the biology of polyploid plants is that the transition from diploidy to polyploidy can occur multiple times within taxa. Over 45 studies have documented the multiple origins phenomenon (reviewed in Soltis and Soltis 1993, 1999), leading to the conclusion of Soltis and Soltis (1993) that the pattern of multiple origins of polyploidy is clearly "the rule, rather than the exception". Over 99% of the hundreds of apomictic plant species examined have proven to be polyploids (Grant 1981; Nogler 1984; Asker and Jerling 1992). Numerous explanations for this tight relationship have been posited, including a greater ecological tolerance of asexuals through polyploidy (Stebbins 1950), the short-term buffering of deleterious mutations in asexual lineages through genomic redundancy (Ohno 1970), non-Mendelian transmission of the genes for apomixis (e.g. Noyes and Rieseberg 2000), as well as lethal action of apomixis alleles in the gametophytic generation (Nogler 1984). Alternatively, it may be that apomictic lineages merely have difficulties in remaining diploid. In absence of sex, mutations in genome copy number might increase in an apomictic lineage with time. These apomicts would gain a unidirectional "ploidy load" with the sole purging mechanism being an upper-bound to ploidy. Additionally, if apomeiosis is accompanied by occasional fertilization events, ploidy-levels would likewise increase through generations. No clear consensus has emerged on whether it is polyploidy that induces the effective expression of apomixis or whether polyploidy offers fitness advantages in an asexual lineage, resulting in this strict correlation of traits. Polyploidy can also facilitate or induce changes in the ability to self-fertilize (Stebbins 1950; Levin 1983; Cook and Soltis 2000; but see Mable 2004). The switch from diploidy to tetraploidy may either directly cause a shift towards greater selfing, through breakdowns in systems of self-incompatibility (Chawla et al., 1997; Stone, 2002), or indirectly cause an increase in selfing, through selection for increased selfing during the establishment of a neopolyploid undergoing a minority cytotype disadvantage (Ramsey and Schemske 1998). As well, in tetraploids, inbreeding depression is reduced (Husband and Schemske 1997) and the deleterious effects of selfing may be tolerated to a greater extent. Selection for increased selfing is often suggested for recolonized landscapes (cf Baker's Law, Baker 1955), areas where many polyploids can be found. Due to its prevalence and 3 potential impact, particularly within the flowering plants (Otto and Whitton 2000), mating system models must incorporate the additional layer of complexity that polyploidy introduces. The frequency of polyploidy increases with latitude in the northern hemisphere (Brochmann et al. 2004). Many reasons for this correlation have been posited. Polyploids may be better-adapted than diploids to extreme climates (Johnson et al. 1965). Due to their higher levels of genetic variability through multiple chromosome copies, polyploids may be at a selective advantage under frequent climate changes (Johnson and Packer 1965). Polyploids may be more likely to colonize formerly glaciated areas due to increased ecological flexibility (Stebbins 1950), possibly resulting from functional diversification of gene copies after chromosome doubling (e.g. Adams and Wendel 2005). However, a relatively recent study found no clear-cut association between polyploidy and the degree of glaciation when measured across the entire arctic flora (Brochmann et al. 2004). However, for 'specialist' taxa with limited northern distributions, the authors found that the frequency of diploids was much higher in Beringia, which remained largely unglaciated during the last ice age, than in the heavily glaciated Atlantic area. This result supports the hypothesis that polyploids are more successful than diploids in colonizing after glaciers recede. Apomixis can likewise impact a plant's geographic distribution. Geographic parthenogenesis (Vandel 1931) describes a pattern wherein asexual organisms have a distribution that is more northerly, more widespread and/or at higher elevations than their sexual counterparts. This common pattern can be found throughout numerous plant and animal taxa (reviewed in Lynch 1984, Bierzychudek 1985). Due to its broad implications for understanding why sex is maintained in nature, geographic parthenogenesis has received theoretical attention (e.g. Glesener and Tilman 1978, Lynch 1984, Gaggiotti 1994, Peck et al. 1998), yet few empirical studies go beyond a basic description of the pattern. Due to the correlation between apomixis and polyploidy, patterns of plant distribution are best approached with a simultaneous consideration of both of these traits. Agamic complexes are groups of species that hybridize, undergo polyploidy, and reproduce through apomixis. As exemplified by well-known systems such as Crepis (Babcock and Stebbins 1938), Amelanchier, (Campbell et al. 1997) and Antennaria (Bayer 1997), studies of agamic complexes can provide insight into broad evolutionary questions regarding the importance of sexual reproduction, gene flow, and genomic redundancy to patterns of diversification. Recombination can act to bind the evolutionary trajectory of these forms: between individuals within populations, among ploidy-levels, and among species that lack strong barriers to gene flow. Townsendia Hook. (Astereae, Asteraceae) provides an interesting contrast to the typical agamic complex. In Townsendia, hybridization appears to occur solely among diploids, only low-level polyploids are found (triploids and tetraploids), and polyploidy has been reported to go hand-in-4 hand with obligate apomixis. Derived within the North American clade of the Astereae (Noyes and Rieseberg 1999), Townsendia consists of perennial, biennial, and annual daisy-like herbs, largely distributed throughout the Rocky Mountain regions of the United States and Canada, and north-central Mexico. Largely unstudied since Beaman's monograph of the genus (Beaman 1957), members of Townsendia have naked, convex receptacles. The pappus consists of reduced scales or bristles. Phyllaries are arranged in 2-7 series and have seriaceous-cilate margins. Disk corollas are yellow, while ray corollas are pink, white, blue or light yellow. Leaves in Townsendia are alternate, entire, and usually pubescent. Twelve of the species are diminutive, prostrate cushion-plants, a condition that Beaman considered an adaptation to harsh environments. He interpreted this phenotype as arising though convergence from independent caulescent (i.e having a well-developed, aboveground stem) "stocks" (i.e. ancestral lineages, Beaman 1957). Most species exhibit a high degree of edaphic specificity, and although a few species of Townsendia are quite widespread, nearly half of the species are localized endemics. Some form of conservation status has been granted to most species (Argus and White 1978; Douglas 1979; White and Johnson 1980; Douglas et al. 1981; Douglas et al. 1998; USDA and NRCS 2006). Through detailed cytogenetic and embryological investigations, Beaman (1957) posited a tight association between polyploidy and obligate apomixis in members of the genus. Within 13 of Townsendia's 26 species, both sexual diploid and apomictic polyploid populations were detected. A l l examined polyploid specimens behaved as cytological autopolyploids, as a lack of bivalent pairing was observed during meiosis. As all apomictic polyploids were morphologically similar to known sexual diploid populations, Beaman did not grant any special taxonomic status to the apomicts. However, sexual diploid and apomictic polyploid plants could be distinguished based on differences in pollen size and viability (Beaman 1954, 1957). Diploid individuals produced small pollen with a very low percentage of inviable grains (Beaman 1957). The percentage of aborted pollen grains was quite high in polyploid plants, with large pollen grains produced. Meiosis in both microsporogenesis and megasporogenesis was found to be irregular in polyploids with many or all of the chromosomes occurring as univalents during metaphase I (Beaman 1957). After this, the meiotic spindle was observed to collapse and a restitution nucleus (an entangled chromosome clump) formed. The subsequent anaphase resembled anaphase ff instead of anaphase I. A pair of unreduced nuclei was derived from this atypical meiosis within the examined polyploids. Townsendia appears to follow the 'Ixeris' type of developmental schema, wherein both daughter nuclei that are derived from the restitution nucleus contribute to the mature 8-celled embryo sac. In this type of development, cell walls form only after the embryo sac has reached the mature 8-celled stage. Thus, the female gametophyte in polyploid Townsendia avoids reduction through diplospory. 5 This appeared to be followed by non-pseudogamous parthenogenesis, in which pollination is neither required for the development of the embryo nor the endosperm. As meiosis appeared to be normal within diploid individuals, with tetrads produced during both microsporogenesis and megasporogenesis, Beaman considered diploid populations to be invariantly sexual. One of the two new species that Beaman described was the Easter daisy, Townsendia hookeri Beaman that comprises both diploid and polyploid populations. The Easter daisy has a distribution that appears consistent with geographic parthenogenesis (Vandel 1931; Bierzychudek 1985). Diminutive and relatively widespread, its distribution straddles the montane regions of the Rocky Mountains, from the front ranges of Colorado in the south, to the northern populations of Alberta and British Columbia. Pollen studies, combined with chromosome counts and embryology, have indicated that sexual diploid populations occur exclusively in the southern part of the species range in Colorado and Wyoming, while apomictic autotetraploids occur northward (Beaman 1957). Since Beaman's work, Easter daisy populations of unknown cytotype and breeding system have been discovered in Canada's Yukon Territory, disjunct by over 1400 km from the rest of the range (Douglas et al. 1981). As the Wisconsin glaciation has played a large role in determining extant plant distribution patterns in northern temperate regions (Hewitt 2000; Abbott and Brochmann 2003), these Yukon populations of T. hookeri provide an opportunity to examine patterns of post-glacial migration in the context of geographic parthenogenesis, as populations can be found in and adjacent to two known ice-free areas: Beringia, the unglaciated region of the Yukon Territory to the north, and the large southern refugium of the continental USA (Dyke and Prest 1987). Whether recolonization of the post-glacial landscape by Easter daisies has occurred from northern or southern ice-free areas has remained a mystery. Beaman's hope was that his monograph on Townsendia would "serve as a reference point from which problems in plant migration, population variability, and phylogenetic modification can be attacked" (Beaman 1957, p.4). The overall objective of this thesis was to team molecular markers with statistical approaches to evaluate genetic structure as pertaining to patterns of mating system and ploidy evolution. These trends were evaluated at four hierarchical levels: (chapter 2) within the genus Townsendia, (chapter 3) within T. hookeri, (chapter 4) within standing populations of this species, and (chapter 5) among progeny from these populations. Specific objectives were as follows: Recurrent evolution of polyploid apomixis. endemism. and acaulescence in the Rocky mountain agamic complex Townsendia (Astereae. Asteraceae): In chapter 2, we construct DNA sequence phylogenies for Townsendia using the internal transcribed spacer (ITS) and the external transcribed spacer (ETS) of the rDNA cistron. These regions have been widely used to resolve phylogenetic relationships for many plant groups at low taxonomic levels (e.g. Baldwin and Markos 6 1998; Clevinger and Panero 2000; Markos and Baldwin 2001; Chan et al. 2002; Roberts and Urbatsch 2003). If multiple origins of polyploidy are "the rule rather than the exception", then we might expect to find evidence of recurrent origins of polyploidy among members of the genus. Trends in polyploid apomixis and endemism are examined. If polyploids have broader ecological tolerances than diploids, we should detect a negative correlation among apomictic polyploidy and endemism within the genus. These molecular phylogenies of Townsendia will form an initial backbone to our understanding of evolutionary patterns and processes within the genus. Patterns of recurrent evolution and geographic parthenogenesis within apomictic polyploid Easter daises (Townsendia hookeri): In chapter 3, we examine the recurrent evolution of ploidy and breeding system throughout the range of the Easter daisy, T. hookeri. Pollen observations are used to assess patterns of ploidy and breeding system throughout the range. Variation in non-recombining, uniparental chloroplast D N A is used to deduce the minimum number of transitions from sexual diploidy to apomictic autoploidy within this species. Patterns of genetic variation and phylogeny are assessed in the context of refugial survival and post-glacial colonization. Geographic patterns of ploidy and breeding system may offer insight into the putative determinants of patterns of geographic parthenogenesis. The detection of clonality and sexuality in diploid and polyploid populations of the Easter daisy. Townsendia hookeri: a comparison of alternative methods: As evidence of cryptic sex may linger in the genomes of predominantly asexual populations, in chapter 4, we test for clonality and sexuality in four putatively apomictic populations of Townsendia hookeri from the Yukon Territory. Pollen and flow cytometry are used to diagnose male-fertility and ploidy. Individuals from each standing population are fingerprinted with amplified fragment length polymorphism (AFLP) markers and multilocus genotypes are subjected to multilocus tests for clonal structure. The A F L P technique can generate a large number of markers, representing a genome-wide survey of broad utility. Use of these markers requires little initial investment relative to other methods, making their use ideal for non-model organisms such as Townsendia. However, AFLP markers also exhibit dominance, and the available analytical methods are generally less well-developed for this type of data. Following Beaman's findings, we expect polyploid populations to be compatible with strict clonality and for diploids to bear the signature of recombination. These results will furnish an understanding of the mating mode as related to ploidy in these unstudied northern disjunct populations. A novel mating system analysis for modes of self-oriented mating applied to diploid and polyploid arctic Easter daisies (Townsendia hookeri): In chapter 5, we develop a procedure for measuring alternative modes of mating in diploids and tetraploids for use with dominant AFLP markers. No methods of mating system analysis exist for the case of dominant markers and 7 polyploidy. Mating hypotheses are examined for progeny arrays from three populations of Townsendia hookeri from the Yukon Territory. Following Beaman's findings, we expect apomixis to be restricted to polyploids and for diploids to be primarily outcrossing. As polyploidy is known to occasionally lead to a break down of self-incompatibility, and as inbreeding depression is usually greater in diploids than in polyploids, the expectation is for less selfing to occur within a diploid population as compared with a polyploid population. This procedure will add to our ability to assess the complexity of mating tactics that diploid and polyploid plants employ in nature. 8 1.1 References Abbott RJ, Brochmann C. 2003. History and evolution of the arctic flora: in the footsteps of Eric Hulten. Molecular Ecology 12: 299-313. Adams K L , Wendel JF. 2005. Novel patterns of gene expression in polyploidy plants. Trends in Genetics 21: 539-543. Argus GW, White DJ. 1978. The Rare Vascular Plants of Alberta. National Museums of Canada: Ottawa, Canada. Asker SE, Jerling L. 1992. Apomixis in Plants. C R C Press: Boca Rotan, USA. Babcock EB, Stebbins G L . 1938. The American species of Crepis. Carnegie Inst. Wash. Publ. No. 504: 1-199. Baker H G . 1955. Self compatibility and establishment of long distance dispersal. Evolution 9: 337-349. Baldwin B G , Markos S. 1998. Phylogenetic utility of the external transcribed spacer (ETS) of 18S-26S rDNA: congruence of ETS and ITS trees of Calycadenia (Compositae). Molecular Phylogenetics and Evolution 10: 449-463. Bayer RJ. 1997. Antennaria rosea (Asteraceae) - a model group for the study of the evolution of polyploid agamic complexes. Opera Botanica 132: 53-65. Beaman JH. 1954. Chromosome numbers, apomixis, and interspecific hybridization in the genus Townsendia. Madrono 12: 169-180. Beaman JH. 1957. The systematics and evolution of Townsendia (Compositae). Contributions from the Gray Herbarium of Harvard University 183: 1-151. Bierzychudek P. 1985. Patterns in plant parthenogenesis. Experientia 41: 1255-1264. Brochmann C, Brysting A K , Alsos IG, Borgen L, Grundt HH, Scheen A - C , Elven R. 2004. Polyploidy in arctic plants. Biological Journal of the Linnean Society 82: 521-536. Campbell CS, Wojciechowski MF, Baldwin B G , Alice L A , Donoghue MJ. 1997. Persistent nuclear ribosomal D N A sequence polymorphism in the Amelanchier agamic complex (Rosaceae). Molecular Biology and Evolution 14: 81-90. Chan R, Baldwin B G , Ornduff R. 2002. Cryptic goldfields: a molecular phylogenetic reinvestigation of Lasthenia californica sensu lato and close relatives (Compositae: Heliantheae sensu lato). American Journal of Botany 89: 1103-1 1 12. Chawla B, Bernatzky R, Liang W, Marcotrigiano M. 1997. Breakdown of self-incompatibility in tetraploid Lycopersicum peruvianum: inheritance and expression of 5-related proteins. Theoretical and Applied Genetics 95: 992-996. Clevinger JA, Panero JL. 2000. Phylogenetic analysis of Silphium and subtribe Engelmanniinae (Asteraceae: Heliantheae) based on ITS and ETS sequence data. American Journal of Botany 87: 565-572. Cook L M , Soltis PS. 2000. Mating systems of diploid and allotetraploid populations of Tragopogon (Asteraceae). II. Artificial populations. Heredity 84: 410-415. Douglas GW- 1979. The Rare Vascular Plants of Saskatchewan. National Museums of Canada: Ottawa. 9 Douglas GW, Straley GB, Meidinger DV. 1998. Rare Native Vascular Plants of British Columbia. BC Environment: Victoria. Douglas GW, Argus GW, Dickson HL, Brunton DF. 1981. The Rare Vascular Plants of the Yukon. National Museums of Canada: Ottawa. Dyke AS, Prest VK. 1987. Late Wisconsin and Holocene retreat of the Laurentide Ice Sheet. Geological Survey of Canada. Ellstrand NC, Roose ML. 1987. Patterns of genotypic diversity in clonal plant species. American Journal of Botany 74: 123-131. Gaggiotti OE. 1994. An ecological model for the maintenance of sex and geographic parthenogenesis. Journal of Theoretical Biology 167: 201-221. Glesener RR, Tilman D. 1978. Sexuality and the components of environmental uncertainty: clues from geographic parthenogenesis in terrestrial animals. The American Naturalist 112: 659-673. Grant V. 1981. Plant Speciation. Columbia University Press: New York, USA. Hamrick JL, Godt MJW. 1989. Allozyme diversity in plant species. In: Brown ADH, Clegg MT, Kahler AL, Weir BS, eds. Plant Population Genetics, Breeding, and Genetic Resources. 43-63. Sinauer Associates: Sunderland, USA. Hewitt G. 2000. The genetic legacy of the Quaternary ice Ages. Nature 405: 907-913. Husband BC, Schemske, DW. 1997. The effect of inbreeding in diploid and tetraploid populations of Epilobium angustifolium (Onagraceae): implications for the genetic basis of inbreeding depression. Evolution 51: 737-746. Johnson AW, Packer JG. 1965. Polyploidy and environment in arctic Canada. Science 148: 237-239. Johnson AW, Packer JG, Reese G. 1965. Polyploidy, distribution and environment. In Wright HE, Frey DG, eds. The Quaternary of the United States.. 407-507. Princton University Press: Princeton, USA. Levin DA. 1983. Polyploidy and novelty in flowering plants. The American Naturalist 122: 1-25. Loxdale HD, Lushai G. 2003. Rapid changes in clonal lines: the death of a 'sacred cow'. Biological Journal of the Linnean Society 79: 3-16. Lynch M. 1984. Destabilizing hybridization, general-purpose genotypes and geographic parthenogenesis. The Quarterly Review of Biology 59: 257-290. Mable BK. 2004. Polyploidy and self-compatibility: is there an association? New Phytologist 162: 803-811. Markos, S, Baldwin BG. 2001. Higher-level relationships and major lineages of Lessingia (Compositae, Astereae) based on nuclear rDNA internal and external transcribed spacer (ITS and ETS) sequences. Systematic Botany 26: 168-183. Maynard Smith, J. 1978. The Evolution of Sex. Cambridge University Press: Cambridge, UK. Mogie M . 1992. The evolution of asexual reproduction in plants. Chapman & Hall: London, UK. Nogler GA. 1984. Gametophytic apomixis. In Johri BM, ed. Embryology of Angiosperms. 475-518 Springer-Verlag: Berlin, DE. 10 Noyes R D , Rieseberg L H . 1999. ITS sequence data support a single origin for North American Astereae (Asteraceae) and reflect deep geographic divisions in Aster S .L. American Journal o f Botany 86: 398-412. Noyes R D , Rieseberg L H . 2000. Two independent loci control agamospermy (apomixis) in the triploid flowering plant Erigeron annuus. Genetics 155: 379-390. Ohno S. 1970. Evolut ion by Gene Duplication. Springer-Verlag: N e w Y o r k , U S A . Otto SP, Lenormand T. 2002. Resolving the paradox o f sex and recombination. Nature Reviews Genetics 3: 252-261. Otto SP, Whitton J. 2000. Polyploid incidence and evolution. Annual Reviews of Genetics 34: 401-437. Parker E D . 1979. Ecological implications of clonal diversity in parthenogenetic morphospecies. American Zoologist 19: 753-762. Peck JR, Yearsley J M , Waxman D . 1998. Explaining the geographic distributions of sexual and asexual populations. Nature 391: 889-892. Ramsey J, Schemske D W . 1998. Pathways, mechanisms, and rates of polyploid formation in flowering plants. Annual Reviews in Ecology and Systematics 29: 467-501. Renner O. 1916. Zur terminologie des pflanzlichen Generationswechsels. Biologisches Zentralblatt 36: 337-374. Roberts R P , Urbatsch L E . 2003. Molecular phylogeny o f Ericameria (Asteraceae, Astereae) based on nuclear ribosomal 3 -ETS and ITS sequence data. Taxon 52: 209-228. Soltis D E , Soltis PS . 1993. Molecular data and the dynamic nature of polyploidy. Cri t ical Reviews in Plant Science 12: 243-273. Soltis D E , Soltis PS . 1999. Polyploidy: recurrent formation and genome evolution. Trends in Ecology and Evolution 14: 348-352. Soltis PS, Soltis D E . 2000. The role of genetic and genomic attributes in the success of polyploids. Proceedings o f the National Academy of Sciences, U S A 97: 7051-7057. Stebbins G L . 1950. Variation and Evolution in Plants. Columbia University Press: N e w York , U S A . Stone JL . 2002. Molecular mechanisms underlying the breakdown of gametophytic self-incompatibility. Quarterly Review o f Biology 77: 17-32. U S D A , N R C S . 2006. The P L A N T S Database, Version 3.5. http://plants.usda.gov. National Plant Data Center: Baton Rouge, U S A . Vandel A . 1931. L a parthenogenese. Doin : Paris, FR . White D J , Johnson K L . 1980. The Rare Vascular Plants o f Manitoba. National Museums of Canada: Ottawa, Canada. Widen B , Cronberg N , Widen M . 1994. Genotypic diversity, molecular markers and spatial distribution of genets in clonal plants, a literature survery. Fol ia Geobotanica & Phytotaxonomica 29: 245-263. 11 2. Recurrent evolution of polyploid apomixis, endemism, and acaulescence in the Rocky mountain agamic complex Townsendia (Astereae, Asteraceae) 2.1 Introduction Phylogenies of agamic complexes (groups of species that hybridize, vary in ploidy, and reproduce asexually) can characterize evolutionary patterns and relationships and can subsequently be used to inform tests of broad evolutionary questions regarding the contributions of recombination, gene flow, and genomic redundancy to diversification. Exemplified by well-known systems such as Crepis (Babcock and Stebbins 1938), Amelanchier (Campbell et al. 1997), and Antennaria (Bayer 1997), agamic complexes exhibit intricate patterns of variation which are generated through gene exchange among diploids and low-level polyploids, to create reticulate high-level polyploids that can exhibit various degrees of asexuality. The Rocky Mountain agamic complex Townsendia Hook. (Astereae, Asteraceae) consists of 28 described species and provides an interesting contrast to the typical agamic complex, as hybridization is suspected to occur only among diploids, only low-level autopolyploids are found, and polyploidy goes hand-in-hand with obligate asexuality through apomixis. This relatively simple suite of traits suggests that Townsendia is apt to provide a tractable system for evolutionary studies. Townsendia is derived within the "North American" clade of the Astereae (Noyes and Rieseberg 1999) and the limits of the genus are well-defined and unambiguous. Townsendia consists of perennial, biennial, and annual daisy-like herbs with naked, convex receptacles (i.e. the disk of tissue which forms the base of the inflorescence, to which the other flower-parts are attached), a pappus (i.e. a group of appendages, above the ovary and outside the corolla which represent a modified calyx) of reduced scales or bristles, and phyllaries (i.e. involucral (i.e. subtending the inflorescence) bracts) with seriaceous-cilate (i.e. silky-haired) margins arranged in 2-7 series. Disk (inner floret) corollas are yellow and ray (outer floret) colloras are white, pink, blue or light yellow. Leaves are alternate, entire, and usually pubescent. Members of the genus are distributed throughout the Rocky Mountain regions of the United States and Canada, and northern Mexico. With the exception of some new taxonomic descriptions, the relationships within Townsendia have remained unstudied since Beaman's 1957 monograph of the genus. In this work, the author dedicates over 75 pages to the discussion of his phylogenetic placement of taxa within the genus. The character states which he considered to be "primitive" are exemplified by the morphology of T. formosa and listed as "a perennial, fibrous rooted, rhizomatous and stoloniferous habit; erect, monocephalous stems; large, thin, glabrate, spatulate leaves; lightly pubescent stems; large heads; a conical receptacle; several rows of broad phyllaries with narrowly scarious and minutely ciliate margins; large, obovate achenes; and a small, unelaborated pappus." Alternate morphologies were 12 believed to have evolved from this "primitive" type. Beaman did not use any particular analytical techniques to reach these conclusions; his work was carried out decades before the broad acceptance of cladistic and numerical taxonomic approaches to phylogenetics. Rather, his ideas were largely derived through experience, intuition, and a "feeling for the organism" (Fox Keller 1993) about trends in morphological evolution and possible colonization patterns throughout the western United States. Anecdotally, most species of Townsendia have a high degree of edaphic specificity, occurring on particular, coarsely-textured arid soils throughout Western North America. Although a few species of Townsendia are quite widespread (e.g. T. hookeri can be found in sporadic pockets from southern Colorado to Alaska, T. exscapa from northen Mexico to Manitoba), nearly half of the species are localized endemics. Taxa have been characterized as "endemic" if they are confined to few popualtions within one or two states (Beaman 1957) and these endemic taxa were hypothesized to be specialized derivatives from more broadly-occuring taxa. Due to the restricted nature of species and/or populations within political regions, some form of conservation status has been granted to most species (Argus and White 1978; Douglas 1979; White and Johnson 1980; Douglas et al. 1981; Douglas et al. 1998; USDA and NRCS 2006). Twelve species of Townsendia are acaulescent (i.e. stemless) prostrate cushion-plants. Beaman considered that species with diminutive growth forms arose as an adaptative specialization to harsh environmental conditions. These reduced forms were posited to be polyphyletic, arising though convergence from independent caulescent (i.e. having an obvious leafy stem and distinct internodes) "stocks" (i.e. ancestral lineages, Beaman 1957). Within 13 of the 26 species recognized at the time, Beaman detected both diploid and polyploid populations (Beaman 1954, 1957). The remaining 13 species of Townsendia consisted exclusively of diploids; no taxa consisted exclusively of polyploids. Al l examined polyploid specimens behaved as cytological autoploids (i.e. bivalent pairing is not observed during meiosis). As diploid and polyploid plants were morphologically identical (except for differences in pollen size and viability) Beaman did not make any taxonomic distinction among the cytotypes, instead keeping the two ploidy-classes as members of the same taxon. He hypothesized that the species with polyploid populations were not a monophyletic assemblage. Through extensive cytogenetic and embryological investigations, Beaman (1954, 1957) discovered a tight association between ploidy and breeding system in Townsendia. He reported that diploid populations of Townsendia were invariantly sexual, while polyploid populations were obligately apomictic. The intensity with which Beaman studied the reproductive biology as correlated with ploidy varied among taxa and with the investigative approach. In his 1954 report (Beaman 1954) Beaman observed microsporogenesis and mitotic chromosome spreads in 1-4 13 populations from each of ten species (total = 20 populations, consisting of 12 diploid and 8 polyploid populations, the exact number of observations for each population was not given). Plants from each population were bagged and examined for seed production (sample size per population is not given). Pollen quality and size was examined from five polyploid populations from four species (range = 329-379 observations per population), as well as from seven diploid populations, each from a different species (range = 348-484 observations per population). The stigmas and stamens were removed from plants from two polyploid populations (sample sizes not given), and these manipulated plants were examined for the production of viable seed. Progeny were grown from seven species (sample sizes not given) and examined for uniformity. In his 1957 monograph of the genus (Beaman 1957), megasporogenesis was assessed in diploids of one to three populations from each of three species, comprising 81-344 (range) ovules per population for a total sample size of 870 ovules. Microsporogenesis was observed in 1-8 diploid populations from each of eight species, although the exact number of observations remains unclear. Megasporogenesis was observed in 1-5 polyploid populations from each of 3 species, with 53-636 (range) ovules observed from each population for a total of 1322 polyploid ovules observed. Microsporogenesis was observed in seven polyploid populations from three species (1-4 populations per species). In addition, Beaman examined pollen size and viability in every collection of Townsendia that he studied, representing 4-120 (range) populations from each of 21 species (although the exact number of pollen grains evaluated per population is not listed). Based upon the results from the sum totality of this work, Beaman concluded that diploid populations were invariantly sexual, while polyploid populations were obligately apomictic, with the female gametophyte in polyploids avoiding reduction by restitutional diplospory (a meiotic abnormality in the megasporocyte), followed by non-pseudogamous parthenogenesis (in which pollination is neither required for the development of the embryo nor the endosperm). Here, we investigate the likelihood of some of Beaman's hypotheses through observations and tests of molecular phylogenies. Specifically, we ask: (1) Are Beaman's ideas about Townsendia's "primitive" characters correct? If so, we would expect T. formosa to be the member of the genus that diverges the earliest, as this taxon possesses all of the morphological characters deemed to be "primitive". (2) Has polyploid apomixis, morphological reduction and endemism evolved recurrently and are there general correlations among these three traits? If taxa that exhibit these traits form monophyletic groups, this would suggest that phylogeny is an important determinant in the evolution of these traits (ie, a key innovation may have occurred). However, if these traits have evolved multiple times, it might suggest that there has been selection for these traits in independent lineages. 