REVEALING HIDDEN DIVERSITY IN A UNIQUE ISLAND ECOSYSTEM: THE EVOLUTION, ECOLOGY AND CONSERVATION OF ANCHIALINE SHRIMP IN THE PACIFIC BASIN by David Andrew Weese A dissertation submitted to the Graduate Faculty of Auburn University in partial fulfillment of the requirements for the Degree of Doctor of Philosophy Auburn, Alabama December 8, 2012 Keywords: Anchialine, Caridea, Hawaiian Island, Ryukyu Islands Copyright 2012 by David A. Weese Approved by Scott R. Santos, Chair, Associate Professor, Department of Biological Sciences Kenneth M. Halanych, Professor, Department of Biological Sciences Michael C. Wooten, Professor, Department of Biological Sciences Jason E. Bond, Professor, Department of Biological Sciences ii Abstract Due to varying anthropogenic pressures, there is an urgency to accelerate research into unique island environments and their biota before they are lost forever. For many islands, one of the most threatened ecosystems may be the anchialine ecosystem. Anchialine habitats are land-locked bodies of water which have no surface connection to the ocean yet contain salt or brackish water that fluctuate with the tides due to subterranean connections to the sea. Unfortunately, these unique habitats have received little attention while experiencing significant negative impacts from anthropogenic activities such as urbanization, groundwater extraction/contamination, and the introduction of invasive species. Given these threats, we risk losing these ecosystems without knowing the nature and extent of the biodiversity contained within them. In this regard, understanding the genetic structure and relationships of taxa endemic to these habitats may help illuminate the processes driving the evolution of anchialine organisms. Such knowledge has important implications in developing sound conservation strategies for these organisms and their imperiled ecosystem. In this context, the research presented here examined the genetic variation, population structure and phylogeographic patterns within and between five species of anchialine Carideans (Halocaridina rubra, Caridina rubella, Antecaridina lauensis, Metabetaeus minutus and Halocaridinides trigonophthalma) endemic to the rapidly vanishing anchialine ecosystems of the Pacific Basin. Overall, analysis of intraspecific mitochondrial sequence data (i.e., cytochrome oxidase subunit I [COI] and large subunit ribosomal [16S-rDNA] genes) revealed iii two main patterns. First, the contrasting patterns of population structure and connectivity exhibited by each ?species? appear to result from complex interactions between intrinsic (i.e., life history traits) and extrinsic (i.e., historical geologic and oceanographic) processes. Second, the diversity of taxa inhabiting the anchialine habitats of the Pacific may be vastly underestimated with many of the species previously described representing cryptic species complexes. This research not only provides additional understanding into the ecology and evolution of anchialine organisms in general, but has application in areas of conservation management; such as identifying source populations of invertebrate species in the wildlife trade as well as populations/habitats warranting conservation management and demonstrating that levels of diversity in subterranean crustaceans are in many cases vastly underestimated. iv Acknowledgements First, I would like to thank my graduate adviser, Dr. Scott Santos, for his continued guidance and support throughout the course of my graduate career. I am also grateful to Drs. Ken Halanych, Michael Wooten, and Jason Bond who have all been especially helpful and patient members of my graduate committee providing invaluable feedback and suggestions during the course of this work, as did my outside reader, Dr. Eric Peatman. I would also like to thank Drs. M. Hidaka and Y. Fujita for their hospitality during my time in Okinawa, Japan and for introducing me to Japanese culture and society. My time in the Okinawa was personally and professionally inspiring, and it will be a trip I will never forget. I owe a special thanks to past and present members of the Molette Biology Laboratory for Environmental and Climate Change Studies (Nathan Kirk, Justin Havird, Damien Waits, Kevin Kocot, Alexis Janosik and Johanna Cannon) for your knowledge and expertise in the lab. This work has been funded in large part by an East Asia and Pacific Summer Institutes Fellowship (OISE-0913667) awarded to me through the National Science Foundation (NSF) as well as an NSF grant (DEB-0949855) awarded to Dr. Scott Santos. Finally, I would like to thank my family for their continued unwavering support and guidance. I would like to dedicate this dissertation to my grandfather, Dr. Thomas Sisterhen who, even though he may not know it, has been a source of motivation and inspiration. v Table of Contents Abstract ...................................................................................................................................... ii Acknowledgments ..................................................................................................................... iii List of Tables ............................................................................................................................ ix List of Figures ........................................................................................................................... xi Chapter 1: Introduction to Dissertation .......................................................................................1 1.1 General introduction and background on anchialine habitats and their biota ...........1 1.2 References ...............................................................................................................11 Chapter 2: Rocks and Clocks: Linking geologic history and rates of genetic differentiation in anchialine organisms .............................................................................................................16 2.1 Abstract ...................................................................................................................16 2.2 Introduction .............................................................................................................17 2.3 Materials and Methods ............................................................................................19 2.3.1 Biological Materials .......................................................................................19 2.3.2 DNA extraction, polymerase chain reaction (PCR) and sequencing .............20 2.3.3 Demographic and genetic structure analyses .................................................20 2.3.4 Divergence time estimation and geologic age ...............................................21 2.4 Results .....................................................................................................................22 2.4.1 Halocaridina from the northeastern coast of Maui ........................................22 2.4.2 Halocaridina from the southeastern coast of Hawai?i ...................................24 2.5 Discussion ...............................................................................................................25 vi 2.5.1 Validation of the Halocaridina molecular clock ...........................................25 2.5.2 Maui and Hawai?i: geologic history and Halocaridina .................................27 2.5.3 Reevaluating Halocaridina evolution in the context of island geology ........29 2.6 Acknowledgements .................................................................................................33 2.7 References ...............................................................................................................34 Chapter 3: Genetic identification of source populations for an aquarium traded invertebrate .46 3.1 Abstract ...................................................................................................................46 3.2 Introduction .............................................................................................................47 3.3 Materials and Methods ............................................................................................50 3.3.1 Biological materials .......................................................................................50 3.3.2 DNA Extraction, polymerase chain reaction (PCR) and sequencing .............50 3.3.3 Genetic analyses .............................................................................................51 3.4 Results .....................................................................................................................52 3.5 Discussion ...............................................................................................................53 3.6 Acknowledgments.....................................................................................................57 3.7 References ...............................................................................................................58 Chapter 4: The long and short of it: Genetic variation and population structure of the anchialine atyid shrimp Caridina rubella on Miyako-jima, Japan ............................................71 4.1 Abstract ...................................................................................................................71 4.2 Introduction .............................................................................................................72 4.3 Materials and Methods ............................................................................................74 4.3.1 Biological Materials ........................................................................................74 4.3.2 DNA extraction, polymerase chain reaction (PCR) and sequencing ...............75 4.3.3 Genetic diversity and parsimony network analyses .......................................76 vii 4.3.4 Population structure, migration, and demographic analysis ..........................76 4.4 Results .....................................................................................................................78 4.4.1 Genetic diversity and parsimony network analyses of Caridina rubella on Miyako-jima ............................................................................................................78 4.4.2 Population structure, migration, and demographic analyses of C. rubella on Miyako-jima ............................................................................................................79 4.5 Discussion ...............................................................................................................81 4.5.1 Morphology and genetic differentiation in C. rubella of Miyako-jima ...........81 4.5.2 Population structure among ?short rostrum? C. rubella of Miyako-jima .........82 4.5.3 Underground dams and conservation aspects ..................................................84 4.6 Acknowledgments.....................................................................................................86 4.7 References ...............................................................................................................87 Chapter 5: Multiple colonizations lead to hidden diversity in a unique island ecosystem: Comparative phylogeography of anchialine shrimp in the Ryukyu Archipelago, Japan .........102 5.1 Abstract .................................................................................................................102 5.2 Introduction ...........................................................................................................103 5.3 Materials and Methods ..........................................................................................106 5.3.1 Taxon sampling, DNA extraction and sequencing ........................................106 5.3.2 Genetic diversity and population genetic analyses ........................................108 5.3.3 Phylogenetic analyses ..................................................................................109 5.4 Results ...................................................................................................................110 5.4.1 Genetic diversity ............................................................................................110 5.4.2 Phylogenetic analyses ....................................................................................111 5.4.3 Population genetic analyses ...........................................................................113 5.5 Discussion .............................................................................................................115 viii 5.5.1 Evidence for multiple colonization of the Ryukyus by anchialine Carideans........................................................................................................116 5.5.2 Population Structure of anchialine Carideans in the Ryukyus Archipelago ..119 5.5.3 Distribution and hidden diversity of anchialine shrimp of the Indo-West Pacific ............................................................................................................123 5.6 References ...............................................................................................................126 Chapter 6: Conclusions .............................................................................................................147 6.1 General Conclusions ...............................................................................................147 6.2 Future Prespcetives ................................................................................................151 6.3 References ...............................................................................................................154 ix List of Tables CHAPTER 2 Table 1: Genetic diversity measures and results of neutrality tests for Halocaridina populations from the northeastern and southeastern coasts of Maui and Hawai?i, respectively, examined in this study ............................................................................40 Table 2: Summary of molecular clock comparisons for Halocaridina .....................................41 CHAPTER 3 Table 1: Numbers of individuals and unique mitochondrial cytochrome c oxidase subunit I (COI) haplotypes recovered for the commercially acquired Halocaridina specimens examined ......................................................................................................................63 Table 2: Numbers of individuals and unique mitochondrial cytochrome c oxidase subunit I (COI) haplotypes for each of the known genetic groups of Halocaridina constituting the ?reference? database utilized .................................................................................64 Table 3: Genetic group identification of commercially acquired Halocaridina specimens examined in this study based on haplotype network analyses with TCS as well as the program Segregating Sites Assigner ...........................................................................65 Supplementary Table 1: Posterior probability (pp) and risk values for the assignment of each commercially acquired Halocaridina specimen to the Kona, Ka?? (islands of Hawai?i) or Kina?u (islands of Maui) genetic groups of these shrimp ........................68 CHAPTER 4 Table 1: Indices of genetic diversity and tests of neutrality for each population of Caridina rubella on Miyako-jima, Ryukyus, Japan. .................................................................96 Table 2: Pairwise ?ST and Snn estimates as measures of genetic differentiation between populations of Caridina rubella on Miyako-jima, Ryukyus, Japan. ..........................97 Table 3: Hierarchical analysis of molecular variance (AMOVA) for northern and southern populations of Caridina rubella on Miyako-jima, Ryukyus, Japan. ...........................98 CHAPTER 5 x Table 1: Measures of genetic diversity and tests of neutrality for populations of Metabetaeus minutus, Antecaridina lauensis and Halocaridinides trigonophthalma from the Ryukyu Archipelago, Japan .......................................................................................136 Table 2: Additional specimens and sequences included in 16S-rDNA analyses.....................138 Table 3: Indicators of demographic expansion events in lineages of of Metabetaeus minutus, Antecaridina lauensis and Halocaridinides trigonophthalma of The Ryukyu Archipelago, Japan.....................................................................................................140 Table 4: Pairwise ?ST and Snn estimates as measures of genetic differentiation between populations of Metabetaeus minutus, Antecaridina lauensis and Halocaridinides trigonophthalma of The Ryukyu Archipelago, Japan ...............................................141 xi List of Figures CHAPTER 1 Figure 1: Global distribution of anchialine habitats. .................................................................15 CHAPTER 2 Figure 1: Map of the high Hawaiian Islands depicting anchialine pools where Halocaridina were sampled for this study .........................................................................................42 Figure 2: Network analyses for the East Maui (A) and Ka?? (B) lineages of Halocaridina from northeastern Maui and southeastern Hawai?i...............................................................43 Figure 3: Posterior distributions of theta (?; A) and divergence time (T; B) of Halocaridina population pairs from northeastern Maui and southeastern Hawai?i ...........................44 Figure 4: Genetic distances and corresponding estimated times since divergence for seven comparisons between populations, genetic groups and/or lineages of Halocaridina across the islands of Hawai?i, Maui, and O?ahu ..........................................................45 CHAPTER 3 Figure 1: Localities of anchialine habitats on the Hawaiian islands of O?ahu, Maui and Hawai?i and the range distributions of the 13 genetic groups of Halocaridina ..........66 Figure 2: Haplotype network depicting the relationships between mitochondrial cytochrome oxidase subunit I haplotypes previously recovered from the (a) Kona, Ka??, and (b) Kina?u genetic groups of Halocaridina on Hawai?i and Maui and haplotypes recovered from commercially acquired Halocaridina specimens examined in this study .............................................................................................................................67 CHAPTER 4 Figure 1: Map of the Southern Ryukyus Archipelago, Japan, indicating the location of Miyako-jima (A) and enlarged map of Miyako-jima depicting anchialine habitats from which Caridina rubella were sampled for this study (B) ...................................99 Figure 2: Networks depicting relationships among cytochrome c oxidase subunit I (COI) haplotypes recovered from the two Caridina rubella lineages on Miyako-jima, Ryukyus, Japan ..........................................................................................................100 xii Figure 3: Posterior probability distributions of migration rates (M) between populations of Caridina rubella on Miyako-jima, Ryukyus, Japan ..................................................101 CHAPTER 5 Figure 1: Map of the Southern Ryukyus Archipelago, Japan, indicating the location of the Ryukyu Archipelago (A) and enlarged maps of the Yaeyama, Miyako and Daito island groups from which individuals of the genus Metabetaeus, Antecaridina, and Halocaridinides were sampled for this study (B) ......................................................143 Figure 2: Phylogenetic relationships inferred for selected species of endemic Caridean shrimp of the Pacific Basin. (A) Map of sampling localities from the Pacific Basin and the Southern Ryukyus. Maximum likelihood (ML) trees for the Alpheid dataset and the Atyid data set. ............................................................................................................144 Figure 3: Networks depicting relationships among cytochrome c oxidase subunit I (COI) haplotypes recovered for three genera of endemic anchialine shrimp sampled from the Ryukyus Archipelago, Japan .....................................................................................145 1 CHAPTER 1. Introduction to Dissertation 1.1 GENERAL INTRODUCTION Beginning over a century ago with Alfred Wallace and Charles Darwin, the study of island biogeography has contributed greatly to our understanding of fundamental evolutionary processes. It was during their travels among the world?s islands that these young naturalists began developing their theories of evolution by natural selection through their observation of how closely related populations on nearby island varied in appearance. Since these times islands have long provided model systems in which theories of evolution have been developed, tested and redefined. In this context, island systems have been considered excellent ?natural laboratories? for studying the fundamental processes of evolution for a variety of reasons (reviewed in Emerson, 2002): (i) they are typically of small size, (ii) are of discrete geographical nature (i.e., isolated), (iii) contain a simplified biota, (iv) despite being geographically small, contain a large diversity of habitats, (v) often have dynamic geologic histories, and (vi) often occur in island groups which can function as experimental replicates. Over the past 150 years these systems have served as templates for studying biogeographic phenomena and it has become obvious that islands harbor a great diversity of organisms. However, understanding the origins of this diversity has not always been so clear (Emerson, 2002). In this context, the examination of phylogenetic relationships and genetic structure between and within island species provides a framework in which to examine the origin of island taxa (Avise, 2004). The ability to easily acquire DNA sequence data and reconstruct the phylogenetic and demographic histories of almost any taxa has revolutionized the field of biogeography and it has now become commonplace to use phylogenetic reconstructions of lineages to interpret the evolutionary histories of islands and their biotas (Brown & Lomolino, 2 2000; Emerson, 2002). In this sense, the examination of phylogenetic relationships and genetic structure between and within island taxa provides a framework to examine the origin of island taxa, determine the processes that have led to species richness within island ecosystems and enable testing of general biogeographic and ecological hypotheses (Avise, 2004). Much like isolated oceanic islands, subterranean ecosystems have been considered natural ecological and evolutionary laboratories (Poulson and White, 1969; Juan et al., 2010). Similar to island systems, subterranean habitats are characterized as having discrete geographical boundaries, dynamic geological histories, simplified communities and temporally and spatially isolated biota. Given these similarities, the study of subterranean organisms can help elucidate the evolutionary processes and historical factors related to biogeography and speciation similar to what has been demonstrated for isolated oceanic islands (Cooper et al., 2007; Page et al., 2008; Juan 2010). However, diversification and speciation processes in these underground worlds are largely influenced by habitat fragmentation and are often times masked by morphological convergence, confounding interpretations leaving many aspects regarding the evolution of subterranean taxa under considerable debate (Lefebure et al., 2006; Cooper et al., 2007). To gain a better understanding of the evolution of subterranean organisms, additional systems with novel attributes are needed (Cooper et al., 2007). Anchialine environments act as an interface between the subterranean freshwater aquifer and the open ocean and are unique from ecological and evolutionary perspectives due to the fact that they represent a habitat where salinity, water levels, and temperature fluctuate daily (Craft et al., 2008). These habitats are land-locked bodies of mixohaline water occupying basins of varying geologic origins (i.e., karst caves, cenotes, fossilized coral reefs or coastal lava fields) with no surface connection to the open ocean yet contain salt or brackish water that fluctuate 3 with the tides due to simultaneous subterranean connections with the ocean and freshwater aquifer (Holthuis, 1973; Maciolek, 1983; Sket, 1996). In general, anchialine habitats can be classified as being either karst or pseudokarst (Mylroie & Mylroie, 2011). The latter can occur in almost any rock material by a variety of mechanisms, with the most significant being those formed by volcanic activities (i.e., lava tubes). Karst habitats, on the other hand, are formed by dissolutional processes and are classified as either stream caves formed by traditional karst processes that are now interacting with the coastal environment (Mylroie & Mylroie, 2011) or coastal caves and shallow basins linked to the marine environment. While, under the right conditions, dissolution can create karst forms in a variety of rock types, anchialine habitats typically form in carbonate rocks (i.e., limestone and dolomite). Given this, it is not surprising that despite being found along coastlines throughout the world from high to low latitudes, anchialine habitats dominate coastlines of tropical and subtropical islands due to historically warm seas and shallow conditions favoring coral reef development. Such habitats (karst or pseudokarst) have been reported from around the world, including the Philippines, Cuba, the South Pacific, the Sinai Peninsula, the Bahamas, Bermuda, Australia, the Yucatan Peninsula, the Canary Islands, the Caribbean, the Ryukyus and Hawai?i (Fig. 1, Holthuis 1973; Maciolek 1983; Maciolek 1986; Iliffe 2000; Becking et al., 2011). Unfortunately, one of the most threatened habitats on many tropical islands may be these anchialine habitats (Sket, 1996; Iliffe, 2002; Santos, 2006). Despite