INFORMATION TO USERS. This manuscript has been reproduced from the microfilm master. UMI films the text directly from the original or copy submitted. Thus, some thesis and dissertation copies are in typewriter face, while others may be from any type of computer printer. The quality of this reproduction is dependent upon the quality of the Copy submitted. Broken or indistinct print, colored or poor quality illustrations and photographs, print bleedthrough, substandard margins, and improper alignment can adversely affect reproduction. In the unlikely event that the author did not send UMI a complete manuscript, and there are missing pages, these will be noted. Also, if unauthorized Copyright material had to be removed, a note will indicate the deletion. Oversize materials (e.g., maps, drawings, charts) are reproduced by ‘sectioning the original, beginning at the upper left-hand corner and continuing from left to right in equal sections with small overlaps. ProQuest Information and Learning 300 North Zeeb Road, Ann Arbor, MI 48106-1346 USA 800-521-0600 ® UMIINTERCONTINENTAL DISPERSAL, BIOGEOGRAPHY, AND SPECIATION IN A FRESHWATER ZOOPLANKTER: INVESTIGATIONS OF THE DAPHNIA OF ARGENTINA A Thesis Presented to ‘The Faculty of Graduate Studies of The University of Guelph By SARAH J. ADAMOWICZ In partial fulfillment of requirements for the degree of Master of Science April, 2002 © Sarah J. Adamowicz, 2002205 Wellington Street 26, nm Wetingion ‘Onawe ON KIA ON ‘Oumem ON KIA ONG Canada Cored The author has granted a non- exclusive licence allowing the copies of this thesis in microform, paper or electronic formats. ‘The author retains ownership of the copyright in this thesis. Neither the thesis nor substantial extracts from it may be printed or otherwise reproduced without the author’s Yar te vor ‘L’auteur a accordé une licence non exclusive permettant 4 la Bibliothéque nationale du Canada de reproduire, préter, distribuer ou vendre des copies de cette thése sous la forme de microfiche/film, de reproduction sur papier ou sur format électronique. ‘L’auteur conserve la propriété du droit d’ auteur qui protége cette thése. Nila thése ni des extraits substantiels de celle-ci ne doivent étre imprimés ou autrement reproduits sans son autorisation. 0-612-71187-0ONVERSTY Aono, Sh on 0101989 Msc.Z00 Zoology GRADUATE PROGRAM SERVICES. CERTIFICATE OF APPROVAL (MASTER'S THESIS) ‘The Examination Committee has concluded that the thesis presented by the above-named candidate in partial fulfilment of the requirements for the degree Master of Science is worthy of acceptance and may now be formally submitted to the Dean of Graduate Studies. Title: __Intercontinental Dispersal, Biogeography, and Speciation in a Freshwater Zooplankter: Investigations of the —— of Argentina Chair, Master's ii P.D.N. Hebert Dr. iL 17 okD2. ate Recsved by tellaheng pw _MAY - 2 2002 for Dean of Graduate Studies v. Optional Graduate Faculty MemberABSTRACT INTERCONTINENTAL DISPERSAL, BIOGEOGRAPHY, AND SPECIATION IN A. FRESHWATER ZOOPLANKTER: INVESTIGATIONS OF THE DAPHNIA OF ARGENTINA Sarah J. Adamowicz Advisor: University of Guelph, 2002 Professor Paul D.N. Hebert This thesis investigates the interactions among long-distance dispersal, environmental conditions, and phylogenetic constraints in shaping the evolutionary diversification of a freshwater zooplankton assemblage. The initial section of the thesis focuses on the use of genetic analyses to resolve species boundaries in the Daphnia fauna of Argentina. New and published DNA sequences from Daphnia from four continents were subsequently used to construct a phylogenetic framework for the genus in order to examine the factors influencing the distribution of polyploid taxa, the role of intercontinental allopatric divergence in the diversification of the genus, and the relative importance of geographic, genetic, and ecological modes of speciation in the evolution of the group. In addressing each of these topics, this study highlighted the importance of long-distance dispersal in promoting speciation and in structuring large-scale biogeographic pattems. While chance historical events have played a fundamental role in the evolution of daphniid diversity, this study revealed that clade-specific properties are also important in determining overall pattems of diversity.Acknowledgements This research was supported by a NSERC research grant and Canada Research Chairs grant to Paul Hebert and by a NSERC post-graduate scholarship to Sarah Adamowicz. Chapter Two was written in collaboration with Ryan Gregory, Maria Cristina Marinone, and Paul Hebert. I thank Teri Crease and two anonymous reviewers for helpful comments on an earlier draft. This paper is currently in press at Molecular Ecology and has been modified to fit the thesis format. The field campaigns would not have been possible without the extensive help and expertise provided by Agustin Bachmann, Paul Hebert, Cristina Marinone, Juan Cesar Paggi, and Silvina Meni Marque. Thank you for sharing in the work and the adventures! Sincere thanks also to Elena Bachmann, Gerald Bachmann, Santiago Echaniz, Alfonso Giudici, Marina Giudici, Ana Maria Marinone, Raquel de Marinone, Susana Jose de Paggi, and Alicia Vignatti. I greatly appreciated their warm hospitality and assistance during my stays in Argentina. I thank Cristina Marinone, Femando Martinez Jerénimo, Cheryl Prokopovich, Veronika Sacherové, Stephen Schwartz, Alcira Villagra de Gamundi, and Horacio Zagarese for graciously providing me with zooplankton samples. Thanks also to John Colboume for generously contributing much sequence data to this project. Thanks to Angela Holliss for her help with sequencing and to Ian Smith for technical assistance. I would like to thank the members of my advisory committee (Teri Crease, Denis Lynn, and Paul Hebert) for their guidance and support and the members of my defence committee (Jinzhong Fu, Paul Hebert, Denis Lynn, and Vernon Thomas) for valuablecomments and criticisms of the thesis. I also thank my former mentor, Cindy Staicer of Dalhousie University, for providing me with my first opportunity to conduct research and for her continuing encouragement. I am deeply grateful to Paul Hebert for the amazing opportunity to work on the zooplankton in South America and for his enthusiasm and support throughout my degree. Heartfelt thanks to all my present and former lab-mates, who have each contributed to my graduate experiences: Shelley Ball, Andrea Cox, Melania Cristescu, Alina Cywinska, Rob Dooh, Ryan Gregory, Dave Hardie, Ed Remigio, James Rhydderch, Veronika Sacherové, Lisa Schieman, Kareen Schnabel, Rachel Sutton, Jeremy deWaard, and Jonathan Witt. Special thanks to Jonathan for his patience in training me in the allozyme lab, Melania for assistance with DNA techniques, and Veronika for being a kindred spirit. I would especially like to thank Ryan for helping me with so many aspects of this thesis, both seen and unseen ~ for all your encouragement and support, for discussions and advice, for countless cups of tea, for boosting me when my resolve was wavering... Thank you! Finally, I would like to thank my mother (Carol), dad (Steve), sister (Emily), and grandmothers (Helen Adamowicz and Catherine Klink) for their love and support.Table of Contents List of Tables. Introduction: The diversification of freshwater life... Invasion of the freshwater realm.. Evolutionary consequences of dispersal via resting eggs. Intercontinental speciation in Daphnia: towards the complete picture. Objectives of the thesi Chapter 1: Species diversity and endemism in the Daphnia of Argentina: a genetic investigation. Comparison with North American species... Results. Diversity in the subgenus Daphnia ivCOI sequence variation. Allozyme variation. Evaluation of relatedness to North American Daphnia species.. Diversity in the subgenus Hyalodaphnia. COI sequence variation. Allozyme variation. Evaluation of relatedness to North American Hyalodaphnia species. Diversity in the subgenus Ctenodaphnia. COI sequence variation. Allozyme variation. Evaluation of relatedness to North American Ctenodaphnia species... Discussior Species diversity of Argentine Daphnia: interpretation of genetic evidence... Subgenus Daphnia. Hyalodaphnia. Ctenodaphnia. Summary .. Continental endemism and intercontinental dispersal in Daphnia. Concluding remarks. Chapter 2: New insights into the distribution of polyploid Daphnia: the Holarctic revisited and Argentina explored.. Abstract...Methods. Specimen collection and identification. Allozyme analysis... ‘The first case of temperate polyploid Daphnia. Polyploid in the north: production and persisten Limits to polyploid distribution: competitive exclusion? Polyploids in the south: fortunate founders?.. Concluding remarks. Chapter 3: Intercontinental phylogeography and allopatricMolecular clock application... Results... Phylogenetic reconstructions. ‘Subgenus Crenodaphnia M2 Global biogeographic patterns: recent migrants and ancient endemics. Concluding Remarks. U7 133 Evidence of speciation mechanisms in Daphnia.. 135 Speciation in allopatry.. 136 Geographic speciation at an intercontinental scale.. oe 136 Intracontinental allopatric speciation. 137Speciation by habitat shifts. . 152 Summary of patterns of speciation: trends across clades and biogeographic regions. 154 ‘Summary: mechanisms of diversification in Daphnia. ‘Comparison of pattems of speciation and biogeography among passive dispersers.. Endemism and intercontinental allopatric speciation.. Intrinsic factors: polyploidy and hybridization. Ecological speciation. 100 million years of speciation in freshwater zooplankton: macroevolutionary noise?. 213 215List of Tables Chapter 1 Table 1.1. Levels of intraspecific variation for nine species of Argentine Daphnia. Table 1.2. Estimates of genetic distance between Argentine species within the subgenera Daphnia and Ctenodaphnia. at Table. 1.3. Mean gene frequencies at seven allozyme loci for five species of the subgenus Daphnia in Argentina. 2 43 Table. 1.4, Allozymic diversity and variability within populations of 12 Daphnia species from Argentina. . 45 Table. 1.5. Average K2P sequence divergence in the COI gene among the five Daphnia ambigua phylogroups. .... 46 Table. 1.6. Genotypic frequencies at six allozyme loci in three populations of Hyalodaphnia from Argentina. .. 47 Table. 1.7. Mean gene frequencies at seven allozyme loci for six species of the subgenus 48 Ctenodaphnia in Argentina. ..Table. 1.8. Tentative species list for the genus Daphnia in Argentina based on genetic evidence. $0 Chapter 2 Table 2.1. Collection sites and codes for populations belonging to the Daphnia pulex complex from Argentina. .. 82 Table 2.2. Genotypes at seven allozyme loci of the two clones of D. Argentine habitats. Chapter 3 Table 3.1. Approximate timeline of relevant continental movements. ..... . 120 Chapter 4 Table 4.1. Summary of the incidence of postulated speciation mechanisms in the three 159 Daphnia subgenera.List of Figures Chapter 1 Fig. 1.1. NJ tree based on COI sequence variation among all unique haplotypes found in Argentine populations of the subgenus Daphnia. 3 Fig. 1.2. UPGMA tree based on allozyme variation at seven loci in three species of the Daphnia obtusa complex from Argentina. .. S54 Fig. 1.3. NJ tree constructed using K2P distances among all unique COI sequences for D. ‘ambigua from North and South America, ... 2 55 Fig. 1.4. Three head morphologies encountered among Argentine populations assigned to the D. laevis complex. a 56 Fig. 1.5. NJ tree based on COI sequence variation among Argentine populations identified as D. laevis and D. gessneri. 57 Fig. 1.6. NJ tree based on COI sequence variation among all unique haplotypes of Argentine populations belonging to the subgenus Crenodaphnia. 58 Fig. 1.7. NJ tree based on COI sequences for North and South American members of the Daphnia similes complex. .. 59Fig. 1.8. NJ tree based on COI sequence variation among a sample of Daphnia spinulata populations from Argentina and D. exilis populations from North America. .... .- 60 Fig. 1.9. UPGMA tree based on allozyme variation at seven loci in Daphnia spinulata populations from Argentina and D. exilis populations from North America. 61 Fig. 1.10. Sampling localities of Argentine Daphnia populations. Photographs are presented for a single individual of each species, where available. 62 Chapter 2 Fig. 2.1. Map of collection sites for Argentine populations of the Daphnia pulex complex. ..85 Fig. 2.2. NJ tree based on COI sequences for members of the Daphnia pulex complex. Fig. 2.3. NJ tree based on NDS sequence variation in the D. pulex complex. Chapter 3 Fig. 3.1. Phylogenetic analyses of species within the subgenus Ctenodaphnia based on 416 bp of the 16S gene. 121 Fig. 3.2. Phylogenetic hypothesis for the subgenus Ctenodaphnia, with putative intercontinental dispersal events and possible vicariance events marked. .... 123Fig. 3.3. Phylogenetic analyses of species belonging to the subgenus Daphnia based on 469 bp of the 12S gene. .... 124 Fig. 3.4. Analysis of two species complexes within the subgenus Daphnia using protein- coding mitochondrial genes. +. 125 Fig. 3.5. Phylogenetic hypothesis for the subgenus Daphnia, with putative intercontinental dispersal events and potential cases of continental vicariance indicated. ... - 126 Fig. 3.6. Phylogenetic analyses of Hyalodaphnia species based on 411 bp of the 12S gene. se 27 Fig. 3.7. Phylogenetic hypothesis for the subgenus Hyalodaphnia, with putative intercontinental dispersal events indicated. . .. 128 Fig. 3.8. Schematic global maps representing the biogeography of species in the genus Daphnia. 129 Chapter 4 Fig. 4.1. Proposed phylogenetic trees for the three Daphnia subgenera, with intracontinental allopatric divergences indicated. .. .. 160Fig. 4.2. Proposed phylogenies of the three Daphnia subgenera, with pairs of hybridizing species and cases of polyploidy indicated. 163 Fig. 4.3. Proposed phylogenies of the three Daphnia subgenera, with habitat occupancy and suspected cases of speciation via habitat shifts indicated. . 166List of Appendices Appendix I. Collection sites for Argentine Daphnia populations. .. Appendix I. Collection sites for North American Daphnia species sequenced for col. 213 Appendix III. Collection sites for North American Daphnia exilis populations sequenced for col. 215 Appendix IV. Mean gene frequencies at seven allozyme loci for Argentine populations belonging to the subgenera Daphnia and Ctenodaphnia. .. 