14 In order to address these questions and to more-formally examine the evolutionary relationships within the genus, we have constructed D N A sequence phylogenies for Townsendia using the internal transcribed spacer (ITS) and the external transcribed spacer (ETS) of the rDNA cistron. These regions have been used to resolve phylogenetic relationships for many plant groups at low taxonomic levels (e.g. Baldwin and Markos 1998; Clevinger and Panero 2000; Markos and Baldwin 2001; Chan et al. 2002; Roberts and Urbatsch 2003). These molecular phylogenies of Townsendia may form a baseline for understanding evolutionary patterns and processes within the genus. 2.2 Materials and methods Materials—A total of 36 accessions of Townsendia, representing 1-2 samples for all described species, most varieties and an as yet unnamed collection (R. Hartman, personal communication), were assembled for this study (Table 2-1). T. aprica, T. gypsophila, T. microcephala, T. nuttallii and T. smithii are taxa which have all been named subsequent to Beamn's monograph (Beaman 1957). Leaf material came from one of three sources: frozen field collections, fresh material from laboratory-grown seed, or dried material from herbarium specimens. Astranthium integrifolium and Dichaetophora campestris were used as outgroup taxa, as these two genera form a highly supported sister clade to Townsendia (Noyes and Rieseberg 1999). ITS sequences for these two taxa came from a published ITS phylogeny of the Astereae tribe (Noyes and Rieseberg 1999). For ETS sequencing, D N A extractions from these two species were kindly provided by R. D. Noyes. D N A extraction, PCR amplification and sequencing—Total D N A was extracted from approximately 0.5 g of leaf tissue using a DNeasy Plant Kit ™ (Qiagen, Mississauga, Canada) according to the manufacturer's protocol. A single fragment containing both ITS regions and the 5.8S cistron was PCR-amplified using the primers ITS-4 and a modification of ITS-5 (White et al. 1990), based on a sequence reported for Glycine, ITS1A (Eckenrode et al. 1985). Single 3'-ETS fragments were amplified using primers 18S-ETS (Baldwin and Markos 1998) and Ast-1 or Ast-8 (Markos and Baldwin 2001). PCR reactions consisted of 25 u.L volumes containing 20ng of genomic DNA, 30mmol/L tricine pH 8.4, 100 u M of each dNTP, 2.0 units of Taq polymerase, and 0.01 nmol of each primer. The ITS amplification conditions were as follows: (1) initial denaturation at 94°C for 3 minutes, (2) 35 cycles of 94°C for 1 minute (denaturation), 48-55°C for 1 minute (annealing), and 72°C for 1 minute (extension), (3) a final extension of 72°C for 7 minutes. The ETS was amplified according to the following schedule: (1) initial denaturation at 97°C for 1 minute, (2) 40 cycles of 97°C for 30 s (denaturation), 55°C for 20 s (annealing), and 72°C for 30 s plus 4 s per cycle (extension), (3) a final extension of 72°C for 7 minutes. Double-stranded PCR products were direct 15 purified using a PCR Purification Kit ™ (Qiagen, Mississauga, Canada) according to the protocols of the manufacturer. Sequencing reactions used the original primers, plus the internal primers ITS-2 and ITS-3 (modifications of primer sequences from White et al. (1990), nucleotide sequences available on request). Purified, double-stranded templates were sequenced in both directions using the Applied Biosystems AmpliTaq Dye Deoxy ™ terminator kit following the manufacturer's instructions, and analyzed on an Applied Biosystems 373A D N A sequencer. Chromatograms were proofread and edited manually. The resulting contigs were assembled and aligned by hand using Se-al v.2.0al 1 (Rambaut 1996) and exported into N E X U S format. The regions successfully sequenced for each accession are given in Table 2-1. Phylogenetic reconstructions— Models of molecular evolution that best fit the data were determined for the ETS region, the ITS region and the entire concatenated alignment using hierarchical likelihood ratio tests (a = 0.01) as implemented with Modeltest 3.7 (Posada and Crandall 2001) and PAUP* vAOb.10 (Swofford 2000). PAUP* was used for all subsequent analyses. Pairwise likelihood distances were calculated according to the model parameters given by Modeltest. An incongruence length difference (ILD) test was performed under parsimony to assess partition homogeneity of the concatenated ETS and ITS data sets, based on 100 repetitions (Farris et al. 1994). Maximum likelihood (ML) trees were constructed for the ETS, ITS, and combined data sets using the model parameters as given by Modeltest. The Rogers-Swofford method was employed for starting branch lengths and the starting parameters for other values were obtained using parsimony-based approximations. Heuristic search options included "as is" stepwise addition of taxa, and TBR branch swapping. Support for monophyletic groups was evaluated using 500 bootstrap replicates (Felsenstein 1985a). Heuristic searches for trees were conducted for the ETS, ITS and combined regions under maximum parsimony. Gaps were treated as a "fifth state". Multiple islands of trees were sought by performing 100 randomized stepwise additions of taxa with TBR branch swapping. Consistency indices (CI, Kluge and Farris 1969), retention indices (RI, Farris 1989), and strict consensus trees were calculated. Support for monophyletic groups was evaluated using 1000 bootstrap replicates (Felsenstein 1985a) using simple stepwise addition of taxa. Hypothesis testing—Fig. 2-1 shows Beaman's phylogeny (Beaman 1957) based on morphology and distributional data, and the presence of polyploid apomixis, acaulescence and endemism in the taxa recognized at that time. T. aprica, T. gypsophila, T. microcephala, T. nuttallii and T. smithii are taxa which have all been described subsequent to Beamn's monograph and are hence not given in Fig. 2-1. Beaman categorized taxa as endemic if they were restricted to a very 16 small area in one or two states (Beaman 1957). To test whether recurrent origins of polyploid apomixis, endemism, and acaulescence were significantly more likely than one origin, additional M L trees were generated, with and without monophyletic constraints for these three conditions as shown in Fig. 2-1. These trees were produced by defining topological constraints then selecting the "enforce topological constraints" option in PAUP* when performing likelihood searches. One-tailed Shimodaira-Hasegawa tests were performed using the resampled estimated log-likelihood (RELL) method (Shimodaira and Hasegawa 1999), with 1000 bootstrap replicates to test for significant differences between the most-likely constrained and unconstrained trees. As ITS data was missing for some important taxa that were included in Beaman's phylogeny (Table 2-1), the ETS data set alone was used (according to the model parameters that best fit the data as given under Modeltest) for all three of these tests. Associations between polyploid apomixis, acaulescence and endemism were evaluated through contingency analyses. 2.3 Results Sequence data attributes—A total of 532 bp from the ETS region and 630 bp from the ITS region were amplified and unambiguously aligned to form a concatenated alignment of 1162 bp. Due to idiosyncratic difficulties with amplification and sequencing among accessions, slightly different combinations of taxa were used for separate ETS and ITS analyses, while combined analyses comprised all samples common to both DNA sequence datasets. Pairwise likelihood distances were moderate within Townsendia and between Townsendia and the outgroups (Table 2-2). Pairwise likelihood distances were higher for the ETS region than for the ITS region, as detected in other genera (Baldwin and Markos 1998; Clevinger and Panero 2000; Markos and Baldwin 2001; Chan et al. 2002; Roberts and Urbatsch 2003). Higher pairwise likelihood distances were found for ETS, ITS and the combination of the two regions between sequences of Townsendia annua (the only annual species) and the remainder of the genus, than within the rest of the ingroup (Wilcoxon Rank Sums, p < 0.0001, Table 2-2). Pairwise sequence divergence among accessions of Townsendia was higher that that found within some other related genera: 1 % in Cirsium (Kelch and Baldwin 2003) and 1.5 % in Lessingia (Markos and Baldwin 2001), versus an average of 2.93% +/- 0.16% (SE) in Townsendia. These comparisons among genera are based on both the combined ITS and ETS regions. Parsimony analyses—Phylogenetic analysis based on the 83 parsimony-informative characters of the ETS region resulted in 440 equally most parsimonious trees of 219 steps. The strict consensus of these trees is given in Fig. 2-2, wherein CI = 0.804 and RI = 0.764. Bootstrap support was quite low for many nodes, and many unresolved polytomies were inferred. Townsendia annua 17 was well-supported as sister to the rest of the genus. This contrasts with Beaman's phylogenetic hypothesis (Fig 2-1) based on morphology and species distribution. Beaman believed that T. formosa's conical receptable and fibrous root system represented the ancestral state and polarized the relationships within the genus accordingly. Other well-supported groupings include the three morphologically similar allopatric species, T. eximia, T. grandiflora and T. texensis, a grouping likewise acknowledged by Beaman (Fig 2-1) as well as a clade that contains T. microcephala, T. minima, T. spathulata, and T. montana. Analyses using the 67 parsimony-informative characters of the ITS region resulted in 133 trees of 171 steps. The strict consensus of these trees is given in Fig. 2-3, where CI = 0.795 and RI = 0. 783. Bootstrap support was generally comparable to that found for the ETS, with little support for most internal nodes. Within the ITS parsimony consensus, T. annua was likewise well-supported as being the earliest-diverged member of the genus. The only strong bootstrap support was provided for the placement of the rare Arizona endemic T. smithii as well as T. strigosa, a biennial species that occurs locally within the Uinta basin and Green River formation of northeastern Utah and southwestern Wyoming. The incongruence length difference (ILD) test returned a highly significant value (p < 0.01), indicating that the data partition between the ETS and ITS regions is non-random and separates significantly incongruent regions. When partitions are highly heterogeneous in their properties, data are incompatible when concatenated under a simple tree-building criterion such as parsimony, which assumes identical molecular evolutionary processes over types of data (Huelsenbeck et al. 1996). A few taxa, such as the varieties of T. jonesii and T. incana (sample 1), are differentially resolved between the ETS and ITS parsimony consensus trees (Figs. 2-2 and 2-3). Homoplasy (1-CI) was slightly higher for the ITS than for ETS parsimony trees. Maximum likelihood analyses—Hierarchical likelihood ratio tests conducted on the ETS, ITS, and combined regions revealed that different models of molecular evolution best fit each data set. A variant of the H K Y model (Hasegawa et al. 1985) which allowed for unequal base frequencies (estimates gave frequency of A = 0.1927, frequency of C = 0.2178, f requency of G = 0.2478, and f requency of T = 0.3417), two categories of rate change (a constant transition : transversion ratio of 1.5067 as estimated from the data), no invariable sites, and a gamma correction of 0.4682 best fit the data from the ETS region. Phylogenetic analyses of Townsendia sequences conducted under this model returned one most likely tree, shown in Fig. 2-4, whose log-likelihood was -1886.22. A variant of the Tamura-Nei (Tamura and Nei 1993) model, which assumed equal nucleotide frequencies, two transversion rates (rate [A-C] =1, rate [A-G] = 2.6421, rate [A-T] =1, rate [C-G] = 1, rate [C-T] = 5.0121, and rate [G-T] = 1), no invariable sites and a gamma distribution of 0.2720 18 best fit the ITS data set. The one most likely tree given by these model parameters is shown in Fig. 2-5, with a log-likelihood of-1835.73. The combined ETS and ITS data set was best fit by a different variant of the Tamura-Nei model (Tamura and Nei 1993). The main difference between this and the former model is that this model allows for unequal base frequencies (estimated at frequency of A = 0.2201, frequency of C = 0.2271, frequency of G = 0.2534, and frequency of T = 0.2994}. Unequal transversion rates (rate [A-C] = 1, rate [A-G] = 2.6738, rate [A-T] = 1, rate [C-G] = 1, rate [C-T] = 4.3544, and rate [G-T] = 1) were estimated, no invariant sites were assumed, and a shape of 0.3109 was applied under this model. The one most likely tree obtained from the analysis of the combined ETS and ITS regions is given in Fig. 2-6, whose log-likelihood was -3780.71. The likelihood trees are similar to the parsimony trees, with respect to both topology and bootstrap support. Insignificant differences between the ETS parsimony consensus tree and the ETS M L tree include a poorly-supported affiliation of T. hookeri, T. exscapa, T. leptotes, T. nuttallii, and T. rothrockii with T. scapigera as a sister taxon under likelihood, while this association was unresolved under parsimony. In addition, T.florifer falls within a polytomy that includes T.formosa, T. glabella and T. strigosa under likelihood, whereas the formerly-mentioned taxon is associated with T. montana, T. parryi and T. condensata under parsimony. As neither of these differential placements is accompanied by strong branch support, it remains unclear whether these alogorithms produce conflicting topologies. The main difference between the parsimony and likelihood trees for the ITS regions is the collapse of nearly all internal nodes in strict consensus under parsimony. None of these nodes were particularly well-supported under likelihood. The M L tree that was based upon a combination of sequenced regions provided the highest levels of bootstrap support for branches and had the highest number of branches with more than 50% bootstrap support. Hypothesis testing—One-tailed Shimodaira-Hasegawa tests (Shimodaira and Hasegawa 1999) were all highly significant in favor of the trees in which monophyletic constraints on polyploid apomixis, endemism and acaulescence were relaxed. The ETS M L tree topology (Fig. 2-4), which indicated multiple origins of polyploid apomixis, was significantly more likely than the tree with just a single origin (p < 0.001). Likewise, multiple origins of endemism were significantly more likely than one origin (p < 0.001). The diminutive, acaulescent and rosulate habit was significantly more likely to have evolved multiple times than to have evolved once (p = 0.025). Although we can state with confidence that multiple origins of these three traits is significantly more likely than one origin, due to the lack of well-supported phylogenetic resolution we quite tentatively give the number of origins of polyploid apomixis, acaulescence and endemism as four, four, and six, respectively. We would also like to make it clear that the number of origins of these traits were not explicitly tested 19 against alternate numbers of origins; just monophyly versus a relaxed constraint was tested. In support of these results from the ETS dataset alone, the findings from analyses of the incomplete ITS dataset were largely congruent (results not shown). Contingency analyses indicated that there is a significant negative association between polyploid apomixis and endemism (%1 = 0.0191), while there was no significant relationship among other pairs of traits (i.e. polyploid apomixis and acaulescence, or acaulescence and endemism). 2.4 Discussion Character evolution within Townsendia—All molecular evidence strongly supports T. annua as the earliest-diverged member of the genus. This contrasts with Beaman's hypotheses of character evolution within the group, which posited that T.formosa was the ancestral type. Concordantly, many of his ideas about character evolution are unsupported by molecular data. Beaman believed that all taprooted species were derived from a fibrous-rooted ancestor. T. formosa is a the only species of Townsendia that is fibrous-rooted. The rest of the species, including T. annua, have taproots. Thus, our results support the notion that fibrous-rooting is a derived character. If not, taproots must have arisen recurrently within the genus. Additionally, the outgroup genus Dichaetophora consists of taprooted plants (Shinners 1946). Astranthium comprises plants with both fibrous and taprooted species, however phylogentic relationships within Astranthium are unknown (DeJong 1965). Hence, based on available data, Beaman's hypothesis on root evolution appears to be incorrect. Beaman also believed that "primitive" achenes were large and obovate (3-4.5mm long, 2-2.5 mm wide) with a tiny, unelaborated pappus, as found in T.formosa. Instead, T. annua has very small (2.0-2.8 long 0.6-1.2 mm wide), delicate and slender achenes with an elaborate pappus of long, plurisetose barbellate bristles. The only other species of Townsendia with a reduced pappus is T. eximia, which is a late-diverging member of the genus according to our molecular data. The outgroup genera have a pappus that is reduced or absent. Thus pappus morphogies have arisen recurrently, regardless of direction. In general, there is a trend toward an increase in the number of rows of phyllaries within Townsendia. This is is contrast to Beaman's hypothesis. T. annua has few (2-3, rarely up to 4) rows of phyllaries. T. smithii, another early-diverging member of the genus likewise has 2-3 rows of phyllaries. Species which have four or greater rows of phyllaries include T. mensana, T.jonesii, T. aprica, T. exscapa, T. hookeri, T. nuttalli, T. leptotes, T. rothrockii, T. eximia, T. texensis, T. grandiflora, T.fendleri, T.formosa, and T. parryi, all of which are later-diverging members of the genus, fn addition Dichaetophora has phyllaries in 2-3 series (Shinners 1946) while Astranthium has 20 phyllaries in 2 rows (DeJong 1965). Therefore, it appears that Beaman's idea that having many rows of phyllaries was an ancestral condition is not well-supported. This featured could be more formally analysed in a morphological reappraisal of the genus. There are other differences among Beaman's ideas of "primitive" and T. annua. T.fomosa is a perennial with erect and monocephalous stems, atopped with with large heads from conical receptacles. T. annua is an annual, with few to numerous prostrate to decumbate branching stems, small heads (8-16 mm) with convex receptacles. However, both T. annua and T.formosa, in addition to the outgroups, are indeed caulescent, congruent with Beaman's idea that the species with reduced, acaulescent habits were derived forms. Recurrent evolution'within Townsendia— A reduced, acaulecent habit has repeatedly evolved from caulescent forms in Townsendia (Fig 2-4). This appears to be concordant with Beaman's ideas. He hypothesized that these reduced forms evolved recurrently and independently, converging from caulescent ancestors as an adaptation to harsh environments. Monophyly of the reduced forms can be ruled out, making a scenario of parallel evolution possible. Further targeted studies may seek to elucidate whether indeed this recurrently evolving morphological syndrome results from parallel evolution. Polyploid apomixis has evolved repeatedly within Townsendia (Fig. 2-4). This is consistent with Beaman's hypothesis (Fig. 2-1). Multiple transitions from diploidy to polyploidy have been documented within numerous plant taxa (Soltis et al. 2003). An increase in genome size through chromosome doubling can be thought of as a frequently occurring class of mutation (Otto and Whitton 2000), however the propensity for the establishment of individuals that undergo genome changes of this nature will vary with the breeding system in question (Ramsey and Schemske 1998, 2002). In Townsendia, this shift from diploidy to triploidy and tetraploidy is apparently accompanied by a shift in reproductive mode from outcrossing to apomixis (Beaman 1957). Although some evidence suggests that components of the apomictic phenotype are under relatively simple genetic control (e.g. Noyes 2000b), the reason for the association between polyploidy and apomixis in Townsendia remains unclear. It is possible that apomixis in Townsendia may be directly induced by a ploidy shift. Alternately, as it has been demonstrated that asexual lineages can accumulate deleterious mutations at a rate of four times that found within their sexual counterparts (Paland and Lynch 2006), perhaps there is selection for polyploidy within asexual individuals, allowing for asexuals to gain a temporary advantage of a genome that is buffered by an increased number of gene copies. In Townsendia, the species that consist of both diploid sexual and apomictic polyploid populations have larger ranges than those that solely include sexual diploids. Consistent with this, we found a strong negative association between polyploid apomixis and endemism within the genus. 21 This is in agreement with results given in Bierzychudeck (1985). However, the association among polyploid apomixis and endemism in our study did not account for phylogeny. The application of a phylogenetic correction method, such as independent contrasts (Felsenstein 1985b) or the likelihood methods developed by Mark Pagel (e.g. Pagel 1992) would remove the unlikely assumption that species are statistically independent entities. These methods were not used in this chapter, as their use is contingent upon phylogenies in which there is full (i.e no polytomies) and robust (i.e well-supported) phylogenetic resolution, a condition that is not met by our data. As the taxa with the various traits are widely distributed across the trees (Figures 2-1 to 2-6), we can speculate that the effects of a phylogenetic correction might produce similar results. Within the species that comprise both sexual diploid and apomictic polyploid populations, a casual observation of distributional information as given in Beaman (1957) and Bierzychudeck (1985) indicates that the sexual diploids are, by and large, significantly more restricted in their distribution compared to the polyploid apomicts. We have not carried out a formal analysis of geographic distributions, as our knowledge of species' distributions has advanced considerably in the past 50 years, with many new populations now discovered, nearly all of unknown cytotype. The necessary field and labwork has not been carried out that would firmly establish the present day distribution of cytotypes within all 13 relevant species of Townsendia that comprise both diploid and polyploid populations. Formal analyses of geographic distributions data among cytotypes would indeed be an interesting future project. Because the effects of polyploidy cannot be disentangled from those of apomixis in Townsendia, it remains unclear whether it is the increase in genome size or attributes of asexuality that are responsible for generating this distribution pattern. Polyploids are often described as having broader ecological tolerances than diploids (Stebbins 1950; Levin 1983; but see Chamcenerion angusdfolium for an exception, Mosquin and Small, 1971). Changes in gene expression after chromosome doubling may result in functional diversification of gene copies (Adams and Wendel 2005), and this may in turn lead to polyploids having broader niches than diploids. However, it may be that asexuality is the factor that leads to this correlation (cf. geographic parthenogenesis Vandel 1931; Bierzychudeck 1985). Most of the species that only have sexual diploid members occur south of the extent of the last glacial maximum. The relationship between ranges of sexuals and asexuals and patterns of glaciation is an interesting factor that may be examined by further, more detailed studies within the genus. The lack of phylogenetic structure in Townsendia—The lack of phylogenetic resolution may be due to rapid ecological radiation, gene flow among species and/or a lack of statistical power. Species of Townsendia were hypothesized by Beaman (1957) to have undergone striking recent 22 independent radiations in substrate preference, with most species endemic to a particular soil type found within a small, localized geographic area. Indeed, we found comb-like topologies of species relationships in our rDNA sequence phylogenies, which is consistent with rapid ecological radiations (e.g Kelch and Baldwin 2003). Beaman (1957) also suggested that geographic isolation may have been the principal factor in speciation. In nature, Townsendia species exhibit a broad range of morphological variation and typically each species is restricted to a different, yet specific, edaphic situation. Limited seed dispersal may have promoted local speciation in Townsendia, resulting in closely-related species having small peripatric ranges in the highly dissected landscapes where Townsendia occurs. The lack of hierarchical phylogenetic structure might also result from gene flow among intercrossable species. In Beaman 1954, experimental F l hybrids were obtained from reciprocal crosses among seven diploid, self-sterile species of Townsendia (Beaman 1957). Al l attempted crosses were successful. Offspring from two of these crosses were grown to reproductive maturity and no abnormalities in meiosis nor pollen fertility were found among these F l plants, suggesting a lack of genetic isolating barriers. Hybridization among diploids has been suggested for numerous natural populations of Townsendia, involving 15 out of the 21 of Beaman's species (Beaman 1957). If gene flow has occurred among species then a lack of phylogenetic structure may result. The lack of phylogenetic resolution within Townsendia may also be due to a lack of statistical power. In addition to the regions analyzed in this chapter, five chloroplast regions, (the trnh intron, the trnTL intergenic spacer, the trnTF intergenic spacer, the trnKl-trnKl region and the ndhF gene) a mitochondrial intron (nad\) and a single-copy nuclear gene with introns (gapdH) were surveyed for nucleotide variation from 4-6 species within the genus Townsendia, including T. annua, with the primers and cycling conditions given in Taberlet et al. (1991), Demesure et al (1995), Jansen (1992), and Olsen and Schaal (1999). Suitable levels of nucleotide variation were not found within these surveys of Townsendia (data not shown), indicating relatively low rates of nucleotide subsitution within the genus as a whole. Availible resources did not permit further D N A sequencing. Future sequencing efforts from additional D N A regions may possibly provide additional data to help elucidate relationships within the genus. Furthermore, divergent ribosomal D N A sequence types within the hybrid genome can also undergo various degrees of recombination leading to chimeric nrDNA sequences (Baldwin et al. 1995). The impact of non-hierarchical rDNA sequence evolution on phylogenetic construction has been largely restricted to discussions of polyploids of hybrid origin (i.e allopolyploids, e.g. Wendel et al. 1995; Rausher et al. 2002), but the same processes likely occur among hybrid diploids and autopolyploids with sufficiently diverged parents. In our study, intragenomic polymorphism of these 23 regions was "averaged", i.e. treated as polymorphisms at nucleotide sites. If divergent repeats have not been homogenized, then the sequencing of multiple clones from the rDNA cistron regions might allow for more precise detection of allelic reticulation (Rausher et al. 2002). Evidence of hybridization may be found by comparing our current phylogenies with those inferred from uniparentally inherited, chloroplast or mitochondrial D N A markers, as well as other nuclear regions. Additional sequencing of replicate individuals from each species for these regions may also furnish further insight into putative cases of rDNA introgression within Townsendia. 