a worldwide distribution (Fig. 1), only ~1,000 habitats fitting this ecosystem definition have been reported globally (Maciolek, 1986). Historically, these extraordinary and vulnerable environments have experienced significant negative impacts from anthropogenic activities such as coastal development, tourism, vandalism, groundwater extraction and contamination, and the 4 introduction of invasive species (Brock et al., 1987; Iliffe, 2002, 2003; Santos, 2006). These problems are exemplified in Hawaiian Islands, which have the single largest concentration (~600; Brock, 1985) of them in the world and is home to the only natural representatives in the United States, where >90% of Hawai?i?s anchialine ecosystems have been degraded or lost by anthropogenic activities in the last 50 years and it is predicted that they will likely disappear altogether in the near future (Maciolek and Brock 1974; Brock & Bailey-Brock 1998; Santos 2006). The destruction of habitats by limestone quarries or construction activities represents another serious environmental concern for anchialine habitats as a number of habitats have been destroyed in Bermuda and the Yucatan Peninsula by quarries producing crushed aggregate for construction purposes (Iliffe & Kornicker, 2009). Given these concerns, we risk losing these unique ecosystems without knowing the nature and extent of the biodiversity contained within them. Thus, there is currently an urgency to accelerate research into these unique island environments and their biota before they are lost forever. In order to manage anchialine habitats and their fauna, it is important to develop an understanding of the biodiversity, ecology, and evolution of organisms from this ecosystem since such knowledge will play a critical role in establishing conservation strategies around the world. To date, a large number of fascinating and unique organisms, including members of the Porifera, Cnidaria, Annelida, Mollusca, Arthropoda, Echinodermata, Pices and Chordata, are known to inhabit anchialine environments (Holthuis 1963; Maciolek & Brock 1974; Iliffe et al., 1984; Iliffe 2002; Becking et al., 2011). However, it appears that the true diversity of these habitats has only recently begun to be recognized as more than 450 new species of anchialine organisms have been discovered and described over the past 25 years (Iliffe & Kornicker, 2009; ISI Web of Science search, June 2012). Given the many threats that anchialine organisms and their habitats 5 face today, coupled with the fact we are only starting to scratch the surface regarding the level of diversity contained in these rare habitats, there is a timely need to determine the spatial scale at which anchialine organisms are potentially isolated since such information is critical in establishing plans to preserve these unique environments. Interestingly, despite the geographic isolation of these habitats, many anchialine taxa show geographically widespread and disjunct distributions (Smith and Williams 1981; Maciolek, 1983; Iliffe, 2000). Given these unexpected distributions, it has been hypothesized that as a result of subterranean connections and marine dispersal, populations of anchialine organisms may experience high levels of gene flow, but low levels of genetic differentiation (Cabezas et al., 2012). However, so far, the few studies exploring the genetic connectivity of anchialine species have yielded sharply different patterns of genetic structure on the basis of mitochondrial DNA variation, from panmixia among populations hundreds of kilometers apart, suggesting frequent dispersal over long distances (Kano & Kase, 2004; Hunter et al., 2008; Page et al., 2008; Russ et al., 2010), to exceptional levels of endemism on the scale of a few kilometers due to low dispersal abilities and/or significant barriers to dispersal (Santos, 2006; Hunter et al., 2008; Craft et al., 2008). Much of the work presented below builds on the previous work of Santos and colleagues regarding the genetic connectivity, phylogeography, and evolution of anchialine organisms in the Hawaiian Archipelago (Santos, 2006; Ivey & Santos, 2007; Craft et al., 2008; Russ et al., 2012). The anchialine biota of these islands are dominated by two species of Caridean shrimp, Halocaridina rubra Holthuis, 1963 (Crustacea: Decapoda: Atyidae) and Metabetaeus lohena Banner and Banner, 1960 (Crustacea: Decapoda: Alpheidea). Despite possessing life history traits thought to be conducive to high colonization and dispersal ability among anchialine 6 habitats (i.e., lecithotrophic larvae; reviewed in Santos, 2006), genetic analysis of mitochondrial (mtDNA) cytochrome oxidase I (COI) gene sequences suggested that H. rubra actually represents a ?cryptic species complex? composed of at least 13 distinct genetic groups belonging to eight discrete lineages, all exhibiting significant levels of genetic structure and restricted levels of gene flow over limited geographic scales across the islands (Santos, 2006; Craft et al., 2008). These results suggested that population differentiation and evolutionary diversification in the anchialine environments of the Hawaiian archipelago may result from extrinsic obstacles to gene flow in the form hydrological barriers imposed by the subterranean aquifers and the open ocean. However, a later study of M. lohena, which produces planktotrophic larvae, revealed little to no population structure among populations sampled from the same anchialine habitats as Halocaridina (Russ et al., 2011). This total lack of population structure at this scale is similar to what has been reported for other anchialine organisms possessing planktotrophic development, such as the neritiliid snail Neritilia, which covers a similar geographic range across the Philippines (Kano & Kase, 2004). From the above-mentioned studies, several hypotheses regarding the evolutionary processes operating on endemic anchialine organisms have been developed. These include: (1) the open ocean can represent a significant isolating barrier, limiting dispersal on and between islands; (2) the compartmentalization of the hypogeal water system of an island acts as barriers to gene flow and restricts dispersal across an island; (3) life history characteristics (i.e., egg size, larval stages, and larval habitat) strongly influence population structure; and (4) larval feeding mode (i.e., lecithotrophy versus planktotrophy) may influence dispersal potential and population structure. The general aim of this thesis is to further test these hypotheses using analyses of mitochondrial DNA while investigating the biogeography and evolutionary relationships within 7 and between anchialine Carideans of the Pacific Basin. Specifically, I am interested in determining how intrinsic (i.e., life history traits) and extrinsic (i.e., geology) factors influence the evolution and ecology of anchialine organisms. Hopefully, this research will not only provide additional understanding into the ecology and evolution of anchialine organisms in general, but will have applied utility in developing management strategies for these organisms and their unique environments for the future. Below, I provide a brief description of each chapter and discuss their insights into the evolution, ecology and conservation of anchialine organisms in the Pacific Basin. The focus of my dissertation research begins with the atyid Halocaridina rubra, which as mentioned above, dominates the anchialine habitats of the Hawaiian Archipelago (Maciolek, 1983). Historically, H. rubra has been considered a single species, but recent studies by Santos (2006) and Craft et al., (2008) suggest that Halocaridina actually represents a ?cryptic species complex? comprise of at least eight potential species. Furthermore, molecular clock estimates calibrated to the geologic age of Kilauea volcano on the island of Hawai?i have suggested that the COI gene of the Halocaridina complex is apparently evolving at an exceptional rate of 20% per million years (Craft et al., 2008). In Chapter 2, the validity of this rapid evolutionary rate is tested by applying it to levels of divergence measured from geographically close, yet genetically distinct, Halocaridina populations on the islands of Hawai?i, Maui and O?ahu. Application of the Halocaridina molecular clock identified a strong correlation between levels of genetic divergence and the geologic age of the region inhabited by those populations on the younger islands of Hawaii and Maui. However, this relationship weakened when applied to Halocaridina populations on the older island of Oahu. Given this, it appears that island age is an important factor that should be taken into consideration when conducting molecular clock analyses on 8 island populations. This chapter has been published as an invited contribution as: Santos, S.R., D.A. Weese (2011) Rocks and Clocks: Linking geologic history and rates of genetic differentiation in anchialine organisms. Hydrobiologia. 677(1):53-64. The strong population structure observed for Halocaridina is utilized in Chapter 3 to examine the utility of molecular techniques for identifying source populations of ornamental invertebrates in the aquarium trade. Halocaridina are popular in the aquarium trade due to their ability to survive in hermetically sealed containers for extended time periods. However, commercial harvesting, coupled with habitat destruction and the strong regional endemism exhibited by these shrimp (Craft et al., 2008), may lead to the depletion/extinction of unique populations. To identify the source populations of Halocaridina being sold in the aquarium trade, mitochondrial gene sequences from commercially purchased individuals were compared to a large database of homologous sequences from previous studies (Santos, 2006; Craft et al., 2008). Commercial specimens were identified as originating from the Kona, Ka?? (west and south coasts, respectively, Hawai?i), or Kina?u (south coast, Maui) genetic groups of these shrimp. Although the majority of individuals originated from the Kona land district as hypothesized, the finding that commercially available Halocaridina originate from three genetic groups spanning two islands suggest that other population also warrant management consideration. This chapter has been published as: Weese, D.A., S.R. Santos (2009) Genetic identification of source populations for an aquarium-traded invertebrate. Animal Conservation. 12:13-19. In the later half of this dissertation, my research focus shifts west to the anchialine organisms of the Ryukyus Archipelago, a continuous chain of islands situated between Japan?s Kyushu Island and Taiwan. The Southern Ryukyus consists of five main islands, Okinawa, 9 Tarama, Miyako, Ishigaki, and Iriomote. These islands, spanning a geographic distance similar to the high islands of the Hawaiian Archipelago (~200 km), offer an excellent opportunity to further test and develop hypotheses regarding the evolutionary diversification and phylogeography of anchialine taxa. This begins with an investigation of the genetic variation and population structure of the anchialine atyid, Caridina rubella, restricted to the island of Miyako-jima in Chapter 4. Having a planktotrophic larval stage and a potentially amphidromous life cycle, populations of C. rubella on the island of Miyako-jima were hypothesized to exhibit little to no structure across the island. Surprisingly, significant genetic structure was exhibited across distances ranging from < 20 m to > 10 km. Additionally, deep genetic divergence correlating with distinct morphological variation was found between closely situated, but apparently isolated, populations, suggesting that ?C. rubella? may actually represent two distinct species on Miyako-jima. Given that this atyid is already listed as a threatened species by the Japanese government, the results presented in this chapter may have significant implications for the formulation and implementation of future conservation plans for C. rubella populations on Miyako-jima. This chapter has been published as Weese, D.A., Y. Fujita, M. Hidaka, S.R. Santos. (2012) The long and short of it: Genetic variation and population structure of the anchialine atyid shrimp Caridina rubella on Miyako-jima, Japan. Journal of Crustacean Biology. 32(1):109-117. Chapter 5 takes a comparative phylogeographic approach to determine the origins of anchialine Caridean diversity in the Ryukyu Islands by investigating the population structure and evolutionary history of three Caridean shrimp from the anchialine niche. Given the geologic history of the islands, it was hypothesized that diversity across the Ryukyus originated within the archipelago and that the phylogenetic history of the three species examined (Antecaridina 10 lauensis, Metabetaeus minutus and Halocaridinides trigonophthalma) would be correlated with the sequential separation of the Ryukyu island groups as sea-levels fluctuated during the Pleistocene. However, given the phylogenetic relationships and geographic distributions of the multiple lineages found for each species coupled with the strong regional oceanographic currents of the Indo-West Pacific, the diversity of anchialine shrimp found in the Ryukyus is likely the result of an accumulation of species dispersing into the Ryukyus from independent source populations. Furthermore, the contrasting patterns of population structure and connectivity exhibited by each ?species? within the Ryukyus appear to result from complex interactions between intrinsic (i.e., life history traits) and extrinsic (i.e., historical geology and oceanography) processes. This chapter will be submitted to The Biological Bulletin as Weese, D.A.,Y. Fujita, S.R. Santos. ?Multiple colonizations lead to hidden diversity in a unique island ecosystem: Comparative phylogeography and conservation of anchialine shrimp in the Ryukyu Archipelago, Japan?. Drawing from the previous four chapters, Chapter 6 offers my concluding thoughts regarding the ecology, evolution and conservation of anchialine Carideans in the Pacific Basin. I present a brief summary of the preceeding chapters and how these studies can bring novel insight into the natural history of anchialine species and how these findings can have important implications for the future management of these species and their habitats. 11 1.2 REFERENCES Avise, J.C. 2004. Molecular markers, natural history, and evolution. Chapman & Hall, New York. Becking, L.E., W. Renema, N.K. Santodomingo, B.W. Hoeksema, Y. Tuti & J. de Voogd, 2011. Recently discovered landlocked basins in Indonesia reveal high habitat diversity in anchialine system. Hydrobiologia 677: 89-105. Brock, R.E., 1985. An assessment of the conditions and future of the anchialine pond resources of the Hawaiian Islands. Prepared for Transcontinental Development Co., Honolulu. Brock, R.E., J.E. Norris, D.A. Ziemann and M.T. Lee. 1987. Characteristics of water quality in anchialine ponds of the Kona, Hawai?i, coast. Pacific Science 41: 200?208. Brock R.E. and J.H. Bailey-Brock, 1998. A unique anchialine pool in the Hawaiian Islands. International Review of Hydrobiology 83: 65-75. Brown, J.H. and M.V. Lomolino. 2000. Concluding remarks: historical perspective and the future of island biogeography theory. Global Ecology and Biogeography 9: 87-92. Cabezas, P., F. Alda, E. Macpherson and A. Machodom. 2012. Genetic characterization of the endangered and endemic anchialine squat lobster Munidopsis polymorpha from Lanzarote (Canary Islands): management implications. ICES Journal of Marine Science. doi:10.1093/icesjms/fss062 Cooper, S.J.B., J.H. Bradbury, K.M. Saint, R. Leys, A.D. Austin and W.F. Humphreys. 2007. Subterranean archipelago in the Australian arid zone: mitochondrial DNA phylogeography of amphipods from central Western Australia. Molecular Ecology 16: 1533-1544. 12 Craft, J. D., A. D. Russ, M. N. Yamamoto, T. Y. Iwai, S. Hau, J. Kahiapo, C. T. Chong, S. Ziegler-Chong, C. Muir, Y. Fujita, D. A. Polhemus, R. A. Kinzie, and S. R. Santos. 2008. Islands under islands: The phylogeography and evolution of Halocaridina rubra Holthuis, 1963 (Crustacean: Decapoda: Atyidae) in the Hawaiian archipelago. Limnology and Oceanography 53: 675-689. Emerson, B.C. 2002. Evolution on oceanic islands: molecular phylogenetic approaches to understanding pattern and process. Molecular Ecology 11: 951-966. Holthuis, L.B. 1963. On red coloured shrimps (Decapoda, Caridea) from tropical land-locked saltwater pools. Zoologische Mededelingen (Leiden) 16: 261?279. Holthuis, L.B. 1973. Caridean shrimps found in land-locked saltwater pools at four Indo-West Pacific localities (Sinia Peninsula, Funafuti Atoll, Maui and Hawai?i islands), with the description of one new genus and four new species. Zoologische Verhandelingen (Leiden) 128: 1?48. Hunter, R.L., M.S. Webb, T.M. Iliffe, and J.R. A. Bremer. 2008. Phylogeny and historical biogeography of the cave-adapted shrimp genus Typhlatya (Atyidae) in the Caribbean Sea and western Atlantic. Journal of Biogeography 35: 65-75. Iliffe, T., M.H.H Wilkens, J. Parzefall and D. Williams. 1984. Marine lava cave fauna: Composition, biogeography, and origins. Science 225: 309?311. Iliffe, T.M. 2000. Anchialine cave ecology, p. 59?76. In H. Wilkens, D. Culver and W. F. Humphreys, [eds.], Subterranean Ecosystems. Elsevier. Iliffe, T.M., 2002. Conservation of Anchialine Cave Biodiversity, in Hydrogeology and Biology of Post-Paleozoic Carbonate Aquifers, J.B. Martin, C.M. Wicks, and I.D. Sasowsky, Editors., Karst Waters Institute: Charles Town, WV. p. 99-102 13 Iliffe, T.M. and L.S. Kornicker. 2009. Woldwide diving discoveries of living fossil animals from the depths of anchialine and marine caves. Smithsonian Contributions to Marine Science 38: 269?280. Ivey, J.L., and S.R. Santos. 2007. The complete mitochondrial genome of the Hawaiian anchialine shrimp Halocaridina rubra Holthuis, 1963 (Crustacea: Decapoda: Atyidae). Gene 394: 35?44. Lefebure, T., C.J. Douady, M. Gouy, P. Trontelj, . J. Briolays and J. Gibert. 2006. Phylogeography of a subterranean amphipod reveals cryptic diversity and dynamic evolution in extreme environments. Molecular Ecology 15: 1797-1806. Juan C., M.T. Guzik, D. Jaume and S.J.B. Coopers. 2010. Evolution in caves: Darwin?s ?wrecks of ancient life? in the molecular era. Molecular Ecology 19: 3865-3880. Kano, Y., and T. Kase. 2004. Genetic exchange between anchialine cave populations by means of larval dispersal: the case of a new gastropod species Neritilia cavernicola. Zoologica Scripta 33: 423?437. Maciolek, J. A. 1983. Distribution and biology of Indo-Pacific insular hypogeal shrimp. Bulletin of Marine Science 33: 606-618. Maciolek, J.A. and R.E. Brock, 1974. Aquatic survey of the Kona coast ponds, Hawaii Island. Sea Grant Advisory Report, UNIHI-SEAGRANT-AR-74-04. Maciolek, J. A. 1986. Environmental features and biota of anchialine pools on Cape Kinau, Maui, Hawai'i. Stygologia 2: 119-129. Mylroie, J. E., and J.R. Mylroie. 2011. Void development on carbonate coasts: creation of anchialine habitats. Hydrobiologia 677: 15-32. 14 Page, T.J., W.F. Humphreys and J.M. Hughes. 2008. Shrimps down under: Evolutionary relationships of subterranean crustaceans from Western Australia (Decapoda: Atyidae: Stygiocaris). PLoS One 3: e1618. Poulson, T.L. and W.B. White. 1969. The cave environment. Science 165: 971?981. Russ, A.R., S.R. Santos, and C. Muir. 2010. Genetic population structure of an anchialine shrimp, Metabetaeus lohena (Crustacea: Alpheidae), in the Hawaiian Islands. Revista de Biologia Tropical 58: 159-170. Santos, S. R. 2006. Patterns of genetic connectivity among anchialine habitats: a case study of the endemic Hawaiian shrimp Halocaridina rubra on the Island of Hawaii. Molecular Ecology 15: 2699-2718. Sket, B. 1996. The ecology of anchialine caves. Trends in Ecology & Evolution 11 221?225. Smith, M.J. and W.D. Williams. 1981. The occurrence of Antecaridina lauensis (Edmondson) (Crustacea, Decapoda, Atyidae) in the Solomon Islands. Hydrobiologia. 85: 49-58. 15 Figure 1. Global distribution of anchialine habitats. 16 CHAPTER 2. Rocks and Clocks: Linking Geologic History and Rates of Genetic Differentiation in Anchialine Organisms 2.1 ABSTRACT The geologic history of a region can significantly impact the development of its flora and fauna, with past events shaping community patterns and evolutionary trajectories of species. In this context, islands are excellent ?natural laboratories? for studying the fundamental processes of evolution due to their discrete geographical nature and dynamic geologic histories. An island system meeting these criteria is the Hawaiian Archipelago, which is ideal for testing how island geologic history influences the processes leading to population genetic variation and differentiation. One Hawaiian endemic whose evolutionary history is closely tied to the geology of the islands is the anchialine atyid shrimp Halocaridina, whose mitochondrial cytochrome oxidase I (COI) gene is hypothesized to be evolving at the rate of 20% per million years. To validate this rapid evolutionary rate, time since divergence estimates between geographically close, yet genetically distinct, populations were calculated for Halocaridina from anchialine habitats on the islands of Hawai?i, Maui, and O?ahu. On the younger (i.e. <1.5 million years) islands of Hawai?i and Maui, where all anchialine habitats occur in basalt, application of the Halocaridina molecular clock identified a strong correlation between levels of genetic divergence and the geologic age of the region inhabited by those populations. In contrast, this relationship weakened when similar analyses were conducted for Halocaridina from limestone anchialine habitats on the older (i.e. >2.75 million years) island of O?ahu. These results suggest geologic age, basin origin and/or composition are important factors that should be taken into consideration when conducting molecular clock analyses on anchialine flora and fauna as well as island populations in general. 17 2.2 INTRODUCTION Islands serve as excellent ?natural laboratories? for understanding the fundamental processes of evolution (reviewed in Emerson, 2002). For example, they have historically been templates for studying biogeographic phenomena, such as the classic hypotheses of colonization and biotic exchange proposed by MacArthur & Wilson (1967). Furthermore, the discrete and active geologic nature of many islands can regionally and significantly impact their flora and fauna, consequently altering the evolutionary trajectory of populations as well as overall community patterns. In this context, examining the genetic structure and phylogenetic relationships of island species, with reference to their specific biology and physical habitat, can provide a framework towards elucidating the forces driving the origin of novel lineages and taxa (i.e., speciation) along with the rate of their genesis (Avise, 2004). One island system with a dynamic geology is the Hawaiian Archipelago. Situated in the central Pacific Ocean, it is an ideal location for testing how the geologic history of islands can influence the demographics and genetic structure of resident populations (e.g., Vandergast et al., 2004). Additionally, the Hawaiian Islands are home to the only natural anchialine habitats in the US as well as having the single largest concentration of them (~600; Brock, 1985) in the world. Globally, such habitats are physically defined as land-locked bodies of mixohaline water with simultaneous subterranean connections to the ocean and groundwater aquifer and occupying basins of varying geologic origins (i.e., karst caves, cenotes, fossilized coral reefs or coastal lava fields) (Holthuis, 1973; Maciolek, 1983; Sket, 1996). Unfortunately, it is estimated that >90% of Hawaii?s anchialine habitats have been historically and contemporarily lost or degraded by anthropogenic activities like coastal development and the spread of exotic species (Maciolek & Brock, 1974; Bailey-Brock & Brock, 1993; Brock & Bailey-Brock, 1998). For those that still 18 remain, it is predicted that they will also succumb to these pressures in the near future if appropriate conservation plans and management actions are not taken (Brock & Kam, 1997). Thus, anchialine habitats potentially represent one of the most threatened ecosystems in the Hawaiian Archipelago. The Hawaiian anchialine ecosystem is home to a limited assemblage of macrofauna dominated by endemic gastropod mollusks and crustaceans (Maciolek & Brock, 1974; Maciolek, 1983). Of these, Halocaridina rubra Holthius, 1963 (Decapoda, Atyidae) is a small (i.e., ~12 mm) microphagous shrimp particularly characteristic of this assembly due to its near ubiquitous distribution in anchialine habitats of the islands (Maciolek, 1983). Based on this, Santos (2006) tested the hypothesis that H. rubra exhibits high levels of gene flow by first sampling populations from the island of Hawai?i. Analyses of mitochondrial (mtDNA) cytochrome oxidase I (COI) gene sequences instead revealed strong subdivision between populations separated by as little as 30 km. A subsequent survey of Halocaridina from 34 populations on Hawai?i, Maui, and O?ahu identified 13 genetic groups belonging to eight (8) lineages qualifying as potential species (Craft et al., 2008). In general, these genetic groups and lineages are restricted to anchialine habitats from a particular region of a single Hawaiian Island, with no individuals being exchanged between them. This phylogeographic pattern is attributed to the low dispersal potential of these atyids due to their life history traits in combination with marine and geologic barriers that compartmentalize the aquifers of each island (Santos, 2006; Craft et al., 2008). Thus, evolution in Halocaridina is driven by population fragmentation, isolation, and subsequent diversification in anchialine habitats of discrete aquifers across Hawai?i. Identifying the widespread ?Halocaridina rubra? as representing a ?species complex? with strong regional endemism highlights the ecological significance of conserving anchialine 19 habitats in the Hawaiian Archipelago. Moreover, these atyids have provided unexpected perspective into the tempo of evolution. Specifically, molecular clock estimates calibrated to the geologic age of Kilauea volcano on the island of Hawai?i lead to the revelation that the COI gene of Halocaridina is apparently evolving at an exceptional rate of 20% per million years (My-1) (Craft et al., 2008), in sharp contrast to the range of 1.7% My-1 (Williams & Knowlton, 2001) to 2.3% My-1 (Brower, 1994) commonly utilized or reported for arthropods or other atyids (e.g., Page & Hughes, 2007; Page et al., 2008). In this contribution, the validity of this rapid evolutionary rate is tested by 1) applying it to levels of divergence measured from geographically close, yet genetically distinct, Halocaridina populations on the islands of Hawai?i and Maui and 2) comparing the resulting time since diversification estimates to the geologic ages of their respective anchialine habitats. 