216 Appendix V. Taxa included in phylogenetic analyses based on the 12S mitochondrial rRNA gene, oo 231Introduction: the diversification of freshwater life Invasion of the freshwater realm All evidence indicates that the diverse metazoan phyla originated in the oceans, and that the diversity of life i the freshwater and terrestrial realms reflects subsequent waves of migration out of the sea. The invasion of inland aquatic habitats by invertebrates began at least 400 million years ago (Gray 1988). Inland waters represent harsh conditions for marine invaders, as they undergo larger fluctuations in ionic concentrations and are far more ephemeral than marine environments. Although their specific mechanisms vary greatly, many groups of freshwater invertebrates have arrived at the same fundamental solution to the challenges of life in a variable environment (Pennak 1985). Most freshwater branchiopods, bryozoans, copepods, rotifers, and sponges, among others, produce some sort of dormant stage that allows survival through unfavourable conditions and also enables dispersal between habitats. It is clear that the evolution of specialized resting stages in many groups of organisms was an important innovation that allowed their invasion and radiation in freshwater settings. Interestingly, a small number of marine copepods and sponges produce resting eggs or gemmules, respectively (Pennak 1985), suggesting that this capability may be considered a “pre-adaptation” to freshwater invasion in some marine taxa (Bayly 1992; Hairston and Cacares 1996). Evolutionary consequences of ia resting eges While the evolution of a dormant stage permitted the invasion of freshwaters, this feature has also been important in structuring the subsequent evolutionary trajectories of freshwater life. Resting eggs are often copiously produced and remain viable for manyyears, withstanding desiccation and extreme temperatures, allowing them to accumulate in lake basins. These “egg banks” can effectively slow the rate of evolution within populations by resurrecting genetic variants that may have been lost from the active population as a consequence of short-term selection or drift (Hairston and De Stasio 1988). However, resting eggs can also be dispersed among habitats, via wind or animal vectors, In fact, many resting eggs possess spines that promote their adherence to fur or feathers, and they have also been shown to remain viable following passage through the digestive tracts of birds (Maguire 1965; Proctor and Malone 1965). Their capacity for movement among insular habitats is likely one of the most important factors structuring the evolution and biogeography of zooplankton taxa. Depending on the rate of dispersal, there are several potential evolutionary outcomes of dispersal among habitats. Very high rates of dispersal over-ride the influences of local adaptation and genetic drift, promoting genetic cohesion over the range of a species. Intermediate rates of ispersal may be important in infusing local populations with new genetic variation, while still allowing gene pool differentiation and local adaptation, enhancing intraspecific variation. On the other hand, rare dispersal events into new habitats may be associated with speciation via founder effects and subsequent allopatric divergence. Empirical studies on freshwater bryozoans (e.g. Freeland et al. 2000), cladocerans (c.g. Crease et al. 1990; Hann and Hebert 1986; Weider et al. 1999a), and copepods (e.g. Boileau and Hebert 1988, 1991) have revealed populations characterized by moderate to strong gene pool differentiation, with little evidence of isolation by distance, even across vast regions, suggesting low levels of dispersal among populations. Indeed, geneticdistances between populations of freshwater copepods are more than ten-fold higher than distances observed in marine copepods (Hairston and Bohonak 1998). Moreover, there is convincing evidence of local adaptation in populations of freshwater zooplankton (De Meester 1996). Thus, the population structure of freshwater invertebrate taxa with resting eggs is typified by a complex mosaic of often locally-adapted subpopulations connected via low levels of dispersal, approximating the metapopulation model (Freeland et al. 2000). This “metapopulation” structure of many passively-dispersed freshwater taxa might be expected to promote divergence, and even speciation, according to Wright's (1932) shifting balance model. A combination of drift and selection acting in numerous, largely-independent subpopulations is expected to enhance the probability of the origin of novelty. This mechanism may approximate the conditions involved in speciation by habitat shifts, as documented in the cladoceran genera Daphnia (Colbourne et al. 1997) and Bosmina (De Melo and Hebert 1994). The ifferent selection regimes encountered upon invading novel habitats, especially ponds vs. lakes, provide a strong force that can promote divergence even under continuous, but low, levels of gene flow. Recent investigations of freshwater taxa, primarily zooplankton, have revealed that a different geographic pattern of intraspecific genetic variation is often superimposed upon this metapopulation structure. Continental-wide studies on broadly-distributed members of the cladoceran genera Daphnia (Taylor et al. 1998; Weider et al. 19992, b; Hebert et al. in review), Sida (Cox and Hebert 2001), and Holopedium (Rowe 2000) have revealed pronounced geographic patterning of genetic diversity, a finding missed by earlier regional surveys. Such taxa tend to consist of largely-allopatric phylogroups,which may each have vast distributions. In some cases the ages of the phylogroups appear to reflect their origin in different glacial refugia during the Pleistocene (Weider et al. 1999a, b; Cox and Hebert 2001; Hebert et al. in review). In other taxa, older divergences seem to have been forged by the rise of mountain ranges, which limited gene flow (Hebert and Wilson 1994). These observations indicate that allopatric divergence can occur in passively-dispersed organisms when there are extrinsic barriers to gene flow, necessitating expanded views of the classic metapopulation view and the role of allopatric speciation in such taxa. Due to the supposed intercontinental distributions of many zooplankton species, it is pertinent to examine phylogeographic structure at a large spatial scale. Detailed morphological work has revealed that “conspecific” populations of cladoceran crustaceans from different continents tend to show consistent morphological differences (Frey 1982a, 1987). Molecular studies on native species in the cladoceran genera Dapiinia (Colbourne et al. 1998; Cerny and Hebert 1999; Schwenk et al. 2000), Sida (Cox and Hebert 2001), and Simocephalus (Hann 1995) have all revealed such large genetic divergences between North American and European populations that they may be considered distinct species. These findings suggest that rare intercontinental dispersal in zooplankton taxa is often followed by divergence in allopatry, and speciation. However, most of these studies have only examined divergence events within morphospecies, constituting the terminal nodes of the phylogenies of these taxa. Consequently, a comprehensive understanding of the role of long distance dispersal via resting eggs in freshwater organisms has remained elusive.Intercontinental speciation in Daphnia: towards the complete picture Daphnia is the most intensively-studied genus of freshwater zooplankton and has become a key taxon for investigating patterns of diversification in passively-dispersed life. This genus is both taxonomically and ecologically diverse, inhabiting varied aquatic environments on all continents, including both polar regions. Furthermore, numerous taxonomic and phylogenetic i westigations have provided the necessary systematic framework for biogeographical investigations of the group (e.g. Colboume and Hebert 1996; Taylor et al. 1996; Schwenk et al. 2000). Detailed studies on members of this genus have begun to shed light on the role of intercontinental dispersal in its evolutionary history. For example, studies on two groups of species, those belonging to the Daphnia pulex complex (subgenus Daphnia) and the subgenus Hyalodaphnia, have revealed that many arctic lineages are young and have vast distributions that span three continents, while temperate members of these assemblages tend to be endemic to a single continent (Colbourne et al. 1998; Weider et al. 1999a, b; Schwenk et al. 2000). This finding suggests that dispersal between the temperate regions of North America and Europe is rare and often followed by allopatric speciation. However, our understanding of global phylogeographic pattems and intercontinental speciation is limited by the restricted taxonomic coverage of prior intercontinental studies and by a lack of genetic information on the daphniid faunas of several major biogeographic regions, namely Africa, Asia, and South America. Argentina is a particularly promising area for beginning an investigation of the daphniids of South America. This country is comprised of a large variety of landscapes,from subtropical lowlands, humid plains, and the arid steppes of Patagonia, to the Andes Mountains and the rugged forested terrain of southern Tierra del Fuego. It thus contains a broad range of aquatic habitats, which are known to harbour a diverse zooplankton fauna. ‘Among the cladocerans, Argentina is home to 87% of the genera and 70% of the total number of species known from the Neotropical region (Paggi 1998). Objectives of the thesis This thesis examines the role of intercontinental dispersal in the evolutionary diversification of the zooplankton genus Daphnia. Firstly, new information is presented on the diversity of this genus in Argentina. Secondly, genetic information on species from four continents is combined to explore the importance of intercontinental allopatric speciation in generating species and lineage diversity and in structuring broad biogeographic patterns. Results of a detailed investigation of the species diversity of daphniids from Argentina are presented in Chapter 1. Allozymes and a protein-coding mitochondrial gene, cytochrome C oxidase subunit I (COI), are used to delimit species boundaries, to screen for hybridization, and to diagnose breeding system diversity. COI sequences from North American and Argentine populations are also compared to assess the taxonomic status of morphologically-similar species on the two continents and to determine the degree of regional endemism. Chapter 2 explores the distribution of polyploid populations belonging to the D. pulex complex in Argentina. Their presence in this temperate region, an unusual setting for polyploid daphniids, is attributed to their origin following long-distance dispersal from the northem hemisphere. This finding is discussed in light of theoretical issuesregarding the origin and maintenance of polyploids, and the role of chance long-dispersal dispersal in determining the evolutionary fate of polyploids is considered. The investigation of the role of intercontinental dispersal in daphniid evolution is, expanded in Chapter 3 by examining broad phylogeographic pattems. A phylogenetic framework for the genus Daphnia is first constructed using mitochondrial rRNA gene sequences for species from four continents. Distributional information is subsequently used to infer the occurrence of intercontinental allopatric speciation events. Furthermore, molecular clock estimates are used to differentiate between dispersalist and vicariance (ie. continental drift) explanations for taxon distributions in this ancient genus. In Chapter 4, the importance of intercontinental allopatric speciation is examined in relation to other mechanisms of speciation in Daphnia. To this end, literature on speciation in Daphnia is reviewed. Geographical, genetic, and ecological factors that promoted diversification are considered in the context of the four-continent phylogenetic framework developed in Chapter 3. The differential operation of different mechanisms of diversification among clades and biogeographic regions is discussed. Finally, the results of this study of diversification in Daphnia are evaluated in light of work on other taxonomic groups, and their significance is considered in a broader evolutionary context.Chapter 1: Species and endemism in the Daphnia of Argentina: a genetic investigation Abstract Although the temperate regions of South America are known to have a diverse daphniid fauna, there has been no genetic evaluation of the existing taxonomic system or the affinities between the North and South American faunas. The present study employs analysis of allozymes and mitochondrial DNA sequences to investigate species diversity in 176 Daphnia populations from Argentina. This work established the presence of at least fifteen species in Argentina, five of which are likely undescribed and three of which represent range extensions of taxa known from North America. Twelve of these species appear endemic to South America, while four other species also occur in North America and show close genetic similarity with these populations, indicating the recent exchange of propagules between the continents. While biological interactions and habitat availability have probably contributed to the observed species distributions, chance dispersal has likely played a dominant role in structuring large-scale biogeographic pattems in the genus. Introduction Our understanding of zooplankton species diversity has undergone a recent paradigm shift from “cosmopolitanism” to “endemism” (Frey 1987). Early investigators, such as Darwin (1859), were impressed by the morphological similarity of freshwater life from different continents, which they attributed to the long-distance dispersal capabilities of organisms adapted to discreet and relatively ephemeral habitats. As such, many speciesof zooplankton that were encountered in freshwater environments of the New World were assigned species names from Europe. However, more detailed comparisons subsequently revealed consistent morphological differences in populations from different continents, leading to the recognition that few, if any, species are cosmopolitan (Frey 1982a, 1987). Genetic analyses have further supported the generality of continental endemism (e.g. Schwenk et al, 2000; Cox and Hebert 2001) and have reinforced the need for detailed regional studies. The systematics and evolution of the cladoceran genus Daphnia have been the subjects of particularly intense study. Species of this genus are dominant members of the zooplankton communities in lakes and ponds throughout the world, except in the lowland tropics (Hebert 1978; Femando et al. 1987). Furthermore, they are important model organisms for ecological, toxicological, and evolutionary studies. However, despite more than 200 years of attention, the taxonomy of this genus is still uncertain. Detailed morphological studies have shown that many variable morphological traits are of limited taxonomic utility because of phenotypic plasticity (Dodson 1988, 1989; FléBner and Kraus 1986; Frey 1982b). To further complicate matters, morphological evolution in the genus has been characterized by stasis and convergence (Colbourne et al. 1997). While morphological analyses have often failed to resolve key aspects of the complex taxonomy of Daphnia, genetic markers have proven useful in delineating species boundaries, detecting hybridization, and clarifying phylogenetic relationships among taxa (e.g. Taylor and Hebert 1992, 1993a; Taylor et al. 1996; Wolf and Mort 1986). The most comprehensive genetic investigation of Daphnia taxonomy has been done in North America. Allozyme studies were initially used to determine both speciesboundaries and incidence of interspecific hybridization (e.g. Taylor and Hebert 1993a), while molecular phylogenetic analyses, employing