24 2.5 References Adams KL, Wendel JF. 2005. Novel patterns of gene expression in polyploidy plants. Trends in Genetics 21: 539-543. Alvarez I, Wendel JF. 2003. Ribosomal ITS sequences and plant phylogenetic inference. Molecular Phylogenetics and Evolution 29: 417-434. Argus GW, White DJ. 1978. The Rare Vascular Plants of Alberta. National Museums of Canada: Ottawa, Canada. Babcock EB, Stebbins GL. 1938. The American species of Crepis. Carnegie Inst. Wash. Publ. No. 504: 1-199. Baldwin BG, Sanderson MJ, Porter JM, Wojciechowski MF, Campbell CS, Donoghue MJ. 1995. The ITS region of nuclear ribosomal DNA—a valuable source of evidence on angiosperm phylogeny. Annals of the Missouri Botanical Garden 82: 247-277. Baldwin BG, Markos S. 1998. Phylogenetic utility of the external transcribed spacer (ETS) of 18S-26S rDNA: congruence of ETS and ITS trees of Calycadenia (Compositae). Molecular Phylogenetics and Evolution 10: 449-463. Bayer RJ. 1997. Antennaria rosea (Asteraceae) - a model group for the study of the evolution of polyploid agamic complexes. Opera Botanica 132: 53-65. Beaman JH. 1954. Chromosome numbers, apomixis, and interspecific hybridization in the genus Townsendia. Madrono 12: 169-180. Beaman JH. 1957. The systematics and evolution of Townsendia (Compositae). Contributions from the Gray Herbarium of Harvard University 183: 1-151. Campbell CS, Wojciechowski MF, Baldwin BG, Alice LA, Donoghue MJ. 1997. Persistent nuclear ribosomal DNA sequence polymorphism in the Amelanchier agamic complex (Rosaceae). Molecular Biology and Evolution 14: 81-90. Chan R, Baldwin BG, Ornduff R. 2002. Cryptic goldfields: a molecular phylogenetic reinvestigation of Lasthenia californica sensu lato and close relatives (Compositae: Heliantheae sensu lato). American Journal of Botany 89: 1103-1 1 12. Clevinger JA, Panero JL. 2000. Phylogenetic analysis of Silphium and subtribe Engelmanniinae (Asteraceae: Heliantheae) based on ITS and ETS sequence data. American Journal of Botany 87: 565-572. DeJong DCD. 1965. A systematic study of the genus Astranthium (Compositae, Astereae). Publications of the Museum, Michigan State University Biological Series: 2L 429-528. Demesure B, Sodzi N , Petit RJ. 1995. A set of universal primers for amplification of polymorphic non-coding regions of mitochondrial and chloroplast DNA in plants. Molecular Ecology 4: 129-131. Douglas GW. 1979. The Rare Vascular Plants of Saskatchewan. National Museums of Canada: Ottawa. Douglas GW, Straley GB, Meidinger DV. 1998. Rare Native Vascular Plants of British Columbia. BC Environment: Victoria. Douglas GW, Argus GW, Dickson HL, Brunton DF. 1981. The Rare Vascular Plants of the Yukon. National Museums of Canada: Ottawa. 25 Eckenrode VK, Arnold J, Meagher RB. 1985. Comparison of the nuceotide sequence of soybean 18S rRNA with the sequences of other small-subunit rRNAs. Journal of Molecular Evolution 21: 259-269. Farris JS. 1989. The retention index and homoplasy excess. Systematic Zoology 38: 406-407. Farris JS, Kallersjo M, Kluge AG, Bult C. 1994. Testing Significance of Incongruence. Cladistics 10: 315-319. Felsenstein J. 1985a. Confidence limits on phylogenies: an approach using the bootstrap. Evolution 39: 783-791. Felsenstein J. 1985b. Phylogenies and the comparative method. American Naturalist 125: 1-15. Keller EF. 1993. A Feeling for the Organism: The Life and Work of Barabara McClintock. Freeman & Company, New York. 235 p. Hasegawa M, Kishino K, Yano T. 1985. Dating the human-ape splitting by a molecular clock of mitochondrial DNA. Journal of Molecular Evolution 22: 160-174. Huelsenbeck JP, Bull JJ, Cunningham CW. 1996. Combining data in phylogenetic analysis. Trends in Ecology & Evolution 11: 152-158. Jansen RK. 1992. Current research. Plant molecular Evolution Newsletter 2: 13. Kelch DG, Baldwin, BG. 2003. Phylogeny and ecological radiation of new world thistles (Cirsium, Cardueae -Compositae) based on ITS and ETS rDNA sequence data. 141-151. Kluge AG, Farris SJ. 1969. Quantitative phyletics and the evolution of anurans. Systematic Zoology 18: 1-32. Markos, S, Baldwin BG. 2001. Higher-level relationships and major lineages of Lessingia (Compositae, Astereae) based on nuclear rDNA internal and external transcribed spacer (ITS and ETS) sequences. Systematic Botany 26: 168-183. Mosquin T and Small E. 1971. An example of parallel evolution in Epilobium (Onagraceae). Evolution 25: 678-682. Noyes RD, Rieseberg LH. 1999. ITS sequence data support a single origin for North American Astereae (Asteraceae) and reflect deep geographic divisions in Aster S.L. American Journal of Botany 86: 398-412. Olsen K M , Schaal BA. 1999. Evidence on the origin of cassava: phylogeography of Manihot esculenta. Proceedings of the National Academy of Science, USA, 96: 5586-5591. Pagel MD. 1992. A method for the analysis of comparative data. Journal of Theoretical Biology 156: 431-442. Paland S, Lynch M. 2006. Transitions to Asexuality Result in Excess Amino Acid Substitutions. Science 311: 990-992. Posada D, Crandall KA. 2001. Evaluations of methods for detecting recombination from DNA sequences: Computer simulations. Proceedings of the National Academy of Science, USA 98: 13757-13762. Rambaut A. 1996. Se-Al: Sequence Alignment Editor. Available at http://evolve.zoo.ox.ac.uk/. Ramsey J, Schemske DW. 1998. Pathways, mechanisms, and rates of polyploid formation in flowering plants. Annual Reviews in Ecology and Systematics 29: 467-501. 26 Ramsey J, Schemske DW. 2002. Neopolyploidy in flowering plants. Annual Reviews in Ecology and Systematica 33: 589-639. Rauscher JT, Doyle JJ, Brown ADH. 2002. Internal transcribed spacer repeat-specific primers and the analysis of hybridization in the Glycine tomentella (Leguminosae) polyploid complex. Molecular Ecology 11: 2691-2702. Roberts RP, Urbatsch LE. 2003. Molecular phylogeny of Ericameria (Asteraceae, Astereae) based on nuclear ribosomal 3-ETS and ITS sequence data. Taxon 52: 209-228. Shimodaira H, Hasegawa M. 1999. Multiple Comparisons of Log-Likelihoods with Applications to Phylogenetic Inference. Molecular Biology and Evolution 16: 1114-1116. Shinners, LH. 1946. The genus Dicaetophora A. Gray and its relationships. Wrightia 1: 90-94. Soltis DE, Soltis PS, Tate JA. 2003. Advances in the study of polyploidy since Plant Speciation. New Phytologist 161: 173-191. Swofford DL. 2000. PAUP*. Phylogenetic analysis using parsimony (*and other methods). Sinauer Associates: Sunderland, USA. Taberlet P, Gielly L, Pautou G, Bouvet G. 1991. Universal primers for amlification of three non-coding regions of chloroplast DNA. Plant Molecular Biology 17: 1105-1109. Tamura K, Nei M. 1993. Estimation of the number of nucleotide substitutions in the control region of mitochondrial DNA in humans and chimpanzees. Molecular Biology and Evolution 10: 512-526. USDAand NRCS. 2006. The PLANTS Database, Version 3.5. http://plants.usda.gov. National Plant Data Center: Baton Rouge, USA. Vandel A. 1931. La parthenogenese. Doin: Paris, FR. Wendel JF, Schnabel A, Seelanen L. 1995. Bidirectional interlocus concerted evolution following allopolyploid speciation in cotton (Gossypium). Proceedings of the National Academy of Sciences, USA 92:280-284. White DJ, Johnson KL. 1980. The Rare Vascular Plants of Manitoba. National Museums of Canada: Ottawa, Canada. White TJ, Bruns T, Lee S, Taylor J. 1990. Amplification and direct sequencing of fungal ribosomal RNA genes for phylogenetics. In Innis MA, Gelfand DH, Sninsky J, White TJ, eds. PCR protocols: a guide to methods and application, 315-322. Academic Press: New York, USA. 27 Table 2-1: List of accessions and DNA regions sequenced. *ITS from Noyes and Rieseberg (1999). Sample name Taxon Source (Herbarium identification) Regions annual T. annua Beaman Coconino Co., Arizona (UTC 204100). ETS, ITS annua2 T. annua Beaman San Juan Co, New Mexico (UTC 186581). ETS, ITS aprica T. aprica Welsh & Sevier Co, Utah. L. Jennings ETS, ITS Reveal condensata T. condensata Parry Prairie Bluff, Alberta. J. Whitton and J. Bain. ETS, ITS ex Gray eximia 71 eximia A. Gray Taos Co, New Mexico (MSU 162690). ETS, ITS exscapa 71 exscapa Douglas Co, Colorado. Rocky Mountain Rare Plants. ETS, ITS (Richards.) Porter fendleri 71 fendleri A. Gray Johnson Village, Colorado. K. Dlugosch, V. ETS, ITS Pasqualetto, and J. Whitton. florifer 71 florifer (Hook.) A. Beaverhead Co, Montana (MSU 289268). ETS, ITS Gray formosa T. formosa Greene Locality unavailable. Rocky Mountain Rare Plants. ETS, ITS glabella 71 glabella A. Gray Pagosa Springs, Archuleta Co, Colorado. K. ETS, ITS Dlugosch, V. Pasqualetto, and J. Whitton. grandifloral 71 grandiflora Nutt. Roxborough State Park, Colorado. A. Chang, V. ETS, ITS Pasqualetto, and J. Whitton. grandiflora2 71 grandiflora Nutt. Laramie Co, Wyoming (MSU 164050). ITS gypsophila 71 gypsophila Lowrey Sandoval Co, New Mexico (MSU 345690). ETS, ITS & Knight hookeri 71 hookeri Beaman Boulder Co, Colorado. S. Thompson and K. Dlugosch ETS, ITS incanal 71 incana Nutt. Colorado National Monument, Glade Park, Utah. A. ETS, ITS Chang, V. Pasqualetto and J. Whitton. incana2 71 incana Nutt. Fremont Co, Wyoming. Rocky Mountain Rare Plants. ETS jonesii jonesii 71 jonesii (Beaman) Sanpete Co, Utah (UTC 134956). ETS Reveal var. jonesii jonesii lutea 71 jonesii (Beaman) Sevier Co, Utah (UTC 209054). ETS, ITS Reveal var. lutea Welsh jonesii 71 jonesii (Beaman) Clark Co, Nevada (UTC 122145). ETS, ITS tumulosa Reveal var. tumulosa Reveal leptotes 71 leptotes (A. Gray) Cumberland Pass, Gunnison Co, Colorado. A. Chang, ETS, ITS Osterhout V. Pasqualetto and J. Whitton. mensana 71 mensana M. E. Duchesne Co, Utah (MSU 156407). ETS, ITS Jones mexicana 71 mexicana A. Gray Nuevo Leon, Mexico (MSU 170642). ETS microcephala 71 microcephala Sweetwater Co, Wyoming (RM 643151). ETS, ITS Dorn minima 71 minima Eastwood Locality unavailable. Rocky Mountain Rare Plants. ETS montanal 71 montana M. E. Big Horn Mountain, Wyoming. Rocky Mountain Rare ETS, ITS Jones Plants. montana2 71 montana M. E. San Pete Co, Utah (MSU 162683). ETS, ITS Jones nuttalli 71 nuttalli Dorn Locality unavailable. Rocky Mountain Rare Plants. ETS, ITS parry i 71 parryi D. C. Eat. Lethbridge, Alberta. Coll. J. Whitton and J. Bain ETS, ITS In Parry rothrockii 71 rothrockii A. Gray Locality unknown. Rocky Mountain Rare Plants. ETS, ITS ex Rothrock scapigera 71 scapigera D. C. Mono Co, California (JEPS 089359). ETS Eat. in S. Wats. 28 Table 2-1: continued Sample name Taxon Source (Herbarium identification) Regions sinbadensis T. spp. "Sinbad Sinbad Valley Colorado. Ronald Hartman ITS Valley" smith ii T. smith ii Shultz & Mohave Co., AZ (UTC 150000). ETS, ITS Holmgren spathulatal T. spathulata Nutt. Locality unknown. Rocky Mountain Rare Plants. ETS, ITS spathulata2 T. spathulata Nutt. Locality unknown. Rocky Mountain Rare Plants. ITS strigosa T. strigosa Nutt. Grand Junction, Colorado. K. Dlugosch, V. ETS, ITS Pasqualetto, and J. Whitton. texensis T. texensis Larsen Randall Co., Texas (UTC 153220). ETS, ITS astranthium Astranthium Arkansas, USA. D. E. Boufford. ETS, integrifolium ITS* (Michx.) Nutt. dichaetophora Dichaetophora Nuevo Leon, Mexico. G. Neesom. ETS, campestris A. Gray ITS* Table 2-2 - Summary of pairwise likelihood distances (range, %). Grouping ETS ITS ETS + ITS Within ingroup (minus T. annua) 0.00-6.78 0.00-5.09 0.00-6.78 Between ingroup (minus T. annua) and T. annua 8.16-12.63 5.01-6.82 8.76-13.30 Between ingroup (minus T. annua) and outgroup 14.60-23.65 9.40-12.81 24.52-44.72 Between T. annua and outgroup 18.13-23.92 10.40-12.22 32.23-46.67 Within outgroup 15.07 3.37 12.38 29 I • — — — — formosa (endemic) I annua ' Strigosa (polyploid apomixis, endemic) ' mexicana (endemic) 1 fendleri (endemic) ^ ' incana (polyploid apomixis) I condensata (polyploid apomixis, reduced) I spathulata (polyploid apomixis, reduced, endemic) I florifer I s c a p i g e r a (polyploid apomixis, reduced) ' parryi (polyploid apomixis) I exscapa (polyploid apomixis, reduced) I hookeri (polyploid apomixis, reduced) ' leptotes (polyploid apomixis, reduced) . I glabella (reduced, endemic) ' rothrockii (polyploid apomixis, reduced, endemic) I I jonesii jonesii (reduced, endemic) I mensana (reduced, endemic) I minima (polyploid apomixis. reduced, endemic) ' m o n t a n a (polyploid apomixis. reduced) . eximia (endemic) _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ grandiflora (polyploid apomixis) texensis (endemic) Figure 2-1. Beamans' hypothesis of phylogenetic relationships within Townsendia based on informal, non-analytical a priori reasoning (Beaman 1957) indicating the species for which polyploid apomixis is known in addition to sexual diploidy (polyploid apomixis), acaulescent, reduced habit (reduced), and a restricted, endemic distribution, defined by Beaman (1957) as being found in a relatively tiny geographic area within one or two states of the United States of America (endemic). 30 100 J 0 P _ 72 59 84 93 84 58 54 90 •annual •annua2 • aprica • jonesii jonesii • jonesii lutea • jonesii tumulosa • mensana jexscapa, hookeri, leptotes, lnuttallii, rothrockii • mexicana • scapigera • eximia • grandifloral -texensis . incana2 • fendleri • incanal -gypsophila .formosa .glabella -strigosa .florifer • montanal •condensata, parryi Jmicrocephala, minima, jspathulatal . montana2 . smithii .astranthium -dichaetophora Figure 2-2. Strict consensus of 440 equally-parsimonious trees of 219 steps, based on the external transcribed spacer (ETS) of Townsendia. Bootstrap support for nodes (a 50%) is indicated above branches. 31 -lqq_ -LOO. _24_ _2fJ_ _7_2_ 35. .51 _54_ •annual •annua2 -aprica -exscapa -hookeri, nuttallii -mensana - leptotes -sinbadensis -condensata -florifer -jonesii lutea -jonesii tumulosa - parry i -glabella -eximia -grandiflora(1,2) -texensis -fendleri -gypsophila -formosa -incanal -microcephala -montanal - m o n t a n a 2 -rothrockii -spathulatal -spathulata2 -strigosa -smithii -astranthium -dichaetophora Figure 2-3. Strict consensus o f 133 equally-parsimonious trees of 171 steps, based on the internal transcribed spacer (ITS) o f Townsendia. Bootstrap support for nodes (a 50%) is indicated above branches. 32 0.01 substitutions/site 100 69 100 53 _ annual annua2 59 511 521 a P r i C a I 'jonesii lutea jonesii jonesii (re) jonesii tumulosa I — m e n s a n a (re) exscapa (ar), hookeri (ar), leptotes (ar), nurtallii, rothrockii (are) scapigera (ar) r - d i 76 mexicana (e) f-eximia (e) texensis (e) Igrandifloral (a) Uncana2 (a) 7 8 | - fend le r i (e) 8 5 L j n c a n a l (a) 53 gypsophila florifer — f o r m o s a (e) glabella (re) strigosa (ae) microcephala, minima (are), spathulatal (are) montana2 (ar) |—condensata (ar), parryi (a) montanal (ar) smithii -astranthium -dichaetophora Figure 2-4. Maximum likelihood tree (-InL = 1886.22) based on the external transcribed spacer (ETS) of Townsendia. Bootstrap support for nodes (& 50%) is indicated above branches. Character states are indicated in parentheses: a = polyploid apomixis is known, r = reduced, acaulescent habit, e = endemic taxon. See text for further descriptions. 100 0.005 substitutions/site 100 56 55. -annual 71 88 annua2 aprica g^exscapa hookeri, nuttalli -mensana jgol leptotes I—sinbadensis -eximia grandiflora(1,2) -texensis •formosa -microcephala rothrockii —• spathulatal spathulata2 condensata ,—florifer — jonesi i lutea jonesii tumulosa L-parryi glabella |-montana1 incanal montana2 • .fendleri gypsophila strigosa smithii .astranthium dichaetophora Figure 2-5. Maximum likelihood tree (-InL = 1835.73) based on the internal transcribed spacer (ITS) of Townsendia. Bootstrap support for nodes (s= 50%) is indicated above branches. 34 0.005 substitutions/site 100 92 . astranthium • dichaetophora 100 c annual annua2 55 7 8 • aprica — mensana PL 86! exscapa hookeri, nuttallii leptotes rothrockii 51 i - microcephala II spathulatal - montana2 Lmontanal 63p condensata I L parryil 14-florifer jonesii lutea 7 9rC 98 8 2 -jonesii tumulosa r eximia texensis rC grandifloral -formosa fendleri incanal gypsophila — glabella , strigosa smithii Figure 2-6. Maximum likelihood tree (-InL = 3780.71) based on the combined analysis of the external and internal transcribed spacer (ETS and ITS) of Townsendia. Bootstrap support for nodes (a 50%) is indicated above branches. 35 3 Patterns of recurrent evolution and geographic parthenogenesis within apomictic polyploid Easter daises (Townsendia hookeri^ 3.1 Introduction Geographic parthenogenesis (Vandel 1931) describes a pattern wherein asexual organisms have a distribution that is more northerly, more widespread and/or at higher elevations than their sexual counterparts. This common pattern can be found throughout numerous plant and animal taxa (reviewed in Lynch 1984,,Bierzychudek 1985). Due to its broad implications for understanding why sex is maintained in nature, geographic parthenogenesis has received theoretical attention (e.g. Glesener and Tilman 1978, Lynch 1984, Gaggiotti 1994, Peck et al. 1998) yet few empirical studies go beyond a basic description of the pattern (e.g. Law and Crespi 2002, Verduijn et al. 2004). Both theoretical and empirical treatments of geographic parthenogenesis must allow that sexual and asexual organisms typically differ in ploidy (Bierzychudek 1985). As approximately 99% of known apomictic plants (which reproduce asexually through seed) and most asexual animals are polyploid (Suomalainen et al. 1987, Asker and Jerling 1992), geographic patterns of breeding system are best approached with consideration of this additional layer of complexity. Polyploidy, the condition of possessing more than two chromosome sets, results from a pervasive and frequently occurring class of mutation in ferns and flowering plants and is sporadically found in other groups of plants and animals (reviewed in Stebbins 1950; Grant 1981; Soltis et al. 1992; Soltis and Soltis 1993; Soltis and Soltis 2000; Otto and Whitton 2000; Soltis et al. 2003). A well-established aspect of the biology of polyploid plants, typified by molecular work on Tragopogon (Roose and Gottlieb 1976; Soltis and Soltis 1989; Soltis and Soltis 1991; Soltis et al. 1995; Cook et al. 1998), is that the transition from diploidy to polyploidy can occur multiple times within taxa. Over 45 studies have documented the multiple origins phenomenon (reviewed in Soltis and Soltis 1993, 1999), leading to the conclusion of Soltis and Soltis (1993) that the pattern of multiple origins of polyploidy is clearly "the rule, rather than the exception". The Rocky Mountain genus Townsendia (Asteraceae) is an interesting system for documenting multiple origins of polyploidy because within this genus autopolyploidy goes hand in hand with asexuality through apomixis (although apomicts are overwhelmingly polyploid, the corollary is not true for plants - most polyploid plants retain their capacity for sexuality). Through detailed cytogenetic and embryological investigations, Beaman (1957) discovered a tight association between polyploidy and apomixis. In polyploid individuals of Townsendia, unreduced female 1 A version of this chapter has been be accepted for publication at Molecular Ecology as: Thompson SL, Whitton J. 2006. Patterns of recurrent evolution and geographic parthenogenesis within apomictic polyploid Easter daises (Townsendia hookeri). 36 gametophytes are produced through a meiotic abnormality in the megasporocyte (diplospory), followed by parthenogenetic development of the embryo that does not require pollen for embryo or endosperm initiation (i.e. non-pseudogamous parthenogenesis; see Nogler 1984). Diploid populations were reported to be invariantly sexual. Sexual diploid and apomictic polyploid plants are nearly identical morphologically, but can be distinguished based on differences in pollen size and viability Sexual diploids have small pollen grains, while apomictic polyploids have large pollen grains and a high percentage of inviable pollen (Beaman 1954; Beaman 1957). Similar observations have been noted in other apomictic polyploid taxa (reviewed in Nogler 1984, Asker and Jerling 1992). The focus of this study, the Easter daisy (Townsendia hookeri), has a distribution that appears consistent with geographic parthenogenesis (Vandel 1931; Bierzychudek 1985). Diminutive and relatively widespread, its distribution straddles the montane regions of the Rocky Mountains, from the front ranges of Colorado in the south, to the northern populations of Alberta and British Columbia. Pollen studies, combined with chromosome counts and embryology, have indicated that sexual diploid populations occur exclusively in the southern part of the species range in Colorado and Wyoming, while apomictic autotetraploids occur northward (Beaman 1957). Since Beaman's work, Easter daisy populations of unknown cytotype have been discovered in Canada's Yukon Territory, disjunct by over 1400 km from the rest of the range (Douglas et al. 1981). As the Wisconsin glaciation has played a large role in determining extant plant distribution patterns in northern temperate regions (Hewitt 2000; Abbott and Brochmann 2003), these Yukon populations of T. hookeri provide an opportunity to examine patterns of post-glacial migration in the context of geographic parthenogenesis, as populations can be found in and adjacent to two known ice-free areas: Beringia, the unglaciated region of the Yukon Territory to the north, and the large southern refugium of the continental USA (Dyke and Prest 1987). Whether recolonization of the post-glacial landscape by Easter daisies has occurred from northern or southern ice-free areas has remained a mystery. Here, we examine patterns of ploidy and breeding system in the Easter daisy using variation in non-recombining, uniparental chloroplast (cp)DNA to address the following questions: (1) What is the minimum number of transitions from sexual diploidy to apomictic autoploidy? (2) Is there evidence that populations survived the Wisconsin glaciation in a northern refugium? (3) Do multiple transitions from sexual diploidy to apomictic polyploidy provide insight into the pattern of geographic parthenogenesis? 3.2 Materials and methods Materials— 40 populations of Townsendia hookeri from throughout its range were sampled for this study (Table 3-1, Fig. 3-1). Leaf samples were collected in the field, frozen immediately in 37 liquid nitrogen, then subsequently stored at -80°C. Additional plants were dried for pollen examination and voucher specimens, and deposited at the University of British Columbia Herbarium (UBC) (accession numbers available upon request). Pollen observations—We used differences in pollen size and viability to characterize ploidy and breeding system of poulations of T. hookeri. Cotton blue (lactophenol) staining (Maneval 1936), a common index of pollen viability that correlates well with vital enzymatic assays (e.g. Mayer 1991), was used to establish the upper bound of pollen viability (Stanley and Linskens 1974). Percent stainable pollen per 200 grains was determined based on deep and even cytoplasmic staining from pooled florets from each of 3-5 individuals from each population. The cytoplasmic diameter of 10 stained pollen grains, if pollen was present, was measured using a Leitz D M R B light microscope with ocular micrometer. Means and ranges of both pollen diameter and stainability were determined and the correlation between the population means of the two variables was assessed by Spearman's Rho using JMP IN 4.0.4 (SAS Institute Inc., Cary, NC, USA). Chloroplast haplofyping—Total D N A was isolated by C T A B (hexadecyltrimethylammonium bromide) protocol (Doyle and Doyle 1987) from approximately 0.1 g of frozen leaf tissue with the following modifications: volumes were reduced for extraction in 1.5 mL microfuge tubes and sodium metabisulfite (1% w/v) was added to the isolation buffer. Isolated DNA was quantified by a Hoefer DyNAQuant™ 200 fluorometer (San Francisco, USA) according to the manufacturer's directions, diluted to 10 ng/uL, then used directly in PCR amplifications. To assess levels of sequence variation, nine candidate regions were screened for 2-5 individuals from 4 populations. These included four chloroplast spacers, one chloroplast coding region, two mitochondrial genes and two non-coding regions of the nuclear ribosomal repeats (details available on request). Based on these results, two chloroplast regions were PCR amplified from a single individual from each populations using the primer pairs trnK.