2.3 MATERIALS AND METHODS 2.3.1 BIOLOGICAL MATERIALS In July 2008, Halocaridina were sampled from previously uncharacterized populations on the northeastern and southeastern coasts of Maui and Hawai?i, respectively (Figure 1). The anchialine habitats at Waikoloa, Maui (WKA; Figure 1A) and Wai?ohinu, Hawai?i (WP; Figure 1B) consist of clusters of 2-3 shallow (i.e., <0.5 m) pools whose basins are comprised of basalt. Temperatures (i.e., 25 ?C), salinities (i.e., 8.0 ppt) and daily tidal fluctuations at these locations were consistent with those from other anchialine habitats in the islands (Craft et al., 2008). At each site, a net was utilized to collect specimens of Halocaridina that were immediately preserved in 100% acetone (Fukatsu, 1999) for genetic analyses. 20 2.3.2 DNA EXTRACTION, POLYMERASE CHAIN REACTION (PCR) AND SEQUENCING For individual Halocaridina, total nucleic acid extraction and polymerase chain reaction (PCR) of a partial COI gene fragment followed procedures outlined in Santos (2006). Briefly, reactions were conducted in 25 ?L volumes containing ~10-30 ng of template DNA, 10 mM Tris-HCl (pH 8.3), 50 mM KCl, 0.001% gelatin, 2.0 mM MgCl2, 200 ?M dNTPs, 1 U Taq DNA polymerase and 0.4 ?M each of primers LCO1490 and HCO2198 (Folmer et al., 1994) using a PTC-100 thermocycler (MJ Research, USA). Purification of amplified products prior to sequencing was done with Montage PCR Filter Units (Millipore, USA) according to the supplier?s protocol. Products were cycle-sequenced in both directions using Big-Dye Terminators v.3.1 and analyzed on a PRISM 3100 (Applied Biosystems, USA). Ambiguities in the chromatograms were corrected by comparison to the complement DNA strand in Sequencher v4.8 (Gene Codes Corporation, USA) and finished sequences of 630 bp manually aligned with SE-AL version v2.0a11 (available at http://tree.bio.ed.ac.uk/software/seal/). Novel COI sequences were deposited into GenBank under accession numbers HM137541-HM137556. 2.3.3 DEMOGRAPHIC AND GENETIC STRUCTURE ANALYSES Demographic measures, such as estimates of genetic diversity (i.e., haplotype (h) and nucleotide (?); [Nei, 1987]) and tests of neutrality (i.e., Tajima?s D [Tajima, 1989]; Fu?s FS [Fu, 1997]), were done for each population using DnaSP v4.10.9 (Rozas et al., 2003) and Arlequin v3.11 (Excoffier et al., 2005). To test for genetic structure between these newly sampled populations and ones previously reported in Craft et al. (2008) that are in close (i.e., ~3.5-10 km) proximity, the nearest-neighbor statistic (Snn; Hudson, 2000) was calculated using DnaSP. Specifically, this test was conducted between Halocaridina populations WKA (this study) and 21 WC (i.e, Waianapanapa Cave, Craft et al., (2008); Figure 1A) and WP (this study) and PU (i.e, Puhi Ula Cave, Craft et al., (2008); Figure 1B) on Maui and Hawai?i (Figure 1), respectively. Additional tests for genetic structure involved calculation of pairwise ?ST statistics in Arlequin under Tamura & Nei?s (1993) model of DNA evolution. Significance of the neutrality tests as well as Snn and ?ST statistics was assessed with 10,000 permutations. Lastly, networks were constructed with TCS v1.21 (Clement et al., 2000) in order to visualize relationships among the COI haplotypes within and between the populations under comparison. Default settings in TCS were utilized for network construction, which provides the 95% parsimoniously plausible branch connections between haplotypes. 2.3.4 DIVERGENCE TIME ESTIMATION AND GEOLOGIC AGE Two different approaches were employed in estimating between-population times since divergence. Firstly, uncorrected (p) genetic distances of COI sequence divergence between population pairs were generated using MEGA v3.1 (Kumar et al., 2004), with values calculated using the ?net between-group means? option so as to correct for within-group polymorphism. These p genetic distances were then subjected to the 20% My-1 diversification rate previously hypothesized for Halocaridina (Craft et al., 2008) to infer a time since divergence between population pairs. Secondly, theta (? = 4Ne?, where Ne is the effective population size and ? is the per-locus mutation rate) and divergence time (T = t/2Ne, where t is generation time) between populations pairs were simultaneously estimated via the Markov Chain Monte Carlo (MCMC) method described by Nielsen & Wakeley (2001) and implemented with the web-based version of the program MDIV (available at http://cbsuapps.tc.cornell.edu/). Analyses were conducted under the finite-site mutation model (i.e., HKY85 [Hasegawa et al., 1985]), which accounts for the 22 possibility of multiple mutations per site, differences in nucleotide frequencies and the presence of transition/transversion bias. A minimum of three independent runs using identical starting conditions (Mmax = 10, Tmax = 1.0 (e.g., for the WC and WKA populations of Maui) or 5.0 (e.g., for the PU and WP populations of Hawai?i), length of Markov chain = 2 X 106 cycles, burn-in time = 5 X 105 cycles) but different random seeds was done to check for consistency in the estimates. Values of ? and T with the highest posterior probabilities were accepted as the best estimates of these parameters and utilized in generating converted estimates of the ?actual years since divergence? (AYSD) with the equation T?/2? (see Nielsen & Wakeley (2001) for details) and a ? of 1.26 ? 10-4 substitutions per sequence per year, which is based on a COI gene fragment length of 630 bp (see above) and the 20% My-1 diversification rate of Craft et al., (2008). Resulting point estimates of divergence time from both of these approaches were subsequently compared to the geologic age of the basalt in which that particular pair of anchialine habitats (and their resident Halocaridina populations) occur. Geologic ages were acquired from the U.S. Geological Survey via geographic information system (GIS) map layer files for the main Hawaiian Islands (Sherrod et al., 2007; available at http://pubs.usgs.gov/of/2007/1089/) that were rendered in ArcGIS v9.3 (ESRI, USA). 2.4 RESULTS 2.4.1 HALOCARIDINA FROM THE NORTHEASTERN COAST OF MAUI Two lineages and genetic groups of Halocaridina occur on Maui, with the Hana genetic group within the East Maui lineage localized to the northeastern coastline of the island (Craft et al., 2008). Genetic analyses found that Halocaridina from the anchialine habitats at Waikoloa (WKA; Figure 1A) also belong to this genetic group and lineage. The only other known 23 population of Halocaridina within this lineage and genetic group inhabits Waianapanapa Cave (WC), approximately 3.5 km northwest of WKA (Figure 1A). Comparisons between the WC and WKA populations revealed similar levels of recovered haplotypes, haplotype (h) and nucleotide (?) diversities (Table 1). In contrast, the WC population possessed significant negative Tajima?s D and Fu?s FS values while these same measures for the WKA population were not significant (Table 1). This reflects that an excess of haplotypes with rare polymorphisms were present at the WC site and is often a signature of recent population expansion (Tajima, 1989; Fu, 1997). Given this, the Halocaridina populations at WC and WKA have distinct demographic histories. Evidence for genetic structure between the WC and WKA populations was absent with either the Snn (0.563; P = 0.096) or pairwise ?ST (0.015; P = 0.199) statistics. Although genetic structure was not evident, only two haplotypes (e.g., the haplotype with the highest outgroup (i.e., ancestral) probability as well as one other) were shared between the populations, with the remaining being both private and typically recovered as a singleton in the network analysis (Figure 2A). This latter pattern suggests additional COI genetic diversity remains to be documented from these Halocaridina populations. Genetic distance between the WC and WKA populations in the form of the ?net between- group mean? yielded an uncorrected p value of 0.0001. In the context of the 20% My-1 diversification rate, this implies that they have diverged in the last ~500 yrs. For the MDIV analyses, the highest posterior probability of ? and T occurred at 4.3 and 0.07, respectively (Figure 3); application of these values to calculate the AYSD (see Materials & Methods) yielded an estimate of 1,194 yrs. Interestingly, the anchialine pools inhabited by Halocaridina at WC and WKA both occur in basalt with a geologic age of 750-1,500 yrs and are intersected by basalt of 200-750 yrs in age (insert in Figure 2A). Thus, the geologic ages of the most recent basalts 24 (i.e., lava flows) in the area strongly coincide with the estimated times since diversification (i.e., ~500/1,194 yrs) of the WC and WKA populations of Halocaridina. 2.4.2 HALOCARIDINA FROM THE SOUTHEASTERN COAST OF HAWAI?I The island of Hawai?i is home to three lineages and five genetic groups of Halocaridina (Santos, 2006; Craft et al., 2008). Of these, the Na?alehu genetic group within the Ka?? lineage is the most restricted in distribution, having only been previously found on the southeastern coast of the island at Puhi Ula Cave (PU). Here, the Halocaridina population of anchialine habitats at Wai?ohinu (WP) was identified as members of the Ka?? lineage, extending the lineage?s range ~10 km to the southwest (Figure 1B). Although the Ka?? lineage is present at both PU and WP, the populations at each site are unique. For example, nearly twice as many haplotypes and 2X more nucleotide (?) diversity were recovered from the WP population relative to that at PU (Table 1). Furthermore, only the population at PU possessed a significant negative Tajima?s D value (Table 1), implying a distinctive demography to the Halocaridina at this site. Lastly, significant genetic structure was detected between the WP and PU populations with both the Snn (0.826; P < 0.001) and pairwise ?ST (0.198; P < 0.001) statistics. The distinct genetics of the populations are also evident in the network analysis, where of the 14 unique COI haplotypes currently identified within the lineage, each population is numerically dominated by private haplotypes while only three are shared (at low frequencies) between the sites (Figure 2B). Thus, the PU and WP populations should be considered distinct genetic groups within the Ka?? lineage of Halocaridina. An uncorrected p value of 0.0008 was calculated for the ?net between-group mean? genetic distance between the WP and PU Halocaridina populations. In this case, application of 25 the 20% My-1 diversification rate resulted in an estimated time since divergence of ~4,000 yrs. Analyses with MDIV identified the highest posterior probability of ? and T as 2.6 and 0.5, respectively (Figure 3), with a resulting AYSD estimate of 5,158 yrs. Similar to the previous example on Maui, this strongly correlates with the geologic age of the area inhabited by these populations. Specifically, the anchialine habitats at WP and PU occupy a contiguous basalt flow whose age ranges between 3,000-5,000 yrs (Figure 2B), which is consistent with the divergence time estimates (i.e., ~4,000/5,158 yrs) for these Halocaridina populations. 2.5 DISCUSSION 2.5.1 VALIDATION OF THE HALOCARIDINA MOLECULAR CLOCK Previously, levels of mtDNA COI gene sequence divergence calibrated to the known geologic age of Kilauea volcano on Hawai?i suggested members of the anchialine atyid genus Halocaridina are diversifying at the rate of 20% My-1 (Craft et al., 2008). This is, to the best of our knowledge, the highest mitochondrial molecular clock estimate currently reported for an invertebrate (see Burridge et al., 2006, 2008; Waters et al., 2007 for comparable estimates from vertebrates). One hypothesis for this is that the mitochondrial genome of Halocaridina is ?unusual? relative to other crustaceans. However, gene order and orientation in the Halocaridina mitochondrial genome is syntenic with nearly all malacostracans examined to date and phylogenetic analyses encompassing the organelle?s 13 protein-coding genes found no ?long- branch? (i.e., potential accelerated evolutionary rate) associated with the genus (Ivey & Santos, 2007). Thus, an alternative interpretation is required towards explaining this extraordinary divergence rate. 26 While the taxonomic status of the eight lineages within Halocaridina currently remains unresolved, numerous lines of evidence imply that they represent ?cryptic species? (Santos, 2006; Craft et al., 2008). Furthermore, many populations within the Halocaridina lineages are on trajectories towards becoming new lineages themselves, such as the South Hilo/Puna and Kalaeloa/Wai?anae genetic groups of Hawai?i and O?ahu, respectively (Santos, 2006; Craft et al., 2008). Thus, the inferred rate of diversification for Halocaridina is an estimate at the interface between populations and ?species?. In this context, there is growing evidence that the derivation and application of molecular clocks at this scale represents special circumstances. Specifically, it has been advocated that molecular clocks exhibit variability in the form of an overall lower, long-term mutation rate preceded by a higher short-term (<1-2 My) substitution rate at the population-species boundary (reviewed by Ho & Larson, 2006). This apparent higher rate is thought to originate from a combination of temporal differences in the population-level retention of mutations based on their ?deleteriousness? and the relative time of their observation (i.e., sampling) in the population under study (reviewed by Penny, 2005). The evolution of Halocaridina across the anchialine habitats of the Hawaiian Islands appears to epitomize this phenomenon, particularly when placed into the perspective of the geologic histories and ages of the islands (see below), and adds to the growing list of examples where failure to recognize this potential would have significantly altered the perceived time-scales at which ecological and evolutionary processes are occurring (see Ho et al., 2008 for case studies). The analyses of previously uncharacterized Halocaridina populations presented here strongly supports the 20% My-1 COI molecular clock first proposed by Craft et al. (2008) for these atyid shrimp. Specifically, divergence time estimates based on this rate for populations in two distinct lineages on two islands were highly consistent with the geologic age of the basalt 27 basins of their anchialine habitats (see Results). This illustrates how the demographics, population structure and evolutionary history of endemic anchialine organisms can be intimately coupled to the particular geologic history of their physical habitat. For Halocaridina, this coupling is due to life history traits such as large egg size, abbreviated development, restricted larval habitat and larval feeding mode effectively isolating individuals on new ?islands? (i.e., distinct aquifers and their anchialine habitats) following successful dispersal and/or an allopatric fragmentation event (Santos, 2006; Craft et al., 2008). Thus, it would not be surprising to find that this scenario, accompanied by exceptional rates of divergence, extends to other anchialine species with similar life history traits or belonging to groups with notoriously poor dispersal abilities (e.g., amphipods; Finston et al., 2007) and for which (most importantly) well- established geologic ages of their habitats are available for calibration. 2.5.2 MAUI AND HAWAI?I: GEOLOGIC HISTORY AND HALOCARIDINA Maui is the second youngest (i.e., <1.5 My) of the current high islands in the Hawaiian Archipelago and is composed of two distinct volcanoes, West Maui and Haleakala. As discussed earlier, the Hana genetic group within the East Maui lineage of Halocaridina is restricted to the northeastern coast of the island, along the flank of Haleakala. In the past 10,000 years, Haleakala volcano has erupted on numerous occasions, including at least ten major eruptions in the last 1,000 years (http://hvo.wr.usgs.gov/volcanoes/haleakala/). Much of this activity was localized along two interconnecting rift zones: the southwest rift zone, which passes from La Perouse Bay on the southwest flank of Haleakala to the summit, and the east rift zone, radiating from the summit along the volcano?s eastern flank to Hana. In fact, the east rift zone of Haleakala is located ~5.5 km south of the anchialine habitats at WC and WKA and is responsible for the 28 relatively young basalts in the area. These include lava flows ranging in age from 750-1,500 yrs, which is the basalt the WC and WKA anchialine habitats occur in, and 200-750 yrs, which traverses the two sites (Figure 1A; Sherrod et al., 2007). Given this geologic history and the divergence time estimates of ~500 or 1,194 yrs, this would imply that the anchialine habitats at these sites are <1,500 yrs in age and, in the interim, that the area?s recent lava flows have potentially impacted the demographics of their local Halocaridina by leading to cycles of contraction and expansion in some populations (i.e., WC) but not in others (i.e., WKA). Such a situation has also been reported for at least one Halocaridina population on the island of Hawai?i, in which the signature of a recent and strong contraction and expansion similarly correlated with a contemporary and localized volcanic episode (Santos, 2006). The youngest (i.e., ~430,000 yrs) island in the Hawaiian Archipelago is Hawai?i, being comprised of five separate volcanoes. Of these, the Ka?? lineage of Halocaridina is restricted to the southeastern flank of Mauna Loa. This particular volcano has erupted 33 times since 1843, making it one of the most active (as well as the largest) on Earth (http://hvo.wr.usgs.gov/maunaloa/). While the oldest dated basalts attributable to Mauna Loa range from 100,000-200,000 yrs, >98% of the volcano's surface is composed of flows younger than 10,000 yrs (http://hvo.wr.usgs.gov/maunaloa/history/main.html). Similar to Haleakala, these lava flows predominately originate from two rift zones radiating from the summit and towards the northeast and southwest, respectively. The southwest rift zone of Mauna Loa is the more active of the two (Macdonald et al., 1983) and the source of basalt for the anchialine habitats at PU and WP. As mentioned previously, both sites occur within a contiguous flow whose age is estimated to range between 3,000-5,000 yrs (Figure 1A; Sherrod et al., 2007), which coincides nearly exactly with the ~4,000 or 5,158 yrs time since divergence estimates for this pair of 29 Halocaridina populations (see Results). In this case, we hypothesize these anchialine habitats (i.e., PU and WP) were colonized by Halocaridina soon after their creation from populations in ?refugia? offered by the 10,000-30,000 yr basalts immediately adjacent to this younger flow (Figure 1A). Given the ~10 km distance between the sites and the fact that peripheral populations of Halocaridina across a range generally exhibit some level of differentiation due to restricted gene flow (Santos, 2006; Craft et al., 2008), this founder effect, reinforced via isolation by distance, then lead to the significant population structure currently evident between the Wai?ohinu and Na?alehu genetic groups of Halocaridina. 2.5.3 REEVALUATING HALOCARIDINA EVOLUTION IN THE CONTEXT OF ISLAND GEOLOGY Validating the 20% My-1 molecular clock for Halocaridina prompts us here to revisit evolution within the genus in the specific context of the geologic histories and ages of the Hawaiian Islands. While a well-resolved phylogeny for the Halocaridina ?species complex? remains to be elucidated (Craft et al., 2008), thus limiting our ability to fully address this topic, seven comparisons either between populations, genetic groups and/or lineages across three islands can be brought into consideration (Table 2; Figure 4). For these, estimated times since divergence were also calculated using the 1.7% My-1 and 2.3% My-1 molecular clocks of Williams & Knowlton (2001) and Brower (1994), respectively (Table 2). In all cases, utilization of these ?generic? clocks resulted in divergence time estimates considerably older than the ?true? age based on geologic evidence (Table 2). For example, the sister lineages of East Hawai?i and Ka?? serve as the basis for the 20% My-1 Halocaridina molecular clock due to their 2% genetic distance average being calibrated to the oldest estimated atmospheric eruption of Kilauea volcano at 100,000 yrs (http://hvo.wr.usgs.gov/kilauea/; dotted line in Figure 4). In contrast, the 30 generic clocks suggest this divergence event occurred somewhere between 0.87-1.2 My ago (#1; Table 2), which is ~2-3X older than the island of Hawai?i itself (~430,000 yrs, Carson & Clague, 1995). For this reason, generic molecular clocks such as these should be applied with caution since they have the potential to greatly mislead divergence time estimates for anchialine organisms or taxa from other ecosystems in general. Within the East Hawai?i lineage of Halocaridina, diversification of the Puna and South Hilo genetic groups was apparently precipitated by an allopatric fragmentation event associated with the East Rift Zone of Kilauea volcano (Santos, 2006). Here, the occurrence of that event is dated at ~60,000 yrs (#2; Table 2; Figure 4). Kilauea is the youngest and most southeastern volcano on Hawai?i and has been in continuous eruption since January 1983. While impressive in itself, this volcano is thought to have undergone no prolonged periods of inactivity (http://hvo.wr.usgs.gov/kilauea/history/), and been in uninterrupted growth, since its emergence from the sea ~100,000 yrs ago. Thus, we hypothesize that 1) the East Hawai?i lineage of Halocaridina colonized the flank of Kilauea at or prior to 60,000 yrs ago; 2) at ~60,000 yrs, establishment of, or successful dispersal around, the East Rift Zone lead to the inception of the Puna and South Hilo genetic groups that have been diverging in allopatry since the event; and 3) their present ranges and anchialine habitats are most likely not those in which the event took place due to Kilauea?s continuous activity and expansion reshaping the coastlines over its subsequent history. O?ahu possesses the greatest known genetic diversity of Halocaridina in the Hawaiian Archipelago, with four lineages and six genetic groups reported from the island (Craft et al., 2008). It is also distinctive in that, unlike Maui and Hawai?i, most of its anchialine habitats occur as a single pool per site and all have basins of fossilized, but porous, limestone materials. This 31 limestone is due to the development of expansive coral reefs along the island?s coasts over its >2.75 My existence (Lau & Mink, 2006). Approximately 131,000 yrs ago, sea levels on O?ahu rose ~2 m at the beginning of the last interglacial period (Szabo et al., 1994), submerging low- lying areas and transforming them into near shore environments conducive to coral reef growth. Overall, this interglacial period lasted for ~17,000 yrs before sea level returned to near-present conditions around 114,000 yrs ago (Szabo et al., 1994), ?stranding? in a terrestrial setting any coral reef environments that had flourished during this interval. Thus, the current anchialine habitats of O?ahu trace their origins back to this particular interglacial event and time period. In three of the lineages on O?ahu, the diversification of Halocaridina genetic groups occurred between 45,000-65,000 yrs ago (#5, #6, #7; Table 2; Figure 4), well after the last interglacial period ended ~114,000 yrs ago. Along with this, the two volcanoes on the island, Wai?anae and Ko?olau, are considered extinct since they have not erupted in the past 500,000 yrs (Macdonald et al., 1983). For these reasons, we cannot confidently identify a specific geologic event of definable age associated with the divergence times of the Kahuku/Lanikai, Wai?anae/Kalaeloa and Kapapa/Kona genetic groups of Halocaridina (Table 2; Figure 4). Instead, the most parsimonious explanation is that dispersal and successful colonization at these estimated times were the events leading to their divergence since they reside allopatrically either in well-compartmentalized, but adjacent, aquifers of the same island or on different islands (Craft et al., 2008). Of these, the Kapapa and Kona/Ka?? genetic groups (#7; Table 2) within the West Hawai?i lineage are particularly interesting. For example, this lineage is the only one to date recorded from more than a single island, being found on O?ahu and Hawai?i (Figure 4). Under the assumption of the ?Progression Rule?, in which populations from older islands act as founders for ones on younger islands (Funk & Wagner 1995), this implies that 1) the arrival of 32 the West Hawai?i lineage on the western coast of Hawai?i is a relatively recent (i.e., ~45,000 yrs ago; Table 2) event; 2) it was the second of the two lineages to colonize that island since divergence in the East Hawai?i lineage predates this arrival time estimate (see above), and; 3) the lineage has been expanding its range along the entire western coast of Hawai?i up to the present while diversifying into the Kona and Ka?? genetic groups (Santos, 2006; Craft et al., 2008), which are the primary source populations of Halocaridina in the aquarium trade (Weese & Santos, 2009). In contrast, the evolutionary history of the Kapapa genetic group within the West Hawai?i lineage is less clear. The reason for this is that Kapapa Island, which this genetic group is restricted to, is a small (i.e., 0.012 km2) limestone islet located ~3.2 km off the eastern coastline and is comparatively young (i.e., 80,000-143,000 yrs; Table 2) in relation to the rest of O?ahu. Thus, from where did the migrants who colonized Kapapa Island originate? Possibilities include an undocumented Halocaridina population from the numerous aquifers that remain to be sampled on O?ahu (or the other islands) or diversification from one of the previously recognized lineages. In the interim, and as answers to such questions are pursued, the genus Halocaridina from the anchialine habitats of the Hawaiian Archipelago has offered exciting insight into the processes and tempo of evolution in this ecosystem and it will be interesting to see if similar patterns are reported from other geographic regions and anchialine taxa in the future. 