mitochondrial DNA (mtDNA) sequences, later provided a systematic framework for the genus (Colbourne and Hebert 1996; Taylor et al. 1996). This work has indicated that North American members of the genus belong to three monophyletic subgenera: Daphnia, Hyalodaphnia, and Ctenodaphnia (Colbourne et al. 1996; Schwenk et al. 2000). Each subgenus has also been divided into a number of species complexes, which were primarily defined by the detection or suspicion of hybridization among member species (Colbourne and Hebert 1996). Genetic studies have established that hybridization is common between many closely-allied species pairs of Daphnia and even occurs between more distant relatives, which may show up to 14% divergence in the mitochondrial 12S rRNA gene (Schwenk et al. 2000). Despite the high incidence of hybrids between some species, introgression is generally limited, and “pure” parental genotypic arrays tend to remain intact (Schwenk and Spaak 1997). However, introgression has been documented and appears to have provoked speciation in at least one case (Taylor and Hebert 1992). Genetic evidence has also elucidated other factors of substantial evolutionary interest, such as breeding system shifts (Crease et al. 1989; Cemy and Hebert 1993) and polyploidy (Dufresne and Hebert 1997), which also contributed to earlier taxonomic confusion in this genus. The key insights gained from such genetic analyses provide a valuable backdrop for investigating species diversity in Daphnia faunas that are unexplored from a molecular point of view, such as that of Argentina. Argentina possesses a large variety of landscapes and harbours a diverse zooplankton fauna. Each of the three Daphnia subgenera is known to occur in this 10country (Paggi 1998). All five species of the subgenus Crenodaphnia (D. dadayana, D. ‘menucoensis, D. notacantha, D. ornithocephala, D. spinulata) known from the that Neotropical region occur in Argentina (Paggi 1998). These are large-bodied speci are usually restricted to fish-free, intermittent habitats in the arid, temperate regions of the country. While the taxonomy of Ctenodaphnia is considered well-resolved, species boundaries in the other two subgenera (Daphnia and Hyalodaphnia) are uncertain (Paggi 1998). The only existing key (Olivier 1962) is known to include many inaccurate records and questionable early descriptions (Marinone pers. comm.; Paggi 1998, pers. comm.) However, itis thought that the subgenus Hyalodaphnia is represented by at least two taxa (D. laevis and D. gessneri) (Paggi 1998), which are small-bodied species found in lakes and reservoirs throughout subtropical South America (Infante 1984; Matsumura-Tundisi 1984). The taxonomy of the subgenus Daphnia is even more problematic, but it is likely that at least 4-5 species are present, including D. ambigua, D. peruviana, several species belonging to the “obtusa” group, and possibly D. pulex (Paggi 1998; Villagra de Gamundi 1986; Villalobos 1994). Members of this subgenus are known to inhabit diverse pond and lake habitats throughout the country (Paggi 1998). In this study, mtDNA sequence data and gene frequencies at nuclear allozyme loci are used to clarify species boundaries, diagnose breeding systems, and screen for interspecific hybrids. The validity of the current taxonomic system for Argentine Daphnia is evaluated by comparing sequence information from Argentine populations with corresponding data from all known North American species. Finally, general patterns of species diversity, dispersal, and endemism are discussed. Subsequent papers will address the phylogenetic position of the species that comprise the Argentine fauna ul(Chapter 3). In addition, formal taxonomic descriptions will be prepared for all newly- detected species (Marinone et al. in prep.). Methods Collections During sampling campaigns in November-December 1999 and January-February 2001, Daphnia populations were collected from 137 water bodies throughout Argentina (Appendix 1). Habitats sampled included ponds, roadside ditches, playa lakes, alpine lakes, rivers, reservoirs, and saline lakes. Each site was given a unique number (see Appendix I for locality details). Zooplankton samples were collected using a 280-ym mesh net. Daphnia were subsequently sorted alive and either flash-frozen in liquid nitrogen for allozyme surveys or preserved in 95% ethanol for DNA analysis. Initial species identifications were made in the field by M.C. Marinone, P. Hebert, and J. Paggi. As there is no modem key for Argentine Daphnia, these identifications were based on original species descriptions and the personal experience of the investigators. In some cases, tentative taxonomic designations were adjusted after genetic analysis. The above collections were augmented with ethanol-preserved specimens of Daphnia peruviana from the province of Tucuman in northern Argentina, provided by A. Villagra de Gamundi. M.C. Marinone provided an additional 16 populations of Daphnia from Argentina, preserved in either ethanol or trehalose, according to the protocol outlined in Taylor et al. (1994). Several frozen and ethanol-preserved North Americansamples from the archived collections held by P. Hebert were also analyzed for comparative purposes (see Appendices Il and Ill for details). DNA sequence lysis Total DNA was extracted from several individuals from each population by placing single animals in 50 uL of proteinase-K extraction buffer, according to the protocol of ‘Schwenk et al. (1998). The extraction mixtures were incubated for 24 hours at 50°C, after which the proteinase-K enzyme was denatured by a 10 min incubation at 94°C. Extractions were subsequently stored at -20°C. A 710-base pair (bp) fragment of the cytochrome ¢ oxidase subunit 1 (COI) mitochondrial gene was amplified via the polymerase chain reaction (PCR) (Saiki et al. 1998) from a single individual from each population using universal primers LCOI490 and HCO2918 (Folmer et al. 1994). Each 50-yL reaction consisted of 3-5 uL of DNA template (from the total volume of 50 uL), $ L of 10x PCR buffer (10mM Tris-HCI, pH 8.3; 50mM KCI), 0.2 uM of each primer, 1.5 mM MgCl, 0.2 mM of each dNTP, and | unit of Tag DNA polymerase. The PCR thermal regime was as follows: one cycle of 1 min at 94°C; 5 cycles of I min at 94°C for 1 min, 1.5 min at 45°C, and 1.5 min at 72°C; 30 cycles of 1 min at 94°C, 1.5 min at 50°C, and 1.5 min at 72°C; and finishing with a final extension at 72°C for $ min. Initially, 5- uL samples of all PCR-reaction products were electrophoresed in 1% agarose gels, stained with ethidium bromide, and visualized with UV light. Successfully-amplified products were then electrophoresed in 2% agarose gels, and similarly stained and visualized. The desired fragment was excised from the gel, purified using the Qiaex I (Qiagen) kit, and subsequently subjected to dye terminator sequencing (25 cycles, 55°C annealing) using the Big Dye Terminator (version 3) sequencing kit (ABI Prism). 13,Products were sequenced in one direction using primer LCOI490. Electrophoresis of sequencing-reaction products was performed on an ABI 377 automated sequencer (Applied Biosystems). Sequence electropherograms were inspected and aligned using the SeqApp 1.9 sequence editor (Gilbert 1992), resulting in a final alignment of 630 bp. Phenetic analyses were performed separately on all unique COI haplotypes for each of the three Daphnia subgenera using the program MEGA 2.1 (Kumar et al. 2001). Pairwise genetic distances were calculated using Kimura's (1980) two-parameter model (K2P) and pairwise deletion of missing sites. Pairwise distance matrices were used to construct phenograms by the neighbour-joining (NJ) algorithm (Saitou and Nei 1987), which does not assume equal evolutionary rates among lineages. North American taxa were included to root the trees, as indicated in the figure legends. Appropriate outgroups were selected based the phylogenetic analyses of Colboumne and Hebert (1996). For example, since subgenera Daphnia and Hyalodaphnia were found to be sister groups by Colbourne and Hebert (1996), a Hyalodaphnia species was used to root the Daphnia tree, and vice versa. Bootstrap values were based on 1000 pseudoreplicates. Distinct mitochondrial clades that corresponded to morphologically-identified groups of populations were tentatively designated as “species” prior to allozyme analysis. Allozyme analysis Allozyme surveys were conducted on all populations for which frozen material was available. Variation was detected by subjecting whole-animal homogenate to cellulose acetate electrophoresis using a Tris-glycine buffer (pH 8.5) (Hebert and Beaton 1993). Depending on the size of the individuals, populations were screened for variation at 4-7 commonly-polymorphic loci, including aspartate amino transferase (AAT) (EC “43.2.1.1), fumarate hydratase (FUM) (EC 4.2.1.2), glucose-6-phosphate isomerase (GPI) (EC 5.3.1.9), glyceraldehyde-3-phosphate dehydrogenase (G3PDH) (EC 1.2.1.12), lactate dehydrogenase (LDH) (EC 1.1.1.27), malate dehydrogenase (ME) (EC 1.1.40), mannose- 6-phosphate isomerase (MPI) (EC 5.3.1.8), and phosphoglucomutase (PGM) (EC 5.4.2.2). Gels to be stained for each locus were electrophoresed for 15 minutes with a voltage of 206, as pilot trials revealed that this run time maximized the separation of allelic variants while still maintaining resolution. Enzyme breakdown during longer trial runs produced smears that complicated scoring. From most sites, 22-44 individuals of each species were analyzed, but in some cases fewer individuals were available (see Appendix IV for sample sizes). During each staining run, two individuals from a clonal stock of North American Daphnia pulicaria (from Lake Washington, Washington state, USA) were used to standardize scoring. All alleles encountered during allozyme screening were named according to their relative mobility compared with this standard. It was known on morphological grounds tha: numerous samples contained multiple Daphnia species, somewhat complicating allozyme analysis. However, as the COI results were obtained first, mitochondrial clusters had already been tentatively designated as “species” prior to allozyme analyses. Pilot allozyme runs revealed that samples that were in Hardy-Weinberg equilibrium, presumably constituting conspecific populations, corresponded with the “species” identified by mtDNA analysis. Furthermore, these allozyme trials revealed fixed allelic differences between each “species” pair within the subgenera Daphnia and Ctenodaphnia. Therefore, for species in these subgenera, separation of individuals belonging to conspecific populations was straightforward, as hybrids were never detected at the diagnostic loci. Thus, allozymedata were analyzed for each coexisting population separately for all Daphnia and Ctenodaphnia species. For the Hyalodaphnia, however, species boundaries were not clear, and so data from all individuals from each collection site were analyzed together. All analyses of allozyme data were performed using GDA software (Lewis and Zaykin 2001). Genotypic frequencies within populations were compared with Hardy- Weinberg (HW) expectations using Fisher's exact test in order to screen for asexually- reproducing populations, mixed-species assemblages, and hybrids. Since multiple statistical tests were performed, a Bonferroni correction was applied by adjusting alpha for each species, such that: a = 0.05 / (No. of polymorphic loci in each population summed over all populations). Levels of genetic diversity and genetic substructure within populations were estimated using three basic statistics: percent polymorphic loci, percent individual heterozygosity, and Fis (the inbreeding coefficient). The extent of genetic divergence among conspecific populations was estimated using Nei’s (1978) genetic distances (D). Fsr statistics (or fixation indices) were used to estimate the degree of genetic structure (ie. gene pool fragmentation) among intraspecific populations. Interspecific divergences were also estimated using Nei’s distances. Allozyme trees were constructed from pairwise distance matrices using the UPGMA method. Comparison with North American species COI sequences from Argentine populations were compared with sequences from all known North American species. Unpublished sequences for most species were provided by J. Colboume, while the remaining sequences were obtained from archived specimens held by P. Hebert (see Appendix Il). For Daphnia ambigua, the extensive North American COI dataset of Hebert et al. (in review) was also used for comparison. 16While preliminary analyses included all North American species, only matches with possibly synonymous or closely-related species are reported in detail here. The phylogenetic relationships among more istantly-related species were later investigated using the mitochondrial 12S rRNA gene (sce Chapter 3). Preliminary genetic screening indicated a close relationship between D. spinulata from Argentina and the North American species D. exilis. Since extensive allozyme surveys had already been conducted (Hebert and Finston 1993), and archived D. exilis were available for DNA sequencing, this species pair was selected for a more detailed investigation of patterns of intraspecific diversity and intercontinental relatedness. COL sequences from 10 populations of North American D. exilis were obtained and analyzed along with sequence data for D. spinulata (see Appendix Ill for list of collection sites). Sequence divergences were estimated using the K2P model and analyzed by phenetic techniques, as described above. Additionally, allozyme variation was compared between D. spinulata populations from Argentina and D. exilis populations from North America. Three archived D. exilis populations from Mexico were electrophoresed in the present study. Live D. exilis populations from Oklahoma (provided by S. Schwartz) were used to standardi allozyme scoring between D. spinulata populations and similar data for D. exilis from Hebert and Finston (1993). Although these authors surveyed 11 allozyme loci, only seven loci were considered in the present study, so that D. spinulata and D. exilis were compared using the same markers. For the D. exilis! D. spinulata dataset, the NJ tree-building method was also used for the COI sequences, and the UPGMA technique was employed to analyze the allozyme data. 7Results Diversity in the subgenus Daphnia Species belonging to the subgenus Daphnia were identified by their prominent medial pecten, Members of this subgenus were collected from 77 habitats (see Appendix 1D. As some habitats contained more than one species, as indicated by both morphology and genetic markers, a total of 8 distinct populations were available for analysis. Initial morphological inspection indicated the presence of at least five species. Four populations identified as D. ambigua were collected, three from lakes or reservoirs and one from a river. A single population of D. peruviana, a darkly melanized species from mountain lakes, was available for analysis. The remaining 83 populations were assigned to either the D. pulex or D. obtusa complexes, which were separated, respectively, by their lack of or possession of elongate setae along the internal