\/trnK2 (Demesure et al. 1995) and ndhF\/ndh¥\4 (Jansen 1992). Single fragments of approximately 3000-4000 bp were obtained from 25 uL reactions containing 20ng of genomic D N A , 30mM tricine pH 8.4, 2mM M g C l 2 , 50mM KC1, 100 u.M dNTPs (equimolar ratio), 2.0 units of taq polymerase, and 0.01 nmol of each primer. Amplification was performed by an MJ Research PTC-100 thermocycler (Watertown, USA) according to the following conditions: initial denaturation at 94°C for 4 minutes; 30 cycles of 92°C for 45 seconds, 50-55°C for 45 seconds, and 72°C for 3 minutes; with a final 10 minute extension at 72°C. Double-stranded PCR products were purified using a PCR Purification Kit™ (Qiagen, Mississauga, Canada) according to the protocols of the manufacturer. Sequencing reactions were performed in the 3'-direction using the original amplification primers trnK.2 and ndh¥\4, amplifying the 3'-rr«K intron and the 3'-end of the ndhF gene and adjacent spacer region, respectively. Internal 38 primers were designed, and 5'-sequencing was accomplished with ndh¥\4R: 5 ' -CCA CCC TTT CTT TCT ATT CCG-3 ' and fr«K2R: 5'-GAT CTA TCT A G C CCT A A A T A G C-3'. A l l sequencing reactions used Applied Biosystems BigDye™ v3.1 terminators and were run on an Applied Biosystems 377 D N A sequencer (Foster City, USA), all according to the manufacturer's directions. Sequence alignment and phytogeny reconstruction— Proof-read sequences were unambiguously aligned by eye using the program SeqApp 1.9a (Gilbert 1994). PAUP* version 4.0b 10 (Swofford 2000) was used in all subsequent phylogenetic analyses. Partition heterogeneity tests were performed to determine the evolutionary compatibility of the ndh¥ and trnK data sets, based on 100 repetitions (Farris et al. 1995) under parsimony, as described subsequently. Exhaustive searches for trees were conducted under maximum parsimony. Gaps in the aligned sequences were treated as "missing data". Consistency Indices (CI) (Kluge and Farris 1969) and Retention Indices (RI) (Farris 1989) were calculated. Support for monophyletic groups was evaluated using 1000 bootstrap replicates (Felsenstein 1985). The model of molecular evolution that best fit the data was determined using MODELTEST (Posada and Crandall 2001). Nucleotide sequence divergence was calculated according to the Hasegawa Kishino Yano (HKY85) model (Hasegawa et al. 1985). A maximum likelihood (ML) tree was constructed using the HKY85 model of sequence evolution (Hasegawa et al. 1985). The starting values for other parameters were obtained by using parsimony-based approximations. No molecular clock was enforced; the Rogers-Swofford method was employed for starting branch lengths. Heuristic search options included random stepwise addition of taxa (100 repetitions), and TBR branch swapping. Support for monophyletic groups was evaluated using 1000 bootstrap replicates (Felsenstein 1985). To test whether multiple origins of polyploid apomixis were significantly more likely than one origin, additional M L trees were generated with each accession categorized as diploid or polyploid, and analyzed with and without a monophyletic topological constraint on sequences from polyploid apomicts. The Kishino-Hasagawa test (KH test) (Kishino and Hasegawa 1989) was performed to test for significant differences between the likelihoods of constrained and unconstrained trees. Analyses of regional diversity—Haplotypes were partitioned into three geographic areas: northern (Yukon territory), central (glaciated southern Canada), and southern (south of the glacial maximum at 18 kya according to Dyke and Prest 1987). Mean number of pairwise nucleotide differences was calculated within each region and a Tukey-Kramer HSD test for all comparisons was performed with JMP IN 4.0.4 (SAS Institute Inc., Cary, NC, USA) to test for significant differences among geographic areas. Nucleotide heterozygosity (the probability that a nucleotide site differs between two individuals from the same region, Tajima 1989) was calculated within each region, using 39 the observed nucleotide frequencies for each region. The statistical error of heterozygosities was found using the bootstrap method, where the 17 variable nucleotide positions were resampled with replacement to create replicate datasets. 100 bootstraps were performed and the distribution of bootstraps used to evaluate the significance of among-region differences. Migration among populations—Our limited sampling of individuals within populations leaves an open question about the structure of haplotype variation among populations. While we were unable to address this issue range-wide, we attempted to gain insight into potential patterns of seed migration using three proximal populations from Jasper National Park (populations 61, 62 and 63). From observed chloroplast haplotype sequence variation, we identified an RFLP polymorphism in ndh¥ that could be used to differentiate haplotype groups identified in these populations. A segment of the ndhY gene was amplified using the primer pairs ndh¥\4 and «d/zF14R, according to the cycling parameters described above, from 20 individuals from each population. PCR products were then digested with the restriction enzyme Hinfl for 6 hours at 37°C in 20uL reactions containing: 5uL of PCR product, 2uL of Buffer #2 (New England Biolabs, Mississauga, Canada), 2u.g of bovine serum albumin, and 1U of Hinfl (New England Biolabs, Mississauga, Canada). Digested fragments were separated by agarose electrophoresis and visualized. 3.3 Results Ploidy and breeding system designation—Measurements of the diameter of pollen cytoplasm (Table 3-1) fell into two non-overlapping sets of values, allowing us to partition the collections into 15 diploid and 25 polyploid populations (designated in Table 3-1 and Fig. 3-1). Diploid populations had mean pollen diameters measuring 20 u M (range = 17-23 uM), while polyploid populations had mean pollen diameters of 30.5 (range = 24-37 uM). This non-overlapping distribution was consistent with Beaman's reports of pollen size differences among cytotypes (Beaman, 1954, 1957). Our ploidy designations for populations based upon pollen diameter were consistent with preliminary results from flow cytometry (S. L. Thompson, personal observation). Eight of our forty populations were collected from locations at or very nearby those previously examined by Beaman (1957) and our cytotype designations for these populations were identical to his. It is noteworthy that two populations that we characterize as diploids (populations 68 and 92) are found in the Yukon Territory, thus adding a disjunct northern element to the known range of diploids in 71 hookeri. Our classification of diploids in the Yukon agrees with a reported chromosome count of 2n = 2x = 18 by T. Mosquin (Love 1968). According to prior observations (Beaman 1954, 1957) pollen produced by sexual diploid plants was of high stainability, while the stainability of pollen from apomictic polyploids was lower. 40 Our results show that the mean pollen stainability of diploids was 72.74 +/- 5.60 (95% range), while the mean pollen stainability of polyploids was 17.49 +/- 2.70 (95% range). Mean pollen diameter (a bimodally-distributed continuous variable, here used to designate ploidy) highly correlated with pollen stainability (a non-normal continuous variable that indicates apomixis) (Fig. 3-2, Spearman's rho = -0.7133, p < 0.0001), supporting a tight association between ploidy and breeding system for Townsendia hookeri. In four populations from the Yukon territory (populations 66, 67; 69 and 71), it was readily observed that plants fail to produce pollen. Instead their anther development is aborted prematurely. These populations have been categorized as apomictic polyploids, and have been confirmed as polyploid through flow cytometry methods as part of a deeper study of genetic variation within Yukon populations (Chapter 4). Pollen from population 18 (the outlier in Fig. 3-2) appears to be another anomaly, with mean pollen diameter within the range of diploids and a high degree of inviable pollen. The diploid status of this population has been confirmed through preliminary flow cytometry results (data not shown). While pollen viability in this population was within the range for populations classified as apomicts, we take a conservative approach here and tentatively classify this population as a sexual diploid. Diploid apomicts are exceedingly rare in plants, and unknown in Townsendia, and further investigation is required to make such a designation (i.e. direct examination of ovule development or genetic analysis of progeny arrays). In categorizing this population as a sexual diploid, we weaken the association between pollen diameter and pollen viability, and thus our categorization is conservative. Chloroplast haplotyping—Our sequence alignment comprised 666 bp of the trnK intron (GenBank accesion numbers AY623667-AY623703) and 623 bp of the ndhY gene, which contained 544 bp of the coding region, and 79 bp of the adjacent spacer region (all variable nucleotides were within the coding region) (GenBank accession numbers AY623706-AY623742). The 40 samples comprised 9 trnK haplotypes and 5 ndhY haplotypes. As the partition heterogeneity test revealed no significant differences between D N A regions (p = 1.000), the sequence data were combined to yield a data set with a grand total of 1289 bp. Ten haplotypes could be distinguished from the combined data based on 17 nucleotide polymorphisms (Table 3-2). Of these 10 haplotypes, one haplotype was found exclusively in one sexual diploid population, 4 haplotypes were found in both sexual diploid and apomictic polyploid populations, while 5 haplotypes were found solely in apomictic polyploid populations. MODELTEST revealed that the HKY85 model of sequence evolution (with a transition to transversion ratio of 2.198, unequal base frequencies with freqA = 0.3914, freqC = 0.1512, freqG = 0.1516, freqT = 0.3058, no invariable sites and equal rates at variable sites, cf. Hasegawa et al. 1985) 41 best fit the data. According to this model, pairwise nucleotide sequence divergence was 0-0.799% within Townsendia hookeri for these chloroplast regions. Phylogenetic analyses— Of 17 variable sites detected, 10 proved to be parsimony informative. Exhaustive searches uncovered one island of 2 most parsimonious trees of 17 steps. The consensus of the two parsimony trees (Fig. 3-4) is identical in topology to one of the two parsimony trees. In the parsimony trees CI=R1=1, revealing no homoplasy in the data set. Under the criterion of maximum likelihood, one island of one most likely tree was found, with an estimated LnL = -1821.4055 (Fig. 3-4). This likelihood tree was identical in topology to the consensus of the two parsimony trees. Under both parsimony and likelihood bootstrap values were low, as found in other intraspecific plant phylogenies (Segraves et al. 1999; Sharbel and Mitchell-Olds 2001; Holderegger and Abbott 2003). Only one highly supported branch with 100% bootstrap support was found. An examination of tree topology reveals that this well-supported branch separates some northern and north-central apomictic polyploid populations (populations 61, 63, 66, 67 and 71) from the rest of the assemblage. A one-tailed K H test was highly significant (p < 0.0001) in favor of the tree wherein the apomictic polyploids were not topologically constrained to be monophyletic, (LnL=-1828.33 vs -1853.95), demonstrating that multiple origins of apomictic polyploidy is significantly more likely than one origin. Diversity analyses of geographic regions— The mean number of nucleotide differences varied among geographic regions. Southern populations had a mean of 1.62 +/- 0.10 (SE), central populations had a mean of 4.05 +/- 0.29, while northern populations had a mean of 4.40 +/- 1.05. Values were found to be significantly different between southern and central regions, and southern and northern regions (Tukey-Kramer HSD test: p = 0.05). Nucleotide heterozygosity for the southern populations was 0.117, for the central populations was 0.223, and for northern populations was 0.209. Bootstrap resampling of the data set showed that the southern region had significantly lower nucleotide heterozygosity (p = 0.01) than both the central and northern populations, while central and northern populations did not differ in levels of heterozygosity. Migration among populations—The two haplotypes found in populations 61-63 from Jasper National Park could be differentiated based on banding patterns resulting from a single restriction site difference (partial results presented in Fig. 3-3). In our deeper survey of 20 individuals per population, no restriction site variation was detected within populations, with all individuals from population 62 lacking a restriction site present in all individuals of populations 61 and 63. The close geographic proximity of these populations (13 km separate population 61 from 62, and less than 4 km separate 62 and 63) suggests that overall levels of seed flow among populations are likely negligible. 42 3.4 Discussion Multiple origins of apomictic polyploidy—Our phylogenetic analysis indicates that a minimum of four transitions to apomictic polyploidy have occurred in the Easter daisy - one for each of four haplotypes shared by both diploids and polyploids. As all apomictic polyploid lineages have likely been historically derived from sexual diploid lineages (i.e., the combined loss of sex and increase in genome size is likely irreversible), the possibility of fewer than four origins is improbable. If there were actually fewer than four independent origins, then identical haplotypes would have originated independently in both sexual diploid and apomictic polyploid lineages, making shared nucleotide polymorphisms cryptically homoplastic. We find no direct evidence of multiple hits in our data set (i.e. no inferred homoplasy; Table 3-2), and while it remains possible that multiple hits have masked additional origins of haplotypes, the low levels of divergence detected suggest that additional mutations are unlikely to have produced cryptic homoplasy. Also, the overall low levels of pairwise sequence divergence (0-0.799%) support the notion that the four haplotypes represent four distinct origins. Additionally, if we compare the best tree allowing for multiple origins (Fig. 3-4) with the best tree in which we force the apomictic lineages to be monophyletic, the tree with multiple origins is significantly more likely (KH testp < 0.0001). This study furnishes yet another example of polyploidy arising repeatedly from diploid progenitors. This phenomenon has been documented in over 45 taxa (Soltis et al., 2003), however, few studies have applied a phylogenetic framework to the documentation of recurrent polyploid origins (e.g. Brochmann et al. 1998; Segraves et al. 1999; Sharbel and Mitchell-Olds 2001, Joly and Bruneau 2004). The vast majority of studies on multiple origins of polyploidy have relied upon population genetic markers such as allozymes and RAPDs to deduce the number of transitions based on extant genotypes in diploids and polyploids. However, several processes can complicate the interpretation of the number of polyploid origins from multilocus genotype data, and this is especially true in sexual polyploids capable of interbreeding following polyploidization. Novel polyploids can produce extensive genetic diversity in very few generations, through mechanisms that are not entirely understood (Osborn et al., 2003), but that include gene conversion, chromosome rearrangements, mitotic recombination and point mutations (Song et al. 1995). Bands that appear or disappear due to these post-polyploidization processes cannot be readily differentiated from those that result from direct inheritance from diploid progenitors. Additionally, a phylogenetic approach leads to more conservative estimates of transition frequency, as only polyploid lineages that have been derived from diploid progenitors, rather than from other polyploid lineages, will be counted. In Townsendia hookeri, the transition to polyploidy is accompanied by a transition to apomixis. Thus, confounding effects resulting from mating among polyploids should be minimized. Even in sexual systems, using 43 non-recombining, maternally inherited chloroplast D N A for phylogenies, we would expect the effect of post-polyploidization gene flow to be limited to chloroplast introgression, which might not affect inferences about the number of origins, unless a rare cpDNA haplotype representing a novel polyploid origin was replaced by a more common haplotype. In this example, loss of haplotypes would again result in an underestimate of the number of polyploid origins. Whether it is better to provide a conservatively low or high estimate of the number of transitions to polyploidy may be context dependent. When evaluating the frequency of this class of mutation, it is our feeling that estimating the minimum number of transitions is most conservative. What is noteworthy here is that even within an autopolyploid complex where the diploids have relatively small ranges, sufficient chloroplast variation occurs to allow us to use phylogenetic approaches to detect multiple origins of apomictic polyploidy (Fig. 3-4). An advantage of the use of chloroplast D N A is that recombination typically does not occur among these molecules (but see Marshall et al. 2001), rendering genetic interpretation of multiple origins straightforward. Several investigations of multiple origins have employed cpDNA markers in allopolyploids (e.g. Soltis and Soltis 1989; Doyle et al. 1990; Song and Osborn 1993; Sharbel and Mitchell-Olds 2001), however, few studies of multiple origins of autopolyploids have used cpDNA variation (but see Soltis et al. 1989; Segraves et al. 1999). Although logic would suggest that searches for intraspecific genetic variation would be most fruitful within non-coding introns and spacer regions (Soltis and Soltis 1998), this study demonstrates that protein-coding genes can contain sufficient variation to address intraspecific questions. The repeated origin of polyploids has several demographic and ecological implications both within and among populations. On a local scale, a neopolyploid initially establishes from within a diploid population, where the minority polyploid cytotype is at a mating disadvantage (Levin 1975). Multiple polyploid origins within populations can play a strictly demographic role: the more times that a switch in ploidy occurs out of a common diploid pool, the more likely the establishment of the novel cytotype, as there will be a higher chance of mating among polyploids. Additionally, to the extent that adaptation relies on genetic variation, when recombination occurs among polyploids that have been independently-derived within a population, the production of novel genetic combinations may facilitate ecological shifts, which in turn promote polyploid establishment (Ramsey and Schemske 2002). On a larger scale, gene exchange can also occur among independently-derived polyploid populations (Doyle et al. 1999), and multiple origins may have broad impacts on the subsequent diversification of polyploid lineages. The mating system of recurrent polyploids will determine their capacity for gene exchange and thus their potential for collective evolution. In the Easter daisy, the transition to polyploidy is 44 accompanied by a switch from sexual to asexual reproduction (although whether this switch is immediate or occurs over time remains unknown). This switch in reproductive mode reduces or eliminates the chances for gene exchange between polyploids and diploids, decreasing minority cytotype effects and thus promoting establishment. As well, for an asexual polyploid, with its decreased capacity for recombination, it is the frequency of recurrent polyploidization that will largely dictate the probability of hitting upon genotypes that prove to be successful for establishment and persistence. Geographic patterns of ploidy and breeding system—The pattern in Easter daisies best fits a story where a formerly broad range of sexual diploids was broken by the Wisconsin glaciation, with a recolonization of the post-glacial landscape by apomictic polyploids, independently arising from both Beringia in the north, and from the southern refugium of the continental United States. In this section, we discuss the disjunct distribution pattern of sexual diploids, the recolonization of formerly-glaciated North America by recurrently-derived asexual polyploids as well as the implications for the broader biological question: what causes patterns of geographic parthenogenesis? This study provides the first documentation of a disjunct distribution pattern of sexual diploid Easter daisies. There are two phylogeographic scenarios that seek to explain patterns of disjunction: vicariance and dispersal (Avise 2000). Based on the evidence at hand, we find vicariance to be the more plausible mechanism for producing the observed disjunction of diploids. Firstly, the distance among northern and southern sexual diploid populations of Easter daisies exceeds 2900 km. While distance alone does not provide evidence for vicariance versus dispersal, when combined with other observations, dispersal seems unlikely to have produced the observed pattern. Easter daisies are though to have highly localized seed migration, due to their diminutive size and stemless, prostrate habit in which the maximum floral height is typically less than 4 cm from the ground. Limited seed dispersal is also supported by our intrapopulational chloroplast haplotype data from geographically proximal populations. Secondly, the sexual diploid haplotype found within the Yukon was undetected among southern sexual diploid populations, and we speculate that it arose further north than the current distribution of southern diploids, perhaps migrating northward following its origins. It is possible that northern sexual diploid haplotypes may have been lost from southern populations or have remained unsampled, although this is not indicated by our data,. Thirdly, if dispersal is responsible for the disjunct pattern of sexual diploids, then it is coincidental that they have come to occupy two non-glaciated areas. A formerly broad distribution of sexual diploids that has become subsequently fragmented, in our view, best fits available information. The range of Easter daisies is likely to have been historically fragmented by the Wisconsin ice sheets. Higher genetic diversity is expected in parts of a species' current distribution that were 45 refugial (Hewitt 1996, 2000). The significantly higher level of chloroplast diversity in the north supports the hypothesis of Beringia being a northern refugium for Easter daisies during the Wisconsin glaciation. This species joins the ranks of botanical taxa for which molecular evidence concurs with Hulten's hypothesis that unglaciated Beringia was a Quaternary refugium for plants (Hulten 1937). Similar genetic signatures have been detected within Yukon and Alaskan populations of Dryas integrifolia (Tremblay and Schoen 1999) and Saxifraga oppositifolia (Abbott et al. 2000). When concordant patterns can be found among independent species, confidence grows that the observed pattern reflects a specific geographic cause. The post-glacial landscape appears to have been exclusively recolonized by asexual polyploid Easter daisies. Although a review of polyploidy in arctic plants (Brochmann et al. 2004) has refuted Stebbins' classic claim (Stebbins 1950; Stebbins 1985) that polyploids were more likely to colonize newly deglaciated arctic regions, in Townsendia hookeri polyploids alone are found on the formerly glaciated landscape, while diploids occur exclusively in unglaciated areas. However, Bierzychudek (1985) detected a positive correlation between plant parthenogenesis and deglaciated areas. As polyploidy is accompanied by apomixis in the Easter daisy, it may be asexuality that is indeed responsible for the observed pattern. Recolonization by polyploid asexual Easter daisies appears to have occurred from recurrently-evolved southern (Boxes 1-3, Fig. 3-4) and northern (Box 4, Fig. 3-4) phylogenetic lineages. Contact zones between descendents of separate glacial refugia are expected to have more haplotypes present, due to multiple colonization events (Hewitt 1996, 2000). Indeed, populations within the formerly-glaciated central region exhibit significantly higher chloroplast diversity than those found in the southern region of the unglaciated United States. Populations within Alberta's Jasper National Park may represent the confluence of two lines of recolonization, by descendents of northern lineages (populations 61 and 63) as well as those from the south (population 62). An ice-free corridor between the Cordilleran and Laurentide glaciers was completely contiguous along the eastern slope of the Canadian Rockies by 12 kya (Dyke and Prest 1987; Pielou 1991), with the southern ice-free area ultimately conjoining with the northern refugium at a location not far from Jasper. Nunataks (unglaciated mountaintop areas) have been documented according to geological criteria within southwestern Alberta (Dyke and Prest 1987; Pielou 1991), an area where high chloroplast diversity has been noted within other plant species (e.g. Golden and Bain 2000). Wisconsin survival within refugia of this sort remains possible for central populations of the Easter daisy. Townsendia hookeri was included as an example in Bierzychudek's review (1985) of geographic parthenogenesis, where asexuals have a distribution that is more northerly, more 46 widespread and/or at higher elevations than their sexual counterparts. However, our discovery of sexual diploids in the Yukon suggests that although asexual Easter daisies remain far more widespread than sexuals, and the distributions of the two types are largely non-overlapping, the Easter daisy deviates from the typical pattern. This may provide some insight into potential causes of patterns of geographic parthenogenesis. Several explanations have been put forth to account for. patterns of geographic parthenogenesis (reviewed in Bierzychudek 1985, reiterated in Verduijn et al. 2004), and these explanations tend to be either demographic or ecological in nature (Law and Crespi 2002). Under ecological scenarios, asexuals are better competitors or have an advantage over sexuals in certain ecological conditions (e.g. where biotic interactions are low), while demographic explanations are linked to the cost of meiosis and finding a mate, giving asexuals a colonization advantage over sexual organisms. In Easter daisies, geographic parthenogenesis appears to be linked with glacial retreat and the subsequent exposure of new habitats for colonization. Whether this repeated pattern is linked with a demographic explanation (e.g. only one asexual individual is required to establish a population in recently exposed territory after the glaciers recede) or an ecological advantage (e.g. asexuals happen to be more r-selected or grow best at low plant densities) remains unclear. Some models which attempt to explain patterns of geographic parthenogenesis make assumptions based on a general latitudinal trend, for example, assuming shorter growing seasons in the north (e.g. Peck et al. 1998), or the relative predominance of abiotic over biotic contributions to selection in the north (e.g. Glesener and Tilman 1978). Both of these assumptions specifically lead to an outcome that favors sexuality in southern environments. However, in this study, sexual diploids are exclusively found within not one, but two non-glaciated areas, both north and south of the Wisconsin ice sheets. In contrast, formerly-glaciated areas are exclusively inhabited by asexual polyploids, derived from independently-arising northern and southern sexual lineages. This suggests that latitudinal factors per se are unlikely to be the predominant contributors to patterns of geographic parthenogenesis. Arguments for the superior colonizing ability of asexuals over sexuals in the generation of patterns of geographic parthenogenesis were presented by Stebbins (1950), Bierzychudek (1985) and Asker and Jerling (1992). These arguments rest in the observation that most apomicts (like T. hookeri) have arisen from outcrossing (self-incompatible or dioecious) progenitors (Stebbins 1950), and thus have the relative advantages of producing phenotypically uniform progeny, and having the potential for establishing new populations from single migrant seeds, and not being reliant on pollinator servicing. A molecular phylogenetic study of parthenogenetic and sexual Timema walking sticks, one of the few empirical studies which specifically aimed to test hypothesis related to geographic parthenogenesis, came to a similar conclusion. In Timema, patterns of geographic 47 parthenogenesis were best explained by a greater colonization rate of asexuals over sexuals, as the asexuals had undergone a range expansion event (Law and Crespi 2002). However, under a largely demographic scenario such as this, it remains possible that genetic and ecological interference through destabilizing hybridization between sexuals and asexuals (Lynch 1984) may block the expansion of sexuals or even erode their range (Mogie 1992). To ultimately determine the contribution of ecological versus demographic factors to patterns of geographic parthenogenesis we need to evaluate whether it is the case that sexuals cannot survive where asexuals are now found, or whether they just fail to ever get there. Studies of fitness in sexuals and asexuals, both alone and in mixtures, under common garden conditions in formerly glaciated areas are required, preferably using organisms of shared ploidy. 