33 2.6 ACKNOWLEDGEMENTS We would like to thank Dr. Bill Gilmartin for allowing sampling of Halocaridina from the anchialine habitats at Wai?ohinu, Skippy Hau (State of Hawai?i Division of Aquatic Resources ? Maui) and John Kahiapo (State of Hawai?i Division of Aquatic Resources ? Hawai?i) for additional assistance in field collections and Dr. Ken M. Halanych for suggestions leading to the title of this manuscript. Dr. Timothy J. Page (Griffith University, Australia) and one anonymous reviewer provided valuable comments that improved this contribution. This work was funded in part by National Science Foundation grant DEB-0949855 (SRS) and represents contributions #73 and #1 to the Auburn University (AU) Marine Biology Program and Molette Biology Laboratory for Environmental and Climate Change Studies, respectively. 34 2.7 REFERENCES Avise, J.C., 2004. Molecular markers, natural history, and evolution. Sinauer Associates Inc.,Sunderland, MA. Bailey-Brock, J.H. & R.E. Brock, 1993. Feeding, reproduction, and sense organs of the Hawaiian anchialine shrimp Halocaridina rubra (Atyidea). Pacific Science 47: 338-355. Brock, R.E., 1985. An assessment of the conditions and future of the anchialine pond resources of the Hawaiian Islands. Prepared for Transcontinental Development Co., Honolulu. Brock R.E. & A.K.H. Kam, 1997. Biological and water quality characteristics of anchialine resources in Kaloko-Honokohau National Historical Park Technical Report 112: Cooperative National Park Resources Studies Unit, University of Hawaii at Manoa. Brock R.E. & J.H. Bailey-Brock, 1998. A unique anchialine pool in the Hawaiian Islands. International Review of Hydrobiology 83: 65-75. Brower, A.V.Z., 1994. Rapid morphological radiation and convergence among races of the butterfly Heliconius erato inferred from patterns of mitochondrial DNA evolution. Proceedings of the National Academy of Science USA 91: 6491-6495. Burridge, C.P., D. Craw & J.M. Waters, 2006. River capture, range expansion, and cladogenesis: the genetic signature of freshwater vicariance. Evolution 60: 1038-1049. Burridge, C.P., D. Craw, D. Fletcher & J.M. Waters, 2008. Geological dates and molecular rates: fish DNA sheds light on time dependency. Molecular Biology and Evolution 25: 624- 633. Carson, H.L. & D.A. Clague, 1995. Geology and biogeography of the Hawaiian Islands. In Wagner, W.L. & V.A. Funk (eds), Hawaiian Biogeography Evolution on a Hot Spot Archipelago. Smithsonian Institution Press, Washington, DC: 14-29. 35 Clement, M., D. Posada & K.A. Crandall, 2000. TCS: a computer program to estimate gene genealogies. Molecular Ecology 9: 1657-1659. Craft, J.D., A.D. Russ, M.N. Yamamoto, T.Y. Iwai, S. Hau, J. Kahiapo, C.T. Chong, S. Ziegler- Chong, C. Muir, Y. Fujita, D.A. Polhemus, R.A. Kinzie & S.R. Santos, 2008. Islands under islands: The phylogeography and evolution of Halocaridina rubra Holthuis, 1963 (Crustacean: Decapoda: Atyidae) in the Hawaiian archipelago. Limnology and Oceanography 53: 675-689. Emerson, B.C., 2002. Evolution on oceanic islands: molecular phylogenetic approaches to understanding pattern and process. Molecular Ecology 11: 951-966. Excoffier, L. & S. Schneider, 2005. Arlequin ver. 3.0: An integrated software package for population genetics data analysis. Evolutionary Bioinformatics Online 1: 47-50. Fletcher III, C.H., C.V. Murrary-Wallace, C.R. Glenn, C.E. Sherman, B. Popp & A. Hessler, 2006. Age and origin of late Quaternary Eolianite, Kaiehu Point (Moomomi), Molokai, Hawaii. Journal of Coastal Research 42: 97-112. Finston, T.L., M.S., Johnson, W.F. Humphreys, S.M. Ebergard & S.A. Halse, 2007. Cryptic speciation in two widespread subterranean amphipod genera reflects historical drainage patterns in an ancient landscape. Molecular Ecology 16: 355-365. Folmer, O., M. Black, W. Hoeh, R. Lutz & Vrijenhoek R, 1994. DNA primers for amplification of mitochondrial cytochrome c oxidase subunit I from diverse metazoan invertebrates. Molecular Marine Biology and Biotechnology 3: 294-299. Fu, Y., 1997. Statistical tests of neutrality of mutations against population growth, hitchhiking and background selection. Genetics 147: 915-925. 36 Fukatsu, T. 1999. Acetone preservation: a practical technique for molecular analysis. Molecular Ecology 8: 1935-1945. Funk, V.A. & W.L. Wagner, 1995. Biogeographic patterns in the Hawaiian Islands. In Funk, V.A. & W.L. Wagner (eds.), Hawaiian biogeography: Evolution on a hot spot archipelago. Smithsonian Institution Press, Washington DC: 379-419. Grossman, E.E. & C.H. Fletcher, III, 2004. Holocene reef development where wave energy reduces accommodation space, Kailua Bay, windward Oahu, Hawaii, U.S.A. Journal of Sedimentary Research 74: 49-63. Hasegawa, M., H. Kishino & T. Yano, 1985. Dating of the human-ape splitting by a molecular clock of mitochondrial DNA. Journal of Molecular Evolution 22: 160-174. Ho, S.Y.W. & G. Larson, 2006. Molecular clocks: when times are a-changin'. Trends in Genetics 22: 79-83. Ho, S.Y.W., U. Saarma, R. Barnett, J. Haile & B. Shapiro, 2008. The effect of inappropriate calibration: Three case studies in molecular ecology. PLoS ONE 3: e1615. Holthuis, L.B., 1973. Caridean shrimps found in land-locked saltwater pools at four Indo-West Pacific localities (Sinai Peninsula, Funafuti Atoll, Maui and Hawai?i islands), with the description of one new genus and four new species. Zoologische Verhandelingen (Leiden) 128: 1-48. Hudson, R.R., 2000. A new statistic for detecting genetic differentiation. Genetics 155: 2011- 2014. Ivey J.L & S.R. Santos, 2007. The complete mitochondrial genome of the Hawaiian anchialine shrimp Halocaridina rubra Holthuis, 1963 (Crustacea: Decapoda: Atyidae). Gene 394: 35-44. 37 Kumar, S., K. Tamura & M. Nei, 2004. MEGA3: Integrated software for molecular evolutionary genetics analysis and sequence alignment. Briefings in Bioinformatics 5: 150-163. Lau L.S. & J.F. Mink, 2006. Hydrology of the Hawaiian Islands. University of Hawaii Press, Honolulu, HI. MacArthur, R.H. & E.O. Wilson, 1967. The Theory of Island Biogeography. Princeton University Press, Princeton, NJ. Macdonald, G.A., A.T. Abbott & F.L. Peterson, 1983. Volcanoes in the Sea: The Geology of Hawaii. University of Hawaii Press, Honolulu, HI. Maciolek, J.A., 1983. Distribution and biology of Indo-Pacific insular hypogeal shrimp. Bulletin of Marine Science 33: 606-618. Maciolek, J.A. & R.E. Brock, 1974. Aquatic survey of the Kona coast ponds, Hawaii Island. Sea Grant Advisory Report, UNIHI-SEAGRANT-AR-74-04. Nei, M., 1987. Molecular Evolutionary Genetics. Columbia University Press, New York. Nielsen, R. & W. Wakeley, 2001. Distinguishing migration from isolation: an MCMC approach. Genetics 158: 885-896. Page, T.J. & J.M. Hughes, 2007. Phylogeographic structure in an Australian freshwater shrimp largely pre-dates the geological origins of its landscape. Heredity 98: 222-231. Page, T.J., W.F. Humphreys & J.M. Hughes, 2008. Shrimps Down Under: evolutionary relationships of subterranean crustaceans from Western Australia (Decapoda: Atyidae: Stygiocaris). PLoS ONE 3: e1618. Penny, D., 2005. Relativity for molecular clocks. Nature 426: 183-184. Rozas, J., J. Sanchez-DelBarrio, X. Messeguer & R. Rozas, 2003. DnaSP, DNA polymorphism analyses by the coalescent and other methods. Bioinformatics 19: 2496-2497. 38 Santos, S.R., 2006. Patterns of genetic connectivity among anchialine habitats: a case study of the endemic Hawaiian shrimp Halocaridina rubra on the Island of Hawai?i. Molecular Ecology 15: 2699-2718. Sherrod, D.R., J.M. Sinton, S.E. Watkins & K.M. Brunt, 2007. Geologic map of the State of Hawai`i: U.S. Geological Survey Open-File Report 2007-1089. http://pubs.usgs.gov/of/2007/1089/. Sket, B., 1996. The ecology of anchialine caves. Trends in Ecology and Evolution 11: 221-225. Szabo, B.J., K.R. Ludwidg, D.R. Muhs & K.R. Simmons, 1994. Thorium-230 ages of corals and duration of the last interglacial sea-level high stand on Oahu, Hawaii. Science 266: 93-96. Tajima, F., 1989. Statistical method for testing the neutral mutation hypothesis by DNA polymorphism. Genetics 123: 585-595. Tamura, K. & M. Nei, 1993. Estimation of the number of nucleotide substitutions in the control region of mitochondrial DNA in humans and chimpanzees. Molecular Biology and Evolution 10: 512-526. Vandergast, A.G., R.G. Gillespie & G.K. Roderick, 2004. Influence of volcanic activity on the population genetic structure of Hawaiian Tetragnatha spiders: fragmentation, rapid population growth and the potential for accelerated evolution. Molecular Ecology 13: 1729-1743. Waters, J.M., D.L. Rowe, S. Apte, T.M. King, G.P. Wallis, L. Anderson, R.J. Norris, D. Craw & C.P. Burridge, 2007. Geological dates and molecular rates: rapid divergence of rivers and their biotas. Systematic Biology 56: 271-282. Weese, D.A. & S.R. Santos, 2009. Genetic identification of source populations for an aquarium- traded invertebrate. Animal Conservation 12:13-19. 39 Williams, S. T. & N. Knowlton, 2001. Mitochondrial pseudogenes are pervasive and often insidious in the snapping shrimp genus Alpheus. Molecular Biology and Evolution 18: 1484-1493. 40 Table 1: Genetic diversity measures and results of neutrality tests for Halocaridina populations from the northeastern and southeastern coasts of Maui and Hawai?i, respectively, examined in this study. n = number of sampled individuals; nh = number of recovered haplotypes; ? = nucleotide diversity; h = haplotype diversity. * = p < 0.05. Diversity indices Neutrality tests Island Region Population n nh ? h Tajima?s D Fu?s Fs Northeastern Maui WC 29 13 0.004 0.702 ? 0.096 -1.64* -5.03* WKA 21 9 0.005 0.795 ? 0.077 -0.72 -1.23 Southeastern Hawai?i PU 30 6 0.002 0.667? 0.055 -1.62* -1.03 WP 30 11 0.004 0.745? 0.082 -1.06 -2.63 41 Table 2: Summary of molecular clock comparisons for Halocaridina. Estimated times since divergence for each comparison were calculated using rates of 1.7% My-1 (Williams & Knowlton 2001), 2.3% My-1 (Brower 1994) and 20% My-1 (Craft et al., 2008). Estimated Time since Divergence Comparison in Figure 4 Island: Coastal region Population/Genetic Group/Lineage Comparison Genetic Distance (% difference)? 1.7% My- 1* 2.3% My-1** 20% My-1*** ?True? Geologic Age #1 Hawai?i:Eastern coast East Hawai?i vs. Ka?? 2.0 1,180,000 870,000 100,000 100,000a #2 Hawai?i:Eastern coast Puna vs. South Hilo 1.2 706,000 520,000 60,000 100,000a #3 Hawai?i:Southeastern coast PU vs. WP 0.08 47,000 35,000 4,000- 5,158 3,000- 5,000b #4 Maui:Eastern coast WC vs. WKA 0.01 5,900 4,300 500- 1,194 750- 1,500b #5 O?ahu: Northern coast Kahuku vs. Lanikai 1.0 590,000 435,000 50,000 114,000- 131,000c #6 O?ahu: Western coast Wai?anae vs. Kalaeloa 1.3 765,000 565,000 65,000 114,000- 131,000c #7 O?ahu/Hawai?i:Northern coast/Western coast Kapapa vs. Kona/Ka?? 0.9 530,000 390,000 45,000 80,000- 143,000d ? calculated based on ?net between-group mean? of uncorrected p distance. See text for additional details. * Williams & Knowlton, (2001) ** Brower, (1994) *** Craft et al., (2008) and this paper (see Results) a Carson & Clague, (1995); http://hvo.wr.usgs.gov/kilauea/ b Sherrod et al., (2007) c Szabo et al., (1994) d Grossman & Fletcher, (2004); Fletcher et al., (2006) 42 Figure 1: Map of the high Hawaiian Islands depicting anchialine pools where Halocaridina were sampled for this study. Inserts: Geologic maps, including ages of local basalt (i.e., lava) flows, for the northeastern coast of Maui (A) and the southeastern coast of the island of Hawai?i (B). Site codes: Waianapanapa Cave (WC), Waikoloa (WKA), Puhi Ula Cave (PU) and Wai?ohinu (WP). 43 Figure 2: Network analyses for the East Maui (A) and Ka?? (B) lineages of Halocaridina from northeastern Maui and southeastern Hawai?i, respectively. Networks depict relationships between cytochrome oxidase subunit I (COI) haplotypes within each lineage. Black dots represent DNA types that might exist in the population but remain unsampled (i.e., missing haplotypes). Rectangles represent DNA types with the highest outgroup probability according to the analysis. The size of circles and rectangles are proportional to the frequency at which a DNA type was recovered. Despite variable lengths, each branch implies a single mutational difference between haplotypes. 44 Figure 3: Posterior distributions of theta (?; A) and divergence time (T; B) of Halocaridina population pairs from northeastern Maui and southeastern Hawai?i. The presented posterior distributions each represent an average of three independent runs from the program MDIV (Nielsen & Wakeley 2001) utilizing identical starting conditions but different random seeds. See text for additional details. 45 Figure 4: Genetic distances and corresponding estimated times since divergence for seven comparisons between populations, genetic groups and/or lineages of Halocaridina across the islands of Hawai?i, Maui, and O?ahu. Genetic distance were calculated as ?net between-group mean? and estimated times since divergence were calculated using the 20% My-1 molecular clock of Craft et al., 2008 calibrated from comparison # 1 (dashed line) to the geologic age of the Kilauea volcano on the island of Hawai?i. Numbers correspond to comparisons in Table 2; see Craft et al., 2008 for site codes. 46 CHAPTER 3. Genetic identification of source populations for an aquarium traded invertebrate 3.1 ABSTRACT Increasingly, wildlife managers are turning to molecular genetics to aid in conservation efforts. While such approaches have been applied to large terrestrial and aquatic vertebrate species, their application to other traded organisms has not been extensively explored. Here, we examined the utility of these techniques for identifying source populations of aquarium ornamental invertebrates, using members of the Hawaiian atyid genus Halocaridina as a study system. These shrimp, restricted to anchialine habitats of the Hawaiian Islands, are popular in the aquarium trade due to their ability to survive in hermetically sealed containers for extended periods of time. However, commercial harvesting, coupled with habitat destruction and strong regional endemism, could lead to the depletion/extinction of unique populations. Because the land district of Kona, along the west coast of the island of Hawai?i, has the state?s highest concentration of anchialine habitats, we hypothesized that commercially available Halocaridina originated from this region. To test this, mitochondrial cytochrome c oxidase subunit I (COI) gene sequences from 96 individuals, obtained from six vendors, were compared to 580 homologous sequences from previous studies covering the known distribution range of Halocaridina. Recovery of identical, regional-specific haplotypes, network analyses, and statistical assignment tests identified these commercially acquired specimens as belonging to either the Kona, Ka??, (western and southern coasts, respectively, island of Hawai?i), or Kina?u (southern coast, island of Maui) genetic groups of these shrimp. Although 39 of the 96 individuals originated from the Kona genetic group as hypothesized, our finding that commercially available Halocaridina are from three genetic groups spanning two islands 47 suggests that other populations also warrant potential management consideration. While this study represents the first application of molecular genetics in identifying source populations of aquarium ornamental species, we feel that these techniques are amenable more broadly since they are dependent on only a few caveats. 3.2 INTRODUCTION The annual trade in wildlife and related products, both legal and illegal, involves hundreds of millions of plant and animal specimens with an estimated value of billions of US dollars (CITES, 2008). This monetary figure includes proceeds from the stocking of public and private farms, hunting ranches, zoo displays and safari parks as well as organisms for biomedical research and teaching, consumption, and exotic pets. Given the varying demands, laws at the international, national, and local levels have been established to help preserve these resources and wildlife managers have turned to molecular genetics as an aid in these pursuits (Millions & Swanson, 2006; Kyle & Wilson, 2007; Russello et al., 2007). In this context, molecular genetics has addressed issues ranging from taxonomic identification (Palumbi & Cipriano, 1998; Hoelzel, 2001; Clapham & Van Waerebeek, 2007) to determining source populations for specific individuals (Frantz et al., 2006) of traded species. While such ?wildlife forensic? approaches have been primarily applied to large terrestrial (e.g., Guglich, Wilson & White, 1994; Frantz et al., 2006; Wasser et al., 2007) and aquatic vertebrate species such as sharks, whales, and seals (e.g., Malik et al., 1997; Cipriano & Palumbi, 1999; Chapman et al., 2003), only a few studies have investigated their utility for other traded organisms, like those popular as aquarium ornamentals (e.g., Stam et al., 2006). This is particularly relevant since it is conservatively estimated that 1,471 fish, 140 stony (i.e., 48 scleractinian) coral, and over 500 other invertebrate species are traded as marine ornamentals on a yearly basis (Wabnitz el al., 2003), with most being stocked from wild caught specimens (Andrews, 1990). Because commerce in aquatic ornamental species is a growing industry (Heerbrandt & Lin, 2006), worth US $200 - $300 million annually (Wabnitz el al., 2003), effective management strategies will need to be implemented to ensure sustainability of these resources into the future. In recent years, shrimps from the family Atyidae have grown in popularity as aquatic ornamentals (Heerdebrant & Lin, 2006). One species, Halocaridina rubra (Holthuis, 1963) (Crustacea: Decapoda: Atyidae), commonly referred to as ??pae ?ula (lit. tiny red shrimp) in the Hawaiian language, has garnered particular attention among hobbyists. This interest stems from the ability of H. rubra to survive in hermetically sealed containers for extended (i.e., > 10 yr) periods of time (Maciolek, 1983) and has lead to an industry where enclosed ecosystems containing these small (i.e., < 12 mm) shrimp are sold around the world as novelty ?self- sustaining and perfect? pets. Additionally, H. rubra is marketed in both pet stores and via the World Wide Web as a live food for the rearing and maintenance of ornamental fish species. These practices raise concerns for two reasons. Firstly, H. rubra is restricted to anchialine habitats of the Hawaiian Islands. Anchialine habitats are land-locked bodies of brackish to marine water with no surface connection to the ocean but whose waters fluctuate with the tides due to subterranean connections (Holthuis, 1973; Maciolek, 1983; 1986). Historically, these habitats have experienced significant negative impacts from a variety of anthropogenic sources (reviewed by Santos, 2006). Secondly, mitochondrial DNA (mtDNA) analyses have identified at least 13 genetic groups of H. rubra across the Hawaiian Archipelago, each of which is confined to a single island or a particular subregion of an island (Craft et al., 2008; Fig. 1). Thus, 49 commercial harvesting, coupled with habitat destruction as well as strong regional endemism, could lead to the depletion and/or extinction of unique Halocaridina populations or genetic groups. An important step towards developing sound conservation strategies for aquarium ornamental species like Halocaridina is to first identify populations warranting management due to potential overexploitation. Here, we examine the effectiveness of ?wildlife forensics? for addressing such a situation. We hypothesized that commercially available Halocaridina originate from the Kona genetic group of these shrimp, which is distributed on the western coast of the island of Hawai?i (Fig. 1). Along with having the highest concentration of anchialine habitats in the Hawaiian Archipelago (Maciolek & Brock, 1974), this area is easily accessible by ground vehicles and in close proximity to an international airport, both of which would facilitate the rapid transport of live animals. To test our hypothesis, animals were purchased from local stores in Hawai?i as well as vendors on the World Wide Web and sequence data from the mtDNA cytochrome c oxidase subunit I (COI) gene of these individuals were compared to a large database of homologous sequences from previous ecological and evolutionary studies covering the known distribution range of the genus (Santos, 2006; Craft et al., 2008). This approach allowed us to assign these ?unknown? individuals to specific Halocaridina genetic groups from particular geographic regions of the Hawaiian Islands and highlights the utility of molecular genetics in identifying source populations of commercially traded aquarium ornamental species. 50 3.3 MATERIALS AND METHODS 3.3.1 BIOLOGICAL MATERIALS Commercial vendors marketing Halocaridina were identified through searches of the World Wide Web. From these, we purchased specimens from five stores on two Hawaiian Islands, Hawai?i (3) and O?ahu (2) (Table 1). Along with this, three enclosed ecosystems originating from a single distributor were acquired from retail stores, or by mail order, within the continental United States (Table 1). Each group of commercially acquired Halocaridina was represented by between 3 and 31 individuals, for a total sample size of 96 animals (Table 1). Cost per individual shrimp ranged from US $0.1, when sold as live fish food in Hawai?i, to ~$22, when acquired as part of an enclosed ecosystem. 3.3.2 DNA Extraction, Polymerase Chain Reaction (PCR) and Sequencing Total nucleic acid was extracted from each Halocaridina individual as described by Santos (2006). Following extraction, ~10 ? 30 ng of DNA was utilized as template to amplify an ~700 bp fragment of the COI gene. Polymerase chain reactions (PCRs) were conducted in 25 ?L volumes containing 10 mM Tris-HCl (pH 8.3), 50 mM KCl, 0.001% gelatin, 2.0 mM MgCl2, 200 ?M dNTPs, 1 U Taq DNA polymerase and 0.4 ?M each of primers LCO1490 and HCO2198 (Folmer et al., 1994). Reactions were performed in a PTC-100 thermocycler (MJ Research, Watertown, MA, USA) following the conditions outlined by Santos (2006). Amplifications were confirmed by electrophoresing an aliquot of the amplified product in a 1% agarose gel, followed by staining and viewing under UV light. Amplified products were purified with Montage PCR Filter Units (Millipore, Billerica, MA, USA) following the supplier?s protocol, cycle-sequenced in both directions using Big-Dye 51 Terminators, and read on a PRISM 3100 Genetic Analyzer (Applied Biosystems, Foster City, CA, USA). Ambiguities in the chromatograms were corrected by comparison to the complementary DNA strand in Sequencher v4.7 (Gene Codes Corporation, Ann Arbor, MI, USA). Completed sequences of 630 bp were aligned manually using SE-AL version v2.0a11 (available at http://tree.bio.ed.ac.uk/software/seal/). 3.3.3 GENETIC ANALYSES In order to assign individuals of the commercially acquired Halocaridina to previously identified genetic groups, their sequences were compared to an existing database of homologous Halocaridina COI sequences (Santos, 2006; Craft et al., 2008). The database contained 580 individuals possessing 221 unique mtDNA COI sequence haplotypes that have been previously identified as clustering into 13 distinct genetic groups (Craft et al., 2008; Table 2). Sample sizes for each of the genetic groups were 6 ? 162 individuals and 2 ? 67 unique mtDNA COI haplotypes, respectively (Table 2). Each genetic group is characterized as having near or complete genetic isolation as measured via pairwise ?ST statistics and distributions restricted to ~30 km on an island (Craft et al., 2008; Fig. 1). Commercially acquired individuals of Halocaridina were assigned to the previously identified genetic groups via three approaches. Firstly, the total dataset of 676 sequences was analyzed with Collapse v1.2 (available from http://darwin.uvigo.es). This allowed binning of the commercially acquired Halocaridina individuals into two categories: region-specific haplotypes that had been sampled in previous studies and unique haplotypes recovered here. Secondly, haplotype networks were constructed with the program TCS v1.21 (Clement et al., 2000) using the total dataset to determine the relationship among all haplotypes, including novel ones that 52 had not been sampled previously. Thirdly, statistical assignment of the commercially acquired individuals to known genetic groups was conducted with the Segregating Sites Assigner, which utilizes a model based, decision theoretic assignment approach, based on the theory of segregating sites probability, to assign newly sequenced individuals to predefined groups and calculates posterior probability and risk values for each assignment (Abdo & Golding, 2007). In cases where approximately equal posterior probabilities were obtained between two groups, such as the Kona and Ka?? genetic groups due to their shallow genetic divergence, assignment of that individual was based on the lower risk value, which represents a smaller probability the assignment to that particular group is incorrect (Abdo & Golding, 2007). 