margin of the carapace (Schwartz et al 1985). Populations belonging to the D. pulex complex were frequently encountered in southem Argentina, while members of the D. obtusa complex were collected throughout the country. There was much morphological variability among populations assigned to the obtusa complex, suggesting the presence of several species. Several populations from the Andean lakes and also from ponds in the southem parts of the country were melanized to varying degrees, but populations from other areas were not. COI sequence variation Phenetic analysis of all unique COI sequences revealed the presence of seven distinct clusters, which were tentatively assigned names based on morphological identifications (Fig. 1.1). Numbers were used to designate the divergent clades that had initially been identified as the same morphospecies. Members of the obtusa complex 18form three distinct clusters (D. obtusa #1, #2, and #3), the third of which contains all of the melanized populations. Populations identified as D. ambigua form two deeply divergent groups. One, D. ambigua #1, consists of two allied sequences detected among three populations from lakes or reservoirs it ifferent regions of Argentina, while the other, D. ambigua #2, contains two identical individuals from the Coronda River (site 9), which is part of the Parana River system in northeastern Argentina. A member of the D. pulex complex, designated D. “pulicaria”, was also detected in 16 habitats in southem Argentina (see Chapter 2 for a detailed analysis of these populations). Finally, D. pervuviana was identified as a divergent lineage. Bootstrap support for all seven clusters is high (100). The topology indicates that D. obtusa #1 and D. obtusa #2 are allied species, whose affinity is supported by a moderately high bootstrap value (74). The relationships among the other species were not resolved with this dataset. However, it is obvious that the two species morphologically identified as D. ambigua are not closely related. COI sequence divergence among individuals assigned to a single cluster is generally small, ranging from a low of 0.3% in D. “pulicaria” to a high of 4.3% in D. obtusa #1 (Table 1.1). By contrast, average pairwise distances between clusters are much larger. The lowest divergence (between D. obtusa #1 and #2) is 16.2%, while other pairwise divergences range from 20.3-28.3% (Table 1.2). Allozyme variation Allozyme data are unavailable for D. ambigua #2 and D. peruviana, as only ethanol-preserved samples were available. However, allozyme analyses partition the other populations into the same five “species” identified by mtDNA analysis. Allelic 19arrays and allele frequencies differ among species, and at least one fixed difference was detected between each pair of putative species (Table 1.3). Nei's genetic distances are generally small among conspecific populations. The ‘maximum intraspecific distances range from 0.0 in D. ambigua to 0.54 in D. obtusa #1 (Table 1.1). By contrast, genetic distances among the five species are large (Table 1.2). The allozyme results confirm that D. obtusa #1 and #2 are most closely allied, with an average genetic distance of 0.78, but average distances between all other pairs range from 0.85-3.46 (Table 1.2). The topology of the UPGMA tree based on allozyme data for the three D, obtusa species (Fig. 1.2) is similar to that based on mtDNA in placing D. obtusa #1 and #2 as sibling species (Fig. 1.1). Membership of obtusa-group populations in one of the three main clusters (i.e. “species”) is consistent with assignments based on COL analysis Genotypic frequencies at allozyme loci are generally concordant with Hardy- Weinberg (HW) expectations, suggesting that most species within the subgenus Daphnia reproduce by cyclic parthenogenesis. Among the 27 populations of D. obtusa #1, only two (40 and 48) displayed significant HW deviations at one locus, due to heterozygote excess in both cases. Of the four screened populations of D. obtusa #2, a significant HW deviation was detected at a single locus in only one (136b), again due to heterozygote excess. No HW deviations were detected in any of the eight populations of D. obtusa #3. The sole polymorphic locus (AAT) in the two D. ambigua #1 populations did not deviate significantly from HW equilibrium. suggesting that these populations are also cyclic parthenogens. On the other hand, the 14 populations of D. “pulicaria” consist of just twogenotypes, or “clones”. This species is treated in detail in Chapter 2, but, briefly, all populations of this species exhibit fixed heterozygosity at four loci Levels of genetic diversity within populations were compared for five species within the subgenus Daphnia using the same seven allozyme loci (AAT, FUM, GPI, G3PDH, LDH, MPI, and PGM), with the exception of D. ambigua, for which only four loci were available (Table 1.4). Observed levels of heterozygosity are similar to the values expected based on allele frequencies in the sexual species. However, levels of polymorphism and heterozygosity for the obligately parthenogenetic D. “pulicaria” are greatly elevated due to the fixed heterozygosity in these populations. Among the remaining species, D. obtusa #1 is the most genetically variable, while D. obtusa #3 is the least (Table 1.4). Fsr statistics, which were calculated for species in which three or more populations were surveyed, range from 0.29 in D. obtusa #2 to 0.81 in D. obtusa #3 (Table 1.1), indicating strong genetic differentiation among conspecific populations. The especially high value for D. obtusa #3 reflects the fixation or near fixation of alternate alleles at AAT in different populations. This one locus dominates the Fsr statistic because polymorphism is limited in this species (Table 1.4). Evaluation of relatedness to North American Daphnia species A comparison of COI sequences from Argentis e and North American Daphnia revealed two cases in which taxa from the two continents are closely affiliated. Firstly, populations of D. ambigua #1 from Argentina and Chile show little sequence divergence from their North American counterparts. The average COI sequence divergence between South and North American sequences is 4.5% (range 2.6-6.5%). NJ analysis including thedata in Hebert et al. (in review) produced a topology in which the South American populations of D. ambigua #1 are a sister clade to all North American populations, although bootstrap support for this result weak (Fig. 1.3). Maximum parsimony analysis based on a combined COI+12S data also supports this sister relationship (Hebert et al, in review). However, 12S data alone (Hebert et al. in review) indicate that the southern South American sequences are nested within the variation present among North American phylogroups of this species, suggesting a paraphyletic relationship. There is substantial variation among North American populations, which form four largely allopatric phylogroups (Hebert et al. in review), but sequence divergence among South American D. ambigua #1 sequences is low (Table 1.1). The average divergence between South American sequences and each of the four North American phylogroups ranges from 3.2% to 5.7% (Table 1.5). The two individuals of D. ambigua #2 from the Coronda River (site 9) show 23- 26% sequence divergence from other D. ambigua from North and South America However, they are only 12% divergent from an individual of D. retrocurva from Mexico, suggesting that Argentine D. “ambigua” #2 belongs to the retrocurva complex. The other possible example of a shared Argentine and North American species involved the D. pulex complex. These populations are investigated in detail in Chapter 2. Briefly, a comparison of South American and Holarctic NDS mitochondrial gene sequences (from Colbourne et al. 1998) indicated that the Argentine pulex-group populations are closely related to the North American species D. pulicaria (see Figs. 2.2 and 2.3). In fact, all Argentine populations form a monophyletic sister group to allHolarctic D. pulicaria clades, from which they show an average of just 2.9% divergence (Fig. 23). Pairwise divergences between the remaining Argentine members of the subgenus Daphnia and all North American species are greater than 17%. Their relationships are investigated in Chapter 3. Diversity in the subgenus Hyalodaphnia Members of the subgenus Hyalodaphnia were obtained from five sites, four of which are reservoirs or lakes and one, a river. Based on head morphology, these individuals were preliminarily identified as either the helmeted species D. gessneri (Herbst 1967) or the unhelmeted species D. laevis (Birge 1878). Three populations (30, 32, and 36) were identified as D. laevis, one population (226) was identified as D. gesseri, while the Coronda River (site 9) appeared to contain both taxa, as well as individuals with intermediate head morphologies (see Fig. 1.4). CO! sequence analysi Three COI haplotypes were found among the eight individuals examined from the five sites (Fig. 1.5). Two individuals of D. gessneri from site 226 and three individuals of D. laevis from sites 30, 32, and 36 possess the same haplotype. A similar sequence (only 0.69 divergent) was found in two individuals from the Coronda River, one with the head morphology of D. laevis, and the other with that of D. gessneri. However, a third individual from the same site, identified as D. gessneri, has a haplotype that shows an average of 13.6% divergence from the other group.Allozyme variation Three of the five Hyalodaphnia populations (30, 32, and 226) were screened for variation at six allozyme loci, three of which are polymorphic (Table 1.6). In all three populations, two loci are fixed, or nearly so, for heterozygotes. Six of the seven polymorphic loci among these populations are severely out of HW equilibrium (p<0.001), due to heterozygote excess (Tables 1.4, 1.6). Evaluation of relatedness to North American Hyalodaphnia species ‘A comparison of COI sequences indicated that both haplotype groups found in Argentina, which perhaps represent two species, are only distantly related to those present in the three North American species belonging to the /aevis complex (D. dubia, D. laevis, D. magniceps) (Fig. 1.5). Pairwise K2P divergences between North and South American sequences range from 18-24%. North and South American D. “laevis” are clearly not sister taxa, as each is more closely related to another Hyalodaphnia species endemic to its home continent (Fig. 1 Diversity in the subgenus Ctenodaphnia A total of 83 Ctenodaphnia populations were sampled from 80 habitats. Preliminary morphological investigation suggested the presence of at least six species. Morphological species assignments were more straightforward than for members of the other subgenera. Daphnia spinulata ( bén 1917) was collected in ponds throughout much of the country. Populations from four sites (73, 204, 206, and 229) were identified as D. notacantha (Birabén 1954) because of the distinctive hump on their heads. However, smaller “notacantha-like” humps were observed in several populations designated as D. spinulata. Twelve populations of D. menucoensis (Paggi 1996) werecollected from shallow saline lakes in the Patagonian steppes, while D. dadayana (Paggi 1999) occurred at 24 sites in southern Argentina. Four populations of the distinctive D. ornithocephala, literally, “bird head” (Birabén 1954), were collected from ephemeral ponds in a restricted arid area of west-central Argentina (sites 299, 231, 233, and 234) Finally, a single population of a Ctenodaphnia species that was morphologically distinct, from any described species was encountered at site 209. COI sequence variation Phenetic analysis of all unique COI sequences revealed six distinct clusters (Fig. 1.6). Four of these genetic groups correspond to recognized morphospecies: D. dadayana, D. menucoensis. D. ornithocephala, and D. spinulata. Populations of D. notacantha possess haplotypes indistinguishable from those of D. spinulata (Fig. 1.6). As well, a single individual from another population (125) was identified by its COI sequence (see below) as D. similis, as redescribed from North America by Hebert and Finston (1993). The morphologically-unique population from site 209, temporarily designated Ctenodaphnia sp. #1, also constituted a distinct lineage. The level of COL sequence divergence within clusters is small, with maximum pairwise divergences ranging from 0.3% within D. ornithocephala to 3.9% in D. spinulata (Table 1.1). By contrast, divergences between clusters range from 20.8-30.1% (Table 1.2). The clusters are strongly supported as evidenced by bootstrap values of 100. However, the relationships among clusters were poorly resolved (Fig. 1.6). Allozyme variation The allozyme results (Table 1.7) support the recognition of all six “species” identified by mtDNA analysis and show that D. spinulata and D. notacantha are notgenetically distinct, based on the surveyed loci. As a result these taxa were pooled for all subsequent analyses and statistics. Allelic arrays of the putative species are distinct, and allelic substitutions are present at two or more loci between each species pair (Table 1.7). Nei’s genetic distances (D) between conspecific populations are generally smaller than distances between species. Maximum D values between conspecific populations range from 0.13 in D. spinulata to 0.87 in D. dadayana (Table 1.1). By comparison, the average distances between pairs of species range from 0.94 (between D. spinulata and D. similis) to 6.7 (Table 1.2). In several other species pairs, genetic distances could not be estimated as no alleles were shared. In general, genotypic frequencies are close to HW equilibrium. Among the 24 populations of D. dadayana, only a single locus in one population (149) is out of HW equilibrium, due to a heterozygote deficit. Likewise, a single locus in one of the four populations of D. ornithocephala (229) differs significantly from HW eq) ibrium, due in this case to an excess of heterozygotes. There are no significant HW departures among the nine populations of D. menucoensis. HW deviations were similarly detected in only two of 27 D. spinulata populations. In one population (66), only a single locus is out of HW equilibrium, while another polymorphic locus in same population is in HWE. The other population (from a road-side puddle, 71) shows fixed heterozygosity at two loci but is monomorphic at other loci. The sole population of Crenodaphnia sp. #1 has genotypic frequencies that are concordant with HW expectations. As only a single individual of D. similis was examined, genotype frequency analysis was not possible for this species.Estimates of within-population variation are similar among most Crenodaphnia species (Table 1.4). Levels of polymorphism and heterozygosity are lowest in D. omithocephala. Observed levels of heterozygosity are close to those expected based on allele frequences (Table 1.4). Evidence of strong local differentiation among conspecific populations was detected in all species, as Fsr statistics range from 0.24-0.52 (Table 1.1). Evaluation of relatedness to North American Ctenodaphnia species The comparison of COI sequences from Argentine Crenodaphnia with those from