48 3.5 References Abbott RJ, Brochmann C. 2003. History and evolution of the arctic flora: in the footsteps of Eric Hulten. Molecular Ecology 12: 299-313. Abbott RJ, Smith LC, Milne RI, Crawford RMM, Wolff K, Balfour J. 2000. Molecular analysis of plant migration and refugia in the Arctic. Science 289: 1343-1346. Asker SE, Jerling L. 1992. Apomixis in Plants. CRC Press: Boca Rotan, USA. Avise JC. 2000. Phylogeography: the history and formation of species. Harvard University Press: Cambridge, USA. Beaman JH. 1954. Chromosome numbers, apomixis, and interspecific hybridization in the genus Townsendia. Madrono 12: 169-180. Beaman JH. 1957. The systematics and evolution of Townsendia (Compositae). Contributions from the Gray Herbarium of Harvard University 183: 1-151. Bierzychudek P. 1985. Patterns in plant parthenogenesis. Experientia 41: 1255-1264. Brochmann C, Brysting AK, Alsos IG, Borgen L, Grundt HH, Scheen A-C, Elven R. 2004. Polyploidy in arctic plants. Biological Journal of the Linnean Society 82: 521-536. Brochmann C, Xiang Q-Y, Brunsfeld SJ, Soltis DE, Soltis PS. 1998. Molecular evidence for polyploid origins in Saxifraga (Saxifragaceae): the narrow arctic endemic S. svalbardensis and its widespread allies. American Journal of Botany 85: 135-143. Cook L M , Soltis PS, Brunsfeld SJ, Soltis DE. 1998. Multiple independent formations of Tragopogon tetraploids (Asteraceae): evidence from RAPD markers. Molecular Ecology 7: 1293-1302. Demesure B, Sodzi N , Petit RJ. 1995. A set of universal primers for amplification of polymorphic non-coding regions of mitochondrial and chloroplast DNA in plants. Molecular Ecology 4: 129-131. Douglas GW, Argus GW, Dickson HL, Brunton DF. 1981. The Rare Vascular Plants of the Yukon. National Museums of Canada: Ottawa, Canada. Doyle JJ, Doyle JD. 1987. A rapid DNA isolation procedure for small quantities of fresh leaf tissue. Phytochemisty. Bulletin 19: 11-15. Doyle JJ, Doyle JL, Brown ADH. 1999. Origins, colonization and lineage recombination in a widespread perennial soybean polyploid complex. Proceedings of the National Academy of Science, USA 96: 10741-10745. Doyle JJ, Doyle JL, Brown AHD, Grace JP. 1990. Multiple origins of polyploids in the Glycine tabacina complex inferred from chloroplast DNA polymorphism. Proceedings of the National Academy of Science 87: 714-717. Dumolin-Lapegue S, Pemonge M-H, Petit RJ. 1997. An enlarged set of consensus primers for the study of organelle DNA in plants. Molecular Ecology 6: 393-397. Dyke AS, Prest VK. 1987. Late Wisconsin and Holocene retreat of the Laurentide Ice Sheet. Geological Survey of Canada. Map 1702A. Farris JS. 1989. The retention index and homoplasy excess. Systematic Zoology 38: 406-407. 49 Farris JS, Kallersjo M, Kluge AG, Bult C. 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. Gaggiotti OE. 1994. An ecological model for the maintenance of sex and geographic parthenogenesis. Journal of Theoretical Biology 167: 201-221. Gilbert D. 1994. SeqApp. University of Indiana, Bloomington, USA. Glesener RR, Tilman D. 1978. Sexuality and the components of environmental uncertainty: clues from geographic parthenogenesis in terrestrial animals. The American Naturalist 112: 659-673. Golden JL, Bain JF. 2000. Phylogeographic patterns and high levels of chloroplast DNA diversity in four Packera (Asteraceae) species in southwestern Alberta. Evolution 54: 1566-1579. Grant V. 1981. Plant Speciation. Columbia University Press: New York, USA. Hasegawa M, Kishino K, Yano T. 1985. Dating the human-ape splitting by a molecular clock of mitochondrial DNA. Journal of Molecular Evolution 22: 160-174. Hewitt G. 2000. The genetic legacy of the Quaternary ice Ages. Nature 405: 907-913. Hewitt GM. 1996. Some genetic consequences of ice ages, and their role in divergence and speciation. Biological Journal of the Linnean Society 58: 247-276. Holderegger R, Abbott RJ. 2003. Phylogeography of the Arctic-Alpine Saxifraga oppositifolia (Saxifragaceae) and some related taxa based on cpDNA and ITS sequence variation. American Journal of Botany 90: 931-936. Hulten E. 1937. Outline of the History of of Arctic and Boreal Biota During the Quaternary Period. Lehre J Kramer: New York, USA. Jansen RK. 1992. Current research. Plant molecular Evolution Newsletter 2: 13. Joly S, Bruneau A. 2004. Evolution in Apios americana (Leguminosae) revealed by genealogical analysis of the histone H3-D gene. Evolution 58: 284-295. Kishino K, Hasegawa M. 1989. Evaluation of the maximum likelihood estimate of the evolutionary tree topologies from DNA sequence data, and the branching order in Hominoidea. Journal of Molecular Evolution 29: 170-179. Kluge AG, Farris JS. 1969. Quantitative phyletics and the evolution of anurans. Systematic Zoology 18: 1-32. Law JH, Crespi BJ. 2002. The evolution of geographic parthenogenesis in Timema walking-sticks. Molecular Ecology 11: 1471-1489. Levin DA. 1975. Minority cytotype exclusion in local plant populations. Taxon 24: 35-43. Love A. 1968. IOPB chromosome number reports. XV. Taxon 17: 91-104. Lynch M. 1984. Destabilizing hybridization, general-purpose genotypes and geographic parthenogenesis. The Quarterly Review of Biology 59: 257-290. Maneval WE. 1936. Lacto-phenol preparations. Stain technology 11: 9-11. 50 Marshall HD, Newton C, Ritland K. 2001. Sequence-Repeat Polymorphisms exhibit the signature of recombination in Lodgepole Pine Chloroplast DNA. Molecular Biology and Evolution 18: 2136-2138. Mayer SS. 1991. Artificial hybridization in Hawaiian Wikstroemia (Thymelaeace'ae). American Journal of Botany 78: 122-130. Mogie M . 1992. The evolution of asexual reproduction in plants. Chapman & Hall: London, UK. NoglerGA. 1984. Gametophytic apomixis. In Johri BM, ed. Embryology of Angiosperms. 475-518 Springer-Verlag: Berlin, DE. Osborn TC, Pires JC, Birchler JA, Auger DL, Chen ZJ, Lee H-S, Comai L, Madlung A, Doerge RW, Colot V, Martienssen RA. 2003. Understanding mechanisms of novel gene expression in polyploids. Trends in Genetics 19: 141-147. Otto SP, Whitton J. 2000. Polyploid incidence and evolution. Annual Reviews of Genetics 34: 401-437. Palmer JD. 1987. Chloroplast DNA evolution and biosystematic uses of chloroplast DNA variation. The American Naturalist 130: S6-S29. Peck JR, Yearsley JM, Waxman D. 1998. Explaining the geographic distributions of sexual and asexual populations. Nature 391: 889-892. Pielou EC. 1991. After the Ice Age: The Return of Life to Glaciated North America. University of Chicago Press: Chicago, USA. Posada D, Crandall KA. 2001. Evaluations of methods for detecting recombination from DNA sequences: Computer simulations. Proceedings of the National Academy of Science, USA 98: 13757-13762. Ramsey J, Schemske DW. 2002. Neopolyploidy in flowering plants. Annual Reviews in Ecology and Systematica 33: 589-639. Roose ML, Gottlieb LD. 1976. Genetic and biochemical consequences of polyploidy in Tragopogon. Evolution 30: 818-830. Segraves KA, Thompson JN, Soltis PS, Soltis DE. 1999. Multiple origins of polyploidy and the geographic structure of Heucheragrossulariifolia. Molecular Ecology 8: 253-262. Sharbel TF, Mitchell-Olds T. 2001. Recurrent polyploid origins and chloroplast phylogeography in the Arabis holboellii complex (Brassicaceae). Heredity 87: 59-68. Shaw J, Lickey EB, Beck JT, Farmer SB, Liu W, Miller J, Siripan KC, Winder CT, Schilling EE, Small R. 2005. The tortoise and the hare II: relative utility of 21 noncoding chloroplast DNA sequences for phylogenetic analysis. American Journal of Botany 92: 142-166. Soltis DE, Soltis PS. 1989. Allopolyploid speciation in Tragopogon: insights from chloroplast DNA. American Journal of Botany 76: 1119-1124. Soltis DE, Soltis PS. 1993. Molecular data and the dynamic nature of polyploidy. Critical Reviews in Plant Science 12: 243-273. Soltis DE, Soltis PS. 1998. Choosing an approach and an appropriate gene for phylogenetic analysis. In Soltis DE, Soltis PS, Doyle JJ, eds. Molecular Systematics of Plants II: DNA Sequencing. 1-42 Kluwer Academic Publishers: Norwell, USA. 51 Soltis DE, Soltis PS. 1999. Polyploidy: recurrent formation and genome evolution. Trends in Ecology and Evolution 14: 348-352. Soltis DE, Soltis PS, Ness BD. 1989. Chloroplast-DNA variation and multiple origins of autoploidy in Heuchera micrantha (Saxifragaceae). Evolution 43: 650-656. Soltis DE, Soltis PS, Tate JA. 2003. Advances in the study of polyploidy since Plant Speciation. New Phytologist 161: 173-191. Soltis PS, Doyle JJ, Soltis DE. 1992. Molecular data and polyploid evolution in plants in Soltis PS, Soltis DE, Doyle JJ, eds. Molecular Systematics of Plants. 177-201. Chapman and Hall: New York, USA. Soltis PS, Plunkett GM, Novak SJ, Soltis DE. 1995. Genetic variation in Tragopogon species: additional origins of the allotetraploids T. mirus and T. miscellus (Compositae). American Journal of Botany 82: 1329-1341. Soltis PS, Soltis DE. 1991. Multiple origins of the allotetraploid Tragopogon mirus (Compositae): rDNA evidence. Systematic Botany 16: 407-413. Soltis PS, Soltis DE. 2000. The role of genetic and genomic attributes in the success of polyploids. Proceedings of the National Academy of Sciences, USA 97: 7051-7057. Song K, Osborn TC. 1993. Polyphyletic origins of Brassica napus: new evidence based on organelle and nuclear RFLP analyses. Genome 35: 992-1001. Song K, Lu P, Tang K, Osborn TC. 1995. Rapid genome change in synthetic polyploids of Brassica and its implications for polyploid evolution. Proceedings of the National Academy of Science, USA 92: 7719-7723. Stanley RG, Linskens HF. 1974. Viability tests. Pollen: biology, biochemistry and management. 67-86. Springer: Berlin, DE. Stebbins GL. 1950. Variation and Evolution in Plants. Columbia University Press: New York, USA. Stebbins GL. 1985. Polyploidy, hybridization, and the invasion of new habitats. Annals of the Missouri Botanical Garden 72: 824-832. Suomalainen E, Saura A, Lokki J. 1987. Cytology and Evolution in Parthenogenesis. CRC Press: Boca Raton, USA. Swofford DL. 2000. PAUP*. Phylogenetic analysis using parsimony (*and other methods). Sinauer Associates: Sunderland, USA. Taberlet P, Gielly L, Pautou G, Bouvet J. 1991. Universal primers for amlification of three non-coding regions of chloroplast DNA. Plant Molecular Biology 17: 1105-1109. Tajima, F. 1989. Statistical method for testing the neutral mutation hypothesis by DNA polymorphism. Genetics 123: 585-595. Tremblay NO, Schoen DJ. 1999. Molecular phylogeography of Dryas integrifolia: glacial refugia and postglacial recolonization. Molecular Ecology 8: 1187-1 198. Vandel A. 1931. La parthenogenese. Doin: Paris, FR. 52 Verduijn MH, van Dijk PJ, van Damme JMM. 2004. Distribution, phenology and demography of sympatric sexual and asexual dandelions (Taraxacum officianale s. /.): geographic parthenogenesis on a small scale. Biological Journal oftheLinnean Society 82: 205-218. 53 Table 3-1. Designation, locality information, and pollen data for populations of Townsendia hookeri included in this chapter. Population designations correspond to collection numbers and corresponding voucher specimens are deposited in the University of British Columbia (UBC) Herbarium. The identification numbers of populations categorized as sexual diploid are in bold and underlined, while those of apomictic polyploids are not. Asterices (*) denote populations that fail to produce pollen, and are polyploid according to flow cytometry (results not included in this chapter). # Locality Latitude/ Pollen Pollen Longitude diameter stainability (UM +/- SE) (% +/- SE) 2 Georgetown, Clear Creek County, Colorado 39°43'N 29.68 +/- 0.53 28.92+/- 3.59 105°40'W 10 Samora Creek, Saguache County, Colorado 38°14'N 20.55 +/- 0.33 78.33+/- 2.96 106°39'W 12 Lincoln's Head, 12 km SE of Laramie, Albany County, 41°17'N 19.35 +/- 0.42 87.00+/- 2.89 Wyoming 105°28'W 13 Grover(lOkmE), Weld County, Colorado 40°51'N 31.33 +/- 0.72 18.67+/- 0.88 104°00'W 18 Mormon Canyon, 8 km S of Glenrock, Converse County, 42°47'N 21.42 +/- 1.43 20.75+/-4.80 Wyoming 105°54'W 19 Arminto (9.5 km N), Natrona County, Wyoming 43°15'N 29.68 +/- 0.66 14.67+/-0.33 107°15'W 21 Francs Fork, 30 km WSW of Meeteetse, Park County, Wyoming44°03'N 33.55 +/- 0.99 42.33+/- 1.45 109°14'W 23 Augusta (10 km NW), Lewis and Clark County, Montana 47°33'N 32.42 +/- 0.74 13.00+/- 1.53 112°28'W 28 Writing-On-Stone Provincial Park, 30 km E of Milk River, 49°05'N 28.44 +/- 1.60 8.00 +/- 2.70 Alberta 111°37'W 29 Radium Hot Springs, British Columbia 50°38'N 35.90 +/- 0.85 11.33+/- 1.20 116°04'W 30 Columbia Lake, 7 km N of Canal Flats, British Columbia 50°05'N 32.72 +/- 1.20 19.20+/- 4.58 115°53'W 34 Canon City (10 km NW), Fremont County, Colorado 38°30'N 22.48 +/-0.35 73.20+/-6.06 105°20'W 35 Florissant Fossil Beds National Monument, Teller County, 38°55'N 21.20 +/- 0.34 78.40+/- 1.57 Colorado 105°17'W 36 National Center for Atmospheric Research (NCAR), 39°59'N 20.85 +/- 0.30 83.70+/- 1.24 Boulder County, Colorado 105°17'W 37 Foothills Trail, City of Boulder Open Space, Boulder 40°04'N 20.38 +/- 0.51 69.67+/- 5.36 County, Colorado 105°18'W 38 Greenbelt Plateau, City of Boulder Open Space, Boulder 39°56'N 21.08 +/-0.38 80.67+/- 1.20 County, Colorado 105°14'W 39 Heil Ranch, Boulder County Open Space, Boulder County, 40°08'N 20.35 +/- 0.30 85.00+/- 1.73 Colorado 105°19*W 40 Six-mile Fold, Boulder County Open Space, Boulder 40°07'N 19.85 +/- 0.40 87.00+/- 2.65 County, Colorado 105°17'W 41 Coalton Trail, City of Boulder Open Space, Boulder County, 39°55'N 20.45 +/-0.24 66.33+/-4.37 Colorado 105°12'W 42 Dixon Reservoir, Fort Collins, Larimer County, Colorado 40°33'N 20.18 +/-0.55 60.20+/-9.98 105°09'W 44 Devil's Playground, Medicine Bow National Forest, Albany 41°08'N 17.80 +/-0.31 93.00+/-2.52 County, Wyoming 105°22'W 49 Blakiston Creek Fan, Waterton Lakes National Park, Alberta 49°05'N 29.58 +/- 0.77 4.67 +/- 1.20 113°55'W 50 Buffalo Paddock, Waterton Lakes National Park, Alberta 49°08'N 30.05 +/- 0.88 15.33+/- 0.88 113°55'W 54 Table 3-1: continued # Locality . Latitude/ ' Pollen Pollen Longitude diameter stainability (UM +/- SE) (% +/- SE) 51 Oldman Reservoir, NW of Brockett, Alberta 49°35'N 113°55'W 30.26 +/- 0.77 26.33+/- 2.96 54 Pavan Park-H, N of Lethbridge, Alberta 49045'N 112°50'W 33.25 +/- 0.22 10.67+/- 0.67 55 Pavan Park-E, N of Lethbridge, Alberta 49°45'N 112°50'W 31.75 +/- 1.17 9.00 +/- 0.00 56 Mud Springs, 12 km SE of Townsend, Broadwater County, 46°16'N 30.92 +/- 0.63 37.50+/- 9.57 Montana 111°40'W 57 Tunnel Mountain Hoodoos, Banff National Park, Alberta 51°11'N 115°29'W 30.42 +/- 0.81 15.67+/- 2.03 58 Yannuska Natural Area, 3 km northwest of Seebe, Alberta 51°06'N 115°03'W 32.18 +/- 0.69 29.67+/- 3.53 59 Columbia River Viewpoint, 1.3 km S of Radium Hot Springs, 50°40'N 35.05 +/- 0.69 35.00+/- 2.65 British Columbia 116°05'W 61 Wabasso Lake Trail, Jasper National Park, Alberta 52°46'N 117°57'W 35.22 +/- 1.74 7.33 +/- 0.88 62 Old Fort Point, Jasper National Park, Alberta 52°52'N 118°03'W 34.60 +/- 0.49 12.33+/- 1.45 63 Pyramid Bench, Jasper National Park, Alberta 52°54'N 118°04'W 36.20 +/- 0.57 21.00+/- 3.21 64 Clayhurst Crossing, Junction of Alces and Peace Rivers, British 56°08'N 33.58 +/- 1.33 26.67+/- 3.53 Columbia 120°03'W 66* Whitehorse, Mile 13 Alaska Highway, Yukon Territory 61°37'N 135°52'W N/A N/A 67* Conglomerate Mountain, Mile 65 Alaska Highway, N of 61°37'N N/A N/A Whitehorse, Yukon Territory 135°52'W 68 Tantalus Butte, N of Carmacks, Yukon Territory 62°07'N 136°15'W 20.15 +/- 0.53 67.67+/- 1.45 69*McCabe Creek, Mile 142 Alaska Highway, South of Minto, 62°32'N N/A N/A Yukon Territory 136°45'W 71* Sheep Mountain, Kluane National Park, Yukon Territory 61°00'N 138°32'W N/A N/A 92 Eagle Rock, Yukon Territory 62°02'N 135°51'W 20.73 +/- 0.85 93.75+/- 1.31 55 Table 3-2. Haplotype designations based on chloroplast D N A sequences from Townsendia hookeri, showing the constituent polymorphisms at respective alignment positions. Indels are denoted by asterisks (*). Dots (.) ' a. Alig ;nment position aplot) ndhV g ene trnK intron X 258 321 352 425 469 529 538 627 704 773 786 860 913 961 1107 1148 1289 a A C G A A C T C A A C G G * C A T b G c T G T G C d T G T T G C e G G T G f g G G . T G T T G G h T G A T G i T T T G G A C T G j T T T G T G C T G 56 Figure 3-1. Localities of sexual diploid and apomictic polyploid Easter daisy (Townsendia hookeri) populations collected from western North America. The known range of T. hookeri is indicated by light grey shading. The extent of the Cordilleran (darkest grey) and Laurentide (medium gray) ice sheets at 18 kya is outlined. Numbers correspond to sexual diploid (bold and underlined) or apomictic polyploid populations as given in Table 3-1. 'North 1 : 1 343 593 (approximate) 57 100 20 25 30 35 mean pollen diameter (IJM) Figure 3-2. Plot of mean pollen diameter by mean pollen stainability for populations of the Easter daisy (Townsendia hookeri), indicating the correlation between ploidy and breeding system within this species (Spearman's rho = -0.7133, p < 0.0001). Putative diploid populations are indicated with black circles, while putative polyploid populations are indicated with clear circles. 58 Figure 3-3. Chloroplast haplotype variation within 3 geographically proximal populations of Easter daisies (Townsendia hookeri). Populations 61, 62 and 63 of Jasper National Park are monomorphic based on the banding patterns of a Hinfl restriction digest of a portion of the ndh¥ gene, indicating negligible migration through seed flow among populations Figure 3-4. Intraspecific chloroplast phylogeny of sexual diploid and apomictic polyploid Easter daisies (Townsendia hookeri). The tree presented is the strict consensus of two most parsimonious trees (17 steps) as well as the maximum likelihood tree (LnL = -1821.41) based on DNA sequence variation in the ndh¥ gene and trnK intron. Bootstrap values (parsimony/likelihood) above 50% are adjacent to branches. Numbers correspond to sexual diploid (bold and underlined) or apomictic polyploid populations as given in Table 3-1. Letters correspond to haplotypes as given in Table 3-2. Independent origins of apomictic polyploidy are indicated by numbered boxes. 1 step (parsimony) 0.0005 substitutions/site (ML) 2, 13, 19,21,23,54, 56, 62, 12, 34, 35, 36 29, 30, 59 J37, 38, 39, 40, 41, 49, 50 '65/66 © 6 1 , 6 3 66, 67, 71 KM 60 4 The detection of clonality and sexuality in diploid and polyploid populations of the Easter daisy, Townsendia hookeri1 4.1 Introduction Clonal reproduction is an important reproductive strategy commonly employed across the flowering plants. It is relatively easy to demonstrate that an organism is primarily sexual or asexual, however inferring strict asexuality often relies on negative evidence. Asexuality can be inferred through direct observation, for example by determining that populations lack males or are missing various reproductive structures. However males may be produced at excessively low frequencies (ie "spanandric" males, Shaeffer et al. 2006), sexual reproductive features may be of minute size, or sexual phases may look extremely different from asexual phases, making our search for evidence of sex unfruitful. Even when sex is cryptic or infrequent, the signature of a rare round of recombination may linger for generations in the genomes of a largely asexual population (Burt et al. 1996). Analyses of patterns of genetic variation can be a reliable and powerful method to assess the extent of asexuality. Asexual populations can harbor much genetic variation, often on par with sexually reproducing relatives (Parker 1979; Ellstrand and Roose 1987; Hamrick and Godt 1989; Widen et al. 1994; Loxdale and Lushai 2003). While it is no longer surprising to find high levels of genetic variation in predominantly asexual populations, clear identification of the sources of this variation remains elusive. Genetic diversity in clonal populations can have many origins, including somatic mutation, multiple origins of clones from sexuals, sporadic episodes of meiotic and mitotic recombination, and migration among genetically differentiated populations (van der Hulst et al. 2000). While disentangling the various contributions of these processes poses significant challenges, a number of approaches have been adopted that distinguish evidence for recombination, sexuality's hallmark, from patterns that are consistent with clonal reproduction (i.e. variation due to mutation accumulation alone). Here we evaluate the presence and extent of clonality and recombination in four standing populations of the Easter daisy (Townsendia hookeri Beaman) from the Yukon Territory, Canada. The Easter daisy is a diminutive, perennial member of the Sunflower family. Within this species, there is a general correlation between polyploidy and asexuality (Chapter 3). Diploid populations have been described as sexual, while autopolyploid populations are reported to reproduce through obligate apomixis (Beaman 1957), achieved through restitutional diplospory (the avoidance of 1 A version of this chapter is in preparation for submition to New Phytologist as Thompson SL, Choe G, Ritland K, Whitton J. The detection of clonality and sexuality in diploid and polyploid populations of the Easter daisy, Townsendia hookeri. 61 reductional division through a meiotic failure of the megasporocyte), followed by non-pseudogamous parthenogenesis (pollination is neither required for the development of the embryo nor the endosperm). Range-wide surveys of pollen size (an indicator of ploidy) and pollen fertility (an indirect indicator of the potential for sexuality), revealed the presence of two diploid populations in the Yukon territory, while no detectable pollen production was found over three survey years among the remaining Yukon populations (Chapter 3). Polyploid populations in the remainder of the range generally produce pollen of low viability (mean 17 %, Chapter 3). Townsendia hookeri has rare status in the Yukon (Douglas et al. 1981), where populations are found exclusively on dry calcareous montane slopes and are isolated by over 1300 km of boreal forest from the nearest populations in southern Alberta. In this chapter, we use a genome-scan with A F L P markers to elucidate the extent of clonality and recombination in Yukon populations of the Easter daisy. Flow cytometry is used to assess plody. The expectation is that male-fertile populations will be diploid and found to frequently undergo recombination, while male-sterile populations will be tetraploid and asexual. Four multi-locus tests infer the presence of clonality and sexuality for populations. The effective long-term rate of sexuality and mutation is jointly estimated for populations. We summarize and integrate our results and evaluate potential sources of genetic variation within sub-arctic Easter daisy populations. 4.2 Materials and methods Collections— Plant material was collected from four Yukon populations of Townsendia hookeri: populations 66 (Mile Thirteen Alaska Highway, 60°59'N 135°10'W), 67 (Conglomerate Mountain, 61°37'N 135°52'W), 68 (Tantalus Butte, 62°07'N 136°15'W) and 71 (Tachal Dhal Kluane, 61°00'N 138°32'W). Easter daisy populations are well-defined, occurring on substrate-exposed hilltops. Individuals were sampled as evenly as possible at each locality to maximize the distance among sampled individuals (typically several meters apart). Young leaves were frozen immediately in liquid nitrogen, then stored at minus 80°C. Fruits were collected from dried capitula (flower heads). Additional plants were dried for pollen examination and voucher specimens deposited at the University of British Columbia Herbarium (UBC, accession numbers available on request). Flow cytometry—Seeds were germinated on moist filter paper in petri dishes, then transferred to seedling trays and allowed to grow for approximately 6 weeks. Approximately 40 mg of young leaf material was prepared for flow cytometry, according to the following protocol, optimized for Epilobium angustifolium (P. Kron and B. Husband, personal communication). Leaf samples from each of 6 plants per population were chopped in petri dishes over ice with a razor blade in 1 mL of the chopping buffer of Bino (1992) with the modifications of Dart et al. (2004). Another 62 0.5 mL of buffer was added to the slurry, the sample was mixed by pipetting, then was filtered through a 30 um CellTrics ® filter (Partec, Munster, Germany). The effluent was centrifuged at 13000 rpm for 10 s, and the supernatant removed. The pellet was resuspended in 0.3 mL of 1.12 mg/mL propidium iodide staining buffer (Arumuganathan and Earle 1991) and stained in the dark for approximately 45 minutes. Pisum sativum cv. Minerva Maple was used as an internal standard for calibration based on its established D N A content of 9.56 pg / 2C nucleus (Johnson et al. 1999). Flow cytometry was performed on a B D FACS Scan benchtop analyzer (BD Biosciences, Mississauga, Canada), at the UBC Multi-user Flow Cytometry Facility (Biomedical Research Centre, University of British Columbia) according to their internal protocols. Peaks were generated on the FL2-A axis for 2000-5000 events for each sample, and peak means were determined using FloJo Software (TreeStar, Inc, Corvalis, USA). Data were analyzed with JMP IN 4.0.4 (SAS Institute Inc, Cary, USA) and K-M E A N S (Pierre Legendre, personal comm.). D N A extraction—Genomic D N A was isolated from 20 plants per population from each of 4 populations by C T A B (hexadecyltrimethylammonium bromide) protocol (Doyle & Doyle, 1987) from approximately 0.1 g of leaf tissue with the following modifications: volumes were reduced for extraction in 1.5 mL microfuge tubes and sodium metabisulfite (1% w/v) was added to the isolation buffer. Isolated D N A was quantified by a Hoefer DyNAQuant™ 200 fluorometer (San Francisco, USA) according to the manufacturer's directions, diluted to 10 ng/uL, then used directly in PCR amplifications. Chloroplast haplotyping—In order to assess the extent of migration by sexual plants into clonal populations through seed flow, all individuals were surveyed for a diagnostic restriction site difference. Yukon populations of Townsendia hookeri are known to comprise 2 chloroplast haplotypes based on D N A sequence data (chapter 3). A segment of the chloroplast ndhV gene was amplified using the primer pairs ndh¥\A (Jansen 1992) and ndh¥\4R (chapter 3) in 25 uL reactions containing 20ng of genomic D N A , 30mM tricine pH 8.4, 2mM M g C l 2 , 50mM KC1, 100 u.M dNTPs (equimolar ratio), 2.0 units of taq polymerase, and 0.01 nmol of each primer. Amplification was performed by an M J Research PTC-100 thermocycler (Watertown, USA) according to the following conditions: initial denaturation at 94°C for 4 minutes; 30 cycles of 92°C for 45 seconds, 53°C for 45 seconds, and 72°C for 3 minutes; with a final 10 minute extension at 72°C. PCR products were then digested with the restriction enzyme Hinfl for 6 hours at 37°C in 20uL reactions containing: 5uL of PCR product, 2uL of Buffer #2 (New England Biolabs, Mississauga, Canada), 2(ig of bovine serum albumin, and 1U of Hinfl. Digested fragments were separated by agarose electrophoresis, visualized under U V after staining with ethidium bromide and scored for the their restriction digest pattern. 