3.4 RESULTS The number of COI haplotypes recovered per group of commercially acquired Halocaridina ranged from 2 ? 18 (Table 1). In total, 51 haplotypes were identified from the 96 individuals examined here (Table 1). Of these, 17 were identical across 630 bp of sequence to haplotypes previously recovered from the Kona, Ka?? (the western and southern coasts, respectively, of the island of Hawai?i), or Kina?u (the southern coast of the island of Maui) genetic groups of Halocaridina (Supplementary Table S1; Craft et al., 2008). The remaining 35 novel haplotypes also belonged to one of these three genetic groups, with network analyses nesting each within one of them (Fig. 2). In all cases, novel haplotypes differed by no more than 4 bp from haplotypes previously recovered in field studies (Fig. 2; Santos, 2006; Craft et al., 2008). Representative sequences of these unique Halocaridina COI haplotypes were deposited into GenBank under Accession Numbers EU784700 ? EU784734. 53 Statistical assignment of the commercially acquired Halocaridina specimens further supported their origin from the Kona, Ka??, or Kina?u genetic groups of these shrimp since all 96 animals were placed within one of these three groups. Overall, the majority of individuals (90 of 96; ~94%) were assigned either to the Kona or Ka?? genetic groups with mean risk values of 0.004 and 0.002, respectively (Table 3). Of these 90 individuals, 39 and 51 belonged to the Kona and Ka?? genetic groups, respectively (Table 3). The remaining six individuals, originating from two of the three enclosed ecosystems, were assigned to the Kina?u genetic group with mean risk values of 0.001 (Table 3). With the exception of two cases, each batch of commercially acquired Halocaridina appeared to be a mixture of haplotypes from two of the three genetic groups, with the majority of individuals coming from a single genetic group (Table 3). Congruent results to the above were obtained from statistical assignment tests conducted with the program BPSI2.0 (Zhang & Savolainen 2008). 3.5 DISCUSSION The data presented here demonstrate the utility of molecular genetics as a tool for identifying source populations of a commercially traded aquarium ornamental species. To the best of our knowledge, this study constitutes the first such application of ?wildlife forensics? to an invertebrate popular in the aquarium trade. Although we focused on Halocaridina from the Hawaiian Archipelago, these techniques can be applied to other organisms, either aquarium ornamental species or broader ranging, since they are dependent on only a few caveats. Firstly, markers possessing the desired level of genetic resolution (i.e., species, population and/or individual) are necessary for the organism of interest. Potential loci for taxonomic classification include genes and spacers of the nuclear rDNA operon (reviewed by Hillis & Dixon, 1991; 54 Coleman, 2003) or organellar ?barcoding? genes such as mitochondrial COI (e.g., Herbert et al., 2003) and chloroplastic matk (Dawnay et al., 2007). The identification of populations and/or individuals, on the other hand, can also be accomplished with organellar data (as done in this study) or via high-resolution fragment-based (e.g., microsatellite loci, amplified fragment length polymorphisms ((AFLP; Vos et al., 1995), etc) methods. Secondly, a database of ?references? is required to facilitate assignment of ?unknown? individuals (i.e., Dawnay et al., 2007) and this may be generated from direct studies and/or obtained from electronic resources such as GenBank. Lastly, knowledge of the spatial distribution of genetic variation for the targeted group is needed for assignments to be done in a statistical framework. In our case, the high genetic diversity and strong population structure previously found in Halocaridina across the Hawaiian Islands (Santos, 2006; Craft et al., 2008) allowed assignment of each individual examined here to a previously recognized genetic group. Thus, similar results should be expected from any species whose populations exhibit some level of genetic structuring between them. In this study, the bulk of commercially acquired Halocaridina specimens originated from genetic groups endemic to the island of Hawai?i. Since ~300 of the > 520 anchialine habitats in the Hawaiian Archipelago occur on this island (Maciolek & Brock, 1974; Brock et al., 1987), we anticipated Hawai?i to be the primary source of Halocaridina in the aquarium trade. However, while 39 specimens belonged to the Kona genetic group, which was consistent with our hypothesis, a nearly equivalent number (51 of 90; ~57%) were assigned to the Ka?? genetic group of these shrimp (Table 3). Likewise, when placed in the context of vendors, Halocaridina from three of the six originate from this genetic group (Table 3). This was unexpected because in contrast to anchialine habitats along the Kona coast (see Introduction), those inhabited by the 55 Ka?? genetic group are located 10 ? 15 km from the nearest major road, typically accessible only by foot or all-terrain vehicle, and > 100 km from an airport. Thus, significant investments of time and resources are being made in acquiring animals from this region of the island of Hawai?i. Although the majority of Halocaridina are being supplied from populations on the island of Hawai?i, individuals from two of the three enclosed ecosystems were assigned to the Kina?u genetic group from the island of Maui. This was surprising since relative to Hawai?i, the number of anchialine habitats on Maui and O?ahu are much fewer and restricted to small areas on these islands? coasts (Fig. 1). For these reasons, the opportunity to harvest Halocaridina on Maui and O?ahu is more limited. Furthermore, most of Maui?s anchialine habitats reside within the boundaries of either parks or natural area reserves administered by the State of Hawai?i. For example, the Kina?u genetic group of Halocaridina has only been surveyed from the Ahihi- Kina?u Natural Area Reserve (Maciolek, 1986) and the immediate area around it (Craft et al., 2008). This suggests that some commercially available specimens of Halocaridina may be originating from habitats designated for conservation and protected by state laws. Though our data implies a potential case of poaching, a cautionary interpretation is warranted due to the fact that this genetic group is also found on properties adjacent to the reserve and collections may be taking place there. Thus, to confirm whether poaching is occurring within the Ahihi-Kina?u Natural Area Reserve will require monitoring of the area for illegal activities. In either case, however, identifying the Kina?u genetic group as a source of Halocaridina in the aquarium trade indicates that commercial harvesting of these shrimp is not restricted to only the island of Hawai?i. Our finding that commercially available Halocaridina are from genetic groups spanning three geographic regions on two islands implies that multiple collectors may be involved in 56 harvesting for the aquarium trade. Multiple collectors also appear to be supplying some vendors, as is the case with animals being marketed as part of the enclosed ecosystems, where a single unit can contain individuals belonging to genetic groups from different islands (Table 3). Although our data suggest that the other vendors are also acquiring Halocaridina from multiple genetic groups (Table 3), this may not be the situation since low levels of gene flow have been measured between the Kona and Ka?? genetic groups in the form of the sporadic recovery of haplotypes outside their respective regions (Santos, 2006; Fig. 2). For this reason, these outliers within a group of animals from a single vendor may represent migrants between the genetic groups rather than collections from multiple source populations. Although overexploitation is considered one of the greatest threats to the survival of vulnerable populations (e.g., Manel, Berthier & Luikart, 2002), the negative impacts of other anthropogenic activities also need to be taken into consideration. For anchialine habitats around the world, these include habitat destruction, vandalism, and groundwater contamination (reviewed by Iliffe, 2002; 2003). Issues such as these are exemplified in the Hawaiian Archipelago, where numerous anchialine habitats have been lost or modified due to urbanization as well as by the introduction of exotic species (Maciolek & Brock, 1974; Bailey-Brock & Brock, 1993; Brock & Bailey-Brock, 1998). To mitigate this, the State of Hawai?i and private entities have established reserves in order to protect anchialine habitats and their endemic biota (reviewed by Santos, 2006). This study provides an additional resource in these efforts by identifying the geographic regions serving as sources of Halocaridina in the aquarium trade. However, while such information can help to focus conservation planning, it must be remembered that regulating the collection and trade of species presents great challenges (e.g., Wasser et al., 2004) and other alternatives may be preferable. While Halocaridina is currently 57 being acquired from wild populations, as evident by the yearly issuing of commercial harvesting permits (State of Hawai?i Division of Aquatic Resources, personal communication), these shrimp can be readily propagated in captivity (Couret & Wong, 1978; Bailey-Brock & Brock, 1993; personal observation). This offers the opportunity for aquarium-traded specimens to one day be solely supplied at sustainable yields via captive husbandry. For this reason, development of commercially viable aquaculture techniques should be pursued for Halocaridina since such a practice would satisfy growing demands and promote commerce while simultaneously alleviating pressures on natural populations and their anchialine habitats in the future. 3.6 ACKNOWLEDGEMENTS We would like to thank Dr. Robert A. Kinzie III (University of Hawai?i at Manoa), Mike N. Yamamoto, Tom Iwai (State of Hawai?i Division of Aquatic Resources - O?ahu ) and three anonymous reviewers for valuable suggestions that improved this work. This is contribution #39 to the Auburn University (AU) Marine Biology Program. 58 3.7 REFERENCES Abdo, Z. & G. Golding, 2007. A step toward barcoding life: A model-based, decision-theoretic method to assign genes to preexisting species groups. Systematic Biology 56: 44-56. Andrews, C. 1990. The ornamental fish trade and fish conservation. Journal of Fish Biology 37: 53?59. Bailey-Brock, J.H. & R. E. Brock, 1993. Feeding, reproduction, and sense organs of the Hawaiian anchialine shrimp Halocaridina rubra (Atyidea). Pacific Science 47: 338?355. Brock, R.E., J.E. Norris, D.A. Ziemann, & T.M. Lee, 1987. Characteristics of water quality in anchialine ponds of the Kona, Hawaii, coast. Pacific Science 41: 200?208. Brock, R.E. & J.H. Bailey-Brock, 1998. A unique anchialine pool in the Hawaiian Islands. International Review of Hydrobiology 83: 65?75. Chapman, D.D., D.L. Abercrombie, C.J. Douady, E.K. Pikitch, M.J. Stanhope, & M.S. Shivji, 2003. A streamlined, bi-organelle, multiplex PCR approach to species identification: Application to global conservation and trade monitoring of the great white shark, Carcharodon carcharias. Conservation Genetics 4: 415?425. CITES (2008). What is CITES? CITES Secretariat, Geneva. http://www.cites.org/eng/disc/what.shtml accessed 02/28/08. Cipriano, F. & S.R. Palumbi, 1999. Genetic tracking of a protected whale. Nature 397: 307?308. Clapham, P. & K. Van Waerebeek, 2007. Bushmeat and bycatch: the sum of the parts. Molecular Ecology 17: 501?513. Clement, M., Posada, D. & K.A. Crandall, 2000. TCS: a computer program to estimate gene genealogies. Molecular Ecology 9: 1657?1659. 59 Coleman, A.W., 2003. ITS2 is a double edged tool for eukaryote evolutionary comparisons. Trend in Genetics 19: 370?375. Couret, C.L. & D.C.L. Wong, 1978. Larval development of Halocaridina rubra Holthuis (Decapoda, Atyidae). Crustaceana 34: 301?309. Craft, J.D., A.D. Russ, M.N. Yamamoto, T.Y. Iwai, S. Hau, J. Kahiapo, C.T. Chong, S. Ziegler- Chong, C. Muir, Y. Fujita, D.A. Polhemus, R.A. Kinzie & S.R. Santos, 2008. Islands under islands: The phylogeography and evolution of Halocaridina rubra Holthuis, 1963 (Crustacean: Decapoda: Atyidae) in the Hawaiian archipelago. Limnolology and Oceanography 53: 675?689. Dawnay, N., R. Ogden, R. McEwing, G.R. Carvalho & R.S. Thorpe, 2007. Validation of the barcoding gene COI for use in forensic genetic species identification. Forensic Science International 1: 1?6. Folmer, O., Black, M., Hoeh, W., Lutz, R. & Vrijenhoek, R. (1994). DNA primers for amplification of mitochondrial cytochrome c oxidase subunit 1 from diverse metazoan invertebrates. Mol. Mar. Biol. Biotech. 3, 294?299. Frantz, A.C., J.T. Pourtois, M. Heuertz, L. Schley, C. Flamand, A. Krier, S. Bertouille, F. Chaumont & T. Burke, 2006. Genetic structure and assignment tests demonstrate illegal translocation of red deer (Cervus elaphus) into a continuous population. Molecular Ecology 15: 3191?3203. Guglich, E.A., P.J. Wilson & B.N. White, 1994. Forensic application of repetitive DNA markers to the species identification of animal tissues. Journal of Forensic Science 39: 353?361. Herbert, P.D.N., A. Cywinska, S.L. Ball, & J.R. de Warrd, 2003. Biological identifications through DNA barcodes. Proceedings of the Royal Society of London B 270: 313?321. 60 Heerbrandt, T.C. & J.D. Lin, 2006. Larviculture of red front shrimp, Caridina gracilirostris (Atyidae, Decapoda). Journal of World Aquaculture Society 37: 186?190. Hillis, D.M. & M.T. Dixon, 1991. Ribosomal DNA- molecular evolution and phylogenetic inferences. Quarterly Review of Biology 66: 410?453. Hoelzel, A.R. 2001. Shark fishing in fin soup. Conservation Genetics 2: 69?72. Holthuis, L.B. 1963. On red coloured shrimps (Decapoda, Caridea) from tropical land-locked saltwater pools. Zoologische Mededelingen 16: 261?279. Holthuis, L.B. 1973. Caridean shrimps found in land-locked saltwater pools at four Indo-West Pacific Localities (Sinai Peninsula, Funafuti Atoll, Maui and Hawai'i Islands) with the description of one new genus and four new species. Zoologische Mededelingen 128:, 1? 48. Iliffe, T.M. 2002. Conservation of anchialine cave biodiversity. In: Hydrogeology and biology of post-paleozoic carbonate aquifers: 99?102. Martin, J. B., Wicks, C. M., Sasowsky, I.D. (Eds.). Charles Town, WV: Karst Waters Institute. Iliffe, T.M. 2003. Submarine caves and cave biology of Bermuda. NSS News: 217?224. Kyle, C.J. & C.C. Wilson, 2007. Mitochondrial DNA identification of game and harvested freshwater fish species. Forensic Science International 166: 68?76. Lahaye, R., M. van der Bank, D. Bogarin, J. Warner, F. Pupulin, G. Gigot, O. Maurin, S. Duthoit, T.G. Barraclough & V. Savolainen, 2008. DNA barcoding the floras of biodiversity hotspots. Proceedings of the National Academy of Science, USA DOI: 10.1073/pnas.0709936105. Maciolek, J.A. 1983. Distribution and biology of Indo-Pacific insular hypogeal shrimp. Bulletin of Marine Science 33: 606?618. 61 Maciolek, J.A. 1986. Environmental features and biota of anchialine pools on Cape Kinau, Maui, Hawai'i. Stygologia 2: 119?129. Maciolek, J.A. & R.E. Brock, 1974. Aquatic survey of the Kona coast ponds, Hawai'i Islands. Sea Grant Advisory Report AR-74-04. Malik, S., P.J. Wilson, R.J. Smith, D.M. Lavigne & B.N. White, 1997. Pinniped penises in trade: A molecular-genetic investigation. Conservation Biology 11: 1365?1374. Manel, S., P. Berthier & G. Luikart, 2002. Detecting wildlife poaching: identifying the origin of individuals with Bayesian assignment test and multilocus genotypes. Conservation Biology 16: 650?659. Millions, D.G. & B.J. Swanson, 2006. An application of Manel's model: Detecting bobcat poaching in Michigan. Wildlife Society Bulletin 34: 150?155. Palumbi, S.R. & F. Cipriano, 1998. Species identification using genetic tools: The value of nuclear and mitochondrial gene sequences in whale conservation. Journal of Heredity 89: 459?464. Russello, M.A., C. Hyseni, J.P. Gibbs, S. Cruz, C. Marquez, W. Tapia, P. Velensky, J.R. Powell & A. Caccone, 2007. Lineage identification of Galapagos tortoises in captivity worldwide. Animal Conservation 10: 304?311. Santos, S.R., 2006. Patterns of genetic connectivity among anchialine habitats: a case study of the endemic Hawaiian shrimp Halocaridina rubra on the island of Hawai?i. Molecular Ecology 15: 2699?2718. Stam, W.T., J.L. Olsen, S.F. Zaleski, S.N. Murray, K.R. Brown & L.J. Walters, 2006. A forensic and phylogenetic survey of Caulerpa species (Caulerpales, Chlorophyta) from the Florida 62 coast, local aquarium shops, and e-commerce: Establishing a proactive baseline for early detection. Journal of Phycology 42: 1113-1124. Vos, P., R. Hogers, M. Bleeker, M. Reijans, T. van de Lee, M. Hornes, A. Frijters, J. Pot, J. Pelemand, M. Kulper & M. Zabeau, 1995. AFLP: A new technique for DNA fingerprinting. Nucleic Acids Research 23: 4407?4414. Wabnitz, C., M. Taylor, E. Green & T. Razak, 2003. From Ocean to Aquarium: the global trade in marine ornamental species UNEP-WCMC, Cambridge, UK. Wasser, S.K., C. Mailand, R. Booth, B. Mutayoba, E. Kisamo, B. Clark & M. Stephens, 2007. Using DNA to track the origin of the largest ivory seizure since the 1989 trade ban. Proceedings of the National Academy of Science, USA 104: 4228?4233. Zhang, A.B. & P. Savolainen, 2008. BPSI2.0: a C/C++ interface program for species identification via DNA barcoding with a BP-neural network by calling the Mathlab engine. Mol. Ecol. Resources DOI: 10.1111/j.1755-0998.2008.02372.x 63 Table 1. Numbers of individuals and unique haplotypes recovered for the commercially acquired Halocaridina specimens examined in this study. Location n nh Eco_1 Georgia 4 4 Eco_2 Alabama 3 2 Eco_3 Arizona 4 4 FUKU Hawai?i, Hawai?i 13 11 MP O?ahu , Hawai?i 14 10 SK O?ahu , Hawai?i 10 6 STPT Hawai?i, Hawai?i 31 18 STY Hawai?i, Hawai?i 17 13 Total 96 52 n = number of sampled individuals; nh = number of recovered haplotypes. 64 Table 2. Numbers of individuals and unique haplotypes for each of the known genetic groups of Halocaridina (modified from Craft et al. (2008)). Island Genetic Group n nh Hawai?i Puna 61 33 South Hilo 32 11 Na?alehu 30 5 Kona 162 67 Ka?? 50 26 Maui Kinau 56 24 Hana 29 13 Oahu Ewa 33 13 Kalaeloa 33 5 Waianae 28 7 Kahuku 30 8 Lanikai 6 2 Kapapa 30 7 Total 580 221 n = number of sampled individuals; nh = number of recovered haplotypes. 65 Table 3. Genetic group identification of commercially acquired Halocaridina specimens examined in this study based on haplotype network analysis and the Segregating Sites Assigner (Abdo and Golding, 2007). Kona Ka?? Kina?u n TCS pp Risk TCS pp Risk TCS pp Risk Eco_1 4 1 0.47175 0.00337 - - - 3 0.71945 0.00116 Eco_2 3 - - - - - - 3 0.71643 0.00120 Eco_3 4 4 0.47504 0.00355 - - - - - - FUKU 13 2 0.41735 0.00374 11 0.47741 0.00243 - - - MP 14 13 0.47788 0.00331 1 0.47338 0.00252 - - - SK 10 2 0.47243 0.00421 8 0.47189 0.00126 - - - STPT 31 2 0.47844 0.00375 29 0.47666 0.00204 - - - STY 17 15 0.48512 0.00334 2 0.47524 0.00295 - - - Total 96 39 0.47972 0.00375 51 0.47595 0.00200 6 0.71794 0.00118 n = number of sampled individuals; pp = posterior probability of assignment 66 Figure 1. Localities of anchialine habitats (shaded gray) on the Hawaiian islands of O?ahu, Maui and Hawai?i and the range distributions of the 13 genetic groups of Halocaridina. See text for additional details and references. 67 Figure 2. Haplotype network depicting the relationships between cytochrome oxidase subunit I (COI) haplotypes previously recovered from the Kona, Ka??, and Kina?u genetic groups of Halocaridina on Hawai?i and Maui (see Santos 2006 and Craft et al. 2008) and haplotypes recovered from commercially acquired Halocaridina specimens examined in this study (see Supplementary Table S1 for exact haplotype identifications). Numbers within circles represent the number of commercial specimens from which that haplotype was recovered if greater than a single individual. Rectangles represent the haplotype with the highest outgroup probability in each network and the size of circles and rectangles is proportional to the frequency at which a haplotype was recovered. Black dots in a network represent unsampled (i.e., missing) haplotypes. Despite variable lengths, each branch implies a single mutational difference between haplotypes. See Santos (2006) and Craft et al. (2008) for additional details on the population structure and evolution of Halocaridina in the Hawaiian Archipelago. 68 Supplementary Table S1. Posterior probability and risk for the assignment of each commercially acquired Halocaridina specimen to the Kona, Ka?? (islands of Hawai?i) or Kina?u (islands of Maui) genetic groups of these shrimp. Previously recovered haplotypes and novel haplotypes are indicated, along with their corresponding Genbank accession numbers; pp = posterior probability. Kona Ka?? Kina?u Sample pp Risk pp Risk pp Risk Identical to field collected haplotype: Acession Number: New haplotype: New Acession Number: ECO1_1 0.47175 0.00337 0.47338 0.00168 0.05487 0.04650 K2 DQ399170 - - ECO1_2 0.21305 0.04005 0.04826 0.04682 0.73869 0.00167 HM20 EF173774 - - ECO1_3 0.23984 0.03740 0.06015 0.04465 0.70001 0.00048 - - HM25 EU784700 ECO1_4 0.22640 0.03806 0.05394 0.04494 0.71966 0.00134 HM5 EF173759 - - ECO2_1 0.23422 0.03897 0.05580 0.00425 0.70998 0.00092 - - HM26 EU784701 ECO2_2 0.22640 0.03806 0.05394 0.04494 0.71966 0.00134 HM5 EF173759 - - ECO2_3 0.22640 0.03806 0.05394 0.04494 0.71966 0.00134 HM5 EF173759 - - ECO3_1 0.47772 0.00417 0.47202 0.00252 0.05026 0.04834 - - K91 EU784702 ECO3_2 0.47175 0.00421 0.47338 0.00252 0.04587 0.04810 K21 DQ399189 - - ECO3_3 0.46812 0.00169 0.46974 0.00169 0.06213 0.04553 K7 DQ399175 - - ECO3_4 0.48255 0.00413 0.46132 0.00257 0.05612 0.04804 - - K92 EU784703 FUKU1 0.47175 0.00412 0.47338 0.00252 0.05487 0.04490 K24 DQ399192 - - FUKU2 0.47432 0.00335 0.46866 0.00169 0.05702 0.04639 - - K93 EU784704 FUKU3 0.46678 0.00425 0.48411 0.00246 0.04902 0.04840 - - K94 EU784705 FUKU4 0.47175 0.00421 0.47338 0.00252 0.05487 0.04810 - - K95 EU784706 FUKU5 0.46354 0.00342 0.48074 0.00165 0.05572 0.04646 K69 DQ399237 - - FUKU6 0.45992 0.00344 0.47699 0.00166 0.06309 0.04451 - - K96 EU784707 FUKU7 0.48123 0.00497 0.46618 0.00340 0.05259 0.04822 - - K97 EU784708 FUKU8 0.47396 0.00589 0.48648 0.00410 0.03956 0.05215 - - K98 EU784709 FUKU9 0.46354 0.00342 0.48074 0.00165 0.05572 0.04646 K69 DQ399237 - - FUKU10 0.47454 0.00335 0.45367 0.00348 0.07179 0.04409 K12 DQ399180 - - FUKU12 0.46354 0.00342 0.48074 0.00165 0.05572 0.04646 K69 DQ399237 - - FUKU13 0.46812 0.00339 0.46974 0.00169 0.06213 0.04297 K83 DQ399251 - - FUKU14 0.47126 0.00592 0.48371 0.00412 0.04503 0.05185 - - K99 EU784710 69 MP1 0.47454 0.00335 0.45367 0.00348 0.07179 0.04409 K12 DQ399180 - - MP2 0.48255 0.00247 0.46132 0.00257 0.05612 0.04644 - - K100 EU784711 MP3 0.47175 0.00337 0.47338 0.00168 0.05487 0.04650 K2 DQ399170 - - MP4 0.48255 0.00413 0.46123 0.00257 0.05612 0.04804 - - K92 EU784703 MP5 0.46812 0.00169 0.46974 0.00169 0.06213 0.04553 K7 DQ399175 - - MP6 0.47454 0.00251 0.45367 0.00261 0.07179 0.04253 - DQ399217 - - MP7 0.47454 0.00335 0.45367 0.00348 0.07179 0.04409 K12 DQ399180 - - MP8 0.51004 0.00548 0.43064 0.00454 0.05932 0.04788 - DQ399235 - - MP9 0.46812 0.00169 0.46974 0.00169 0.06213 0.04553 K7 DQ399175 - - MP10 0.47175 0.00421 0.47338 0.00252 0.05487 0.04810 - - K95 EU784706 MP11 0.47454 0.00335 0.45367 0.00348 0.07179 0.04409 K12 DQ399180 - - MP12 0.48600 0.00410 0.46458 0.00256 0.04947 0.04838 - - K101 EU784712 MP13 0.47053 0.00507 0.46492 0.00341 0.06454 0.04602 - - K102 EU784713 MP14 0.47454 0.00251 0.45367 0.00261 0.07179 0.04253 K49 DQ399217 - - SK1 0.47053 0.00422 0.46492 0.00256 0.06455 0.04602 - - K103 EU784714 SK2 0.47432 0.00419 0.46866 0.00254 0.05702 0.04799 - - K104 EU784715 SK3 0.46354 0.00342 0.48074 0.00165 0.05572 0.04646 K69 DQ399237 - - SK4 0.46812 0.00254 0.46974 0.00084 0.06213 0.04455 K13 DQ399181 - - SK5 0.46354 0.00342 0.48074 0.00165 0.05572 0.04646 K69 DQ399237 - - SK6 0.46812 0.00254 0.46974 0.00084 0.06213 0.04455 K13 DQ399181 - - SK7 0.47053 0.00376 0.46492 0.00170 0.06455 0.04444 - - K105 EU784716 SK8 0.46812 0.00391 0.46974 0.00169 0.06213 0.04614 K85 DQ399253 - - SK9 0.46812 0.00254 0.46974 0.00084 0.06213 0.04455 K13 DQ399181 - - SK10 0.46812 0.00254 0.46974 0.00084 0.06213 0.04455 K13 DQ399181 - - STPT1 0.46354 0.00342 0.48074 0.00165 0.04657 0.04657 K90 DQ399258 - - STPT2 0.47432 0.00352 0.46866 0.00169 0.05702 0.04639 - - K93 EU784704 STPT3 0.47126 0.00507 0.48371 0.00329 0.04503 0.05227 - - K106 EU784717 STPT4 0.46812 0.00254 0.46974 0.00084 0.06213 0.04455 K13 DQ399181 - - STPT5 0.47772 0.00417 0.47202 0.00252 0.05260 0.04834 - - K107 EU784718 STPT7 0.48443 0.00492 0.46928 0.00384 0.04629 0.05016 - - K108 EU784719 STPT8 0.48255 0.00329 0.46132 0.00171 0.05612 0.04644 - - K109 EU784720 STPT9 0.47999 0.00582 0.47614 0.00418 0.04385 0.05192 - - K110 EU784721 STPT10 0.46812 0.00254 0.46974 0.00084 0.06213 0.04455 K13 DQ399181 - - STPT11 0.46812 0.00254 0.46974 0.00084 0.06213 0.04455 K13 DQ399181 - - STPT12 0.48255 0.00330 0.46132 0.00171 0.05612 0.04644 - - K109 EU784720 STPT13 0.46812 0.00254 0.46974 0.00084 0.06213 0.04553 K13 DQ399181 - - 70 STPT14 0.46354 0.00342 0.48074 0.00165 0.05572 0.04646 K69 DQ399237 - - STPT22 0.47432 0.00419 0.46866 0.00254 0.05702 0.04799 - - K104 EU784715 STPT23 0.46678 0.00425 0.48411 0.00246 0.49110 0.04840 K70 DQ399238 - - STPT24 0.46812 0.00254 0.46974 0.00084 0.06213 0.04455 K13 DQ399181 - - STPT25 0.46812 0.00391 0.46974 0.00169 0.06213 0.04614 K85 DQ399253 - - STPT26 0.46354 0.00342 0.48074 0.00165 0.05572 0.04657 K69 DQ399237 - - STPT27 0.48255 0.00330 0.46132 0.00171 0.05612 0.04644 - - K111 EU784722 STPT28 0.46354 0.00342 0.48074 0.00165 0.04646 0.04646 K69 DQ399237 - - STPT29 0.45215 0.00613 0.50846 0.00392 0.03934 0.05216 - - K112 