all known North American species revealed two shared taxa. The individual of D. similis from Argentina is closely related to North American populations, showing just 1.9% and 1.4% divergence from D. similis from Nevada and Washington, respectively. By comparison, the two North American sequences are 2.3% divergent from one another (Fig. 1.7). ‘The commonest Crenodaphnia in Argentina, D. spinulata, is closely allied with the North American species D. exilis (Fig. 1.7). Two mitchondrial clusters were identified within D. spinulata, showing an average of 3.1% COI divergence (Figs. 1.6, 1.8). The more common of these Argentine clusters, Group A, is only 1.6% divergent from haplotypes in North American populations of D. etilis, while Group B shows an average of 3.1% divergence from D. exilis (Figs. 1.8). By way of comparison, the maximum. divergence among Argentine COI sequences is 3.9%, while maximum divergence among the North American D. exilis sequences is only 1.9% (Table 1.1). The NJ analysis grouped all North American haplotypes together. but with low boostrap support (24) (Fig. Lg).In contrast to the mtDNA results, UPGMA analysis of allozyme variation revealed a clear genetic separation between D. spinulata and D. exilis populations (Fig. 1.9). The mean Nei's genetic distance between populations from these two taxa is 0.46. The maximum genetic distance among North American D. exilis populations is 0.45, compared with D. spinulata’s maximum of 0.15, based on the same seven loci (Table 1). The remaining Ctenodaphnia species (D. dadayana, D. menucoensis, D. ornithocephala, and Ctenodaphnia sp. #1) have no close North American allies, as they are 18-28% divergent from all North American Ctenodaphnia species. However, an affiliation between D. menucoensis from Argentina and D. salina from North America, which show 23% divergence in the COI gene, was detected by phylogenetic analysis using the much slower-evolving 12S gene (see Chapter 3). Discussion Species interpretation of genetic evidence This study has provided the first genetic evaluation of species boundaries in the Daphnia of Argentina, A relatively comprehensive survey of the country was undertaken, with samples collected from both a wide geographic range (Fig. 1.10) and a large variety of habitats (Appendix I). Fifleen species were detected among the 176 populations examined. While additional rare species likely await discovery, the results presented here consti ite a solid foundation for a taxonomic revision of the genus in Argentina and, furthermore, provide the necessary framework for biogeographic and evolutionary studies.In general, interpretation of the genetic results was straightforward. With the exception of the polyploid populations belonging to the D. pulex complex (see Chapter 2), the genotypic characteristics of all other species showed that they are apparently diploid and reproduce by cyclic parthenogenesis. Moreover, intra- and interspecific divergences among COI sequences were consistent with divergences previously reported in Daphnia using COL (Schwenk et al. 2000). Although allozyme divergences were generally larger than most values previously reported for Daphnia, due to the deliberate selection of a small number of polymorphic loci, divergences among supposedly conspecific populations were always smaller than interspecific divergences, supporting the present taxonomic conclusions. Subgenus Daphnia The present genetic analyses suggest that the taxonomy of the Argentine members of this subgenus is simple. This fact is remarkable considering the confused taxonomic history of this group in South America. The older literature is rife with dubious varieties of D. obtusa, D. pulex, and other poorly-described species that are mistrusted by modem investigators (Marinone, pers. comm.; Paggi, pers. comm.). By contrast, the genetic evidence provides clear support for the recognition of seven species. The genetic analyses show that specimens historically identified as D. obtusa actually comprise three species. The congruence between the mitochondrial and allozyme results suggests that each taxon possesses both monophyletic mitochondrial lineages and an isolated nuclear gene pool. Although the three species formerly assigned to D. obtusa are allied, their large interspecific divergences suggest a long period of evolutionary independence. Furthermore, no evidence of interspecific hybridization was detectedamong these species, as heterogyzotes were never detected at loci exhibiting fixed differences between these species. Reproductive isolation was even apparent in the case of sympatry of D. obtusa #1 and #3 (site 159). Thus, the genetic evidence strongly supports the conclusion that these three groups warrant recognition as separate species. The melanized species (D. obtusa #3) corresponds to a taxon that has been previously called “D. middendorffiana” (Paggi 1973). However, D. middendorffiana ss. is a member of the D. pulex complex, while its Argentine namesake possesses elongate setae along its internal carapace margin, indicating its membership in the obtusa complex (pers. obs.; Paggi 1998). D. obtusa #3 is morphologically distinctive, but differences between D. obtusa #1 and #2 are more subtle, although individuals of D. obtusa #2 rily possess smaller heads than the other taxon (Fig. 1.10a). The three Argentine species in the D. obtusa complex are not closely related to any North American members of this group, as they show more than 17% divergence at COI from all northern species (D. cheraphila, D. obtusa, D. neoobtusa, D. pileara, D. prolata). While phylogenetic relationships among North and South American species require investigation, it is clear that populations from the two continents are not conspecific and that new names should be assigned to the three Argentine species. The proper identification of D. obtusa #1 is likely D. brasiliensis (Lubbock 1855). It is also possible that species names assigned by Daday (1902) or varietal names listed in Olivier (1962) correspond to one of the three taxa and should perhaps be resurrected. Although members of the obsusa complex are dominant, a member of the D. pulex complex is also common in southern Argentina. Mitochondrial evidence suggests that these populations are closely allied to North American D. pulicaria. As such, this 30represents a new species record for Argentina. However. as allozyme evidence indicates that these are tetraploid asexuals of hybrid origin (see Chapter 2), the classification of these populations is problematic. These lineages are likely to be D. pulex X D. pulicaria hybrids, but the paternal species is not known with certainty (Chapter 2). Despite this complication, the designation “D. pulicaria” is used to describe these populations in the remainder of this thesis. Individuals morphologically identified as D. ambigua were shown to belong to two divergent lineages. One group of populations is closely related to North American D. ambigua. Thus, this species is widely distributed in southern South America, as well as in North America. However, the clustering of mitochondrial haplotypes suggests that populations of D. ambigua on the two continents may represent a case of incipient divergence (Fig 1.3). The second group of South American D. “ambigua” includes just two individuals collected from the Coronda River, which may have been flooded out of reservoirs located upstream in the Parana River system. This lineage is more closely allied to D. retrocurva, from which it showed only 12% COI divergence, than to D. ambigua. Thus, D. “ambigua” #2 likely represents a new species belonging to the retrocurva complex. As this species was collected from a subtropical region of northeastern Argentina, it is possible that some of records for D. ambigua from Brazil and Venezuela also refer to this taxon. Its morphological similarity to D. ambigua suggests convergent evolution, or perhaps both species possess a conserved ancestral body form.Finally, D. peruviana, a highly-melanized mountain species described from Peru (Harding 1955), was confirmed to represent a divergent lineage. This species does not have any close North American relatives, and is apparently a South American endemic. Hyalodaphnia Although the taxonomy of the Hyalodaphnia was not clearly resolved in the present analysis, some important insights were obtained. All Hyalodaphnia populations encountered were identified as either D. gessneri or D. laevis, which are morphologically similar barring differences in head shape. Previous studies on members of this subgenus have shown that head shape is an unreliable taxonomic feature, as it is highly plastic and seasonally variable (Dodson 1988, 1989). The present study has again revealed incongruences between morphological identifications and mitochondrial lineages. Prior work has also revealed that interspecific hybridization often underlies phenotypic variability in the Hyalodaphnia (Taylor and Hebert 1993a; Wolf and Mort 1986), and this also appears likely in Argentina. Both populations of morphological “D. laevis” and one of “D. gessneri” show extremely high levels of heterozygosity, suggesting that these populations are comprised largely of hybrids. A more detailed investigation of these populations, combining morphological, mtDNA, and allozyme analyses on the same individuals, is necessary to verify the occurrence of hybridization and to clarify their taxonomy. It would also be desirable to survey further populations. Former studies have shown that the detection of “pure” parental populations, as evidenced by their conformance to HW expectations, is useful for elucidating species boundaries and the incidence of introgression in hybridizing species complexes (Taylor and Hebert 1993a, b, 1994). Although the current study examined only a few populations, the detection of twomitochondrial lineages showing 13% COI divergence, along with elevated levels of allozyme heterozygosity, does suggest that the /aevis complex in Argentina includes two divergent “species” that frequently hybridize. Although Argentine “D. laevis” may yet be shown to be a valid species, this name is not appropriate, since the South American populations are highly divergent (>17% at CON from North American populations collected near the type locality. Paggi (1977) first described Argentine D. laevis from a population from the province of Tucumén, in the northem part of the country. He noted several subtle differences between the Argentine animals and their North American counterparts (Brooks 1957), including a longer tail spine, more extensive setation along the valve margin, and a slightly different head shape. Since these characters are known to be variable and subject to change during cyclomorphosis, he concluded that the Argentine populations should be assigned to D. laevis. However, genetic evidence indicates that Argentine and North American lineages of D. “laevis” are highly divergent, suggesting the need for description of a new species from Argentina. ‘The name D. gessneri is appropriate for the second lineage in the Argentine laevis complex, since this species was described from South America (Herbst 1967). However, it is clear that further genetic work and a re-examination of diagnostic morpholo; features are necessary. Although based on a small sample size, it is interesting to note that several individuals with D. gessneri head morphology possessed “/aevis"-type mitochondrial haplotypes, but not the reverse. Taken alone, this result could indicate that helmeted forms represent either pure D. gessneri or D. gessneri X D. laevis hybrids, while unhelmeted forms represent pure D. “laevis”, an interpretation that would be 33consistent with results reported for hybridizing Hyalodaphnia in the Northem Hemisphere (Taylor and Hebert 1993a). However, the allozyme data suggest a more complex situation, since both D. “gessneri” and D. “laevis” populations display elevated levels of heterozygosity, suggestive of a hybrid origin. Cyclomorphosis may confound the interpretation of these genetic results, as it is possible that helmets are seasonally present in some of the populations that were found to be “unhelmeted” during the present study. A complete understanding of the taxonomy of hybridizing Hyalodaphnia species complexes is possible, but generally requires an intensive approach, involving repeated sampling and genetic and morphological analyses. “ten yhnia Of the three subgenera, the Crenodaphnia have been the “kindest” to the traditional taxonomist. As opposed to the situation in the Daphnia and Hyalodaphnia, the different Argentine Ctenodaphnia species are generally separated by large and concordant morphological, mitochondrial, and allozymic divergences. Genetic evidence supports the species status of D. dadayana, D. menucoensis, D. ornithocephala, and D. spinulata, which constitute four of the five Crenodaphnia species previously recognized by morphologists (Paggi 1998). However, neither mtDNA nor allozyme evidence support the recognition of a fifth morphospecies, D. notacantha. Daphnia notacantha (Birabén 1954) is morphologically similar to D. spinulata. barring a conspicuous hump on the back of its head. However, the results presented here have shown that individuals identified as D. notacantha and D. spinulata often possess identical mitochondrial haplotypes, and are allozymically indistinguishable. Moreover, some individuals of D. spinulata possess a hint of a notacantha-like hump, suggesting 34that there is a gradation of morphological forms. Since juvenile D. spinulata tend to resemble the notacantha-like form, this distinct head shape in adults may represent a neotenic maintenance of the trait, which could be either environmentally induced or genetically determined. Interestingly, one of the collection sites of D. notacantha was from a locality near Viedma, close to the location from which the species was first described nearly fifly years ago (Birabén 1954). The fact that other Daphnia species possess inducible head structures, such as helmets and neckteeth, suggests the possibility that the “notacantha” head protuberance is simply a predator-induced morphological form of D. spinulata. If so, its spatial distribution may coincide that of a particular predator. Genetic analysis suggests that the taxonomic status of D. spinulata itself requires examination because of its close affiliation with the North American species, D. evilis. As these taxa show just 1.5-3% COI sequence divergence, they are clearly closely allied. Moreover, the fact that all COI sequences for D. exilis are nested within the greater diversity present in D. spinulata suggests that one or a few dispersal events from South America could have established the North American populations (Fig. 1.8). However, allozyme variation suggests a different history. North and South American populations appear to be differentiated at nuclear markers, and, furthermore, there is greater nuclear diversity among North American populations (Fig. 1.9). The maximum genetic distance among North American populations of D. exilis is 0.45, compared to just 0.15 for D. spinulata, while the mean genetic distance between these groups is only 0.46. This discrepancy