63 Amplified fragment length polymorphisms (AFLPs)—-All individuals were A F L P fingerprinted based on the method of Vos et al. (1995) with modifications developed by R. D. Noyes (Noyes and Rieseberg 2000). Denatured reaction products were run on an Applied Biosystems 377 D N A sequencer (Foster City, USA) at UBC's Nucleic Acid and Protein Sequencing Unit (NAPS), according to their internal protocols. Eight primer combinations (all permutations of Eco+ (ACC / ACG) and Mse+ (AGC / A C G / A C C / AAC)) were twice screened for three individuals from each of the four populations, and two primer combinations were selected according to the optimality criteria of consistency across all individuals, repeatability for each individual, and level of polymorphism. Al l individuals were fingerprinted using EcoACC / MseAGC and EcoACG / MseAGC primers then scored for band presence/ band absence using Genographer 1.6.0 (Benham 2001). Over-representation analyses— The excess of a particular multilocus genotype within a population is often the most robust and significant evidence of clonal reproduction (Tibayrenc et al. 1991). To characterize this excess of representation, we compute the probability of observing at least n individuals with the identical multilocus genotype assuming unlinked loci and panmixis. For dominant markers this is: where n is the observed number of individuals with the identical multilocus genotype, Wis the sample size (identical and non-identical genotypes), / is the number of loci and fi'is the frequency of the /th banded genotype in the population. As Psex indicates the probability that repeated multilocus genotypes in the sample result from random mating, Psex is expected to be low under clonal reproduction and high under sexuality. The significance of this value was determined by randomizing loci according to the given band frequencies among all individuals, then recomputing the Psex value for 10000 replicate populations of size N. This was done to generate a distribution of Psex values under panmixis. Although Psex gives a probability, the Psex value should be viewed as a test statistic (as t-values in a t-test) as its values are discretely distributed and it is possible to obtain identical Psex values for different multilocus genotypes. Since exact calculations of critical values are analytically intractable with many variable markers, the distribution was simulated with the program MLGsim (Stenberg and Lundmark 2002). A significant p-value from the randomizations indicates that a repeated multilocus genotype is unlikely to have arisen under sexuality, implicating clonality within the population. Phylogenetic analyses— Tree-building methods were tested for the presence of sexuality. For a strictly clonal population, where mutation is the sole method of generating genetic variability, the expectation is a strictly hierarchal tree, devoid of homoplasy. If recombination occurs among loci, N 64 we expect trees to be poorly resolved, have homoplasy present, and the best tree would not differ significantly from a randomized distribution of trees (Burt et al. 1996). PAUP* version 4.Ob 10 (Swofford, 2000) was used to perform phylogenetic analyses on each population, coding each A F L P band as present or absent, and treating characters as unordered. Depending on the number of AFLP genotypes per population, exhaustive (ten or fewer multilocus genotypes) or heuristic (greater than ten multilocus genotypes) searches were used to search for trees under maximum parsimony. For heuristic searches, starting trees were obtained by simple stepwise addition of taxa, and TBR branch swapping was used on best trees. Tree lengths, homoplasy indices (HI) and the number of equally parsimonious trees were noted. Strict consensus trees were constructed when more than one most parsimonious tree was found for each population. In order to determine whether the most parsimonious tree(s) differed significantly from a randomized data distribution, permutation tail probability tests (Faith and Cranston 1991) were conducted based on 1000 replicates. Compatibility analyses—- Compatibility analysis, originally a systematic method based on the LeQuesne test (LeQuesne 1969; Meacham and Estabrook 1985), has the demonstrated ability to outperform other methods for detecting recombination from D N A sequence data (Posada 2002). The method works on the assumption that if reproduction is strictly clonal, then for pairs of biallelic loci no more than 3 of the 4 possible combinations should be observed. If all four combinations are observed for a pair of binary loci, recombination is a more parsimonious explanation than 3 independent mutation events. The number of incompatibilities in all pairwise of multilocus genotypes is then tallied; a significant difference between observed and permuted values indicates a lack of recombination, and therefore clonality is likely within the population. Compatibility analyses were performed for each population using PICA 4.0 (Wilkinson 2001), with monomorphic loci excluded and with only one representative of each A F L P genotype included. Matrix incompatibility and permutation tail probabilities (PTP) based on 10000 random data permutations were determined for each population using the M A T R I X program. In order to determine the number and identity of genotypes that are implicated in recombination events within populations, the amount of matrix incompatibility caused by each AFLP genotype was estimated by jackknifing the genotypes with the program J A C T A X . After jackknifing, the most incompatible genotypes were successively removed, then matrix incompatibility and jackknifing calculations were repeated after each removal, until no incompatibility remained within the dataset following the approach of van der Hulst et al. (2000). Association analyses— Non-random association among alleles at different loci (linkage disequilibrium) is expected if recombination is absent within a population. Here we assume that 65 significant associations among loci are due to departures from sexuality. To evaluate the association of genotypes between loci due to clonality, the following measure of association was used: where fi is the frequency of banded genotypes at locus is the frequency of banded genotypes at locus j, bjk is an indicator variable for the presence (b,k =1) or absence =0) of bands at locus i in individual k, and likewise for bjk. Loci with extreme genotype frequencies (outside 0.01 to 0.99) were excluded. This formula is equivalent to Agapow and Burt's (2001) index of multilocus linkage association, fd, which they show is independent of the number of loci compared. This measure is also analogous to the measure of normalized linkage disequilibrium of Hill and Robertson (1968), except that presence/absence of bands replaces observed presence/absence of alleles. The value of our r2 is proportional to the amount of linkage disequilibrium within populations and will be zero with zero linkage disequilibrium. Its maximum value is one, although, like the co-dominant measure of r 2, it can only attain a value of one when allele frequencies are equal among loci (all normalized measures of linkage disequilibrium suffer from a varying dependence upon allele frequencies). Significance values were calculated through 100 permutations of genotypes among individuals, sampling with replacement at each locus. Statistical significance was ascertained by a lack of overlap of the original estimate with the distribution of permuted estimates. The variance of the permuted values was used to estimate the standard error of the original estimate. In a fully recombining population, there will be a lack of distinction between observed and permuted values of 2 r . Estimation of the frequency of recombination— When the number of sexual recombination events per generation is low, mutation becomes an important consideration in clonal populations. The long-term frequency of recombination was indirectly inferred according to the method outlined in Thompson (submitted) which extends the approach of Burt et al. (1996). Mutation, which is a single-locus source of multilocus clonal non-identity, is incorporated into the estimation procedure. At N equilibrium: 2f 66 where Ns is the number of sexual events per generation within the entire population (ie no particular population size need be assumed), g is the probability that individuals have different band states at a single locus,/is the multilocus probability that two randomly selected individuals from a population are identical clones, and n is the number of marker loci surveyed. The mutation rate among allelic states (banded to unhanded, and vice versa) per generation within the population, Nu, was also estimated as Nu = (l-g)/(8g-) when Nun is of approximate order Ns (Thompson, submitted). Al l loci were included in calculations and standard errors of estimates were found by jackknifing across individuals. 4.3 Results Ploidy and pollen -Genome sizes fell into three distinct clusters of values (K-means clustering of all flow cytometry values according to the second-order Calinski-Harabasz criterion, Legendre and Legendre 1998). One cluster comprised samples from population 68, the second comprised values obtained from populations 66, 67 and 71, while the third cluster of values was exclusively found in population 71. Mean D N A contents for these clusters (by population) are shown in Fig. 4-1, and indicate distinct differences in D N A content per 2C nucleus among Yukon populations (Fig. 4-1, Tukey-Kramer HSD test: a = 0.05). These clusters likely correspond to diploid, tetraploid and triploid values respectively. Population 68 had a mean D N A content of 13.22 +/- 0.37 (SE) pg / 2C nucleus, and this mean differed significantly from the means of all other populations. Independent chromosome counts characterize population 68 as 2n = 18 (Love 1968; T. Mosquin personal communication), the diploid number for Townsendia (Beaman 1954; Beaman 1957). Populations 66 and 67 had mean genome sizes of 25.93 +/- 0.51 and 25.68 +/- 0.45 pg D N A / 2C nucleus, respectively. These population means did not differ significantly from one another and represent an approximate two-fold increase in D N A content relative to population 68, suggesting that these populations consist of tetraploid plants. However, in population 71 (mean 23.15 +/- 1.01 pg D N A / 2C nucleus) there are two discrete clusters of 2C values (Fig. 4-1). If the two discrete clusters of values are separated into two categories, one category corresponds to approximately the tetraploid level (mean 25.45 + /- 0.86 pg D N A / 2C nucleus), and one that corresponds to an approximately triploid value (mean 21.56 + /- 0.29 pg D N A / 2C nucleus). The mean of the tetraploid grouping does not differ significantly from the means from populations 66 and 67, while the triploid grouping mean differs significantly from the tetraploid grouping as well as all other population means. These are the first indications of triploidy in T. hookeri. Chloroplast haplotyping—After restriction digest, all individuals surveyed from population 68 had a diagnostic 400 bp fragment (type I). Populations 66 and 67 were monomorphic for the 67 alternate restriction site pattern (type II), in which the 400 bp fragment is cleaved into 2 fragments (190 and 210 bp). However, in population 71, one individual, 71q, (5% of the population sample) had the type I restriction site pattern, while the remaining 19 samples (95% of the population sample) all had the type II banding pattern. A F L P fingerprinting — 58 loci were scored that were variable among the four populations: 32 using primer EcoACC and 26 using primer EcoACG. Two samples (one from population 67 and one from population 68) did not amplify consistently and were excluded from further analyses. Within the total 78 samples, 39 multilocus genotypes were detected. Within population 66, 5 loci were polymorphic and 6 genotypes were detected; in population 67, 20 polymorphic loci defined 8 genotypes; population 68 had 16 genotypes based on 18 loci, while population 71 had 13 genotypes and 27 polymorphic loci. Over-representation of genotypes—Over-representation statistics are given in Table 4-1. Our results indicate that all three of the repeated multilocus genotypes found in population 66, as well as the most common multilocus genotypes in populations 67 and 71, were significantly over-represented. Additionally, the most common multilocus genotypes from populations 66 and 67 were identical. The recurrence of improbable genotypes among distinct localities is best interpreted as a consequence of clonal reproduction (Tibayrenc et al. 1991). None of the repeated multilocus genotypes within population 68 was significantly over-represented. Parsimony analyses—Most parsimonious trees from each population are shown in Figure 4-2. Phylogenetic analyses of populations 66 and 67 each resulted in one fully resolved tree (homoplasy index, HI = 0) of 5 and 20 steps, respectively, consistent with genetic variation being attributable to mutation alone. Permutation tail probability (PTP) tests indicate that the tree length for population 66 does not differ significantly from a randomized distribution of trees, likely due to low levels of genetic variation detected within the population. Population 67,.however, had a tree length that was significantly shorter than a randomized distribution of trees (PTP = 0.002). Contrastingly, phylogenetic analysis of population 68 yielded 556 equally most parsimonious trees with a length of 24 steps and an HI of 0.25. Population 71 was likewise rife with homoplasy (HI = 0.27), with 130 equally parsimonious trees uncovered with a tree length of 37 steps. For both populations 68 and 71, the length of optimal trees did not differ from a randomized distribution, as expected if recombination had occurred within these populations. The comb-like topology of the strict consensus of each set of poorly resolved trees is shown in Figure 4-2. Compatibility analyses— Results for compatibility analyses are shown in Figure 4-3. Population 66 had a matrix incompatibility count of zero, although this result is non-significant, likely due to a paucity of polymorphic loci within the population. Population 67 likewise had a matrix 68 incompatibility count of zero, and this was highly significant, indicating a lack of recombination within this population. Populations 68 and 71 had non-zero matrix incompatibility counts with all four possible combinations of band presence and absence being detected for pairs of loci within populations, consistent with the occurrence of recombination. After jackknifing the genotypes and recalculating the matrix incompatibility, it was determined that in order for population 68 to be fully compatible with genetic differentiation through mutation alone, 5 out the 16, or 31.3%, of the multilocus genotypes required successive deletion, indicating that 26.3% of the total population sample results from recombination. Within population 71,4 out of 14, or 28.6%, of the multilocus genotypes were successively deleted in order to make the dataset fully compatible with clonality. This represents 4 out of 20 or 20% of the population sample. Associations among loci— Table 4-2 gives the results of the associations, as measured by r2. Population 67 had the strongest association between loci with a value of 0.217, followed by population 71, which showed about 75% of the association of population 67. Population 66 showed the next lowest association (0.097), but this value did not significantly differ from zero, probably because of the low number of polymorphic loci (5) in this population. Population 68 had the lowest estimate of association, at 0.073. Permutation tests indicate that estimates of association among loci were significant in populations 67, 68 and 71.; Estimation of the frequency of long-term recombination—Table 4-2 shows that populations 66 and 67 have low rates of long-term sexuality, on the order of zero to one-half of one sexual events in the entire population per generation (i.e., one every two generations within the entire population),while population 68 has a significantly higher amount of sexuality within the population -40-100 times larger than populations 66 and 67. Population 71 exhibits a higher degree of recombination, with six outcrossing events for every five generations. Values of Nu were approximately an order of magnitude less than Ns. AA Discussion Clonality and sexuality within Easter daisy populations—Our results indicate that sub-arctic Easter daisies exhibit both clonality and recombination. The extent of clonality and recombination varies with population, male-fertility, and ploidy level. Below we summarize the results of each test for each population and interpret the likely causes of the observed patterns. Population 68 consisted of diploid plants that make viable pollen (mean pollen stainability of % 67.67%, Chapter 3) and undergo recombination. Population 68 exhibits no evidence of clonality by over-represented multilocus genotypes, as the two multilocus genotypes that were represented by more than one individual do not differ significantly from expectations for unlinked loci and panmixis. 69 Phylogenetic methods show that the most parsimonious tree has an unresolved comb-like structure and is not significantly shorter than randomized trees, as expected under recombination. Compatibility analyses concur that recombination is likely ongoing within this population. Although the association among loci indicates some degree of significant linkage disequilibrium, this population had the lowest value among the examined populations. This linkage disequilibrium may persist in the population due to the effects of selection, random drift within a small population, pre-equilibrium conditions, or some degree of non-random mating (e.g. selfing, biparental inbreeding). When considered together, our five tests give a picture of a population that is largely sexual. In complete contrast, population 67 consisted of male-sterile tetraploid plants with significant clonal structure. The most common multilocus genotype was significantly over-represented, as expected under clonality. A parsimony tree of population 67 was fully resolved, with no homoplasy detected, suggesting a lack of sex. Multilocus genotypes were completely compatible, indicating a significant lack of recombination within this population. This population also showed the highest estimate of association among loci, reflecting the highest amount of linkage disequilibrium. Collectively, these results indicate that although there may be much genetic variability within this population (e.g. 8/19 or 42% of sampled individuals were distinguishable), the structure of the multilocus genotypes appears consistent with an accumulation of mutation within clonal lineages. No convincing evidence of sexuality exists within population 66 and this population is likely clonal. This population also consisted of male-sterile tetraploid plants, and 3 of the 6 multilocus genotypes were significantly over-represented within this population, suggesting asexuality. One of these overrepresented multilocus genotypes was also significantly overrepresented in population 67; this geographic pattern being an expectation under clonality (Tibayrenc et al. 1991). There was no incompatibility among the multilocus genotypes, and these produced a non-homoplasious tree. A lack of significance for either test is likely to have resulted from an overall lack of variable loci (ie 5 variable loci were detected) within this population,. Association analysis indicated that linkage disequilbirium was present but insignificant, likely for the above stated reason. A more complex situation exists within population 71, which exhibits evidence of both clonality and recombination. The sample of plants examined here was male-sterile, however both tetraploid and triploid plants occurred in this population. The most common multilocus genotype was significantly over-represented within this population, indicating that these individuals were likely of clonal origin. Parsimony analyses yield a homoplastic tree, indicative of sexuality. Compatibility analysis likewise suggests sexuality, as four genotypes are recombinant within population 71. When these four individuals are removed from the parsimony analysis, a fully resolved tree lacking homoplasy results (data not shown), demonstrating that the remaining individuals within the 70 population are consistent with clonal reproduction. Association analyses provide the second highest value among the four populations, indicating some significant degree of linkage disequilibrium. Taken together, these results portray a picture of a population that comprises both sexual and clonal organisms. Additionally, a low level of chloroplast haplotype variation was detected, with 1/20 (5%) of the examined samples having the identical, haplotype to that within the diploid population. This could be a result of recent immigration through seed flow from the sexual population or incomplete lineage sorting of an ancestral polymorphism. It would be most interesting to know whether correspondence exists between the two ploidies and two chloroplast types and to know the ploidy levels of the recombinant genotypes. Unfortunately sampling was independent and destructive, i.e., we expected to find uniform tetraploidy in male sterile populations, as triploidy has never before been observed in Towensendia hookeri, and all available leaf material from seedlings was used for flow cytometry. Residual sexuality appears to be below the threshold of observation in this population as fertile males, if indeed present, were not detected in our sampling. This emphasizes the power of surveying the genome with molecular markers for evidence of sexuality: marker analyses indicate recombination, whereas morphological evidence indicates male-sterility and thus asexuality. Genetic markers that can be used to characterize clonal reproduction include isozymes, microsatellites, RAPDs (randomly amplified polymorphic DNA), ISSRs (intra-sequence simple repeats) and AFLPs (amplified fragment length polymorphisms). In terms of providing evidence for recombination, the ideal data set would include a large number of polymorphic co-dominant markers. In practice, isozyme and microsatellite data sets typically include a modest number of loci. In addition, when combined with low numbers of polymorphic loci, low numbers of alleles per locus may not allow fine-scale identification of clonal structure. Van der Hulst (2003) found that both microsatellite and A F L P data sets detected similar and consistent numbers of clones (33 and 32, respectively) in a single stand of dandelions, while isozymes discerned far fewer clones (18). When combined with small numbers of loci, the high mutation rate, and high potential for homoplasy of microsatellite alleles can further complicate detection of recombination and clonality. The alternative use of a large number of dominant markers, as typically obtained when using RAPDs, ISSRs or AFLPs remains attractive in many situations, as these data sets provide a genome-wide survey of broad utility that requires little initial investment relative to other D N A markers. However, these markers also exhibit dominance, a condition that warrants consideration when testing for clonality especially considering that 99% of known agamospermous plants and most asexual animals are polyploid (Suomalainen et al. 1987; Asker and Jerling 1992). The loss of information from individual 71 loci is partially offset by the large number of loci and decreased potential for homoplasy in AFLPs relative to microsatellites (Mariette et al. 2001). In a comparison of the utility of microsatellites and AFLPs, Gaudeul et al. (2004) concluded that A F L P data sets provided more accurate estimates of genetic diversity than microsatellites, provided that at least four times as many polymorphic A F L P fragments are used. This is especially true when patterns of gene flow among populations are low (Mariette et al. 2002). Alternate origins of clonal diversity -Mutation and occasional bouts of recombination are the most commonly invoked mechanisms to explain the levels of variation often uncovered within clonal populations. Of course our tests for clonality and sexuality, which allow for mutation and recombination, represent a simplification. Alternative sources of genetic variation include immigration via seed or pollen movement, as well as multiple origins of clones from sexuals. An advantage of this study of sub-arctic populations of the Easter daisy is that these alternate sources of genetic variation can be examined. Chloroplast haplotype results indicate limited immigration by sexuals into asexual populations. The one individual from polyploidy population 71 which had the "sexual" haplotype (type I, monomorphic within diploid population 68) was also one of the individuals that was responsible for patterns of incompatibility (i.e. recombination) within the population. Although the distance between these two populations is relatively large, recent migration from recombining population 68 could have been responsible for this low level of polymorphism, and thus could be contributing to genetic variation. This could also explain the occurrence of triploids, which could have arisen from crosses between diploids and tetraploids. Alternatively, the chloroplast polymorphism within population 71 may be much older, representing a lineage that remains as yet unsorted. This possibility is less likely, as Yukon populations of the Easter daisy tend to be quite small, typically under 100 individual plants, meaning that the time to fixation for chloroplast markers would be roughly 200 generations, thus requiring balancing selection to maintain polymorphism over the long term. In addition to seed migration, genetic variation due to pollen immigration from diploid populations could also occur. However, Easter daisies flower immediately upon spring thaw, a time of the year when pollinators are typically scarce. It is not unusual to find Easter daisies flowering under snow. As well, known diploid populations are several hundred kilometers from the examined tetraploid populations, although it is possible that there are undetected diploid populations elsewhere in the Yukon Territory. If pollen flow is a contributor to overall patterns and levels of genetic variation in Yukon populations is is likely to occur as a low rate. 72 Multiple origins of clones from sexuals can result in genetic variation within clonal populations. For the Easter daisy, a phylogenetic analysis of chloroplast DNA sequences indicates that the three polyploid populations examined in this study represent only one origin of polyploid asexuality (Chapter 3). Though this is a conservative estimate, there is no evidence to date that suggests that multiple origins of clones are rampant and widespread in the Yukon. As discussed in Eckert et al. (2003), dominance of marker polymorphisms can obscure minor genetic variation, as the rarer recessive genetic variants tend to be present only in heterozygotes,and hence are not detected. Polyploidy further compounds this problem, as the rarer recessive variants are detected only when completely homozygous, which can be quite rare for tetraploids. Hence for the same actual levels of genetic variation, levels of apparent variation and clonality should be generally less in tetraploids compared to diploids. This is not a significant problem with the over-representation analysis, as this is a test for presence/absence of clonality. With this test, dominance results in less power of detection of clonality in tetraploids compared with diploids, so that the trend of higher clonality in tetraploids is conservatively estimated. However, the prediction that genotypes in sexual diploid populations should exhibit greater incompatibility (MIC) than clonal tetraploid populations is confounded by dominance in the opposite direction. In the absence of any difference of clonality between diploid and tetraploid populations, the presence of dominance would reduce the frequency that both loci are recessive, and hence spuriously increase the compatibility of tetraploids, compared to diploids. The magnitude of this bias is difficult to predict though with multiple loci it should be less of a problem (Eckert et al. 2003). Since we do not compare raw MIC counts among population, dominance per se does not bias the inference of presence/absence of clonality, it only reduces the power to detect clonality in tetraploids, which again, makes inferences about greater clonality in tetraploids more conservative. Arguments can likewise be stated for homoplasy in phylogenies and the linkage-disequilbrium association metric. Future studies should seek broader explorations of the mechanisms through which clonal diversity can come to exist. Many such examinations exist within the microbial, fungal and animal literature (see Tibayrenc et al. 1990; Tibayrenc et al. 1991; Burt et al. 1996; Anderson and Kohn 1998; Halkett et al. 2005 and the entire volume 79(1) of the Biological Journal of the Linnean Society for some examples). The botanical literature, however, only furnishes studies from three taxa that adopt some, not all, of the methods advocated here: Taraxacum (Mes 1998; van der Hulst et al. 2000; van der Hulst et al. 2003), Allium vineale (Ceplitis 2001), and Butomus umbellatus (Eckert et al. 2003). 73 4.5 References Agapow PM, Burt A. 2001. Indices of multilocus linkage disequilibrium. Molecular Ecology Notes 1: 101-102. Anderson JB, Kohn LM. 1998. Genotyping, gene genealogies and genomics bring fungal population genetics above ground. Trends in Ecology and Evolution 13: 444-449. Arumuganathan K, Earle ED. 1991. Estimation of DNA content of plants by flow cytometry. Plant Molecular Biology Reporter 9: 229-233. Asker SE, Jerling L. 1992. Apomixis in Plants. CRC Press: Boca Rotan, USA Barraclough TG, Birky CW, Burt A. 2003. Diversification in sexual and asexual organisms. Evolution 57: 2166-2172. Beaman JH. 1954. Chromosome numbers, apomixis, and interspecific hybridization in the genus Townsendia. Madrono 12: 169-180. Beaman JH. 1957. The systematics and evolution of Townsendia (Compositae). Contributions from the Gray Herbarium of Harvard University 183: 1-151. Benham JJ. 2001. Genographer. http://hordeum.oscs.montana.edu/genographer. Bino RJ, De Vries JN, Kraak HL, Van Pijlen JG. 1992. Flow cytometric determination of nuclear replication stages in tomato seeds during priming and germination. Annals of Botany 69: 231-236. Burt A, Carter DA, Koenig GL, White TL, Taylor JW. 1996. Molecular markers reveal cryptic sex in the human pathogen Coccidioides immitis. Proceedings of the National Academy of Science, USA 93: 770-773. Ceplitis A. 2001. The importance of sexual and asexual reproduction in the recent evolution of Allium vineale. Evolution 58: 1581-1591. Dart S, Kron P, Mable BK. 2004. Characterizing polyploidy in Arabidopsis lyrata using chromosome counts and flow cytometry. Canadian Journal of Botany 82: 185-197. Diggle PK, Lower S, Ranker TA. 1998. Clonal diversity in alpine populations of Polygonum viviparum (Polygonaceae). International Journal of Plant Science 159: 606-615. Douglas GW, Argus GW, Dickson HL, Brunton DF. 1981. The Rare Vascular Plants of the Yukon. National Museum of Natural Sciences: Ottawa, Canada. Doyle JJ Doyle JD. 1987. A rapid DNA isolation procedure for small quantities of fresh leaf tissue. Phytochemisty Bulletin 19: 11-15. Eckert CG, Lui K, Bronson K, Corradini P, Bruneau A. 2003. Population genetic consequences of extreme variation in sexual and clonal reproduction in an aquatic plant. Molecular Ecology 12: 331-344. Ellstrand NC, Roose ML. 1987. Patterns of genotypic diversity in clonal plant species. American Journal of Botany 74: 123-131. Fager EW. 1972. Diversity: a sampling study. American Naturalist 106: 293-310. Faith DP, Cranston PS. 1991. Could a cladogram this short have arisen by chance alone?: on permutation tests for cladistic structure. Cladistics 7: 1-28. 