EU784723 STPT30 0.45590 0.00347 0.47282 0.00168 0.07128 0.04255 - - K113 EU784724 STPT31 0.46954 0.00342 0.48074 0.00165 0.05572 0.04646 K69 DQ399237 - - STPT32 0.46354 0.00342 0.48074 0.00165 0.05572 0.04646 K69 DQ399237 - - STPT33 0.47658 0.00586 0.48189 0.00413 0.41523 0.05205 - - K114 EU784725 STPT34 0.46354 0.00342 0.48074 0.00165 0.05572 0.04657 K69 DQ399237 - - STPT35 0.46354 0.00342 0.48074 0.00165 0.05572 0.04646 K69 DQ399237 - - STPT36 0.46812 0.00254 0.46974 0.00084 0.06213 0.04553 K13 DQ399181 - - STPT37 0.47772 0.00417 0.47202 0.00252 0.05026 0.04834 - - K115 EU784726 STPT38 0.46354 0.00342 0.48074 0.00165 0.05572 0.04646 K69 DQ399237 - - STPT39 0.46329 0.00601 0.49635 0.00402 0.05211 0.05211 - - K116 EU784727 STY1 0.47454 0.00251 0.45367 0.00261 0.07179 0.04253 K49 DQ399217 - - STY2 0.46812 0.00254 0.46974 0.00084 0.06213 0.04455 K13 DQ399181 - - STY3 0.47454 0.00335 0.45367 0.00348 0.07179 0.04409 K12 DQ399180 - - STY4 0.49540 0.00403 0.44699 0.00353 0.05762 0.04957 - - K117 EU784728 STY5 0.46354 0.00256 0.48073 0.00248 0.05572 0.04646 - - K118 EU784729 STY6 0.49140 0.00406 0.44338 0.00444 0.06522 0.04599 - - K119 EU784730 STY7 0.48596 0.00410 0.46458 0.00255 0.04947 0.04838 - - K120 EU784731 STY8 0.47454 0.00419 0.45367 0.00261 0.07179 0.04409 K8 DQ399176 - - STY9 0.47454 0.00419 0.45367 0.00261 0.07179 0.04409 K8 DQ399176 - - STY10 0.47175 0.00337 0.47338 0.00168 0.05487 0.04650 K2 DQ399170 - - STY11 0.50184 0.00477 0.44091 0.00536 0.05724 0.04798 - - K121 EU784732 STY13 0.47772 0.00417 0.47202 0.00252 0.05026 0.04834 - - K122 EU784733 STY14 0.49140 0.00487 0.44338 0.00355 0.06522 0.05989 K18 DQ399186 - - STY17 0.47454 0.00335 0.45367 0.00348 0.07179 0.04409 K12 DQ399180 - - STY18 0.49140 0.00487 0.44338 0.00355 0.06522 0.05989 K18 DQ399186 - - STY19 0.49140 0.00487 0.44338 0.00355 0.06522 0.05989 K18 DQ399186 - - STY20 0.50580 0.00553 0.42706 0.00457 0.06714 0.04432 - - K123 EU784734 71 CHAPTER 4. The long and short of it: Genetic variation and population structure of the anchialine atyid shrimp Caridina rubella on Miyako-jima, Japan 4.1 ABSTRACT One of the most threatened ecosystems on many islands may be anchialine habitats, or coastal land-locked water bodies with no surface connection to the sea yet containing brackish water that fluctuates with the tides. To better manage these habitats, it is important to develop a broader understanding of the biodiversity within them since such knowledge plays critical roles in establishing conservation strategies. In this study, the genetic variation and population structure of an anchialine atyid shrimp, Caridina rubella Fujino and Shokita, 1975, was investigated in the Southern Ryukyu Archipelago, Japan. Given a planktotrophic larval stage and its potential amphidromous life cycle, populations of C. rubella on the island of Miyako- jima (to which it is apparently restricted) were hypothesized to have little to no structure across the island. To test this, 61 individuals were collected from four anchialine caves and sequence variation examined at the cytochrome c oxidase subunit I (COI) region of the mitochondrial DNA (mtDNA). Surprisingly, significant genetic structure was exhibited across distances ranging from < 20 m to > 10 km. Additionally, deep (~17% p-distance) genetic divergence correlating with distinct variation in rostrum lengths was found between closely situated, but more-or-less completely isolated, populations. This implies ?C. rubella? may actually represent two distinct species on Miyako-jima. Given that this atyid is already listed as a threatened species by the Japanese government, the results presented here are useful in the formulation and implementation of future conservation plans for C. rubella populations on Miyako-jima. 72 4.2 INTRODUCTION Global climate change, the introduction of invasive species, and habitat destruction have become significant threats to biodiversity around the world. This loss of biodiversity is occurring at such a rate that there is the potential risk of losing numerous ecosystems without knowing the full nature and extent of the biodiversity contained within them. One of the clearest demonstrations of ecosystem degradation by human impact can be seen on island systems, which intrinsically have much higher extinction rates than continental areas (Gerlach, 2008; Gillespie et al., 2008). Given this, it is not surprising that nine of the top 25 biodiversity ?hotspots? for conservation priorities are predominately, or completely, made up of islands (Myers et al., 2000). One of most threatened habitats on many tropical islands may be the anchialine ecosystem (Sket, 1996; Iliffe, 2002; 2003; Santos, 2006). Anchialine habitats are coastal land- locked bodies of water with no surface connection to the sea yet contain salt or brackish water that fluctuate with the tides due to subterranean connections to the ocean (Holthuis, 1973; Maciolek, 1983). While such environments occur in the Sinai Peninsula, Bermuda, the Caribbean, the Hawaiian Islands, the South Pacific, the Philippines, and the Ryukyus, only ~1,000 habitats fitting this ecosystem definition have been reported globally (Maciolek, 1986). Unfortunately, many of these habitats have experienced significant negative impacts worldwide from a variety of anthropogenic sources such as habitat destruction, vandalism, and groundwater contamination (Iliffe, 2002, 2003; Santos, 2006). This is alarming since despite the large number of organisms inhabiting anchialine environments, such as mollusks and crustaceans (Brock & Kam, 1987), only a few of the ~560 currently described species endemic to this ecosystem (Iliffe, 2002; ISI Web of Science search, January 2011) have been examined to date from a population genetic perspective. For some of these cases, little to no genetic structure was 73 identified among populations separated by ~200 km (Kano & Kase, 2004; Hunter et al., 2008; Russ et al., 2010), implying frequent dispersal over long distances. In contrast, a number of other anchialine species have been found to possess exceptional levels of endemism on the scale of a few km (Santos, 2006; Craft et al., 2008; Hunter et al., 2008; Page et al., 2008), presumably due to low dispersal abilities and/or significant barriers to dispersal. Thus, given the potential for strong regional endemism along with ongoing modification and destruction of anchialine habitats around the world, there is now an urgency to accelerate research into these unique island environments and their biota before they are possibly lost forever. The islands of East Asia are among the richest in the world in terms of biodiversity (Meijaard, 2003). Pertinent among the region?s archipelagos, the Ryukyus are a continuous chain of islands situated between Japan?s Kyushu Island and Taiwan. Decapod crustaceans dominate the anchialine biota of the Ryukyus, with at least eleven shrimp and crab species having been documented or formally described (Komai & Fujita, 2005; Cai & Shokita, 2006; Naruse & Tamura, 2006; Fujita, 2007; Fujita & Sunagawa, 2008). Of these, the atyid shrimp, Caridina rubella Fujino & Shokita, 1975 (Decapoda: Atyidae), is thought to have a disjunct distribution throughout the Indo-West Pacific, inhabiting anchialine pools and caves in the Philippines (Cai & Anker, 2004), the Cook Islands (T.J. Page personal communication, 2010), Java in Indonesia (T. von Rintelen personal communication, 2010) as well as the Ryukyus (Cai & Shokita, 2006). In the Ryukyus, however, C. rubella has been reported from just four of the ~55 islands in the archipelago: Okinoerabu-jima, Okinawa-jima, Minamidaito-jima, and Miyako- jima (Fujino & Shokita, 1975; Shokita & Nishijima, 1976). This restricted distribution along with continued habitat loss has led to the listing of C. rubella as a threatened species by the Japanese Ministry of the Environment and Okinawa Prefectural Government (Fujita & Shokita 74 2005; Shokita, 2006). Given this situation, it is important to develop a better biological understanding of C. rubella in the Ryukyus since such information is vital for developing management plans to preserve the species and, by extension, the anchialine ecosystem of these islands. Here, the genetic variation, population structure and demography of C. rubella from Miyako-jima in the Southern Ryukyus were investigated via sequence analyses of mitochondrial DNA (mtDNA). Specifically, given that its larvae are planktotrophic and possessing a possible amphidromous life cycle (Shokita, 1979; Fujita, unpublished data), both life history traits thought to be significant in facilitating population connectivity of anchialine taxa (Kano & Kase, 2004; Craft et al., 2008; Russ et al., 2010; Cook et al., 2006, 2008), it is hypothesized that C. rubella populations are homogenized among anchialine habitats on the island due to marine and/or groundwater dispersal of larvae and/or adults. 4.3 MATERIAL AND METHODS 4.3.1 BIOLOGICAL MATERIALS Between July and August 2009, ~42 anchialine habitats were surveyed from nine islands (including Okinawa-jima, Minamidaito-jima, and Miyako-jima) belonging to the Okinawa, Miyako, and Yaeyama island groups of the Ryukyus (Fig. 1). From these surveys, adult specimens of C. rubella were only found to inhabit anchialine habitats on Miyako-jima. All sites surveyed were classic examples of anchialine habitats occurring within porous limestone basins, being located < 0.95 km from the coast and showing tidal fluctuations with water depths, salinities, and surface areas ranging between 0.001-2 m, 1-15?, and ~3 -> 50 m3, respectively. Of these, the species was found at five of seven sites. Due to the close proximity (~25 m) of two of these five habitats, only four were sampled for this study (Fig. 1). However, at one locality 75 (Ana Ga; ANA), samples were obtained from two ponds separated by < 20 m (ANA1 and ANA2). Notably, individuals of C. rubella from ANA2 possessed noticeably longer rostrums relative to individuals from the other populations sampled on the island (see below). Given this, rostrum length, as measured from the end of the eye stock to the end of the rostrum, was recorded for all collected individuals. Unfortunately, due to damage following preservation and transport, only five out of the eight individuals collected from the ANA2 population could be measured for rostrum length. Between eight (8) and 16 individuals of C. rubella were collected from each population (Table 1) using baited traps or small aquarium nets and immediately preserved in 99% ethanol for genetic analyses. 4.3.2 DNA EXTRACTION, POLYMERASE CHAIN REACTION (PCR) AND SEQUENCING Total genomic DNA was extracted from each individual following procedures outlined in Santos (2006) and utilized as template to amplify a ~670 base pair (bp) fragment of the mtDNA cytochrome c oxidase subunit I (COI) gene via the polymerase chain reaction (PCR). Reactions were carried out in 25 ?L volumes containing ~10-30 ng of template DNA, 10 mM Tris-HCl (pH 8.3), 50 mM KCl, 0.001% gelatin, 2.0 mM MgCl2, 200 ?M of each deoxynucleotide triphosphate (i.e., dATP, dCTP, dGTP, and dTTP), 1 U Taq polymerase, and 0.4 ?M each of primers LCO1490 and HCO2198 (Folmer et al., 1994). Reactions were performed in a PTC-100 thermocycler (MJ Research, Watertown, MA, USA) following Santos (2006) and amplicons purified with Montage PCR Filter Units (Millipore, Billerica, MA, USA) according to the supplier?s protocol. Purified amplicons were then cycle-sequenced in both directions using Big- Dye Terminators, and read on a PRISM 3100 Genetic Analyzer (Applied Biosystems, Foster City, CA, USA). Ambiguities in the chromatograms were corrected by comparison to the 76 complementary DNA strand in Sequencher v4.7 (Gene Codes Corporation, Ann Arbor, MI, USA). Completed sequences of 630 bp were aligned manually using SE-AL version v2.0a11 (available at http://tree.bio.ed.ac.uk/software/seal/). 4.3.3 GENETIC DIVERSITY AND PARSIMONY NETWORK ANALYSES To assess the genetic diversity within each population of C. rubella, nucleotide (?) and haplotype (h) diversity indices were calculated according to the methods of Nei (1987) using DnaSP v5.10.00 (Rozas et al., 2009). The relationships among COI haplotypes of C. rubella were visualized via statistical parsimony networks constructed with TCS v1.21 (Clement, Posada & Crandall, 2000). This program utilizes the cladogram estimation algorithm of Templeton et al. (1992) and was conducted using the default settings, which provide the 95% parsimoniously plausible branch connections between haplotypes. Reticulations in the network, representing ambiguous connections, were resolved using the criteria outlined in Crandall et al. (1994). 4.3.4 Population Structure, Migration, and Demographic Analyses To test for genetic differentiation between our populations, pairwise ?ST statistics (based on haplotype frequency and molecular divergence) were conducted with Arlequin v3.11 (Excoffier & Schneider, 2005). Hudson?s (2000) nearest-neighbor statistic (Snn) was also calculated to test for genetic differentiation between populations using DnaSP. This statistic is a measure of how often the ?nearest neighbors? (in sequence space) are from the same locality and performs well in cases of low sample size (Hudson, 2000). Statistical significance of the pairwise ?ST and Snn analyses were assessed with 10,000 and 1,000 permutations in Arlequin and DnaSP, respectively. Arlequin was also used to perform an analysis of molecular variance 77 (AMOVA; Excoffier et al. 1992) to determine the manner in which genetic variation was partitioned within C. rubella of Miyako-jima. For the AMOVA, ?-statistics were used to estimate the relative contribution of molecular variance at three hierarchical levels: among geographic regions (?CT), among populations within a region (?SC) and within populations (?ST) and significance assessed by 10,000 permutations. For the pairwise ?ST statistics and AMOVA, the Tamura & Nei (1993) model of evolution [selected by the Akaike information criterion (AIC) using the FindModel web version of Modeltest (Posada & Crandall, 1998; available at http://darwin.uvigo.es/)] was utilized. To estimate potential migration between populations of C. rubella, the web version of MDIV (Nielsen & Wakeley, 2001; available at http://cbsuapps.tc.cornell.edu) was utilized under a finite-site mutation model, which accounts for multiple mutations per site, nucleotide frequencies differences and transition/transversion biases potentially present in sequence data. For these calculations, three independent runs using identical starting conditions (Mmax = 50, Tmax = 10, length of Markov chain = 2 ? 106 cycles, burn-in time = 5 ? 105 cycles) were done with different random start seeds to check for consistency in the estimates. The M values with the highest posterior probabilities were accepted as the best estimates of migration rate per generation, as per Nielsen & Wakeley (2001). Lastly, Mantel tests (Mantel, 1967), implemented in Alleles In Space (AIS; Miller 2005), were performed to test for correlations between geographic and genetic distance (isolation by distance) and significance assessed with 10,000 permutations. Tajima?s D (Tajima, 1989) and Fu?s Fs (Fu, 1997) tests were also conducted in DnaSP to determine whether patterns of mitochondrial COI sequence variation in C. rubella were consistent with predictions under the neutral model of evolution. In the absence of selection, 78 both methods can provide potential information on the recent demographic forces affecting a population (bottlenecks or expansions: Tajima, 1989; Fu, 1997, Akey et al. 2004). 4.4 RESULTS 4.4.1 GENETIC DIVERSITY AND PARSIMONY NETWORK ANALYSES OF C. RUBELLA ON MIYAKO-JIMA Direct sequencing of mtDNA COI amplicons yielded a 630 bp fragment from each of the 61 C. rubella individuals included in this study. From these, 24 unique haplotypes were identified (Table 1) and deposited into GenBank under accession #?s JF926126-JF926149. Of the 24 haplotypes recovered, 11 occurred as singletons while the remaining 13 were sampled more than once. One hundred and twenty-one (19.2%) of the 630 bp analyzed were found to be variable across haplotypes and 117 (18.6%) of these were parsimoniously informative. Translation of the nucleotide sequences into amino acids found no stop codons and while the majority of substitutions were ?silent?, one was found to be non-synonymous, with the resulting change to an amino acid with similar biochemical properties (data not shown). This implies that all sequences were derived from mitochondrial copies of COI and not nuclear copies of mitochondrial derived genes (numts; Lopez et al., 1994), which can be common in arthropods (Buhay, 2009). Overall nucleotide and haplotype diversity indices were similar for all populations (Table 1). Parsimony network analysis of the complete COI sequence data set produced two discrete networks (Fig. 2). One corresponded to ?long rostrum? (5.06 ? 2.4 mm, n = 5) C. rubella and consisted of four haplotypes and nine individuals. Eight of these nine individuals were collected from ANA2, while one individual was obtained from ANA1. The second network represented ?short rostrum? (2.55 ? 1.19 mm, n = 53) C. rubella and encompassed 20 haplotypes recovered 79 from 52 individuals collected at the ANA1, BUT, ZUZ, and TAG sites (Fig. 1). Uncorrected (p) distances among haplotypes within each network ranged from 0.20-1.7% while the p-distances between haplotypes in the two networks ranged from 16.3-17.5%. 4.4.2 POPULATION STRUCTURE, MIGRATION, AND DEMOGRAPHIC ANALYSES OF C. RUBELLA ON MIYAKO-JIMA Eight of the possible ten pairwise comparisons revealed significant genetic differentiation between the examined populations of C. rubella. For example, the two close proximity populations of ANA1 and ANA2 exhibited structure approaching the maximum limit of 1.0 (Table 1), implying that nearly no individuals are exchanged between these caves despite being separated by < 20 m (the exception being an apparent dispersal event by a single individual from the ANA2 to the ANA1 site; see above). Along with this, the ANA2 (?long rostrum?) population exhibited similar and significant genetic structure relative to all other populations of C. rubella on Miyako-jima (Table 2). Due to the large genetic divergence of the ANA2 population as well as its occurrence at only a single site, ANA2 was excluded from subsequent comparative analyses. For the remaining populations of ?short rostrum? C. rubella, significant genetic differentiation was observed between sites separated by > 10 km (northern sites compared to southern sites, Table 2) while no significant genetic differentiation was found between populations separated by < 5 km (among northern or southern sites, Table 2). For the AMOVA, populations were grouped by geographic region (northern and southern) based on their proximity to one another. In this context, the AMOVA identified a significant amount of genetic variation as partitioned within individual populations (73%, ?ST = 0.273, P < 0.001) while a large, but non- significant component of genetic variation occurred between geographic regions (25%, ?CT = 80 0.250, P = 0.319). Likewise, the partition of genetic variation among populations within each region was not significant (2.24%, ?SC = 0.030, P = 0.169). To assess potential migration (M) between populations, estimates were generated via MDIV in a pairwise fashion for ?short rostrum? C. rubella (Fig. 3). For populations in close proximity to one another (< 5 km; ZUZ-BUT and ANA1-TAG; Fig. 3A and B), plateaus in the posterior probability distributions of M were readily apparent, implying high migration between these sites and consistent with the lack of genetic structure between them (see above). In contrast, populations separated by geographic distances of ~11-13 km possessed discernable peaks in their posterior probability distributions of M, indicative of lower levels of migration. For example, estimates of M = 0.9 and M = 0.6 were recovered for comparisons of ZUZ/ANA1 and ZUZ/TAG (Fig. 1, Fig. 3A), respectively. A similar result (M = 0.7) was obtained for BUT and TAG (Fig. 3B) while the estimate for BUT and ANA1 (~11 km) was higher (M = 2.5; Fig. 3B). Additionally, the Mantel test identified a significant correlation between the genetic and geographic distances of populations (r = 0.255; P < 0.001), further supporting the conclusions derived from MDIV. Taken together, these results suggest that isolation by distance drives the genetic structure of ?short rostrum? C. rubella on Miyako-jima, with dispersal occurring over short (< 5 km) geographic distances but decreasing with increasing distance between populations. Lastly, both Tajima?s D and Fu?s F were not significant for any of the C. rubella populations examined in this study (Table 1). 81 4.5 DISCUSSION Contrary to the proposed hypothesis, most populations of C. rubella of Miyako-jima exhibit significant genetic structure across scales ranging from < 20 m to > 10 km. Along with this, deep (~17% p-distance) genetic divergence correlating with distinct variation in rostrum lengths was identified between closely situated, but more-or-less completely isolated, populations of this anchialine atyid ?species.? The significance and implications of these results are discussed below. 4.5.1 MORPHOLOGICAL AND GENETIC DIFFERENTIATION IN C. RUBELLA OF MIYAKO-JIMA The relatively high (~17% p-distance) genetic divergence, along with the ease of visually distinguishing ?long rostrum? and ?short rostrum? morphotypes, implies ?C. rubella? of Miyako- jima is actually comprised of two species. Here, the ?short rostrum? morphotype aligns with the original description of the species established by Fujino & Shokita (1975) and suggests that the ?long rostrum? morphotype represents either an undescribed species of Caridina or an already described Caridina sp. not previously recorded from the anchialine habitats of Miyako-jima. While additional morphological studies are required to distinguish between these possibilities, the restricted distribution of the ?long rostrum? morphotype to just a single site on one island should be taken into consideration when developing conservation and management strategies for ?C. rubella? and anchialine habitats on Miyako-Jima and the Ryukyus in general. Although the exact factors leading to the spatial segregation of ?long rostrum? and ?short rostrum? morphotypes have yet to be determined, several hypotheses can be proposed that might explain such a pattern. First, despite being found in close proximity to ?short rostrum? C. rubella, the ?long rostrum? morphotype may be physically restricted to the ANA2 pond due to 82 the geologic characteristics of its particular habitat. Compartmentalization of hypogeal water systems by underground barriers has been shown to be a strong isolating mechanism over short distances for anchialine fauna in the Hawaiian Islands (Santos, 2006; Craft et al., 2008) and may play a similar role on Miyako-jima. If this is the case, studies incorporating isotopic tracer methods (Scholl et al., 1996) could help determine the extent to which the anchialine habitat at ANA2 may be physically disconnected from the rest of the subterranean water system of the island. Second, the nearly complete isolation of the ?long rostrum? morphotype to ANA2 could be due to slight, but significant, differences in reproductive traits. For example, small variations in egg size have been shown to strongly correlate with distinct levels of population structure in other Caridina spp., with large-egged species having more pronounced structure (Page & Hughes, 2007). Thirdly, the two morphotypes could be segregating due to one or more ecological factors, such as dietary differences, other resource partitioning and/or competition. Thus, additional studies designed to specifically test these (or alternative) hypotheses are needed as they may help clarify the biology and taxonomic status of ?C. rubella? on Miyako-Jima. 4.5.2 POPULATION STRUCTURE AMONG ?SHORT ROSTRUM? C. RUBELLA OF MIYAKO-JIMA The isolation by distance revealed here for ?short rostrum? C. rubella was unexpected given the fact that this atyid possesses life history traits conducive to dispersal (see Introduction). For example, C. rubella produces planktotrophic larvae and this reproductive mode is typically thought to contribute to higher connectivity and less structured populations in marine environments due to the ability of larvae to feed and delay metamorphosis while dispersing as plankton (reviewed by Palumbi 1994). Such a scenario apparently holds true for other anchialine species like the neritiliid snail Neritilia cavernicola and alpheid shrimp Metabetaeus lohena 83 Banner & Banner, 1960, both of which are planktotrophic and exhibit little to no genetic differentiation, presumably due to high gene flow, over ranges of ~200 km in the Philippines and Hawaiian Islands, respectively (Kano & Kase, 2004; Russ et al., 2010). Likewise, amphidromy, defined as the active or passive movement of larvae from freshwater taxa to estuarine or marine areas for development before migrating back to adult freshwater habitats, is also thought to facilitate long distance dispersal for a number of organisms, including many atyids (Page et al., 2005, 2008; Cook et al., 2006, 2008, 2009). Taken together, the basic predictions of these life history traits, particularly in their most general forms, are violated by the strong population structure of ?short rostrum? C. rubella from Miyako-jima. Although amphidromous species are thought to possess high dispersal potential, local selection for larval retention could be strong for such species if suitable adult habitat (i.e., anchialine habitats) is rare (Sponaugle et al., 2002; Strathmann et al., 2002). Cases of local retention have been suggested for a number of amphidromous taxa, including gobies (Sorensen & Hobson, 2005) and snails (Haynes, 2000) as well as for the larvae of an amphidromous shrimp (Benstead et al., 2000). In the case of ?short rostrum? C. rubella, retention of larvae in the aquifer system immediately below a particular anchialine habitat and/or in nearshore coastal waters may limit gene flow to local scales, ultimately leading to patterns of isolation by distance. Additionally, the strong population structure seen for ?short rostrum? C. rubella may again be a consequence of egg size (see above). Despite ?short rostrum? C. rubella possessing relatively small eggs (~0.52 mm; Shokita, 1979), this value falls within the range of other Caridina spp. with significant genetic structure between populations in relatively close proximity (Page & Hughes, 2007). Taken together, the local retention of larvae coupled with relative egg size may 84 limit dispersal and contribute to the patterns reported here for Miyako-jima?s ?short rostrum? C. rubella populations. Another potential factor influencing the population structure of ?short rostrum? C. rubella may be the strong oceanic currents circulating past the island. Specifically, Miyako-jima separates two western boundary currents of the North Pacific Subtropical Gyre: the Kuroshio Current of the East China Sea to the west and the Ryukyu Current to the east (Jin et al., 2010). In this context, mesoscale hydrographic features of these currents such as fronts and eddies may act as mechanisms for larval retention, thereby limiting dispersal of species with otherwise high dispersal potential (Sabate`s & Olivar, 1996). For example, dynamic interactions between the Kuroshio Current and shelf waters of the East China Sea have been shown to prevent offshore larval fish dispersal (Okazaki & Nakata, 2007). Alternatively, the strong Kuroshio and Ryukyu Currents may quickly sweep larvae offshore and away from the island, in turn limiting successful dispersal of ?short rostrum? C. rubella around Miyako-jima and producing a similar pattern of isolation by distance. 