between levels of mitochondrial and allozyme diversity might be explained by the following hypothetical scenario: a) the common ancestor of all extant D. spinulata 35and D. exilis populations originated in South America; b) Groups A and B diverged in South America, possibly in allopatry, as suggested by their current distributions; c) one or a few propagules of a Group A member reached North America; d) North and South American populations diverged from one another; ¢) a second migration event from South to North America involving Group A members occurred; and f) mixing of North ‘American and Group A lineages from South America produced the elevated allozyme diversity. While the actual evolutionary history of these populations remains unknown, it is clear that D. spinulata and D. exilis are closely related. However, given their divergence at nuclear loci, along with their allopatric distributions, it is likely that populations on the two continents are now on independent evolutionary pathways. Thus, it seems appropriate to maintain recognition of the Argentine and North American populations as distinct species. ‘The present study is the first to document the presence of D. similis in Argentina. This species is clearly rare as it was only detected in a single habitat. However, the sole Argentine individual sequenced shows less than 2% COI divergence from two North American isolates, suggesting that populations from the two continents are conspecific. The other individual possesses allozyme alleles that are consistent with those reported among North American populations of D. similis (Hebert and Finston 1993). It is worth noting that D. similis is not an appropriate name for this New World species, as individuals of D. similis from Europe and Israel, near its type locality, are deeply genetically divergent from North American populations (J. Colboume, pers. comm.; Chapter 3), 36A single population of a new, apparently narrowly-endemic Crenodaphnia species was detected (“Ctenodaphnia sp. #1”). Recognition of this species is supported by both allozyme and mitochondrial results. It appears to constitute an old lineage that may be endemic to South America. Preliminary morphological study indicated that individuals of this species possess morphological features (e.g. very large antennular mounds, weak tail spines) that distinguish them from other Crenodaphnia from Argentina. ‘Summary This study presents the first genetic investigation of species diversity in Daphnia from Argentina. Fifteen species were identified by genetic criteria among the Argentine populations examined. The genetic results revealed that some species are consistent with previously-described morphospecies. However, other species are inadequately described, are incorrectly named, or were previously unrecognized in Argentina. A list of the species that were genetically characterized in this study is presented in Table 1.8, along with recommendations for future work and taxonomic revision. Given that at least four undescribed or inadequately-described species were revealed during this survey, as well as two species that had not been documented in Argentina, it is likely that additional species await discovery. Studies in the high alpine and northern regions of the country may be particularly fruitful. Continental endemism and intercontiner dispersal in Daphnia Among the fifteen species identified in the this study, eleven appear to be endemic to South America, while four (D. ambigua, D. “pulicaria”, D. similis, D. spinulata) are conspecific or very closely allied with North American species. A population of a fifth shared species, D. pulex, has been detected in Chile (pers. obs.). 37Another North American species, D. parvula, has also been reported from South America (Paggi 1998), but was not detected in the present survey. Molecular work is necessary to confirm the status of North and South American populations assigned to this taxon. The present study also failed to examine two species currently included in the genus Daphniopsis, D. chilensis and D. marcahuasensis, which are narrow endemics of the high Andes (Hann 1986; Valdivia and Burger 1989). As members of this genus are properly assigned to the Ctenodaphnia based on morphological (Hrbaéek 1987) and molecular (see Chapter 3) evidence, these species also require characterization. Although further sampling in other regions of South America is desirable, it is likely that most species of Daphnia in South America are now known. Thus, although additional species undoubtedly await discovery, current evidence suggests that 13 of 19 species now known from South America (68%) are endemic to this continent. North America harbours nearly twice the number of Daphnia species that are known from South America, as 34 (vs. 19) taxa have been recorded (Hebert 1995). Part of this difference is likely attributable to the higher intensity of sampling in North America. However, the difference likely also reflects a real difference in species richness. The lower diversity of South America may reflect the fact that much of this continent lies. in the tropics, a setting where daphniid diversity is low. Interestingly, a similar proportion of North American species (66%) is endemic to this continent, while other species are shared with South America or Eurasia (see Chapter 3 for detailed analysis). Australia, a much smaller continent, harbours approximately the same number of species as South America. About 21 Australian species are currently known (Hebert in prep.) all belonging to the subgenus Crenodaphnia. Five of these species have also been reported 38from Asia or Africa (see Chapter 3), suggesting that Australia’s fauna contains a higher Proportion (at least 76%) of endemics, likely reflecting the greater isolation of this continent. Concluding remarks This genetic investigation has significantly advanced our knowledge of the taxonomy and biogeography of the genus Daphnia. While formal taxonomic descriptions await completion, and further genetic surveys will likely reveal additional species, a solid foundation now exists for future studies on the South American daphniids. Moreover, with the delineation of species boundaries, investigations of interesting evolutionary phenomena and patterns, such as phylogenetic relationships, phylogeography, processes of speciation, and the evolutionary consequences of long-distance dispersal, are possible. The present taxonomic evaluation has already revealed some interesting biogeographic pattems. While most Argentine daphniids are endemic to South America, migration events between North and South America are surprisingly common given the great distances involved. Such intercontinental migration is apparently also of substantial evolutionary significance. Phylogenetic analyses of the shared North and South American taxa suggest that many of these populations from the two continents are in the early stages of divergence, suggesting that allopatric speciation at a global scale is an important mechanism of speciation in this genus. The taxonomic framework established in this study provides the opportunity for the exploration of the role of long-distance dispersal in the evolutionary diversification of a prominent zooplankter. 39Table 1.1. Levels of intraspecific variation for nine species of Argentine Daphnia. Maximum genetic distances for COI and Nei’s distance (based on allozyme variation) are used to characterize intraspecific diversity, while Fs values indicate the level of intraspecific genetic fragmentation. Nei’s genetic distances and Fsr values are based on seven allozyme loci for all species (AAT, FUM, GPI, LDH, MPI, PGM plus G3PDH for species in the subgenus Daphnia and ME for Ctenodaphnia species), with the exception of D. ambigua #1, for which four loci were used (AAT, GPI, LDH, PGM). Statistics for D. exilis from North America are included for comparison with D. spinulata. D. exilis data are from the present study and from Hebert and Finston (1993) (see text). ‘% Maximum pairwise Maximum Nei's K2P distance for COI _genetic distance Species (No. of populations) __(No. of populations) _Fsr Subgenus Daphnia D. ambigua #1 0.9 (4) 0.00 (2) * D. obtusa #1 43(41) 0.54 (27) 0.50 D. obtusa #2 138) 0.04 (5) 29 D. obtusa #3 239) 0.128) 081 D. “pulicaria” (polyploid) 0.3 (8) 0.01 (16) cS Subgenus Ctenodaphnia D. dadayana 3.6 (21) 0.87 (24) 0.52 D. menucoensis 1.4(12) 0.33 (9) 0.33 D. ornithocephala 03 (4) 0.18 (4) 0.49 D. spinulata 3.9 (38) 0.13 (37) 0.24 D. exilis 191) 0.36 (16) 0.43 * Fst was not calculated, as only two D. ambigua populations were available for allozyme surveys. ** Fer statistic was not calculated for this obligate parthenogen, as this statistic assumes sexual reproduction. 40= sl ww 0c ste vst CST 4. Punoynd,, hae 7 Cle £87 Vez fe puntansad 180 wu 90e viz St £4 psmyo sso wu Lol a ra Vw ZH psmigo ool wu £60 80 e Ow Ly 2siiyo wu. wu eu eu cu - ZH onBiquin Sve eu wi ort wi eu La onBiquie vunoynd, —punianiad gH DSO. —_ZHDSMO —[YSMYO ZH YNBYUD —[y MBIqaD yuydog (euo8eip anoge) aouersip s,19N / (IeuoTeIp Mojaq) % UI doURISIP JZ AAeIDAY uoxey, pupjansad G40 7 ONSIqUD ‘CJ 405 aIqe|teAt 10U 219m eiEp aUKzol|Y ‘set9ads wruydypouDry 105 AW pur vruydngy srudfiqns oy uI4ILN soI9ads 40) HED std (Dd PU {fa ‘HO ‘1d WA ‘LVY) 1901 xis outes ayn uodn paseq st q_‘so!sods yo sited soypo [Te JO “14 PUSIgn ‘g Hutafoaut suostredtos 40) (Dd Pte ‘HG ‘Id ‘LVV) 190] snoy Sjuo uo poseq st. “xuFetu ay) Jo sey dor ay) Adnad0 sowzo][e Uo paseg (q) s9oUueisip ouad $,10N aSesaAy “XUTEU dtp Jo JfeY toH}og a4) UF pauasord oue UIT [OD O41 UI SsoUATLOAIP (ZH) Jo}OWELEd-Z-cANLUTy astnuIE afeiony “viuydypouar) pue vinydog exsuofqns amp uiyuar soisods auntay wooaag aoueIsIp onoudH yo sareuNSy "Zt IGP, 4‘poueys 219M S9f9}]e OU Se POIEIND]E dq OU PINoD D9UIEISIP DNOUIH y oC sve se ut ta lids 8st 60 901 09 ivjnurds we 60z 19% oe leds 802 we 09% od pyonuids ~ Iz ole oe sygtats * eS roe ww -vyoydavoyinuo zl 6x9 a 987 siswaoanuatu 81 . ee7 _ pundopop syjuis —nyoydaooynuuo —_stswooonuatu puncopop juydopourry (jeuo3eip 2noge) 2ouersip s,10N / ({euoferp moj9q) aoueIstp ZW adeI0Ay woxey, panunuos °Z'1 3148,Table 1.3. Mean allele frequencies at seven allozyme loci for five species of the subgenus Daphnia in Argentina. Alleles are named according to their relative mobility compared with a laboratory clone of D. pulicaria. The number of populations analyzed for each species is indicated in parentheses. Not all individuals were scored for all loci (see Appendix IV for detailed data). Allele frequencies may not always total 1 due to rounding. Abbreviations are: npops = number of populations; n = number of individuals. Daphnia: ambigua #1 obtusa #1 obtusa #2 obtusa #3. “pulicaria”* Alleles (npops=2) __(npops=27) _(npops=4) (npops=16) AAT (n=64) (n=632) 32) (n=328) 0.73 0.87 = = 0.75 - 0.03 ~ 0.88 = = - 1.00 0.13 0.97 . 1.00 1.09 = <0.01 ~ = FUM (a=0) (a=737) in=194) —(n=328) 1.00 Wa 0.78 ~ 0.50 Lun wa 0.22 . = 14 Wa ~- ~ 0.50 1.20 Wa = . 1.00 = GPI (0-622) (n=139) (n=328) 88 <0.01 0.08 - 0.50 0.92 = = = = 1.00 1.00 0.73 1.00 0.50 1.10 - - 0.19 - - Lig - - 0.01 - - G3PDH —(n=0) (n=585) (n=328) 0.87 wa 1.00 ~ 1.00 wa 1.00 a3Table 1.3 continued. Daphnia: Alleles ambigua #\__obtusa#\___obtusa#2_ obtusa #3_“pulicaria* LDH (n=37) (n=576) —(n=126) (n=234) —(n=328) 0.79 1.00 - - ogi 1.00 - 0.83 - ~ 0.50 1.00 - 1.00 0.50 112 - <0.01 - - - MPL (a=0) (n=781) (n=360) 0.86 Wa 0.04 0.47 0.92 wa 0.16 0.03 0.96 Wa - - 1.00 wa 0.40 0.47 1.03 va 0.17 ~ 1.09 Wa 0.19 0.03 119 va 0.03 - 1.30 Wa <0.01 _ - PGM (n=44) (n=778) (n=328) 0.89 - <0.01 - - 1.00 - 0.03 ~ 1.00 1.06 - 0.45 0.04 - 12 1.00 0.51 0.96 - 1.22 - <0.01 - - - * These populations are tetraploid asexuals, with fixed heterozygosity at four loci (see Chapter 2 for details). 4Table 1.4. Allozymic diversity and varial y within populations of 12 Daphnia species from Argentina. Statistics for D. ambigua #1 are based on four loci (AAT, GPI, LDH, and PGM). The same six loci (AAT, FUM, GPI, LDH, MPI, and PGM) are used for all other species, with the addition of G3PDH for some species of Daphnia and ME for most of the Ctenodaphnia. Ho = observed heterozygosity; He = expected heterozygosity; Fis = inbreeding coefficient. ‘Average % loci ‘Average % individual Species (No. of populations) polymorphic Per population Ho (He) Fis ‘Subgenus Daphnia D. ambiguattt (2) D. obtusa #1 (27) D. obtusa #2(4) D. obtusa #3 (8) polyploid D. “pulicaria” (16) Subgenus Hyalodaphnia Dilaevis/gessneri (3) Subgenus Ctenodaphnia D. dadayana (24) D. menucoensis (9) D. ornithocephala (4) D. similis (1)* D. spinulata (36) Ctenodaphnia sp. #1 (No. of loci) 25.0(4) 29.6 (7) 25.0(7) 5.4(7) $7.1(7) 38.9 (6) 29.4(7) 21.5(7) 12.5 (7) 06) 29.0(7) 28.6 (7) heterozygosity per population: 6.7(5.9) “0.14 11.4 (11.4) 0.001 6.7 (5.8) -0.17 1.41.9) 0.26 $7.1 30.1) -1.00 32.5 (17.8) 0.85 10.0 (10.3) 0.03 8.3(8.2) -0.02 6.2 (5.8) -0.08 10.2 (9.3) -0.10 10.3 (8.7) -0.19 * Only a single individual of D. similis was available for analysis.Table 1.5. Average K2P sequence divergence in the COI gene among the five Daphnia ambigua phylogroups identified by phenetic analysis (adapted from Hebert et al. in review). Phylogroup Average K2P distance (in %) East Central North West South West ‘South American East 3.82 478 5.03 5.70 Central North West South West 4.10 - 2.59 4.46 - 3.17 4.89 4.07Table 1.6. Genotypic frequencies at six allozyme loci in three populations of Hyalodaphnia from Argentina, Populations 30 and 32 were identified as D. laevis, while population 226 was identified as D. gessneri, based on head morphology. Genotype 30 32 226 AAT (n=19) (n=21) (n=21) Homozygotes: 0.79/0.79 1.00 = - Heterozygotes: 0.79/0.90 - 1.00 1.00 PGI (0-87) (n=21) (n=22) Homozygotes: 0.85/0.85 0.06 0.09 1.00 Heterozygotes: 0.73/0.85 0.94 0.91 - PGM (n=85) (n=21) (n=42) Homozygotes: 0.92/0.92 0.01 = 0.07 1.07/1.07 0.01 0.90 ~ Heterozygotes: 0.92/0.97 0.82 - - 0.92/1.07 0.14 0.10 0.93 0.97/1.07 0.01 - - FUM (n=22) (n=18) (n=20) 1.00-a 1.00 1.00 1.00 LDH (n=22) (n=22) (n=16) 0.87-a 1.00 1.00 1.00 MPI (n=21) (n=21) (n=8) 127-2 1.00 1.00 1.00 7Table 1.7. Mean allele frequencies at seven allozyme loci for six species of the subgenus Ctenodaphnia in Argentina. Alleles are named according to their relative mobility compared with a laboratory clone of D. pulicaria. The number of populations analyzed for each species is indicated in parentheses. Detailed data for each population are presented in Appendix IV. (npops = number of populations; n = number of individuals.) Crenodaphnia: spinulata similis menucoensis dadayana ornithocephala sp. #1 Alleles (npops=36) (npops=1) (npops=1) AAT (=1303) (n=1) 0.77 - 0.85 - = 0.96 - 1.00 1,00 = 1.03 0.72 - 1.05 - 112 0.28 _ (n=39) FUM = (n=1153) (n=1) 0.70 - - 0.76 0.85 0.91 0.98 1.00 1.03 1.23 GPI 0.79 0.92 - - 0.95 - - 1.00 - 1.00 Lu Loo - LDH (n=384)— (n=) 0.72 <001 0 = 0.80 - = 0.84 - - 0.87 1.00 1.03 = 1a = = 48Table 1.7. continued. Crenodaphnia: Alleles spinulata similis __menucoensis_dadayana_ornithocep! ME (n=532) (u=0) 0.78 Wa 0.90 wa - 1.00 - 0.93 wa 1.00 - 0.95 - na - - (n=157) (n=64) (n=11) MPI 0.92 0.97 1.06 LiL 112 116 118, 1.23 1.30 PGM 0.92 0.98 Log Lut 118 0-awsso9ou st uorstad aquwouoxes Inq, OUKA a4 Jo au0 jwasaidos