74 Gabrielsen TM, Brochmann C. 1998. Sex after all: high levels of diversity detected in the arctic clonal plant Saxifraga cernua using RAPD markers. Molecular Ecology 7: 1701-1708. Goddard MR, Charles H, Godfray J, Burt A. 2005. Sex increases the efficacy of natural selection in experimental yeast populations. Nature 434: 636-640. Halkett F, Simon JC, Balloux F. 2005. Tackling the population geneics of clonal and partially clonal organisms. Trends of Ecology and Evolution 20: 194-201. Hamrick JL, Godt MJW. 1989. Allozyme diversity in plant species. In: Brown ADH, Clegg MT, Kahler AL, Weir BS, eds. Plant Population Genetics, Breeding, and Genetic Resources. 43-63. Sinauer Associates: Sunderland, USA. Hill WG, Robertson A. 1968. Linkage disequilibrium in finite populations. Theoretical and Applied Genetics 38:226-231. Jansen RK. 1992. Current research. Plant Molecular Evolution Newsletter 2: 13. Johnson JS, Bennett MD, Rayburn AL, Galbraith DW, Price HJ. 1999. Reference standards for determination of DNA content of plant nuclei. American Journal of Botany 86: 609-613. Johnson SG. 2000. Population structure, parasitism, and survivorship of sexual and autodiploid parthenogenetic Campeloma iimum. Evolution 54: 167-175. Jokela J, Lively CM, Dybdahl MF, Fox JA. 2003. Genetic variation in sexual and clonal lineages of a freshwater snail. Biological Journal of the Linnean Society 79: 165-181. Law JH, B. J. Crespi BJ. 2002. The evolution of geographic parthenogenesis in Timema walking-sticks. Molecular Ecology 11: 1471-1489. Legendre, P., AND L. Legendre. 1998. Numerical Ecology. Elsevier Science BV: Amsterdam. LeQuesne WJ. 1969. A method of selection of characters in numerical taxonomy. Systematic Zoology 18: 201-205. Love A. 1968. IOPB chromosome number reports. XV. Taxon 17: 91-104. Loxdale HD, Lushai G. 2003. Rapid changes in clonal lines: the death of a 'sacred cow'. Biological Journal of the Linnean Society 79: 3-16. Maneval WE. 1936. Lacto-genol preparations. Stain technology 11: 9-11. Mayer SS. 1991. Artificial hybridization in Hawaiian Wikstroemia (Thymelaeaceae). American Journal of Botany 78: 122-130. Maynard Smith, J. 1978. The Evolution of Sex. Cambridge University Press: Cambridge, UK. Meacham CA, Estabrook GF. 1985. Compatibility methods in systematics. Annual Review of Ecology and Systematics 16: 431-446. Mes THM. 1998. Character compatibility of molecular markers to distinguish asexual and sexual reproduction. Molecular Ecology 7: 1719-1727. Noyes RD, Rieseberg LH. 2000. Two independent loci control agamospermy (apomixis) in the triploid flowering plant Erigeron annuus. Genetics 155: 379-390. 75 Otto SP, Lenormand T. 2002. Resolving the paradox of sex and recombination. Nature Reviews Genetics 3: 252-261. Parker ED. 1979. Ecological implications of clonal diversity in parthenogenetic morphospecies. American Zoologist 19: 753-762. Pielou EC. 1969. An Introduction to Mathematical Ecology. Wiley-Interscience: New York. Posada D. 2002. Evaluation of methods for detecting recombination from DNA sequences: empirical data. Molecular Biology and Evolution 19: 708-717. Posada D, Crandall KA. 2001. Evaluations of methods for detecting recombination from DNA sequences: Computer simulations. Proceedings of the National Academy of Science, USA 98: 13757-13762. Stanley RG, Linskens HF. 1974. Viability tests. In: Pollen: biology, biochemistry and management. Springer: Berlin. 67-86. Stenberg P, Lundmark M . 2002. MLGsim v.1.0. http://www.molbiol.umu.se/forskning/saura/software.htm. Suomalainen E, Saura A, Lokki J. 1987. Cytology and Evolution in Parthenogenesis. CRC Press: Boca Raton, USA. Swofford DL. 2000. PAUP*. Phylogenetic analysis using parsimony (*and other methods). Sinauer Associates: Sunderland, USA. Tibayrenc M , Kjellberg F, Arnaud J, Oury B, Breniere SF, Darde ML, Ayala FJ. 1991. Are eukaryotic microorganisms clonal or sexual? A population genetics vantage. Proceedings of the National Academy of Science, USA 88: 5129-5133. Tibayrenc M , Kjellberg F, Ayala FJ. 1990. A clonal theory of parasitic protozoa: the population structures of Entamoeba, Giardia, Leishmania, Naegleria, Plasmodium, Trichomonas, and Trypanosoma and their medical and taxonomical consequences. Proceedings of the National Academy of Sciences, USA 87: 2414-2418. van der Hulst RGM, Mes THM, den Nijs JCM, Bachmann K. 2000. Amplified fragment length polymorphis (ALFP) markers reveal that population structure of triploid dandelions (Taraxacum officinale) exhibits both clonality and recombination. Molecular Ecology 9: 1-8. van der Hulst RGM, Mes THM, Falque M, Stam P, den Nijs JCM, Bachmann K. 2003. Genetic structure of a population sample of apomictic dandelions. Heredity 90: 326-335. Vos P, Hogers R, Bleeker M, Reijans M, van de Lee T. 1995. AFLP: a new technique for DNA fingerprinting. Nucleic Acids Research 23: 4407-4414. Widen B, Cronberg N, Widen M. 1994. Genotypic diversity, molecular markers and spatial distribution of genets in clonal plants, a literature survery. Folia Geobotanica & Phytotaxonomica 29: 245-263. Wilkinson M. 2001. PICA 4.0: software and documentation. The Natural History Museum: London, UK. 76 Table 4-1. Overrepresentation statistics for four sub-arctic populations of the Easter daisy (Townsendia hookeri). n indicates the number of individuals with the identical multilocus genotype (for n > 1) found within each population Population n p 1 sex P-value 66 12 2.85 x IO"14 0.00004 3 1.77 x 10"3 0.047 2 1.83 x 10"6 0.00034 67 11 5.22 x 10"4 0.014 2 0.292 0.54 68 3 0.330 0.71 2 0.162 0.53 71 7 1.67 x IO - 4 0.031 2 4.05 x IO"2 0.50 Table 4-2. Association statistics, measures of identity, and estimates of effective rates of long-term sexuality and mutation within four populations of the Easter daisy (Townsendia hookeri) Populatio n number of polymorphic loci, k associations among loci, r2 (SE) multilocus clonal identity,/(SE) single locus identity, g (SE) Ns (SE) Nu (SE) 66 5 0.097 (0.029) 0.368 (0.028) 0.048 (0.003) 0.497 (0.082) 0.006 (0.000) 67 20 0.217(0.020)* 0.327 (0.031) 0.147 (0.014) -0.206 (0.111) 0.022 (0.002) 68 18 0.073 (0.009)* 0.023 (0.005) 0.172 (0.008) 20.812 (6.735) 0.026 (0.001) 71 27 0.157(0.012)* 0.116 (0.016) 0.271 (0.017) 1.212 (0.589) 0.047 (0.004) 77 30n t o | 25H C O CM 20H < 10-66 67 68 71 Population Figure 4-1. DNA content per 2C nucleus as determined through flow cytometry for Yukon populations of the Easter daisy, Townsendia hookeri. Box plots give the mean and quartiles for each population. For populations 66,68, and 71, n = 6, while n = 4 for population 67. 78 B 66a, 66b, 66c, 66d, 66e, 66f, 66g, 66h, 66i, 66j, 66k, 661 • 66m • 66n 66o, 66p, 66q •66r 67a, 67b, 67c, 67d, 67e, 67f, 67g, 67h, 67i, 67j, 67k •671 67m •67n 67o, 67p 67q • 66s, 66t •67r Obs. = 5, PTP = 1.000 Obs. = 20, PTP = 0.002 •67s Obs. = 24, PTP = 0.276 68a, 68b, 68d 68c, 68f 68e 68g 68h 68i 68j 68I 68m 68n • 681 68o • 68p • 68q 1 68r 68s ,71a, 71b, 71c, 71d, 71e, 71f, 71p •71g •71m •71q •71h • 71i •71j •71k, 711 • 71n • 71o •71r •71s •71t Obs. = 37, PTP = 0.209 Figure 4-2. Most parsimonious trees (A, B) and consensus trees (C, D) from Yukon populations of the Easter daisy, Townsendia hookeri, based on AFLPs. The observed number of steps (Obs.) in the most parsimonious tree(s) and permutation tail probabilities (PTP) of the number of steps for a randomized distribution of trees are given. (A) population 66, (B) population 67, (C) population 68, (D) population 71. 79 B 100-50-Obs. = 0 PTP = 1.00 incompatibility count 30 -25-Obs. = 0 PTP = 0.001 0 1 2 3 4 5 6 7 8 9 10 incompatibility count 15^ £ 10-1 Obs. = 12 PTP = 0.237 II... 5 7 g 11 13 15 17 19 21 23 25 incompatibility count a 6 Obs. = 37 PTP = 0.932 "47" incompatibility count Figure 4-3. Incompatibility distributions for four Yukon populations of the Easter daisy, Townsendia hookeri. Observed incompatibility counts (Obs.) are indicated by arrows, while permutation tail probabilities (PTP) give levels of significance from 10000 randomizations. (A) population 66, (B) population 67, (C) population 68, (D) population 71 80 5 A novel mating system analysis for modes of self-oriented mating applied to diploid and polyploid sub-arctic Easter daisies (Townsendia hookeri)1 5.1 Introduction The mating system governs the transmission of gametes among generations, hence virtually all investigations of the dynamics of genetic change are concerned, directly or indirectly, with the mating process (Clegg 1980). Traditionally, estimation of mating systems using genetic markers has assumed a mixture of selfing and random outcrossing (Ritland 2002). This model stems from the conceptual advance of Allard and his coworkers (cf. Brown and Allard 1970) who demonstrated that plant populations could practice a mixture of mating types. Yet many asexual species continue to be treated as exhibiting one fixed mode of reproduction: either obligate or facultative asexuality. The criteria used to discriminate between these two categories varies from study to study. Additionally, selfing and asexuality have rarely been jointly considered as alternative deviations from sexual outcrossing. Mating patterns are best interpreted as a complex continuum, ranging from random outcrossing to selfing to complete asexuality (Bayer et al. 1990). "Self-oriented mating" is defined here as a transmission bias so that a greater proportion of an organism's genome is passed to its progeny relative to that expected under random outcrossing. Self-oriented mating systems include the reproductive phenomena of self-fertilization, automixis and apomixis. Self-fertilization (autogamy) is the union of products from different meioses from the same individual, and causes heterozygosity to be theoretically reduced by half within progeny for each generation. This differs from automixis (sometimes known as automictic parthenogenesis, reviewed in Mogie 1986), which is the fusion of two products of the same meiosis. Automixis is known to occur rarely among eggs in animals, among fungal meiospores, and among derivatives of the megasporocyte (egg sac precursor) in plants. Under automixis, heterozygosity is likewise lost among progeny, yet this loss occurs at a slower rate than under selfing, decreasing by one-third with each generation. Automixis has yet to be incorporated into mating system estimation methods. Apomixis (or agamospermy) is parthenogenetic reproduction through seed and can occur via a myriad of embryological pathways (see Nogler, 1984 for authoritative descriptions). Apomixis results in the transmission of an exact copy of the maternal genotype. Shifts in breeding system may occasionally be facilitated or induced by polyploidy (Richards, 1997). Polyploidy, the possession more than two complete chromosome sets, is a common class of 1 A version of this chapter has been published as Thompson SL, Ritland K. 2006. A novel mating system analysis for modes of self-oriented mating applied to diploid and polyploid arctic Easter daisies (Townsendia hookeri). Heredity 97 : 119-126. 81 genome change throughout plants and animals (reviewed in Otto and Whitton 2000). Polyploidy is known to impact selfing ability in some organisms (Stebbins 1950; Levin 1983; Cook and Soltis 2000; but see Mable, 2004), may lead to a breakdown of self-incompatibility (Chawla et al. 1997; Stone 2002) and may be accompanied by relaxed inbreeding depression (Husband and Schemske 1997) although increased inbreeding depression may also occur depending on the mode of gene action (Ronfort 1999). Polyploidy is also associated with asexual reproduction, in fact, 99 % of all known apomictic plants are polyploid but usually for asymmetric or odd ploidy levels (Nogler 1984; Asker and Jerling 1992). Due to its prevalence and potential impact, particularly within the flowering plants (Otto and Whitton 2000), mating system models must incorporate the additional layer of complexity which polyploidy introduces. The goal of the present study was to develop a procedure for measuring alternative modes of self-oriented mating in diploids and tetraploids for use with dominant molecular markers and to test hypotheses about mating patterns using one diploid and two polyploid populations of the sub-arctic Easter daisy {Townsendia hookeri, Asteraceae) from Canada's Yukon Territory. For each population, progeny arrays were assayed using dominant amplified fragment length polymorphism (AFLP) markers, and following joint estimation of the alternative modes of self-mating, likelihood ratio tests were conducted to test for presence/absence of mating modes. We expected less self-oriented mating to occur within the diploid population, as inbreeding depression is usually greater in diploids, and also expected apomixis to be mixed with selfing or automixis within polyploid populations, as apomixis is not necessarily a fixed condition in many plant populations (Bayer et al. 1990). 5.2 Materials and Methods A new estimation model for self-oriented matings—Estimation of mixed mating systems with tetraploidy and dominant markers has not been previously performed. Ritland (1990) presented a program, "mldt", for the case of diploid dominant markers, while Murawski et al. (1994) presented a model and program, "mltet", for tetraploids under co-dominance (both programs are available at http://www.genetics.forestry.ubc.ca/ ritland/programs.html). In addition, estimation of mating systems with potential automixis has never been considered with any type of marker or inheritance mode. Here we describe the procedure for using dominant markers to jointly estimate outcrossing and the three modes of self-oriented mating (apomixis, automixis and selfing) under either diploid or tetraploid inheritance. For the sexual modes (i.e. outcrossing and selfing), with no double-reduction in the tetraploid, gametes are sampled from parents without replacement. Double reduction (sampling with replacement) is unlikely as multivalents, a necessity for double reduction, were rarely observed in 82 tetraploid Townsendia hookeri (univalents are formed, Beaman 1957), and double reduction is generally not found in tetraploids (Julier et al., 2003). For the general case of parent genotype A,AyAfcA/, where subscripts index alternative alleles, there are six possible gametes, each of equal probability (1/6): A,A„ A/A*, A/A/, A/A*, A/A/, and A A A/. The probabilities of progeny gametes at a diallelic locus, conditioned on parent genotype, can be calculated. These are given in Table 5-1, for both codominance and dominance. From these gametic probabilities, the basic probabilities of the mating system model (the probabilities of offspring genotypes) are obtained, separately for outcrossing and selfing. These are given in Table 5-2 for the case of dominance, where q is the frequency of the recessive allele in the population. With selfing, the progeny frequencies are the squares of the gametic phenotype frequencies. With random outcrossing, the progeny frequencies are functions of the pollen pool frequencies of the dominant gamete (1 -q2) and of the recessive gamete, q2. These outcrossing probabilities are approximations that ignore the presence of inbred gametes (there are two copies of an allele in each gamete). To obtain probabilities of progeny for the parthenogenetic modes (i.e. apomixis and automixis), we assume no double reduction during meiosis, which means the maternal genotype is sampled without replacement. With apomixis, the meiotic product is the complete maternal genotype, e.g., the progeny phenotype is identical to the parental phenotype. With automixis the situation is more complicated. There are three equally probable ways that the alleles of the maternal genotype A / A / A A A / , can be distributed among meiotic products: (1) A/A /7A/A //AiA//AjfcA/, (2) A/A* /A/A* /A/A/ /A/A/, and (3) A/A/ /A,A/ /A/A* / A / A * . There are six potential matings within each of these three segregation patterns, leading to 18 possible automictic outcomes. Of these, 12 (or 3/4) will be tetra-allelic (same as the maternal genotype), while the remaining 6 will be bi-allelic (A,A,A/A/, A/A/A^A*, A,A/A/A/, AjAjAicAk, A/A/A/A/, A^A^A/A/), each with the probability of 1/18. These probabilities then lead to the offspring probabilities, given in Table 5-2 also for the case of dominance. For diploids, the probabilities of dominant progeny under automixis are 1, 5/6 and 0 for parent genotypes A A , Aa and aa respectively. Probabilities of dominant progeny under selfing are 1, 3/4 and 0, respectively, while those for outcrossing are 1, \-q/2 and l-q, respectively. In the estimation procedure below, we need to specify the probabilities of progeny, given the parent genotype, for an arbitrary combination of mating event probabilities. Denote the probabilities given in Table 5-2 with four arrays, T, S, t/and A corresponding to outcrossing, selfing, automixis and apomixis, and each array is indexed by two subscripts, corresponding to parent genotype and progeny phenotype, respectively. With t, s, u and a being the rates of outcrossing, selfing, automixis 83 and apomixis in the population (t+s+u+a=\), the probability that a progeny is a dominant phenotype, given parent j, is the mixture Pjx =tTjl+sSjl+uUJl+aAji and likewise with the subscript "2" for the recessive phenotype. Now, to estimate the rates of selfing, outcrossing, automixis and apomixis, we require several progeny to be collected from each of several parents, then genotyped. The progeny of a mother is termed a "progeny array". To analyze this progeny array data, the following assumptions are made: independence of mating events among progeny, constant gene frequencies, and constant probabilities of mating events. There are two major steps for progeny array analysis: inference of maternal parentage, and estimation of mating frequencies given the parentage. In the first step, the genotype of the maternal parent is inferred probabilistically from the progeny array as follows (cf. Ritland 1986). If in family i, Ni dominant offspring are observed and N2 recessive offspring are observed, the likelihood of the array of phenotypes, is Ly-itT^+sSj! +uUJ]+aAJ])N'(tTJ2 +sSj2+uUJ2 + aAj2)N> The probability of the array across all possible parents is the sum of the likelihoods over alternative parent genotypes, weighted by the frequency of parent genotypes in the population j where we assume fj = (\-qf, 4q(\-qy, 6q2(\-qf, 4q3(\-q), and qA for genotypes A A A A , A A A a , AAaa, Aaaa and aaaa, respectively. These are approximations that ignore inbreeding. We also assume the pollen gene frequencies equal the population gene frequencies (obtained as the fourth root of the recessive phenotype frequency for tetraploids; the square root for diploids) In the second step, estimates of mating frequencies are obtained by maximizing the likelihood function of the entire sample. This likelihood is the product of the Z, across arrays, and the Newton-Raphson method (see Ritland 1986) was used to maximize it. The likelihood ratio test was used to detect significant deviations of selfing, outcrossing, automixis and apomixis from zero. A computer program, tsu 'nami (f, s, u, 'n a mating mference) implementing this procedure and written in F O R T R A N 95 is available from KR upon request. Evaluation of model properties—To investigate the theoretical statistical properties of jointly estimating automixis with other mating modes, a F O R T R A N 90 program calculated the variance-covariance matix of the estimates of outcrossing rate t, selfing rate s and automixis rate u. For simplicity, apomixis was omitted as the focus was on the properties of u relative to 5 and t. Also for simplicity, maternal genotype and pollen gene frequency were assumed known. For recessive gene frequencies of g=0.5 and q=0.15, simulated data sets were generated for sample sizes /V=50, 100, and 84 200, then the Fisher information matrix (second derivatives of the log-likelihood function) was calculated and inverted to give the variance-covariance matrix, which was then multiplied by Nto standardize comparisons. This was replicated 100,000 times for each combination of N and q to obtain exact values. We also evaluated the ability of tsu 'nami to recover correct estimates by generating data under a uniform distribution of gene frequencies from 0.2 to 0.8, for various numbers of families and progeny array sizes, and using tsu 'nami to obtain estimates. We considered the four cases of r=l, 5=1, u-\ and a=\. For the diploid population, the same approach was used. Sampling of sub-arctic Easter daisy populations—In Canada's Yukon Territory, the perennial and diminutive Easter daisy, Townsendia hookeri, has rare status (Douglas et al. 1981) and is known to comprise disjunct diploid and autopolyploid populations (chapter 3, chapter 4). Diploid plants make viable pollen, while polyploid plants are male-sterile and putatively apomictic (Beaman 1957, chapter 3, chapter 4). Apomixis in Townsendia is of the Teen's' type (Beaman 1957, Nogler 1984) and is reported to be obligate within the genus (Beaman 1957). A peculiar meiotic abnormality within the megasporocyte (egg sac precursor) results in the formation of a collapsed restitution nucleus before the development of a 2N egg. The segregation pattern of multiple chromosome copies upon division of the restitution nucleus will dictate whether parthenogenesis is automictic (independent segregation) or apomictic (non-independent segregation leading to an outcome that is genetically equivalent to a mitosis). Pollination is neither required for the development of the embryo nor the endosperm in apomictic Townsendia (Beaman 1957). Plant material was collected from 3 Yukon populations of Townsendia hookeri: Mile Thirteen (60°59'N 135°10'W) and Tachal Dhal (61°00'N 138°32'W), both of which consist of tetraploids, and Tantalus Butte (62°07'N 136°15'W) which is a diploid population (chapter 4). Unopened flower heads were collected from 5-10 plants per population so that anthers could be dissected and male-fertility assessed. Fruits were collected from intact mature capitula from 10 maternal plants for each population. After 3 months of cold treatment, seeds were submerged in 7% sodium hypochlorite for 15 s, washed in three rinses of distilled water, placed on dampened filter paper, sealed within sterile petri plates, then imbibed for 3 days in the dark at 4 °C. Germinants were then transferred to 16 h daylength and room temperature conditions. Seedlings were reared for 5 weeks, until approximately 100 mg of leaf material could be obtained. A F L P assays—Genomic D N A was isolated from 6 progeny from each of the 30 maternal families by the C T A B (hexadecyltrimethylammonium bromide) protocol (Doyle and Doyle 1987) with the following modifications: volumes were reduced for extraction in 1.5 mL microfuge tubes and sodium metabisulfite (1% w/v) was added to the isolation buffer. Isolated D N A was quantified by a 85 Hoefer DyNAQuant™ 200 fluorometer (San Francisco, USA) according to the manufacturer's directions. Al l individuals were A F L P fingerprinted based on the method of Vos et al. (1995) with modifications developed by R. D. Noyes (Noyes and Rieseberg 2000). Denatured reaction products were run on an Applied Biosystems 3100 Avant D N A sequencer (Foster City, USA) according to the manufacturer's protocols. Eight primer combinations, (EcoACC, EcoACG) x (MeAGC, MseACG, MseACC, MseAAC),' were twice screened for 3 individuals from each of the 4 populations, and two primer combinations were selected according to the optimality criteria of consistency across all individuals, repeatability for each individual, and level of polymorphism. A l l individuals were fingerprinted using EcoACC / MseAGC and EcoACG / MseAGC primers then scored for band presence/ band absence using Genographer 1.6.0 (Benham 2001). 5.3 Results Properties of the estimation procedure—The results of the calculations of the theoretical variance-covariance properties of outcrossing rate t, selfing rate s and automixis rate u are given in Table 5-3. The estimates of selfing and automixis are very strongly negatively correlated, with values of -0.84 to -0.95, meaning that the statistical information to separate these two components of inbreeding is very slight. This accords with the simulations of estimates, and with the results of the field study, both of which indicated that selfing and automixis were highly confounded. Table 5-3 also shows that the variances of s and u are comparable, and are much larger than that for t, probably because s and u are confounded (if only s and t are estimated, their variances are equal since then 5=1-t). In addition, Table 5-3 shows that when sample sizes is small (/V=50), variances of estimates are significantly inflated above their asymptotic values (represented by N-200), particularly when the recessive allele is of lower frequency. Simulated data indicated that tsu 'nami correctly estimated the approximate true values when either selfing or automixis were omitted from the model. However, these simulations indicated that selfing and automixis are very difficult to distinguish. In the simulations, even with large sample sizes (20 families of size 10, and 1000 loci), datasets consisting of pure selfers gave estimates of about 50% selfing and 50% automixis, and conversely, datasets of pure automicts give similar 50-50 proportions of selfing and automixis. Pollen observations and AFLP results—Upon dissection, it was readily observed that plants from the Mile Thirteen and Tachal Dhal populations failed to produce viable pollen. The stamens of plants from these populations experience early abortion of their anthers. This was confirmed with plants collected in two other field seasons (SL Thompson personal observations). 86 A total of 107 A F L P loci were scored for all progeny. Of these 107, 60 loci were from the primer combination EcoACC / MseAGC and 47 were from the primer combination EcoACG / MseAGC. The distribution of gene frequencies in the three study populations, estimated by the procedure above, is given in Figure 5-1. As a general rule, loci are informative when the gene frequency is between about 0.2 and 0.8 (Ritland 1986). At frequencies outside of these "intermediate" values, most mating events involve parents and progeny homozygous for the same alleles, hence different types of matings cannot be distinguished. Figure 5-1 shows that the Mile Thirteen site had the most informative array of gene frequencies, with 35 of intermediate frequency. Tantalus Butte showed only 16 intermediate frequency loci, while Tachal Dhal showed 23 intermediate frequency loci. Thus we expect the highest precision of estimates in the Mile Thirteen population, and the lowest in the Tantalus Butte population. Mating patterns in three Easter daisy populations—Table 5-4 gives estimates of mating patterns and tests of various hypothesis. For each population, the first line is the case of all four parameters estimated (rates of outrossing, selfing, automixis and apomixis). The next four lines are hypothesis tests (1) no outcrossing, (2) no selfing, (3) no automixis, and (4) no apomixis. In these tests, three parameters are estimated the fourth set to zero. The last line is the test for (5) no selfing and no automixis; e.g., only outcrossing and apomixis was estimated. When all parameters were simultaneously estimated (line 1, Table 5-4), all populations showed a predominance of apomixis, from 65-72%, and lower levels of self-fertilization and automixis, ranging from 3 to 20%. Rates of outcrossing were low in the tetraploid populations, 2-4%, and moderate (23%) in the diploid population, Tantalus Butte. To determine the confidence interval for these estimates, log-likelihoods were found across a range of outcrossing rate t, for other parameters (s, u, a) jointly estimated (Figure 5-2). Note the sharp crest of likelihoods at low t for the tetraploid populations, while the diploid population (Tantalus Butte) has a much broader crest at a higher t, reflecting a higher uncertainty of estimate. To determine whether these differences are significant, we can estimate the confidence interval using the likelihood ratio test. From the point of maximum likelihood (Lm), the point / of lower likelihood at which -2(Lm -Li) equals 3.84 (chi-square with one degree of freedom) defines the upper and lower confidence interval (Venzon and Moolgavkar 1988). The confidence intervals are approximately ±0.02 for both tetraploid populations, and ±0.06 for the diploid population. Thus the higher outcrossing rate found in the diploid population is statistically significant. Table 5-4 also gives hypothesis tests about the presence of each mode of mating. The first four likelihood ratio tests in the last column of Table 5-4 involved comparisons between models with 4 parameters (t, s, u and a all estimated) vs. 3 parameters (either t, s, u and a set to zero), thus testing 87 for the presence of significant levels of these mating types. In these tests, -21n(L]/Lo), is asympotically distributed as chi-square with one degree of freedom, hence values greater than 3.84 are significant at the 95% level. A final test involved the constraint u=a=0 (no apomixis and no automixis), and this test has two degrees of freedom, with values greater than 6.99 being significant. In Table 5-4, the tests for t=0 and a=0 were both extremely significant, indicating that outcrossing and apomixis are present in all populations. In contrast, in all three populations, the tests clearly indicate that either automixis or selfing does not exist, as chi-square values were nearly zero. However, the test for s=u=0 was highly significant for the tetraploid populations, but not the diploid population (a value of 5.34 was found; 6.99 is needed for significance). Hence we conclude that either selfing or automixis exists in the tetraploid population, but that the data are not sufficiently informative to distinguish between these two reproductive modes. 