4.5.3 UNDERGROUND DAMS AND CONSERVATION ASPECTS Approximately 57% of the total land area of Miyako-jima is utilized for agriculture (field crops and animal husbandry). Thus, like most oceanic islands, freshwater is a valuable resource and commodity. The annual rainfall for the island is ~2,300 mm, but in the past the majority of this drained through the porous limestone of the island and either contributed to groundwater resting on the impermeable layer of clay that makes up the structural foundation of Miyako-jima or flowed out to sea (Wisniewski, 2004). To further develop groundwater resources for agriculture, the Japanese government implemented construction of two major subsurface dams 85 (Ishida et al., 2003) in 1987 to simultaneously prevent groundwater from flowing into the sea while raising levels of the underground aquifers to facilitate water extraction by pumping (Fig. 1B). This project was completed in 2001, with a total capacity of ~20,000,000 m3 and providing up to 50,000 m3 of freshwater per day, making it one of the largest subsurface dam systems in the world (Ishida et al., 2003). Given the novelty of subsurface dams, little is known regarding the biological impacts such structures may have on aquifer-associated organisms like those endemic to the anchialine ecosystem. However, the ecological consequences of surface dams have been well documented, including dramatic changes in sedimentation, channelization, flooding and temperature regimes (Proff & Hart, 2002; Hall et al., 2011). Similar effects have already been reported for the subterranean dams on Miyako-jima in the context of overall water quality (increased temperature, depositional rates and dissolved inorganic carbon; Kano et al., 2007; Ishida et al., 2011) due to significant alterations in both groundwater flow and levels. Thus, such alterations could have negative impacts on the already threatened C. rubella of Miyako since increased sedimentation and flow rates have been shown to reduce habitat suitability and alter community assemblages in other carideans (Iwata et al., 2003). However, one direct effect of dams and other impoundments is habitat fragmentation via the erection of artificial barriers. For aquatic organisms in general, such fragmentation can decrease access to suitable habitat, restrict long distance dispersal events and impede seasonal and/or reproductive related migrations, all of which (among others) can contribute to the extirpation of populations (Dehais et al., 2010). For amphidromous species like C. rubella, habitat fragmentation due to dam construction and water withdraw is of particular concern as it can result in a reduction or loss of connectivity between larval (marine) and adult (anchialine) habitats (Cook et al., 2009). In this respect, these 86 underground dams should be recognized as possible (and formidable) barriers to the amphidromous life cycle of C. rubella that may impact the long-term viability of the ?species? on Miyako-jima in multiple ways. 4.6 ACKNOWLEDGEMENT This work was funded in part by a National Science Foundation (NSF) East Asia and Pacific Summer Institutes Fellowship OISE-0913667 (DAW) and NSF DEB-0949855 (SRS). We thank Dr. Timothy J. Page (Griffith University, Australia) and one anonymous reviewer for providing valuable comments that improved the manuscript and Morgan McCarthy for assistance with sample collection. This manuscript represents contributions #81 and #2 to the Auburn University (AU) Marine Biology Program and Molette Biology Laboratory for Environmental and Climate Change Studies, respectively. 87 4.7 REFERENCES Akey, J. M., M.A. Eberle, M.J. Rieder, C.S. Carlson, M.D. Shriver, D.A. Nicerson & L. Kruglyak, 2004. Population history and natural selection shape patterns of genetic variation in 132 Genes. PLoS ONE 2: e286 Banner, A.H. & D.M. Banner, 1960. Contributions to the knowledge of alpheid shrimp of the Pacific Ocean. Part VII. On Metabetaeus Borradaile, with a new species from Hawaii. Pacific Science 14: 299-303. Benstead J.P., J.G. March & C.M. Pringle, 2000. Estuarine larval development and upstream post-larval migration of freshwater shrimps in two tropical rivers of Puerto Rico. Biotropica 32: 545-548. Brock R.E. & A.K.H. Kam, 1987. Biological and water quality characteristics of anchialine resources in Kaloko-Honokohau National Historical Park Technical Report 112: Cooperative National Park Resources Studies Unit, University of Hawaii at Manoa. Buhay, J.E, 2009. ?COI-Like? sequences are becoming problematic in molecular systematic and DNA barcoding studies. Journal of Crustacean Biology 29: 96-110. Cai, Y. & A. Anker, 2004. On a collection of freshwater shrimps (Crustacea Decapoda Caridea) from the Philippines, with descriptions of five new species. Tropical Zoology 17: 233- 266. Cai, Y & S. Shokita, 2006. Atyid shrimps (Crustacea: Decapoda: Caridea) of the Ryukyu Islands, southern Japan, with descriptions of two new species. Journal of Natural History 40: 2123-2172. Clement, M., D. Posada & K.A. Crandall, 2000. TCS: a computer program to estimate gene genealogies. Molecular Ecology 9: 1657-1659. 88 Cook, B.D., A.M. Baker, T.J. Page, S.C. Grant, J.H. Fawcett, D.A. Hurwood & J.M. Hughes, 2006. Biogeographic history of an Australian freshwater shrimp, Paratya australiensis (Atyidae): the role of life history transition in phylogeographic diversification. Molecular Ecology 15: 1083-1093. Cook, B.D., C.M. Pringle & J.M. Hughes, 2008. Molecular evidence for sequential colonization and taxon cycling in freshwater decapod shrimps on a Caribbean island. Molecular Ecology 17: 1066-1075. Cook, B.D., S. Bernays, C.M. Pringle & J.M. Hughes, 2009. Marine dispersal determines the genetic population structure of migratory stream fauna of Puerto Rico: evidence for island-scale population recovery processes. Journal of the North American Benthological Society 28: 709-718. Craft, J.D., A.D. Russ, M.N. Yamamoto, T.Y. Iwai, S. Hau, J. Kahiapo, C.T. Chong, S. Ziegler- Chong, C. Muir, Y. Fujita, D.A. Polhemus, R.A. Kinzie & S.R. Santos, 2008. Islands under islands: The phylogeography and evolution of Halocaridina rubra Holthuis, 1963 (Crustacean: Decapoda: Atyidae) in the Hawaiian archipelago. Limnology and Oceanography 53: 675-689. Crandall, K.A., A.R. Templeton & C.F. Sing, 1994. Intraspecific phylogenetics: Problems and solutions, pp. 273?297. In, R. W. Scotland, D. J. Siebert and D. M. Williams (eds.), Models in Phylogeny Reconstruction. Clarendon Press, Oxford. Dehais C., R. Eudeline, P. Berrebi & C. Argillier, 2010. Microgeographic genetic isolation in chub (Cyprinidae: Squalius cephalus) population of the Durance River: estimating fragmentation by dams. Ecology of Freshwater Fish 19: 267-278. 89 Excoffier, L., G. Laval & S. Schneider, 2005. Arlequin (version 3.0): An integrated software package for population genetics data analysis. Evolutionary Bioinformatics 1: 47-50. Folmer, O., M. Black, W. Hoeh, R. Lutz & R. Vrijenhoek, 1994. DNA primers for amplification of mitochondrial cytochrome c oxidase subunit I from diverse metazoan invertebrates. Molecular Marine Biology and Biotechnology 3: 294-299. Fu, Y., 1997. Statistical tests of neutrality of mutations against population growth, hitchhiking and background selection. Genetics 147: 915-925. Fujino, T. & S. Shokita, 1975. Report on some atyid shrimps (Crustacea, Decapoda, Caridea) from the Ryukyus Islands. Bulletin of Science and Engineering Division, University of the Ryukyus, (Mathematics and Natatural Science) 18: 93-113. Fujino, T. & S. Shokita, 2005. Caridina rubella, Fujino & Shokita, 1975, p. 203-204 (in Japanese). In: Okinawa Prefectural Government (Nature Conservation Division, Department of Cultural and Environmental Affairs) (ed.), Threatened wildlife in Okinawa, 2nd edit. (Annimals): Red Data Okinawa. Okinawa Prefectural Government (Nature Conservation Division, Department of Cultural and Environmental Affairs), Okinawa. Fujita, Y., 2007. Decapod crustaceans inhabiting springs on Miyako Islands, the Ryukyus, Japan. Bulletin of the Miyakojima City Museum. 11, 89-110. [in Japanese with English abstract] Fujita, Y. & H. Sunagawa, 2008. Cavernicolous and terrestrial decapod crustaceans of Tarama- jima Island, the Ryukyu Islands, southwestern Japan. Bulletin of the Miyakojima City Museum. 12, 53-80. [in Japanese with English abstract] 90 Gillespie, R.G., E.M. Claridge & G.K. Roderick, 2008. Biodiversity dynamics in isolated island communities: interaction between natural and human-mediated processes. Molecular Ecology 17: 45-57. Gerlach, J., 2008. Preliminary conservation status and needs of an oceanic island fauna: the case of Seychelles insects. Journal of Insect Conservation 12: 293-305. Hall, C.J., A. Jordaan, & M.G. Frisk, 2011. The historic influence of dams on diadromous fish habitat with a focus on river herring and hydrologic longitudinal connectivity. Landscape Ecology 26: 95-107. Haynes, A., 2000. The distribution of freshwater gastropods on four Vanuatu islands: Espiritu Santo, Pentecost, Efate and Tanna (South Pacific). Annales de Limnologie-International Journal of Limnology 36: 101?111. Holthuis, L. B., 1973. Caridean shrimps found in land-locked saltwater pools at four Indo-West Pacific localities (Sinai Peninsula, Funafuti Atoll, Maui and Hawai?i islands), with the description of one new genus and four new species. Zoologische Verhandelingen (Leiden) 128: 1-48. Hudson, R. R., 2000. A new statistic for detecting genetic differentiation. Genetics 155: 2011- 2014. Hunter, R.L., M.S. Webb, T.M. Iliffe & J.R.A. Bremer, 2008. Phylogeny and historical biogeography of the cave-adapted shrimp genus Typhlatya (Atyidae) in the Caribbean Sea and western Atlantic. Journal of Biogeography 35: 65-75. Iliffe, T.M., 2002. Conservation of anchialine cave biodiversity, pp. 99-102. In, J. B. Martin, C. M. Wicks, and I. D. Sasowsky (eds.), Hydrogeology and Biology of Post-Paleozoic Carbonate Aquifers. Karst Waters Institute, Charles Town, WV. 91 Iliffe, T.M., 2003. Submarine caves and cave biology of Bermuda. NSS News: 217-224. Ishida, S., T. Tsuchihara, S. Yoshimoto, and M. Imaizumi. 2011. Sustainable use of groundwater with underground dams. Japan International Research Quarterly 45: 51-61. Ishida, S., M. Kotoku, E. Abe, M.A. Fazal, T. Tsuchihara & M. Imaizumi, 2003. Construction of subsurface dams and their impact on the environment. Materials and Geoenvironment 50: 149-152. Iwata, T., M. Inoue, S. Nakano, H. Miyasaka, A. Doi & A.P. Covich, 2003. Shrimp abundance and habitat relationships in tropical rain-forest streams, Sarawak, Borneo. Journal of Tropical Ecology 19: 387-395. Jin, B., G. Wang, Y. Liu & R. Zhang, 2010. Interactions between the East China Sea Kuroshio and the Ryukyu Current as revealed by the self-organizing map. Journal of Geophysical Research 115: C12047, doi:10.1029/2010JC006437 Kano, A., R. Hagiwara, T. Kawai, M. Hori & J. Matsuoka, 2007. Climatic conditions and hydrological change recorded in a high-resolution stable isotope profile of recent laminated tufa on subtropical island, Southern Japan. Journal of Sedimentary Research 77: 59-67. Kano, Y. & T. Kase, 2004. Genetic exchange between anchialine cave populations by means of larval dispersal: the case of a new gastropod species Neritilia cavernicola. Zoologica Scripta 33: 423?437. Komai, T. & Y. Fujita, 2005. A new stygiobiont species of Macrobrachium (Crustacea: Decapoda: Caridea: Palaemonidae) from an anchialine cave on Miyako Island,Ryukyu Islands. Zootaxa 1021: 13-27. 92 Kumar, S., K. Tamura & M. Nei, 2004. MEGA3: Integrated software for molecular evolutionary genetics analysis and sequence alignment. Briefings in Bioinformatics 5: 150-163. Lopez, J.V., N. Yuhki, R. Masuda, W. Modi & S.J. O?Brien, 1994. Numt, a recent transfer and tandem amplification of mitochondrial DNA to the nuclear genome of the domestic cat. Journal of Molecular Evolution 39: 174-190. Maciolek, J.A., 1983. Distribution and biology of Indo-Pacific insular hypogeal shrimp. Bulletin of Marine Science 33:606-618. Maciolek, J.A., 1986. Environmental features and biota of anchialine pools on Cape Kinau, Maui, Hawai'i. Stygologia 2: 119-129. Mantel, N. A., 1967. The detection of disease clustering and a generalized regression approach. Cancer Research 27: 209-220. McDowall, R.M., 2007. On amphidromy, a distinct form of diadromy in aquatic organisms. Fish and Fisheries 8: 1-13. Meijaard, E., 2003. Mammals of south-east Asian islands and their Late Pleistocene environments. Journal of Biogeography 30: 1245-1257. Miller, M.P., 2005. ALLELES IN SPACE (AIS): computer software for the joint analysis of interindividual spatial and genetic information. Journal of Heredity 96: 722?724. Myers, N., R.A. Mittermeier, C.G. Mittermeier, G.A.B. da Fonseca & J. Kent, 2000. Biodiversity hotspots for conservation priorities. Nature 403: 853-858. Naruse, T. & H. Tamura, 2006. A first record of anchialine crab of the genus Orcovita Ng and Tomascik, 1994 (Decapoda: Brachyura: Varunidae) from Japan, with description of the species. Limnology 7: 147-151. Nei, M., 1987. Molecular Evolutionary Genetics. Columbia University Press, New York. 93 Nielsen, R. & W. Wakeley, 2001. Distinguishing migration from isolation: an MCMC approach. Genetics 158: 885-896. Okazaki, Y. & H. Nakata, 2007. Effect of the mesoscale hydrographic features on larval fish distribution across the shelf break of East China Sea. Continental Shelf Research 27: 1616-1628. Page, T.J., W.F. Humphreys &J.M. Hughes, 2008. Shrimps down under: Evolutionary relationships of subterranean crustaceans from western Australia (Decapoda: Atyidae: Stygiocaris). PLoS ONE 3: e1618. Page, T.J., K. von Rintelen & J.M. Hughes, 2007. An island in the stream: Australia?s place in the cosmopolitan world of Indo-West Pacific freshwater shrimp (Decapoda: Atyidae: Caridina). Molecular Phylogenetics and Evolution 43: 645?659. Page, T.J., A.M. Baker, B.D. Cook & J.M. Hughes, 2005. Historical transoceanic dispersal of a freshwater shrimp: the colonization of the South Pacific by the genus Paratya (Atyidae). Journal of Biogeography 32: 581-593 Palumbi, S.R., 1994, Genetic divergence, reproductive isolation, and marine speciation. Annual Review of Ecology and Systematics 25: 547?572. Posada D. & K.A. Crandall, 1998. Modeltest: testing the model of DNA substitution. Bioinformatics 14: 817-818. Poff N.L. & D.D. Hart, 2002. How dams vary and why it matters for the emerging science of dam removal. Bioscience 52: 659?668. Rozas, J., J. Sanchez-DelBarrio, X. Messeguer & R. Rozas, 2003. DnaSP, DNA polymorphism analyses by the coalescent and other methods. Bioinformatics 19: 2496-2497. 94 Russ, A.R., S.R. Santos & C. Muir, 2010. Genetic population structure of an anchialine shrimp, Metabetaeus lohena (Crustacea: Alpheidae), in the Hawaiian Islands. Revista de Biologia Tropical 58: 159-170. Sabate`s, A. & M.P. Olivar, 1996. Variation of larval fish distributions associated with variability in the location of a shelf-slope front. Marine Ecology Progress Series 135: 11- 20. Santos, S.R., 2006. Patterns of genetic connectivity among anchialine habitats: a case study of the endemic Hawaiian shrimp Halocaridina rubra on the island of Hawai?i. Molecular Ecology 15: 2699-2718. Scholl, M.A., S.E. Ingebritsen, C.J. Janik & J.P. Kauahikaua, 1996. Use of precipitation and groundwater isotopes to interpret regional hydrology on a tropical volcanic island: Kilauea volcano area, Hawai?i. Water Resources Research 32: 3525?3537. Shokita, S., 1979. The distribution and speciation on the inland water shrimps and prawns from the Ryukyu Islands I. Bulletin of the College of Science, University of the Ryukyus 28: 193-278. Shokita, S., 2006. Caridina rubella Fujino & Shokita, 1975, p. 61 (in Japanese). In, Ministry of the Environment (ed.), Threatened Wildlife of Japan- Red Data Book. 2nd ed.- Volume 7, Invertebrate (except Insecta and Mollusca), Wildlife Research Center, Tokyo Japan. Shokita, S. & S. Nishijima, 1976. Aquatic animals of the ?Shiokawa Salty Spring,? with a discussion on the flow mechanism of salty water, pp 68-91. In, Okinawa Prefectural Board of Education (ed) Survey report on natural monuments of Okinawa Prefecture, 6th series. Survey report on dynamics of the Shiokawa Salty Spring, II (in Japanese). Okinawa Prefectural Board of Education, Okinawa. 95 Sket, B., 1996. The ecology of anchialine caves. Trends in Ecology and Evolution 11: 221-225 Sorensen P.W. & K.A. Hobson, 2005. Stable isotope analysis of amphidromous Hawaiian gobies suggests their larvae spend a substantial period of time in freshwater river plumes. Environmental Biology of Fishes 74: 31-42. Sponaugle S., R.K. Cowen, A. Shanks, S.G. Morgan, J.M. Leis, J.S. Pineda, G.W. Boehlert, M.J. Kinsford, K.C. Lindman, C. Grimes & J.L. Munro, 2002. Predicting self-recruitment in marine populations: biophysical correlates and mechanisms. Bulletin of Marine Science 70: 341-375. Strathmann R.R., T.R. Hughes, A.M. Kuris, K.C. Lindeman, S.G. Morgan, J.M. Pandolfi & R.R. Robert, 2002. Evolution of local recruitment and its consequences for marine populations. Bulletin of Marine Science 70: 377-396. Tajima, F., 1989. Statistical method for testing the neutral mutation hypothesis by DNA polymorphism. Genetics 123: 585-595. Tamura, K. &`M. Nei, 1993. Estimation of the number of nucleotide substitutions in the control region of mitochondrial DNA in humans and chimpanzees. Molecular Biology and Evolution 10: 512-526. Wisniewski, M., 2004. A translation of the folklore and natural science of Miyako Island, Okinawa. Miyakojima City Museum, pp. 1-26. 96 Table 1. Indices of genetic diversity and tests of neutrality for each population of Caridina rubella on Miyako-jima, Ryukyus, Japan. n, number of sampled individuals; nh, number of unique haplotypes; ?, nucleotide diversity; h, haplotype diversity. Values in parentheses in n and nh columns represent the actual number of individuals or recovered haplotypes, respectively, from that site. Differences in values represent individuals within a population belonging to a different lineage then what is endemic to that site. Outliers were excluded from the diversity measures and neutrality tests of that particular site. See text for additional details. Diversity indices Neutrality tests Population Site Code n nh ? h Fu?s Fs Tajima?s D Ana Ga 1 ANA1 15 (16) 8 (9) 0.006 0.867 -1.134 -0.877 Ana Ga 2 ANA2 8 4 0.002 0.750 -0.903 -1.030 Tomori Ama Ga TAG 12 9 0.008 0.939 -2.368 -0.672 Butura Ga BUT 12 6 0.005 0.803 0.027 -0.881 Zuza Ga ZUZ 13 6 0.005 0.782 0.083 0.942 Total 60 (61) 24 0.045 0.945 97 Table 2. Pairwise ?ST (above diagonal) and Snn (below diagonal) estimates as measures of genetic differentiation between populations of Caridina rubella on Miyako-jima, Ryukyus, Japan. *P < 0.05, ** P < 0.001 ANA 1 ANA 2 TAG BUT ZUZ ANA 1 - 0.980** 0.056 0.204** 0.224* ANA 2 1.000** - 0.976** 0.982** 0.982** TAG 0.465 1.000** - 0.337** 0.317** BUT 0.745** 1.000** 0.833** - -0.006 ZUZ 0.850** 1.000** 0.859** 0.666* - 98 Table 3. Hierarchical analysis of molecular variance (AMOVA) for northern and southern populations of Caridina rubella on Miyako-jima, Ryukyus, Japan. **P < 0.001 Source d.f. Sum of squares Variance component % Variation ? statistic Among geographic regions 1 19.566 0.651 Va 25.03 0.250 ?CT Among populations within regions 2 5.292 0.058 Vb 2.24 0.030 ?SC Within populations 48 90.857 1.893 Vc 72.73 0.273** ?ST Total 51 115.715 2.603 99 Figure 1. Map of the Southern Ryukyus Archipelago, Japan, indicating the location of Miyako- jima (A) and enlarged map of Miyako-jima depicting anchialine habitats from which Caridina rubella were sampled for this study (B). In panel B, grey dashed lines represent underground aquifer boundaries and thick dark lines represent the location of major subsurface dams of Miyako-jima. Site labels are listed in Table 1 and geographical coordinates of sampling sites are available from the corresponding author upon request. 100 Figure 2. Networks depicting relationships among cytochrome c oxidase subunit I (COI) haplotypes recovered from the two Caridina rubella lineages on Miyako-jima, Ryukyus, Japan. (A) ?short? and (B) ?long? rostrum morphotypes of C. rubella. For each network, black dots represent unsampled (missing) haplotypes while a rectangle represents the haplotype with the highest outgroup probability according to the analysis. The size of circles and rectangles are proportional to the frequency at which a specific haplotype was recovered. Despite variable lengths, each branch implies a single mutational difference between haplotypes. 101 Figure. 3. Posterior probability distributions of migration rates (M) between populations of Caridina rubella on Miyako-jima, Ryukyus, Japan. Presented posterior distributions are the averages of three independent runs with MDIV (Nielsen and Wakeley 2001) (see text for additional details). Panels A and B represent all possible pairwise comparisons between populations of ?short rostrum? C. rubella and are presented as two graphs for readability. 102 CHAPTER 5. Multiple colonizations lead to hidden diversity in a unique island ecosystem: Comparative phylogeography of anchialine shrimp in the Ryukyu Archipelago, Japan 5.1 ABSTRACT Geological events (i.e., tectonics, volcanism, sea level changes) can significantly impact an area?s flora and fauna, consequently shaping the structure and evolutionary trajectory of populations. In this context, the archipelagos of the Indo-West Pacific are among the richest in the world in terms of biodiversity due to the region?s dynamic geological past and stable climatic conditions. To explain this high species diversity, recent molecular studies have supported one of two alternative hypotheses: the center of origin hypothesis or the center of accumulation hypothesis. To differentiate between these two hypotheses, we took a comparative phylogeographic approach to determine the origins of the anchialine diversity in the Ryukyu Islands by investigating the genetic variation, population structure and evolutionary history of three shrimp species endemic to the anchialine niche. Given the dynamic geologic history of the islands, it was hypothesized that the anchialine diversity of the Ryukyus originated within the archipelago and that the phylogenetic history of these shrimp would be correlated with the sequential separation of the Ryukyu Islands as sea-levels fluctuated during the Pleistocene (i.e., Center of Origin). Accordingly, sequence analyses of two mitochondrial (i.e., cytochrome oxidase subunit I [COI] and large subunit ribosomal [16S-rDNA]) genes found multiple genetic lineages with considerable genetic diversity and population structure for each species sampled. However, given the phylogenetic relationships and geographic distributions of the multiple lineages found and the strong regional oceanographic currents of the Indo-West Pacific, the diversity of the anchialine Carideans 103 of the Ryukyus is likely the result of an accumulation of species dispersing into the Ryukyus from independent source populations. Furthermore, the contrasting patterns of population structure and connectivity exhibited by each ?species? within the Ryukyus appear to result from complex interactions between intrinsic (i.e., life history traits) and extrinsic (i.e., historical geology and oceanography) processes. 