Aout Sot90ds SIN “SAA “(Z961) 421A110 Ut pars] S: “{SS81) Y90qqN] Aq KensNup| Woy paquasop saroads, v ‘sysuaysoag “99 Kew [4 Bsmigo “q_“(¢ s0Idey9 998) vsmigo ‘qj weadomns] Jo ULOLIAWY YON ay YIM d1yt9adsuOd SI so1dads ary) asay Jo TUON, tuoisiaa4 astnbos xoqduuoa wsiigo aip Jo saisads ueouioury yInos aq [IV “SHA “(¢ sardey, 998) xajduios munsosja4 ayy Jo aqui v AyeOr yeyp payeoiput sisKqeur. atrauodol4yd “paquasopun aq 01 steadde satsads a4t]-,onSiqu,, SIU, “SAA “(LS61 SYM 40861 Sorowyy) adosng 0} uonanponut ouaBodospue ue juosasdas 0} 1YTNOY si sotoods si -puejSug ut suoneindod woy (Z461) Playsnoog £q poquosop Jsay Sea yoIyes (aMatAad UF "Ye 19 HOgoH) YnTiquun “G saIoads WeILDUY YON, ‘941 ypta aytoadsuoo 9q 01 swodde suoneindod aunuaday “Axess920u LON ‘[hressooou UONeANSSAUT SA SaA SAA ON TpepuDUAGDST adueyo we a tt nyuydog snusdqng TOs womreRisontT) ——_{papuauntuosas Apmis sa poreutisap se 50(0661) 188eq Aq paquosap KyBno1oyp Ajuo994 sem saroods S14 (ON (6661 188eq) paquasap 1jam puo994 sem so}oods 514, (ON -pouureua1 2q plnoys snip pue ‘uoxel uesLOWTy YLON ayt yt ayroadsuo9 You any, “q_ Se pareusap suorreindod aunuaBay ‘9amoy ‘posn aq ou! 8919}) aIuiapus UedLAIUY YIMOg v Se paquosop sear vausso8 ‘q Sy ‘woHeZIpUgty Jo auapIoUt ay jarop pure AttioUoXe) a4 ‘Aju 0} Aaessooou st sisKyeue anausd pue earfojoydiow pouiquo3 “$3 bos st uonenysoaut sony “spugcy xapnd “GX put projdjod asoyy, "eunusdy wt paruauna0p ‘ua9q 10u 9Avy Xojdutoo xaynd “q7 Oyp Jo soatvewuosoiday “SIA (9861 punwey op ede!) vunuafiry ut paruoumnsop Aysnoraaid uaaq sey pur (¢¢61 Buypsey4) nu2q wo4) paquosop sem sotoods sip, ‘Sunsozout sdeysad nq ‘A1ess999U LON, -uonnduasap eunoy smigo ayy ut saivads Mau e st if se nq “(EL6I Hed “F'9) ON ON ON ON SHA nuydopouary boussa8 qs muydopoosyy .munoynd,, cH vsmiqo“q st-sowpaid e 01 aunsodxo &q poany 2q Kou youn ozoynugds ‘g Jo wo} jeorSojoydious e syuasosdas Ajaypy 1} “PeaIsu “satgads piqea e 10U S1 DysuavIOU “Gg ‘APMIS SIy) Jo S){NS9s BY WO PIseg “SAA pastnbai st siséyeue peorSojoydiow se jjom se ‘uorfos siyy toss Suyjduses Joyyng “soleods s1tuapus Sjmoueu ‘paquosapun ue 9q 0) steadde stM, SAA -sozoads om asayy {Jo snyeis orwouoxe) ayp Sutssasse ur ryasn aq pjnoo sy1x9 “g_Jo suoneindod UROLOWY YON YIM Uostedwiod jeorBojoydioyy ‘souIs paredrysaaut u99q rou sey ing “(L 161) uaqeste Aq poquiosap sem sotoods SIY], “SA “(¢ sardoyg 998) jaeasy ut ‘Qouls pourtuexa Ajqeoryu9 w39q Jou sey Inq “(ps6I) UaqestE Aq Poquosop sem soloads siti], Zunsoiaqut 99 pnom sdeysod inq ‘Axess900u LON SHA SA ON SAA ON pysuvaviou g inynuads a yoydooor92. 91 204, 206, 210, 212 400 105 FL4Haplotype "a" rt 100 is D. obtusa #1 L Haplotype “8° —29 Li} a 43 iar 243, ESE 135, 136 ze 18 D. obtusa #2 16 bern 248 ———_______- _ p i . peruviana 8 199) Dz “ambigua”’ #2 188 46 162 191 /a9, 93, 94 D. obtusa #3 174 100/283 251 —2a7 24 sl i | Polyploid D. “pulicaria” 159 100 237, 242_ |} D. ambigua #1 (—__________—p mendotae 0.05 Fig. 1.1. NJ tree based on COI sequence variation among all unique haplotypes found in Argentine populations of the subgenus Daphnia. This tree was constructed using K2P distances and pairwise deletion of missing sites. D. ‘mendotae, a North American species belonging to the subgenus Hyalodaphnia, ‘was included to root the tree. Bootstrap values (based on 1000 pseudoreplicates) for major clusters and among clusters are presented. D. obtusa #1 “haplotype A” was found at sites 1, 13, 15, 16, 17, 18, 20, 22, 26, 27, 40, 42, 46, 48, 122, 243, 250, and 256; haplotype “B” was found at sites 132b, 169, 172, 183, 193, 194, and 195. D. “pulicaria” haplotype °C” was found at sites 135, 156, 171, 202, and 205. Site codes are listed in Appendix I. 325 st fees D. obtusa #1 210 206 212 163 178 193 195 8 f 118 fae ] D. obtusa #2 at 188 89 ot 165 i D. obtusa #3 199 162 v3 0.4 Fig. 1.2. UPGMA tree based on allozyme variation at seven loci in three species of the Daphnia obtusa complex from Argentina. The scale bar represents Nei's genetic distance.mt AR Central NE Central A [se ORB Central TX1-A 92] MEX2 MEX1 36) [plat Lax1e 1x2 ra] Lr cate a3 4 Southwest NY 74) — Mat [—RI FLA FL2-AFL7-A I FL2B FL7-B 7] pr FL3FL6 Eastern [Maz fF FL4-A, FLS-B FL4-B FLS FLE-A \0-ORA CATA __] Northwest oof 237, 242 —*[Fenne ‘J South American 286 0. pulex 102 Fig. 1.3. NJ tree constructed using K2P distances among all unique COL sequences for D. ambigua from North and South America. Another member of the subgenus Daphnia, D. pulex from North America, was used to root the tree. Bootstrap values from 1000 pseudoreplicates are presented for major clusters. Site codes for all North American populations and the single Chilean population, as well as the cluster designations, are from Hebert et al. (in review). Haplotype “Central A” was detected at sites LA2, MO2, OKI, OK2. and TX3. “Central B” was found at AR, CA2, [D1, [D2, 1D3, MOL, NE, OHI, TX2, and TX3. Site codes for Argentine populations are listed in Appendix I. The scale bar designates K2P genetic distance. 38D. gessneri form Fig. 1.4. Three head morphologies encountered among Argentine populations assigned to the D. laevis complex.to { 2: laevis (8), D. gessneri (8) 100. D. laevis (30, 32, 36), D. gessneri (226) D. gessneri (9) D. dubia (NA) BL. magniceps (NA) D. laevis (NA) D. mendotae (NA) 0.02 Fig. 1.5. NJ tree based on COI sequence variation among Argentine populations identified as D. laevis and D. gessneri. The identification based on head morphology is indicated in this tree, followed by the collection site numbers. Sequences from the North American members of the D. laevis complex (D. dubia, D. laevis, and D. magniceps) were included for comparison. D. mendotae, a Hyalodaphnia species belonging to a different species complex (Colbourne and Hebert 1996), was used to root the tree. Bootstrap values based on 1000 replicates are presented, The scale bar represents K2P distance. 3716,2338, 49. 58,73 20 6 7 ez. 125,250 23206 A ts D. spinulata / D. notacantha 39,112 100 Mes 105 204.206, 209. 210,213 J B D.z similis J D. ornithocephala Ctenodaphnia sp. #1 D. dadayana D. menucoensis 0.02 Fig. 1.6. NJ tree based on COI sequence variation among all unique haplotypes of Argentine populations belonging to the subgenus Crenodaphnia. Two members of the subgenus Daphnia (D. obtusa and D. pulex) were included to root the tree. Bootstrap values based on 1000 pseudoreplicates are presented for major clusters. K2P distances were used, which are indicated by the scale bar. and pairwise deletion of missing sites was employed. The collection site of each individual is indicated by its population code (see Appendix 1). 3899 | D. spinulata Group A (102, 103) D. exilis (Texas) D. spinulata Group B (67) 1007 — D. similis (Nevada) eee ee similis (128) 68 — 0. similis (Washington) 0.02 Fig. 1.7. NJ tree based on COI sequences for two populations of North American and one population of South American Daphnia similis. Populations of Argentine D. spinulata and North American D. exilis are included for comparison. The scale bar represents K2P distance. 2»71,74 [— 229 Les 66 bal Lyp— 113 85 Argentine D. spinulata 14 f 105 | Group A 204 [—— §6 Ly— 60 7 wat fot cent alleA-2 = mex ok a stor North American D. exilis 91 | \———— chan ord mata ly — alles amar 100 f 102 67 | Argentine D. spinulata —— 0.005 Fig. 1.8. NJ tree based on COI sequence variation among a sample of Daphnia spinulata populations from Argentina and D. exilis populations from North America. The scale bar represents K2P distance. The codes for Argentine populations are provided in Appendix I. while D. exilis codes are found in Appendix Il.60, 102 85 7 7 49 128 18 37, 403 114 23 4 42 229 4 D. lk . spinulata (Argentina) 2 pi i 198 288 43 ny D. exilis (North America) 01 Fig. 1.9. UPGMA wee based on allozyme variation at seven loci in Daphnia spinulata populations from Argentina and D. exilis populations from North America. Data for most of the D. exilis populations are from Hebert and Finston (1993), but trimmed to the same seven loci surveyed in the Argentine populations. The scale bar represents Nei’s genetic distance. Codes in capital letters are from Hebert and Finston (1993), while codes in small letters represent new data (alle = pond near Allende, Mexico; amar = Amarillo. Texas: mata = Mata, Mexico; santo = Santo Domingo, Mexico). oa) @ D. obtusa #3 wi} te Diobrusa #1 and re a #3 co-habiting 400km, Fig. 1.10. Sampling localities of Argentine Daphnia populations. Photographs are presented forasingle individual of each species, where available. Maps are presented for: a) members of the D. (Daphnia) obtusa complex, b) all other species within the subgenus Daphnia, c) populations belonging to the subgenus Hyalodaphnia, d) members of the D. (Ctenodaphnia) Similis complex, and e) all remaining species within the subgenus Ctenodaphnia. 62b) & D.ambigua #1 D. “ambigua” #2 (retrocurva complex) WD. perwviana © D. “pulicaria” 63°) “34 D. gesineri form “a7 (genetically same as faevis) ‘& D laevis form é *e Both and intermediate forms \ Pe Gnitochondil lineagesDD. notacantha © D spinutaa & D. similis BE da) 65e) adult female @ D. memucoensis ° BD omiceptas ~~ 66Chapter 2: New insights into the distribution of polyploid Daphnia: the Holarctic revisited and Argentina explored Abstract It has long been known that polyploid plants and animals are more prevalent in arctic than in temperate environments. Past explanations for this geographical trend have focused on the role of glacial cycles in generating polyploids and the influence of abiotic factors in favouring polyploidy in the north. In combination, these mechanisms probably suffice to explain the observed geographical cline in ploidy levels in members of the Daphnia pulex complex in the Holarctic. While only diploid members of the D. pulex complex are found in the temperate regions of North America and Europe, allozyme and DNA quantification analyses indicate that the southem Argentine pulex-complex fauna is dominated by polyploids. Indeed, the present study is the first to document the presence of polyploid members of the D. pulex complex in any temperate climate. Phylogeographic analyses suggest that this difference in polyploid distribution between the northem and southern hemispheres is based more on ecological and historical contingencies than direct selection for polyploidy. Specifically, competition with diploid relatives likely limits the lower latitudinal range of polyploids in the north, but appears not to have occurred in Argentina, Because of these differences, the present study provides important insights into the diverse factors that determine the distributions and evolutionary fates of polyploid organisms. 67Introduction Polyploidy is a phenomenon of particular interest in evolutionary biology, as ii provides an opportunity for instantaneous genome reconfiguration and speciation. Not surprisingly, considerable attention has been paid to the observation that arctic and alpine environments harbour a substantially higher proportion of polyploid organisms than temperate regions. Such an elevated incidence of polyploidy n the north has long been documented in plants (Stebbins 1950), and is also apparent in animals such as insects (Suomalainen et al. 1987) and zooplankton (Beaton and Hebert 1988; Little et al. 1997). Notably, the dominant freshwater fish family in arctic Canada, the Salmonidae, is also polyploid (Gregory and Hebert 1999). Although several hypotheses have been advanced to account for these patterns, a comprehensive understanding of the relationship between ploidy level and environmental conditions has remained elusive, largely due to the fact that polyploidy and parthenogenesis tend to be acquired together (Glesener and ilman 1978; Bell 1982; Suomalainen et al. 1987; Beaton and Hebert 1988). However, studies on plants have been particularly useful in demonstrating the adaptive significance of polyploidy in the arctic, as comparisons of closely-related diploid and polyploid sexual species have permitted the effects of ploidy level to be examined independently of variation in breeding system (reviewed in Bierzychudek 1985). It is a freshwater zooplankton species complex that has provided the converse study system, in which geographical variation in ploidy level has also been examined under a constant, this time asexual, breeding system. The Daplinia pulex group in North America contains obligately parthenogenetic diploids and polyploids, as well as sexual diploids. Surveys of clonal Daphnia populations 68have revealed that high arctic environments are dominated by polyploids, while both diploids and polyploids occur at low arctic sites, but only diploids are found in temperate zones (Beaton and Hebert 1988). Studies on Daphnia from other geographic regions (Ward et al. 1994) and on other cladoceran genera (DeMelo and Hebert 1994; Little et al. 1997) further suggest that polyploidy is prevalent among zooplankton inhabiting the arctic, but is absent in temperate regions. Thus, geographic patterns of ploidy level in zooplankton suggest that polyploidy is adaptive in the arctic. Members of the D. pulex complex have a broad Holarctic distribution, and while they are absent from the tropical regions of both North and South America (Paggi 199! Hebert and Finston, 2001), they have been reported from southern and alpine regions of the latter continent (e.g. Villalobos 1994). Argentina, which boasts a diverse landscape and extends into the southernmost reaches of South America, provides an excellent opportunity to explore breeding system diversity and ploidy levels in the D. pulex complex along a latitudinal gradient in the south. This study presents the first genetic investigation of members of the D. pulex complex in South America. Breeding system, clonal diversity, ploidy levels, and the origin of Argentine populations were investigated using allozyme, mtDNA, and DNA ‘quantification techniques. Previous explanations for the geographic patterns of polyploidy and asexuality in the D. pulex complex are re-evaluated in light of the present results, and a new synthesis involving the roles of abiotic, ecological, and historical factors in the generation and maintenance of polyploids is proposed.Material and Methods Specimen collection and identification During sampling campaigns in November-December 1999, and January-February 2001, zooplankton collections were made from more than 250 water bodies throughout Argentina, 176 of which contained Daphnia (see Appendix 1). Habitats sampled included ponds, roadside ditches, playa lakes, alpine lakes, rivers, reservoirs, and salt lakes. Initial species assignments were made in the field based on morphology. Animals were sorted alive and either flash-frozen in liquid nitrogen for allozyme surveys or preserved in 95% ethanol for DNA analysis. Taxonomic assignments were later revised based on the results of mtDNA analysis and more detailed morphological investigations. Species of the subgenus Daphnia were identified by their prominent medial pecten, and members of the D. pulex complex were separated from those of the D. obrusa complex by their lack of clongate setae along the internal margin of the carapace (Schwanz et al. 1985). Populations belonging to the D. pulex complex were encountered in 16 ponds in extra-Andean regions of the southernmost provinces of Argentina (Santa Cruz and Tierra del Fuego), numbered ARG 1-16 (Table 2.1; Fig. 2.1). Allozyme analys Allozyme surveys were conducted on all Argentine populations by cellulose acetate electrophoresis using a Tris-glycine buffer (pH=8.5) (Hebert and Beaton, 1993). Populations were screened for variation at seven commonly-polymorphic loci: aspartate amino transferase (AAT) (EC 3.2.1.1), fumarate hydratase (FUM) (EC 4.2.1.2), glucose- 6-phosphate isomerase (GPI) (EC 5.3.1.9), glyceraldehyde-3-phosphate dehydrogenase (G3PDH) (EC 1.2.1.12), lactate dehydrogenase (LDH) (EC 1.1.1.27), mannose-6- 70phosphate isomerase (MPI) (EC 5.3.1.8), and phosphoglucomutase (PGM) (EC 5. 