5.4 Discussion Results of the estimation model—Empirical and theoretical treatments of mixed mating systems in plants typically model data as a mixture of selfing and outcrossing, ("mixed mating", reviewed in Goodwillie et al. 2005). Studies of asexuality, however, often employ discrete categories of obligate vs. facultative asexuality vs. random mating. Studies of outcrossing-selfing are rarely combined, or for that matter compared, to studies of sexuality-asexuality (Richards 1997). This study is the first, to our knowledge, where hypotheses of outcrossing and several types of self-oriented mating have been tested and quantified in tandem. Our procedure for the inference of mating systems involved assaying progeny arrays for genetic markers. This is more robust than earlier approaches that estimate population gene frequency and departures from expected heterozygosity under panmixis. The analyses of the segregation of genetic markers among progeny of a common mother are more direct and conclusive (Clegg 1980). We have taken the unique approach of testing several mating hypotheses to evaluate the strength of support for each type pf mating in relation to other modes of mating. In the two tetraploid populations of the Easter daisy, we detected significant levels of either automixis or selfing, however we could not ultimately distinguish between the two. Their lack of separation was theoretically confirmed by an extremely negative statistical correlation between the estimates of automixis and selfing in calculations involving the Fisher information index (Table 5-3). Evidently the segregation patterns do not sufficiently differ between these two modes of self-oriented mating for the case of tetraploidy and dominant markers. This confounding of the estimates of automixis and selfing can probably be greatly reduced by using co-dominant markers, and also by a multilocus approach. With co-dominant markers, the 88 segregation patterns of selfing vs. automixis are more distinct. At the multilocus level, "true" selfing occurs simultaneously at all loci within a single individual, while automixis would occur randomly across loci (with apomixis occurring at the other loci). The joint estimation of automixis with apomixis using these two alternative approaches is an area worth investigating further. Regardless, through the use of our new method of analysis, we conclude that apomixis is the dominant mode of reproduction in tetraploid populations of the Easter daisy (72% and 65%). The diploid population was also found to reproduce predominantly through apomixis (68%), a condition only rarely reported for diploid plants (Nogler 1984; Asker and Jerling 1992; Roy 1995). Relation to previous studies of apomicts—The breeding system of putative naturally-occurring apomicts has been the subject of several molecular studies involving progeny arrays. Classically, apomixis is detected as fixed heterozygosity among progeny, but the lack of suitable levels of isozyme variation and the complexity of interpreting banding patterns in high-level polyploids have been impediments in many cases (Roy 1995), while the expense in developing and assaying microsatellite markers can be prohibitive for others. Bayer et al. (1990) estimated the degree of apomixis vs. outcrossing for subpopulations of Antennaria media using codominant isozyme markers. Apomixis rates of 59% and 14% for .4. media led to the designation of subpopulations as obligately apomictic and partially apomictic, respectively. Roy (1995) used progeny arrays and isozymes to detect the presence of fixed heterozygosity (species labeled as apomicts) or homozygosity (species labeled as selfers) in six species of Arabis. Ford and Richards (1985) found that in several Taraxacum agamospecies, up to 62% of offspring had isozyme profiles that differed from their mothers. Based on microsatellite gel patterns, Robertson et al. (2004) found that progeny of Sorbus arranensis were all identical to their mothers and denoted as obligate, while in S. pseudofennica, 17.5%o of seeds differed in marker phenotype from their mothers. This species was deemed a facultative apomict. Bartish et al. (2001) classified Cotoneaster scandinavicus as apomictic based on identical R A P D (randomly amplified polymorphic DNA) profiles among progeny, but classified C. canescens as either selfing or apomictic based on estimates of Jaccard's coefficients of similarity of 0.973 and 0.979 for two accessions. Kollmann et al. (2000) detected low levels of A F L P variation among progeny from pseudogamous Rubus armeniacus and R. bifrons, leading to their characterization as facultative apomicts. Outcrossing or automixis was suspected in Rubus based on the observation that 14-17% of seedlings were genetically distinct from the maternal genotype. If selfing is occurring within sub-arctic polyploid populations of Easter daisies, pollen must be produced at unobservable low levels, as male-sterility was observed in natural populations over three field seasons (S L Thompson, personal observation). We favor automixis over selfing, as 89 embryological observations also suggest that automixis is a possibility for polyploid Townsendia (Beaman 1957). Automixis has been reported as a possible contributor to high levels of genetic variation within standing populations of apomictic Taraxacum (van der Hulst et al. 2003). Cytological investigations of meioses in natural and synthetic lines of Taraxacum confirm that automixis could be occurring (van Baarlen et al. 2000). The authors also speculate that automixis may be advantageous in the sense that it may limit the accumulation of mutations in parthenogenetic lines, as recessive deleterious mutations can become homozygous and exposed to natural selection. For example, the genotype Aaaa produces aaaa progeny 1/6 of the time (Table 5-2). The lack of selfing in the diploid Townsendia population suggests that i f selfing happens to occur in the tetraploids, which are derived from diploids (albeit perhaps through a triploid intermediate), then it is a derived condition. The transition from diploidy to tetraploidy may either directly cause a shift towards greater selfing, through breakdowns in systems of self-incompatibility (Chawla et al. 1997; Stone 2002), or indirectly cause an increase in selfing, through selection for increased selfing during the establishment of a neopolyploid undergoing a minority cytotype disadvantage (Ramsey and Schemske 1998). As well, in tetraploids, it is predicted that inbreeding depression should be reduced (Husband and Schemske 1997), although this has been rarely tested. If it is reduced, the deleterious effects of selfing should be tolerated to a greater extent. Also, the potentially selfing Townsendia tetraploids exclusively inhabit formerly glaciated regions of the Yukon Territory, while the non-selfing diploids are located in non-glaciated Beringia. Population recolonization can favor increased selfing, on both demographic grounds (only one individual required to form a new population, Baker 1955) and ecological grounds (more rapid fixation of successful genotypes within a selfing population, cf. Richards 1997). In the two tetraploid Townsendia populations, we found low but highly significant levels of outcrossing (Mile Thirteen: 2%, Tachal Dhal: 4%). This result is unexpected, as anthers are aborted before the development of pollen within the polyploid Easter daisy populations in the Yukon Territory (chapter 3, chapter 4). Plants flower immediately upon spring thaw, before many pollinators become active, and it is not uncommon to find them flowering under snow. It is possible that the low rates of outcrossing we detected are due to genotyping error. Bonin et al. (2004) have investigated sources of genotyping error and report rates of 2.6% for dwarf birch leaves, however, error rates vary with the organism. Possible sources of error include contamination (here reduced through sterile seedling and growth conditions), material used (we used constant tissues of similar age and development, as well as similar amounts of genomic DNA), amplification artifacts, and scoring errors. In future studies, scoring error might be reduced through consistent use of blind samples and 90 automation. At least, genotyping error should not introduce spurious evidence of selfing nor apomixis, as the errors normally introduce non-maternal alleles. In summary, this study demonstrates that plant breeding systems in sub-arctic environments can consist of many components, from outcrossing to the self-oriented modes of self-fertilization, automixis and apomixis. Dominant molecular genetic markers can help identify these modes, but cannot adequately separate automixis from selfing. Rather than thinking of species as strictly following one mode of reproduction (e.g. obligate apomixis, facultative apomixis) we should focus on the frequencies of various mating strategies within populations, which may ultimately lead to understanding the adaptive significance of these mixtures. 91 5.5 References Asker SE, Jerling L. 1992. Apomixis in Plants. CRC Press: Boca Rotan, USA. Baker HG. 1955. Self compatibility and establishment of long distance dispersal. Evolution 9: 337-349. Bartish IV, Hylmo B, Nybom H. 2001. RAPD analysis of interspecific relationships in presumably apomictic Cotoneaster species. Euphytica 120: 273-280. Bayer RJ, Ritland K, Purdy BG (1990). Evidence of partial apomixis in Antennaria media (Asteraceae: Inuleae) detected by the segregation of genetic markers. American Journal of Botany 77: 1078-1083. Beaman JH. 1957. The systematics and evolution of Townsendia (Compositae). Contributions from the Gray Herbarium of Harvard University 183: 1-151. Benham JJ. 2001. Genographer. v. 1.6.0 http://hordeum.oscs.montana.edu/genographer. Bonin A, Bellemain E, Eidesen PB, Pompanon F, Brochmann C, Taberlet P. 2004. How to track and assess genotyping errors in population genetics studies. Molecular Ecology 13: 3261-3273. Brown AHD, Allard RW. 1970. Estimation of the mating system in open-pollinated maize populations using isozyme polymorphism. Genetics 66: 133-145. Chawla B, Bernatzky R, Liang W, Marcotrigiano M. 1997. Breakdown of self-incompatibility in tetraploid Lycopersicumperuvianum: inheritance and expression of 5-related proteins. Theoretical and Applied Genetics 95: 992-996. Clegg MT. 1980. Measuring plant mating systems. Bioscience 30: 814-818. Cook L M , Soltis PS. 2000. Mating systems of diploid and allotetraploid populations of Tragopogon (Asteraceae). II. Artificial populations. Heredity 84: 410-415. Douglas GW, Argus GW, Dickson HL, Brunton DF. 1981. The Rare Vascular Plants of the Yukon. National Museums of Canada: Ottawa. Doyle JJ, Doyle JD. 1987. A rapid DNA isolation procedure for small quantities of fresh leaf tissue. Phytochemisty Bulletin 19: 11-15. Ford H, Richards AJ. 1985. Isozyme variation within and between Taraxacum agamospecies in a single locality. Heredity 55: 289-291. Goodwillie C, Kalisz S, Eckert CG. 2005. The evolutionary enigma of mixed mating systems in plants: occurrence, theoretical explanations and empirical evidence. Annual Review of Ecology and Systematics 36: 47-79. Husband BC, Schemske, DW. 1997. The effect of inbreeding in diploid and tetraploid populations of Epilobium angustifolium (Onagraceae): implications for the genetic basis of inbreeding depression. Evolution 51: 737-746. Julier B, Flajoulot S, Barre P, Cardinet G, Santoni S, Huguet T, Huyghe C. 2003. Construction of two genetic linkage maps in cultivated tetraploid alfalfa (Medicago saliva) using microsatellite and AFLP markers. BMC Plant Biology 3: 9-28. Kollmann J, Steinger T, Roy, BA. 2000. Evidence of sexuality of European Rubus (Rosaceae) species based on AFLP and allozyme analysis. American Journal of Botany 87: 1592-1598. 92 Levin DA. 1983. Polyploidy and novelty in flowering plants. The American Naturalist 122: 1-25. Mable BK. 2004. Polyploidy and self-compatibility: is there an association? New Phytologist 162: 803-811. Mogie M . 1986. Automixis: its distribution and status. BiologicalJournal of the Linnean Society 28: 321-329. Murawski DA, Fleming TH, Ritland K, Hamrick, JL. 1994. Mating system in Pachycereuspringlei: an autotetraploid cactus. Heredity 72: 86-94. Nogler GA. 1984. Gametophytic apomixis.in B. M. Johri, ed. Embryology of Angiosperms. Springer-Verlag: Berlin, DE. 475-518 Noyes RD, Rieseberg LH. 2000. Two independent loci control agamospermy (apomixis) in the triploid flowering plant Erigeron annuus. Genetics 155: 379-390. Otto, SP, Whitton J. 2000. Polyploid incidence and evolution. Annual Reviews of Genetics 34: 401-437. Ramsey J, Schemske DW. 1998. Pathways, mechanisms, and rates of polyploid formation in flowering plants. Annual Reviews in Ecology and Systematics 29: 467-501. Richards AJ. 1997. Plant Breeding Systems. Chapman and Hall: London, UK. Ritland K. 2002. Extensions of models for the estimation of mating systems using n independent loci. Heredity 88: 221-228. Ritland K. 1990. A series of FORTRAN programs for estimating plant mating systems. Journal of Heredity 81: 235-237. Ritland K. 1986. Joint maximum likelihood estimation of genetic mating structure using open-pollinated progenies. Biometrics 42: 25-43. Robertson A, Newton AC, Ennos RA. 2004. Breeding systems and continuing evolution in the endemic Sorbus taxaonArran. Heredity 93: 487^195. Ronfort J. 1999. The mutation load under tetrasomic inheritance and its consequences for the evolution of the selfing rate in autotetraploid species. Genetical Research 74: 31-42. Roy BA. 1995. The breeding systems of six species of Arabis (Brassicaceae). American Journal of Botany 82: 869-877. Stebbins GL. 1950. Variation and Evolution in Plants. Columbia University Press: New York, USA. Stone JL. 2002. Molecular mechanisms underlying the breakdown of gametophytic self-incompatibility. Quarterly Review of Biology 77: 17-32. van Baarlen P, van Dijk PJ, Hoekstra RF, de Jong JH. 2000. Meiotic rcombination in sexual diploid and apomictic triploid dandelions (Taraxacum officinale L.). Genome 43: 827-835. van der Hulst RGM, Mes THM, Falque M, Stam P, den Nijs JCM, Bachmann K. 2003. Genetic structure of a population sample of apomictic dandelions.' Heredity 90: 326-335. Venzon DJ, Moolgavkar SH. 1988. A method for computing profile-likelihood based confidence intervals. Applied Statistics 37: 87 -94. Vos P, Hogers R, Bleeker M, Reijans M, van de Lee T. 1995. AFLP: a new technique for DNA fingerprinting. Nucleic Acids Research 23: 4407-4414. 93 Table 5-1. Probabilities of gametes from auto tetraploid parents assuming double reduction, under co-dominance and dominance, respectively. ' Gamete AAAA AAAa Parent genotype AAaa Aaaa aaaa codominance AA 1 1/2 1/6 0 . 0 Aa 0 1/2 2/3 1/2 0 aa 0 0 1/6 1/2 1 dominance A_ 1 1 5/6 1/2 0 aa 0 0 1/6 1/2 1 Table 5-2. Probabilities of tetraploid offspring genotypes with dominance and assuming no double reduction, under outcrossing, selfing, automixis and apomixis, respectively. Offspring Parent genotype genotype AAAA AAAa AAaa Aaaa aaaa outcrossing A 1 1 0-<72/6) (W 2 /2) aaaa 0 0 q2/6 q2/2 selfing A 1 1 35/36 3/4 0 aaaa 0 0 1/36 1/4 i automixis A 1 1 17/18 5/6 0 aaaa 0 0 1/18 1/6 1 apomixis A 1 1 1 1 0 aaaa 0 0 0 0 1 Table 5-3. Theoretical variances and correlations of estimates per individual sampled when selfing rate, automixis and outcrossing rate are simultaneously estimated, for two frequencies of the recessive marker q, and three levels of sample size (N). N Var(i) Var(w) Var(f) Coxx{s,u) Corr(.?, i ) Corr(«,f) <7 = 0.5 50 93.4 218.4 69.64 -0.84 0.34 -0.78 100 62.5 72.7 7.44 -0.94 0.06 -0.37 200 56.8 63.2 5.04 -0.95 0.04 -0.32 9 = 0.75 50 59.8 58.5 1.29 -0.98 -0.14 -0.0009 100 55.9 54.7 1.18 -0.98 -0.14 -0.0023 200 54.3 53.2 1.14 -0.98 -0.14 -0.0041 94 Table 5-4. Estimates of outcrossing /, selfing s, automixis u, and apomixis a, for each of the three populations, under various hypotheses. Also given are log-likelihoods for each model, and likelihood ratio tests for absence of each component of the mating system relative to the full model (negative values are due to slight errors in numerical convergences of likelihood functions). Model t s u a LnL -2Ln(L,/L0) Mile Thirteen (tetraploid) full 0.02 0.06 0.20 0.72 -2226.88 t=0 0.00 0.37 0.13 0.51 -2507.05 560.34 s=0 0.02 0.00 0.29 0.69 -2226.71 -0.35 u=0 0.02 0.21 0.00 0.77 -2227.50 1.23 a=0 0.02 0.00 0.99 0.00 -2287.06 120.36 s=u=0 0.07 0.00 0.00 0.93 -2309.58 165.39 Tachal Dhal (tetraploid) full 0.04 0.16 0.15 0.65 -1568.06 t=0 0.00 0.95 0.04 0.01 -1954.38 772.65 s=0 0.04 0.00 0.37 0.59 -1568.53 0.94 u=0 0.04 0.26 0.00 0.70 -1567.78 -0.56 a=0 0.04 0.00 0.96 0.00 -1597.42 58.72 s=u=0 0.09 0.00 0.00 0.91 -1626.28 116.45 Tantalus Butte (diploid) full 0.23 0.03 0.08 0.68 -1943.71 t=0 0.00 0.99 0.01 0.00 -2526.75 1166.08 s=0 0.21 0.00 0.12 0.67 -1943.71 -0.01 u=0 0.21 0.08 0.00 0.71 -1943.71 0.00 a=0 0.19 0.00 0.81 0.00 -1983.01 78.59 s=u=0 0.25 0.00 0.00 0.75 -1946.38 5.34 95 35 30 25 20 15 10 5 0 Mile Thirteen 0.0 0.2 0.4 JZL, 0.6 0.8 1.0 35 30 -I 25 20 -I 15 10 5 0 Tachal Dhal o.o 0.2 0.4 35 30 25 20 15 10 5 0 n , n 0.6 0.8 Tantalus Butte (diploid) o.o 0.2 0.4 0.6 0.8 1.0 1.0 Frequency of recessive allele Figure 5-1. Gene frequency distribution for A F L P markers from one diploid (Tantalus Butte) and two tetraploid (Mile Thirteen and Tachal Dhal) sub-arctic populations of the Easter daisy, Townsendia hookeri. 96 -1400 -1600 4 -1800 4 -2000 -2200 -2400 4^  -2600 A -2800 Tachal Dhal Tantalus Butte Mile Thirteen o.o 0.1 0.2 0.3 0.4 Outcrossing rate Figure 5-2. Log-likelihoods across a range of outcrossing rate /, for other parameters (s, u, a) jointly estimated. Note the sharp crest of likelihoods at low t for the tetraploid populations, while the diploid population (Tantalus Butte) has a much broader crest at a higher t. 97 6 Conclusions A l l aspects of evolutionary change are governed, either directly or indirectly, by an organism's mating system. In turn, the mating system evolves in response to reproductive challenges, environmental variation and genetic background. Species that differ both in mating system and ploidy offer a fascinating opportunity to study these interactions. In this thesis, molecular and statistical techniques were used to examine genetic structure and hence infer patterns of mating and ploidy in the Townsendia agamic complex. The structure was examined at four hierarchical levels: (chapter 2) within the genus, (chapter 3) within T. hookeri, (chapter 4) within standing populations of this species, and (chapter 5) among progeny arrays from these populations. Chapter 2: Phylogenetic analyses of the external transcribed spacer (ETS) and the internal transcribed spacers (ITS) of the rDNA cistron demonstrated that polyploid apomixis evolved multiple times within Townsendia and these traits were negatively associated with endemism. This was paralleled with the recurrent evolution of a diminutive, reduced acauelscent habit as well as repeated evolution of localized edaphic endemics. Endemism was significantly associated with taxa that solely comprise sexual diploids, although this association was not corrected for phylogeny. The relationships among many of the species of Townsendia were poorly resolved by these data. Analyses at this broad level indicated that transitions to polyploidy represent a frequently occurring class of mutation and that the complex series of genetic changes that allow for the simultaneous expression of diplospory, parthenogenesis, and autonomous endosperm formation can evolve repeatedly within a group of plants. Chapter 3: Geographic patterns of parthenogenesis and the number of transitions from sexual diploidy to apomictic autopolyploidy were examined for 40 populations of the Easter daisy, Townsendia hookeri. Pollen diameter and stainability characterized 15 sexual diploid and 25 apomictic polyploid populations from throughout the plant's western North American range. Sexual diploids were restricted to two Wisconsin refugia: Colorado/Wyoming, south of the ice-sheets, and northern Yukon/Beringia. Phylogenetic analyses of 10 chloroplast D N A haplotypes indicated that five exclusively polyploid haplotypes were derived from four haplotypes, which are shared among ploidies, conservatively inferring a minimum of four origins of apomictic polyploidy. Three of these apomictic polyploid origins were from southern sexual diploids, while the fourth origin was from northern sexual diploids. Analyses of regional diversity were suggestive of a formerly broad distribution for sexual diploids that has become subsequently fragmented, likely due to the last round of glaciation. As sexual diploids were exclusively found north and south of the glacial maximum, while formerly-glaciated areas were exclusively inhabited by asexual polyploids derived from both northern and southern sexual lineages, it is more likely that patterns of glaciation, as opposed to a 98 particular latitudinal trend, played a causal role in the establishment of the observed pattern of geographic parthenogenesis in Easter daisies. Chapter 4: Pollen studies, flow cytometry, and multilocus tests on AFLP marker genotypes characterized the extent of clonality and sexuality within four populations of 71 hookeri from the Yukon Territory. These northern disjunct populations were discovered subsequent to Beaman's monograph of the genus and were of unknown ploidy and breeding system. Results indicated sexuality in one male-fertile diploid population, clonality in two male-sterile tetraploid populations, and a combination of sexual and clonal reproduction in one male-sterile population. This latter population of mixed sexual and clonal mating consisted of triploids and tetraploids, and provided evidence that cryptic sex may linger in the genomes within a morphologically asexual population. Chapter 5: A new method for mating system analysis, which jointly estimates the rates of outcrossing, selfing, automixis and apomixis, was developed and applied to dominant A F L P marker genotypes from progeny whose mothers arose from three Yukon populations of T, hookeri, one consisting of male-fertile diploids, the other two of male-sterile tetraploids. Apomixis was the predominant mating mode in all populations. Moderate levels of outcrossing were detected in the diploid population but not in the tetraploids, while selfing/automixis was absent in the diploids and moderate in tetraploids. These findings suggest that the correlation between ploidy and apomixis is not strict when observed on a fine scale, that polyploidy alone does not induce apomixis, and that the transition to polyploidy is accompanied by an increase in self-oriented mating. The overarching conclusion of this thesis, as we progress down the hierarchy of population structure, from the genus, to the species, to the population, to the progeny array, is that well-defined patterns break down, and a complexity emerges. While the dogma that diploids are sexual and polyploids are apomitic seems to hold at higher levels (chapter 2, 3), more detailed studies of the genome indicated that the classic dichotomy of sexual diploids and apomictic polyploids is not so strict within Townsendia (chapters 4, 5). When averaged across populations throughout the species range, polyploids were correlated with the low levels of male-fertility found in apomicts (chapter 3). However, when male-sterile standing populations were dissected, polyploid genomes showed evidence of cryptic sex (chapter 4). As we delved deeper by assessing the progeny structure among polyploids, we found evidence of selfing or automixis in addition to apomixis (chapter 5). This indicates that polyploid individuals can produce offspring in various ways, not just apomictically. As diploidy is an extraordinarily rare condition in apomictic plants, it is noteworthy that diploid progeny faithfully reproduce the maternal genotype, in addition to being derived through random mating (chapter 5). If we reconsider evidence from diploid standing populations (chapter 4), association analyses indicate a low but significant level of linkage disequilibrium. This may be due to 99 apomixis among diploids. However, tests for recombination (i.e., compatibility analyses and phylogenetic analyses) show that sex occurs as well. It could be that even a low to moderate outcrossing rate is sufficient to eradicate the signature of clonal structure in standing populations, as multilocus geneotypes were not overrepresented in the diploid population. As slow-growing progeny were lab-reared until seedlings were merely large enough to extract suitable levels of DNA, it remains to be seen whether apomictic diploids live to maturity and reproduce successfully in nature. It could be that apomictic diploids are excluded from standing populations via competition with sexually-produced diploids. Of mention is the outlier in Fig 3-2 from population 18, which comprised diploid individuals that produced pollen of very low viability (chapter 3). This may have resulted from "male-meltdown" within a diploid population that is primarily apomictic. As our fine-level studies of genetic variation in standing populations (chapter 4) and progeny arrays (chapter 5) were conducted exclusively from northern lineages of T. hookeri (chapter 3), it remains to be seen whether southern diploids reproduce apomictically. Studies of the breeding system within southern diploid populations would demonstrate whether the propensity to reproduce apomictically is universal in Easter daisies, occurring among diploids of different glacial refugia, or whether apomictic reproduction is a strategy that is restricted to northern diploids. Apomictic reproduction may have particular advantages in northern environments, where plants experience shortened and unpredictable growing seasons, cold temperatures, water and nutrient limitations, as well as low and variable rates of pollination. These conditions favor the reproductive assurance that apomixis provides. The detection of apomixis within diploid individuals suggests that polyploidy is not necessary for the expression of the trait. However, polyploids still remain more broadly distributed than diploids (chapter 3), regardless of mating system. In light of this, the conclusion might be that polyploidy is of advantage to an apomict, by offering the benefit of increased genome copies as an alternative to recombination. This could lead to broader ecological tolerances and a genetic load that is more-buffered over the short term. Perhaps the shift toward less outcrossing that we found in polyploids played a role in the recolonization of post-glacial habitats. The genus Townsendia offers many future opportunities for investigations into the origin and maintenance of polyploidy and apomixis. Co-dominant markers such as single nucleotide polymorphisms (SNPs) and microsatellites (SSRs), in addition to nuclear gene sequences, may provide greater statistical power in order to improve the resolution of population relationships and mating systems. However, these types of markers require a significant financial investment for their development. Greenhouse-based experiments are largely impractical at this time due to long generation times. Further research that relies upon field observations and collections remains feasible. 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0092998/manifest

Comment

Related Items