2): All gels were electrophoresed for 15 minutes at a voltage of 206, as pilot trials revealed that this run time maximized allele separation while maintaining crisp bands. From most populations, 18-44 individuals were analyzed, but only two individuals were available from ARG 13. During each staining run, two individuals from a clonal stock of North American Daphnia pulicaria (from Lake Washington, Weshington state, USA) were used to standardize scoring. A population of D. pulex (from a pond in Guelph, Ontario, Canada) was also used in several runs for comparison. Particular attention was paid to the comparison of alleles at LDH, as this locus is useful for diagnosing species in the D. pulex complex (Hebert et al. 1989). Allozyme surveys were used to estimate levels of genetic diversity in Argentine populations and also to determine their breeding systems, as this approach has been shown to provide reliable diagnoses in studies that coupled allozyme analysis with subsequent breeding experiments (e.g. Hebert and Crease 1983). As well, allozyme phenotypes were used to gain an initial indication of ploidy levels. Evidence of polyploidy includes the detection of three or more alleles at a locus or asymmetrical banding patterns, which generally indicate uneven (and hence multiple) allelic copy numbers (e.g. Ward et al. 1994; Dufresne and Hebert 1995). Mitochondrial DNA analysis Total DNA was extracted from several individuals from each population in 50 uL of proteinase-K extraction buffer, according to the protocol of Schwenk et al. (1998). A 710-base pair (bp) fragment of the cytochrome ¢ oxidase subunit I (COI) mitochondrial gene was amplified via the polymerase chain reaction (PCR) (Saiki et al. 1998) from a single individual from seven populations (ARG 1-4, 8, 9, and 15) using universal primers nLCO1490 and HCO2918 (Folmer et al. 1994). A 550-bp fragment of the NADH dehydrogenase subunit 5 gene (NDS) was amplified for one individual from ARG 9 and ARG II using primers designed for Daphnia (Colbourne et al. 1998). Each 50-L reaction consisted of 3-5 uL of DNA template, 5 uL of 10x PCR buffer (10mM Tris-HCI, pH 8.3; 50 mM KCI), 0.2 uM of each primer, 1.5 mM MgCl, 0.2 mM of each dNTP, and 1 unit of Tag DNA polymerase. The PCR thermal regime for both genes was as follows: one cycle of 1 min at 94°C; 5 cycles of 1 min at 94°C for | min, 1.5 min at 45°C, and 1.5 min at 72°C; 30 cycles of 1 min at 94°C, 1.5 min at 50°C, and 1.5 min at 72°C; and finishing with a final extension at 72°C for 5 min. PCR products were gel-purified using the Qiaex II (Qiagen) kit and sequenced in one direction (with primer LCO1490 for COI or NDS-A) using the Big Dye Terminator (version 3) sequencing kit (ABI Prism), Sequencing-reaction products were electrophoresed on an ABI 377 automated sequencer (Applied Biosystems). ‘Sequence electropherograms were inspected and aligned using the SeqApp 1.9 sequence editor (Gilbert 1992), resulting in a final alignment of 630 bp for COT and 499 bp for NDS. All unique sequences are available from GenBank (accession numbers: ‘AF489523-AF489527). Pairwise genetic distances were calculated using Kimura's (1980) two-parameter model (K2P) in MEGA 2.1 (Kumar et al. 2001) and used to construct phenograms by the neighbour-joining (NJ) algorithm (Saitou and Nei 1987). In all cases, trees were constructed using pairwise deletion of missing sites, and bootstrap values were based on 1000 pseudoreplicates. COI sequences for North American members of the Daphnia pulex complex were also included in the analysis. Unpublished D. pulicaria and D. tenebrosa sequences werecontributed by J. Colbourne, while the D. pulex sequence was from Crease (1999). Much of the past work on the Holarctic members of the D. pulex complex also employed the NDS gene. Consequently, the two NDS sequences obtained in the present study were incorporated into the data set of Colboume et al. (1998) to examine the relationship between North and South American members of the D. pulex complex. DNA quantification Whole bodies of ethanol-preserved Daphnia from Argentina, along with a diploid laboratory culture of D. pulicaria, were stained in a standard Feulgen protocol (Beaton and Hebert 1988). Briefly, the animals were placed in glass tubes with perforated caps that allowed their simultaneous submersion in the various solutions. Daphnia from the live culture were first placed in 95% EtOH and allowed to dehydrate, and then all specimens were post-fixed in an MFA solution (85 MeOH: 10 formalin : 5 glacial acetic acid, v:v:v) for 30 min, Specimens were next rinsed in running tap water for 10 min, hydrolyzed in SN HCI for 30 min, and then dipped in 0.1N HCI to prevent the carry-over of strong acid. Hydrolyzed specimens were then stained in freshly-prepared Schiff reagent for 60 min before being passed through three rinses in fresh bisulfite solution. A second 10-min tap- water rinse was followed by three changes in distilled water, and the animals were then transferred to 20% EtOH for storage. Daphnia species display pronounced endopolyploidy in most tissues, but the exopodites of the thoracic limbs show no evidence of this phenomenon, and were therefore selected for analysis (Beaton and Hebert 1988, 1989). Within 48 hours of staining, exopodites were removed by dissection in distilled water and allowed to air-dry onto microscope slides. Relative nuclear DNA contents were determined using the Bioquant True Color Windows 98 v3.50.6 image analysis software package (R&M BBiometrics Inc., Nashville, TN) along with an Optronics DEI-750 CE three-chip CCD colour camera connected via a BQ6000 frame-grabber board to a Pentium If 300MHz PC (Hardie et al. 2002). Specimens were visualized under a 100X oil-immersion objective (No = 1.515) on a Leica DM/LS microscope. The exopodites were mounted in immersion oil with a refractive index of 1.520 to reduce glare caused by exoskeletal chitin (Beaton and Hebert 1988). A 546-nm interference filter was used to increase the contrast between the stained nuclei and the background. Integrated optical densities (IODs) were measured from six individuals from ARG 11, for a total of 65 nuclei. As well, 70 nuclei were measured from individuals of the diploid laboratory culture. Similar, though less intensive, measurements were performed on individuals from three other populations (ARG 2, 4, and 6). OD measurements were consistent across individuals of the same population, generally showing coefficients of variation of less than 10% when the [OD data were pooled. The DNA contents of the purported polyploid clones were evaluated by comparing the ratios of the mean IOD of each Argentine population to that of the diploid D. pulicaria culture. Allozyme surveys revealed only two multi-locus genotypes in the 16 Argentine populations, which are subsequently referred to as clones A and B (Table 2.2). Only a single clone was detected in each habitat. Clone B was found at two sites (ARG I1 and 13), while clone A occupied the remaining 14 Argentine habitats. Both of these clones were heterozygous for the same alleles at FUM, GPI, and LDH, and were homozygous for ”the same allele at AAT, G3PDH, and PGM. While FUM and LDH heterozygotes exhibited normal phenotypes, the staining patter of all GPI heterozygotes was unbalanced (with the faster allele staining more heavily), suggesting that both clones were polyploid. Further evidence for the polyploid state of clone B was observed at the final locus, MPI, which is a monomer. Although clone A was a heterozygote with two roughly equally-staining alleles, clone B displayed a phenotype suggesting that it possessed four different alleles. Given the observed fixed heterozygosity, all sixteen populations were assumed to reproduce by obligate parthenogenesis. COI and NDS sequence diversity There was very limited sequence divergence among the seven COI sequences examined from the Argentine populations (all containing allozyme clone A). Only three nucleotide positions were variable among the sequences, and the maximum K2P pairwise distance was 0.32%. Analysis of NDS sequences from a single individual of each clone indicated that they were also closely allied, as their sequence divergence was only 0.20%. Phylogenetic ret ionships Mitochondrial DNA analyses indicated that Argentine populations assigned to the pulex complex were closely related to North American members of this complex, specifically D. pulicaria (Figs. 2.2, 2.3). D. pulicaria consists of at least three distinct clades in the Holarctic (Dufresne and Hebert 1997; Colboume et al. 1998). NDS sequence variation indicates that the Argentine populations of D. pulicaria form a group distinct from all of the known Holarctic clusters (Fig. 2.3). Its average NDS sequence divergence from all members of the three D. pulicaria clades was 2.94% Allozymic profiles were consistent with the mtDNA results. In tetrameric enzymes such as LDH, the precise relative mobilities of individual alleles are difficult to measure 5when in heterozygous condition, due to the complex banding pattern. However, the faster allele at LDH in the Argentine populations appeared to have the same mobility as the “F™ allele considered diagnostic for D. pulicaria in North America, while the slower LDH allele in Argentine populations appeared to have the same mobi as the “S” allele that is diagnostic for D. pulex (Hebert et al. 1989). Relative DNA content The mean IOD of exopodite nuclei from individuals of clone B (site ARG 11) was 163.7 (standard error + 1.4), while the lab culture of D. pulicaria showed a mean [OD of 83.8 ( 2.0). Thus, the ratio of DNA contents in exopodite nuclei from the Argentine clones and diploid D. pulicaria is approximately 1.9. Individuals from allozyme clone A (sites ARG 2, 4, and 6) had somewhat dispersed DNA in their exopodite nuclei, which complicated measurements, but they showed a similarly elevated DNA content relative to the diploid D. pulicaria (ratio ~1.7). Discussion The first case of temperate polyploid Daphnia The present study confirms that the Daphnia pulex complex is represented in South America. Moreover, all Argentine populations were asexual tetraploids, as evidenced by their allozymic phenotypes and nuclear DNA contents. Given their high heterozygosities, these clones are likely of hybrid origin. Hybrid polyploid clones of the D. pulex complex also dominate the most northerly portions of the range of this group (Weider et al. 1987; Beaton and Hebert 1988; Dufresne and Hebert 1994), suggesting that similar processes may have produced this apparent geographic symmetry. However, there 16are important distinctions between the environments occupied by polyploids in the northern and southern hemispheres. In North America and Europe, polyploids are found at a minimum of 58°N and only achieve dominance at about 70°N (Beaton and Hebert 1988; Ward et al. 1994). By contrast, Argentine polyploids were found within the much lower latitudinal range of 46- 54°S, where they experience longer and warmer growing seasons than their Holarctic counterparts. The southern polyploids occupy habitats in a region with a “temperate or cool-temperate” climate (Paruelo et al. 1998), having mean January (i.c. summer) temperatures between 9°C and 16°C (National Meteorological Service of Argentina). On the other hand, arctic polyploids only achieve dominance north of the 10°C July isotherm (Beaton and Hebert 1988; Ward et al. 1994). In fact, and despite exhaustive sampling efforts, polyploids have never been detected in any temperate region of North America or Europe (e.g. Cemy and Hebert 1993; Hebert et al. 1993; Ward et al. 1994; Hebert and Finston 2001). Thus, the detection of polyploids in relatively warm environments in Argentina challenges existing theories regarding the relationship between polyploidy and ‘environment and may provide insight into the processes responsible for the generation and maintenance of polyploids. Polyploids in the north: production and persistence Some authors have suggested that the elevated incidence of polyploidy in the far north is attributable to the indirect action of the Pleistocene glaciations in generating polyploid lineages (Stebbins 1984; Dufresne and Hebert 1997). Specifically, it is argued that during glacial advances, species’ ranges are reduced to isolated refugia where gene pool divergence occurs. Hybridization of divergent genomes during secondary contact following glacial retreat is thought to promote (allo)polyploidization. However, while this, 7glacial cycling mechanism may promote the production of polyploids via hybridization, it does not explain their persistence in arctic environments. In Daphnia in particular, it is clear that hybridization alone is not sufficient to account for the presence of polyploid lineages, as interspecific hybridization is common between many Daphnia species inhabiting temperate climates where polyploids are never found (Taylor and Hebert 1993; Colbourne and Hebert 1996; Hebert and Finston 2001). Consequently, other authors have emphasized the role of selection to explain the differential maintenance of polyploids in arctic vs. temperate environments (Beaton and Hebert 1988; Ward et al. 1994) Some authors have proposed that the success of polyploids in the north is due to the adaptive value of parthenogenesis (e.g. Bell 1982), primarily since asexuals are considered superior colonizers (Stebbins 1950) and are also expected to perform well under conditions of low biotic complexity (Glesener and Tilman 1978). However, in the case of the D. pulex complex, the fact that diploid asexual populations are found as far south as Mexico (Hebert and Finston 2001) suggests that it is polyploidy itself that is especially relevant in the arctic. The proposed adaptive significance of polyploidy in cold conditions relates primarily to the effects of increased DNA content on cell size (Gregory 2001). In plants, larger DNA contents in northern species may reflect the necessary shift in emphasis from cell division to cell enlargement as the former is impeded in the cold (Grime and Mowforth 1982; Mowforth and Grime 1989). Similarly, a comparison of laboratory-reared clones revealed that polyploid members of the D. pulex complex mature faster than their diploid counterparts under cold conditions (Dufresne and Hebert 1998). Because of their increased DNA content, polyploid Daphnia also produce larger (but fewer) eggs and offspring than diploids (Dufresne and Hebert 1998) -- a fact directly 78