5.2 INTRODUCTION Geological events (i.e., tectonics, volcanism, sea level changes) can significantly impact an area?s flora and fauna, consequently shaping the structure and evolutionary trajectory of populations (e.g., Vandergast et al., 2004; Liggins et al., 2008; Santos and Weese, 2011). In this respect, oceanic islands, with their discrete geography and generally high levels of endemism, are ideal systems for studying how such events influence organismal diversification. The archipelagos of the Indo-West Pacific have received considerable interest in this regard (Liu et al., 2008) since these island groups are among the richest in the world in terms of species diversity (Myers et al., 2000; Meijaard, 2003). Although a number of hypotheses have been proposed to explain this high species diversity (reviewed in Carpenter et al., 2011), phylogenetic studies generally support one of two alternatives: (i) the center of origin hypothesis, in which the majority of the region?s biodiversity is endemic (i.e., due to repeated bouts of isolation by geologic events) and a source of species for surrounding areas (Benzie, 1989, Briggs 2000, Barber et al., 2006), or (ii) the center of accumulation hypothesis, whereby this high biodiversity is due to species accumulation via dispersal from surrounding areas (i.e., Indian, Pacific and Australasian biotas; Murphy & Austin, 2005, Page et al., 2007, de Bruyn & Mather, 104 2007). However, most studies conducted to date have focused on single species; often leading to an inability to differentiate between these two possibilities (see de Bruyn & Mather, 2007). The Ryukyu Archipelago, located in the Indo-West Pacific and stretching between Japan and Taiwan, has experienced an extensive range of geological events over its history. Specifically, the backbone of the islands formed during the late Miocene (~10 MYA) following subduction of the Philippine Plate by the Eurasian Plate (Konishi & Sudo, 1972; Koba, 1980; Lee et al., 1980). During low sea level stands associated with glacial periods of the late Pliocene (~2 MYA), a landbridge is thought to have extended to southeastern China, connecting the central Ryukyus with Taiwan and continental Asia (Kimura, 2000). Furthermore, this land bridge was submerged in the interglacial periods of the early Pleistocene (~1.5 MYA), creating islands corresponding to the present groups (Hikida & Motokawa, 1999) that experienced sequential periods of connections and oceanic isolation due to fluctuating sea levels throughout the Pleistocene. These latter events are believed to have significantly influenced the diversification and distributions of fauna in the Ryukyus and the phylogenetic relationships of several terrestrial taxa reflect these paleogeographic patterns. For example, Lin et al. (2002) found that the phylogeny of Takydromus grass lizards strongly correlated with the sequential separation of islands during the late Pleistocene. Similar patterns have also been reported for Salganea and Panesthia wood-feeding cockroaches (Maekawa et al., 1999), Parachauliodes fishflies (Liu et al., 2008), and Anomala beetles (Muraji et al., 2008). Another group whose evolutionary history is predicted to correlate with the geologic history of the Ryukyu Archipelago are organisms from the anchialine 105 ecosystem. This ecosystem is comprised of land-locked bodies of salt or brackish water in proximity to the sea that fluctuate with the tides due to subterranean connections to both the ocean and aquifer system (Holthuis, 1973; Maciolek, 1983). Such habitats have been reported from around the world, including the Sinai Peninsula, Bermuda, the Caribbean, Hawaii, the South Pacific, the Philippines, as well as the Ryukyus (Maciolek, 1986). With a few exceptions (Kano & Kase, 2004; Hunter et al., 2008; Russ et al., 2010), population genetic studies have revealed exceptional levels of endemism and strong structure on the scale of only a few km for a number of anchialine organisms (Santos, 2006; Craft et al., 2008; Hunter et al., 2008; Page et al., 2008). For example, populations of the anchialine shrimp Caridina rubella Fujino & Shokita, 1975 (Decapoda: Atyidae) on the island of Miyako in the Ryukyus are significantly structured across distances ranging from <20 m to >10 km (Weese et al., 2012). Additionally, anchialine habitats in general have been heavily influenced by sea-level changes during the Quaternary (Mylroie & Mylroie, 2011), and for anchialine organisms in the Ryukyus, such migrating coastlines may have lead to cycles of isolation and connection as suitable habitats (i.e., the subterranean aquifer system) contracted and expanded with fluctuating sea-levels. In this context, comparing the phylogeography of different taxa endemic to this ecosystem provides an opportunity to understand how past environmental changes (i.e., fluxes in sea levels) may have impacted patterns of anchialine biodiversity in the Ryukyus and the species diversity of Indo-West Pacific archipelagos in general. Decapod crustaceans are the dominant macrobiota of the anchialine ecosystem in the Ryukyu Archipelago, with at least eleven species of shrimp and crabs having been recorded (Komai & Fujita, 2005; Cai & Shokita, 2006; Naruse & Tamura, 2006; Fujita, 106 2007; Fujita & Sunagawa, 2008). Of these, three caridean shrimp species, Metabetaeus minutus Whitelegge, 1897 (Decapoda: Alpheida), Antecaridina lauensis Edmondons, 1935 (Decapoda: Atyidae) and Halocaridinides trigonophthalma Fujino & Shokita, 1975 (Decapoda: Atyidae), have disjunct distributions both in the Ryukyus as well as across the Pacific Basin (Cai & Shokita, 2006; Anker, 2010). If M. minutus, A. lauensis and H. trigonophthalma exhibit genetic differentiation in the Ryukyus and this diversification was driven by geological events directly impacting the archipelago, it is hypothesized that the phylogeography of these species will reflect the history of the islands. Alternatively, it can also be hypothesized that any observed ?differentiation? in these species is due to the accumulation of distinct lineages into the Ryukyu Archipelago from elsewhere via dispersal. To differentiate between these two hypotheses, the genetic variation, population structure and evolutionary history of these three shrimp species were investigated via sequence analyses of two mitochondrial (i.e., cytochrome oxidase subunit I [COI] and large subunit ribosomal [16S-rDNA]) genes. 5.3 MATERIAL AND METHODS 5.3.1 TAXON SAMPLING, DNA EXTRACTION AND SEQUENCING Approximately 42 anchialine habitats were surveyed from nine islands belonging to the Daito, Miyako, and Yaeyama island groups of the Ryukyu Archipelago (Fig. 1) between January and August 2009. From these, specimens of Antecaridina lauensis, Halocaridinides trigonophthalma and Metabetaeus minutus were acquired from 11 anchialine habitats on the islands of Ishigaki, Taketomi, Tarama, Miyako, Irabu and Minami-daito (Fig. 1). From each habitat, 6 -17 individuals were collected (Table 1) 107 using baited traps or small aquarium nets and immediately preserved in 99% ethanol for genetic analyses. For phylogenetic analyses, an individual representing ?short rostrum? Caridina rubella (Weese et al., 2012) was collected from Miyako and additional Metabetaeus and Antecaridina specimens from outside the Ryukyus were supplied as 75% ethanol preserved materials by Dr. Arthur Anker (Florida Museum of Natural History). Total genomic DNA was extracted from each individual using 2X cetyltrimethyl ammonium bromide/chloroform (CTAB) according to procedures outlined in Santos (2006). Between 10-30 ng of DNA were utilized as template to amplify an ~670 base pair (bp) fragment of the mtDNA cytochrome c oxidase subunit I (COI) gene via polymerase chain reaction (PCR). Reactions were conducted in 25 ?L volumes containing 10 mM Tris?HCl (pH 8.3), 50 mM KCl, 0.001% gelatin, 2.0 mM MgCl2, 200 ?M dNTPs, 0.4 ?M each of primers LCO1490 and HCO2198 (Folmer et al., 1994), and 1 U Taq polymerase. Reactions were conducted in a PTC-100 thermocycler (MJ Research, Watertown, MA, USA) under the conditions outlined in Santos (2006). Additionally, sequence data from the mtDNA large subunit ribosomal (16S-rDNA) gene were obtained from one to two individuals for each divergent genetic lineage within the three ?species? (see below) for phylogenetic analyses. The PCRs for an ~850 bp fragment of the 16S- rDNA were conducted in 25 ?L volumes with the ?touchdown? thermocycling profile outlined in Craft et al., (2008) and containing 0.4 ?M each of primers CRUST16SF and CRUST16SR (Ivey & Santos 2007) along with the reaction constituents outlined above. Amplicons were purified with Montage PCR filter units (Millipore, Billerica, MA, USA) according to the supplier?s directions, cycle-sequenced in both directions 108 using Big-Dye Terminators v3.1 and read on a PRISM 3100 Genetic Analyzer (Applied Biosystems, Foster City, CA, USA). Any ambiguities in the chromatograms were corrected by comparison with the complementary DNA strand in SEQUENCHER v4.7 (Gene Codes Corporation, Ann Arbor, MI, USA). Finished COI and 16S-rDNA sequences were aligned manually using SE-AL v2.0a11 (available at http://tree.bio.ed.ac.uk.software/seal/). 5.3.2 GENETIC DIVERSITY AND POPULATION GENETIC ANALYSES Population-level analyses utilized mtDNA COI sequence data since this gene has proven to be informative for other alpheid and atyid species (Cook et al., 2006; Page et al. 2007; Russ et al., 2010). Nucleotide (?) and haplotype (h) diversity estimates were calculated according to the methods of Nei (1987) using DnaSP v5.10.00 (Rozas et al., 2009). To assess genetic differentiation, pairwise ?ST statistics (which incorporates information from both haplotype frequencies and molecular divergence) were calculated with Arlequin v3.11 (Excoffier & Schneider, 2005). For these comparisons, the Tamura & Nei (1993) model of DNA evolution, selected by the Akaike Information Criterion (AIC) with the FindModel web version of Modeltest (Posada & Crandall, 1998; available at http://darwin.uvigo.es/), was utilized. As another measure of genetic differentiation, Hudson?s (2000) nearest-neighbor statistic (Snn; which measures how often the ?nearest neighbors? (in sequence space) are from the same locality) was calculated using DnaSP. To visualize relationships among COI haplotypes, networks were constructed via TCS v1.21 (Clement et al., 2000), which utilizing the cladogram estimation algorithm of Templeton et al., (1992). The analyses were conducted under default settings, providing 109 95% parsimony plausible branch connections between haplotypes. Reticulations in the networks, representing ambiguous connections, were resolved using the criteria outlined in Crandall et al., (1994). For each of the resulting networks, historical demography was inferred using Tajima?s D (Tajima, 1989) and Fu?s Fs (Fu, 1997) neutrality tests conducted in Arlequin. Both methods provide information regarding population history in the absence of selection, with significant negative or positive values generally suggesting expansions or bottlenecks (Tajima, 1989; Fu, 1997; Akey et al., 2004), respectively. Statistical significance in the pairwise ?ST and neutrality tests was assessed in Arlequin by 1,000 permutations and the Snn in DnaSP by 1,000 permutations. 5.3.3 PHYLOGENETIC ANALYSES Phylogenetic analyses utilizing mtDNA 16S-rDNA sequence data were used to infer evolutionary relationships among divergent genetic lineages recovered from the three shrimp ?species? (see Results) examined in this study. For these analyses, two data sets were created and analyzed in the same manner. The first consisted of alpheid 16S-rDNA sequences, including those of Metabetaeus ?minutus? from this study and an additional Metabetaeus spp. and Betaeus harrimani sequence downloaded from GenBank (Table 2). The second dataset consisted of atyid 16S-rDNA sequences, including the Antecaridina ?lauensis?, Halocaridinides ?trigonophthalma?, and Caridina spp. sequences generated here along with additional Antecaridina spp. and Halocaridina rubra sequences from GenBank (Table 2). Both datasets utilized Macrobrachium japonicum as an outgroup. The 16S-rDNA sequences were aligned using ClustalX (Thompson et al., 1997) and checked manually using SE-AL. For each data set, 110 evolutionary relationships among species were inferred via Maximum Likelihood (ML) analyses with PHYML v3.0 (Guindon & Gascuel, 2003) under the appropriate model of evolution chosen by the AIC in Modeltest v3.7 (Posada & Crandall, 1998). For each phylogeny, the transition/transversion ratio and proportion of invariable sites were estimated, with the starting tree determined by BioNJ (default settings). Branch supports were estimated by 1,000 bootstrap replicates. The resulting phylogenetic trees were viewed with FigTree v1.3.1 (available at http://tree.bio.ed.ac.uk/software/figtree/). 5. 4 RESULTS 5.4.1 GENETIC DIVERSITY Direct sequencing of mtDNA COI amplicons yielded a 630 bp fragment from each of the 33, 61, and 58 individuals of Antecaridina lauensis, Halocaridinides trigonophthalma, and Metabetaeus minutus analyzed in this study. From these, 23 A. lauensis, 33 H. trigonophthalma, and 27 M. minutus unique COI sequences were identified and deposited into GenBank under assession #?s ###-###. While nuclear copies of mitochondrial derived genes (numts: Lopez et al., 1994) can be common among arthropods (Buhay, 2009), translation of these COI nucleotide sequences into amino acids found no stop codons and any non-synonymous changes identified were to residues with similar biochemical properties (data not shown). This implies that sequences analyzed here were derived from mitochondrial copies of COI. Specific sample sizes and number of haplotypes recovered for each species are summarized in Table 1. 111 5.4.2 PHYLOGENETIC ANALYSES From inspection of the COI sequence alignments, it was evident that A. ?lauensis?, H. ?trigonophthalma? and M. ?minutus? from the Ryukyus were each potentially comprised of 2-3 divergent genetic lineages. To test whether each ?species? was monophyletic as well as to infer relationships among lineages, 16S-rDNA sequences were generated for one to two individuals per lineage, as well as from specimens of the same and/or closely related species from outside the Ryukyus (Table 2), for analyses in a phylogenetic context. Maximum likelihood (ML) trees, inferred under the GTR model of evolution (as chosen by the AIC), for both datasets were well resolved and are presented in Figs. 2B and 2C. The ML tree for the alpheid data set found Metabetaeus to be monophyletic with strong (100%) bootstrap support (Fig. 2B). Within Metabetaeus, M. ?minutus? was also monophyletic to the exclusion of M. lohena (Hawai?i) and M. mcphersonae (Moorea). For M. ?minutus? of the Ryukyus, two lineages were recovered. One of these, collected exclusively from the southern Ryukyu Islands of Ishigaki (IS), Miyako (MY), and Tarama (TA), was sister to M. ?minutus? from Christmas Island (CI) in the Indian Ocean (Fig. 2B). The second lineage of M. ?minutus? is apparently confined to the island of Minami-daito (MD), > 600 km east of the southern Ryukyus. The phylogenetic analysis placed this lineage with relatively strong (i.e., 87%) bootstrap support as sister to a Metabetaeus sp. (a potential member of M. ?minutus?) from New Caledonia (NC) in the southwest Pacific Ocean (Figs. 2A and 2B). Thus, both groups of M. ?minutus? in the Ryukyus are more closely related to lineages collected from other geographic regions then they are to each other. Furthermore, morphological variation also supports the 112 inferred phylogenetic relationships. Specifically, Anker (2010) noted differences in rostrum length and the angle of orbital teeth between M. ?minutus? collected from Minami-daito and Christmas Island. Here, specimens from the Minami-daito lineage conformed to the description of materials from the same locality, while specimens from the southern Ryukyus lineage closely resemble the description of the Christmas Island specimen (Supplementary Figure 1) presented in Anker (2010). Similar phylogenetic patterns were recovered from the atyid ML analysis. In this case, Antecaridina and Halocaridinides were each monophyletic with strong (98% and 100%, respectively) bootstrap support (Fig. 2C). For Antecaridina, one of the two genetic lineages, localized to Ishigaki (IS) and Tarama (TA) in the southern Ryukyus and Minami-daito (MD), was sister to A. ?lauensis? from Christmas Island (CI) and Hawai?i (HI) in the central Pacific Ocean, with < 1% sequence divergence in 16S-rDNA across this geographic range (Figs. 2C). The second lineage in the Ryukyus also occurs on Ishigaki (IS) but, unlike the other (and more widespread) A. ?lauensis? lineage, appears restricted to one anchialine habitat on this single island. Phylogenetically, this lineage is sister to an Antecaridina sp. from East Timor (ET), which lies between the South China Sea and Indian Ocean (Fig. 2A). Just as in the case of M. ?minutus?, both groups of A. ?lauensis? are more closely related to lineages from outside the Ryukyus then to one another. Lastly, three genetic lineages were recovered from H. ?trigonophthalma?, with two of the lineages restricted to Taketomi (TK) and Tarama (TA) forming a monophyletic group sister to a third lineage exclusive to the islands of Ishigaki (IS), Miyako (MY), and Irabu (IR) (Fig. 2C). Unfortunately, no specimens or genetic data for H. ?trigonophthalma? from outside the Ryukyus could be obtained for this study. 113 5.4.3 POPULATION GENETIC ANALYSES Metabetaeus ?minutus?: Estimates of haplotype diversity (h) for M. ?minutus? ranged from 0.93-1.0, with nucleotide diversity (?) being consistent among populations (Table 1). Significant genetic structure, in the form of pairwise ?ST and/or Snn values approaching, or at, their upper limits of 1.0 (Table 4), was observed between the Southern Ryukyus and Minami-daito, implying no individuals are exchanged between these populations and distinct genetic lineages separated by > 600 km (Figs. 3A and 3B). On the other hand, an absent of structure was evident among M. ?minutus? populations of the southern Ryukyus lineage spanning a distance of ~ 200 km (Table 4). This homogenization of M. minutus inhabiting anchialine habitats on islands, isolated by expanses of ocean of ~50 km, implies an absence of isolating barriers at this scale, thus leading to panmixia between populations on the three islands. The total lack of population structure at this scale is similar to what has been reported for M. lohena in the Hawaiian Islands (Russ et al. 2011), which covers a similar (~ 260 km) geographic range across the three main Hawaiian islands. Two discrete networks (Fig. 3C), corresponding to the two lineages identified in the phylogeny (Fig. 2B), were recovered from parsimony analysis of the COI sequences. One represented M. ?minutus? from the Southern Ryukyus, consisting of 14 haplotypes from 21 individuals collected on Ishigaki (IS), Miyako (MY) and Tarama (TA) islands (Table 3). The second network was of M. ?minutus? from Minami-daito (MD), encompassing nine haplotypes recovered from 12 individuals (Table 3). Uncorrected (p) distances among haplotypes within each network ranged from 0.16-1.4% while those for haplotypes between the two networks ranged from 9.7-10.6%. Both Tajima?s D and Fu?s F values were significantly negative for M. 114 ?minutus? populations in the Southern Ryukyu (Table 3), implying recent population expansion for this lineage. Antecaridina ?lauensis?: With the exception of one outlier (i.e., Yoshino Cave (YOSH) on Ishigaki [see below]), estimates of h and ? were consistent across populations of A. ?lauensis? (Table 1). Unlike M. ?minutus?, no significant genetic structure was detected between A. ?lauensis? populations in the southern Ryukyus and Minami-daito (Table 4). However, the A. ?lauensis? population at YOSH was significantly different from all others in the Ryukyus (Table 4), due to being the only site having both lineages of A. ?lauensis? (see above). Parsimony analysis of the COI sequences produced two discrete networks, representing the two lineages of A. ?lauensis? in the Ryukyus, with individuals from the YOSH population occurring in both (Fig. 3D). One network consisted of 23 haplotypes and 48 individuals from the islands of Ishigaki (IS), Tarama (TA) and Minami-daito (MD) whereas the other included four haplotypes from 10 individuals specific to the island of Ishigaki (IS) (Fig. 3D). Uncorrected p-distances among COI haplotypes were 0.16-1.4% within a network and 12.7-13.5% between the two networks. Tajima?s D and Fu?s F values were found to be significant and negative for the two lineages (Table 3), suggesting recent population expansion in both. If COI sequences from A. ?lauensis? and Antecaridina sp. sampled in Hawai?i, Christmas Island (both supplied by Dr. Timothy J. Page, unpublished), and East Timor (EF173843 - EF173847) are included in the parsimony analysis, those from Hawai?i (1) and Christmas Island (2) are identical to the ancestral haplotype of the Ishigaki (IS), Tarama (TA) and Minami-daito (MD) lineage while a third network, comprised of five haplotypes from 115 eight East Timor individuals (i.e., the Antecaridina sp.), is also recovered (data not shown). Halocaridinides ?trigonophthalma?: For H. ?trigonophthalma?, estimates of h spanned 0.371-0.983 and, with the exception of the one site on Taketomi (TK) (? = 0.028), ? values were consistent for all populations (Table 1). Significant genetic differentiation and population structure was found between populations on the islands of Tarama (TA) and Taketomi (TK) relative to those on the islands of Miyako (MY), Irabu (IR) and Ishigaki (IS) (Table 4). Three discrete networks were identified in the parsimony analysis of COI sequences for H. ?trigonophthalma? (Fig. 3E). The first corresponded to 9 haplotypes and 29 individuals from Ishigaki (IS) and Miyako (MY) islands while the second was composed of 24 haplotypes and 30 individuals from Tarama (TA) and Taketomi (TK) (Fig. 3A). The third network, consisting of two individuals and one haplotype, was exclusive to the single Taketomi (TK) population (Fig. 3E). Here, p- distances among haplotypes in the same network ranged from 0.16-1.2% while those between the three networks were 17.1-19.2%. Two of the three H. ?trigonophthalma? lineages exhibited recent population expansion as evident from significantly negative Tajima?s D and Fu?s Fs values (Table 3). 5.5 DISCUSSION Here, we took a comparative phylogeographic approach to determine the origins of the anchialine diversity in the Ryukyu Islands by investigating the population structure and evolutionary history of three Caridean shrimp from the anchialine niche. Following the Center of Origin hypothesis (see Introduction), it was hypothesized that the anchialine 116 diversity of the Ryukyus originated within the archipelago and that the phylogenetic history of A. lauensis, M. minutus and H. trigonophthalma would be correlated with the sequential separation of the Ryukyu island groups as sea-levels fluctuated during the Pleistocene. Accordingly, multiple genetic lineages with considerable genetic diversity and population structure were found to occur within the archipelago for each of the three ?species? investigated. Similar patterns have been seen for a number of marine species in the Coral Triangle supporting the center of origin hypothesis for the Indo-West Pacific (reviewed in Benzie, 1989; Carpenter et al., 2012). However, taken together, the phylogenetic relationships and geographic distributions of the multiple lineages found for each species coupled with the strong regional oceanographic currents of the Indo-West Pacific suggest that the diversity of anchialine shrimp found in the Ryukyus is the result of an accumulation of species dispersing into the Ryukyus from independent source populations followed by subsequent diversification within the archipelago. 5.5.1 EVIDENCE FOR MULTIPLE COLONIZATION OF THE RYUKYUS BY ANCHIALINE CARIDEANS Whether the occurrence of related taxa on an island/archipelago reflects in situ diversification or multiple colonizations can be reconciled through the reconstruction of phylogenetic relationships; if island taxa are monophyletic, a single colonization is favored, whereas paraphyly suggests multiple colonizations (Emerson, 2002). For example, in the Hawaiian Islands, phylogenetic monophyly of the specieous Hawaiian Drosophila (Grimaldi et al., 1990) and silverswords (Baldwin et al., 1991) revealed that these spectacular radiations were each the result of a single colonization event. Here, the 117 occurrence of multiple lineages of each species and their similar phylogenetic relationships to populations outside the archipelago (Fig. 2), suggests that the anchialine fauna of the Ryukyus is the result of an accumulation of species dispersing into the archipelago from multiple source populations around the Indo-West Pacific. For Metabetaeus and Antecaridina, the sympatric lineages of each species occurring in the Ryukyus are not sister taxa, but rather exhibit paraphyletic relationships to Metabetaeus/Antecaridina lineages from outside the archipelago (Christmas Island, East Timor or New Caledonia). Furthermore, given these patterns, it is likely that the multiple lineages of Halocaridinides found in the Ryukyus are also the result of multiple colonizations from sources outside the archipelago. Similar phylogentic patterns attributed to multiple colonizations have recently been elucidated for a number of terrestrial (Harbaugh and Baldwin, 2007; Swenson et al., 2007; Nattier et al., 2011) and marine (Holland et al., 2004 Burridge et al., 2006) taxa. While similar phylogenetic patterns could arise from anchialine species diversifying in the Ryukyus and subsequently dispersing out of the archipelago, this seems unlikely given the phylogenetic relationships of the multiple lineages recovered here to Carideans sampled from islands of the south Indo-West Pacific (i.e.,, Christmas Island, East Timor and New Caledonia) and the oceanographic patterns of the Indo-West Pacific. The currents of the area are dominated by the, historically stable, Kuroshio Current which diverges from the North Equatorial Current in the area east of the Philippine Islands and flows northward into the Okinawa Trough, past the Ryukyu Islands and out into the Pacific Ocean through the Tokara Strait (Ujiie et al. 2003). As the worlds second-largest ocean current (Shen et al., 2011), the Kuroshio Current has 118 played a major role in generating the species diversity of the Indo-West Pacific seen today (e.g., Kojima et al., 1997; Mukai et al., 2009; Soeparno et al., 2011). For example, the North Equatorial and Kuroshio Currents have been shown to contribute to the dispersal and migration of marine taxa from the Central/Southern Indo Pacific to the Japan and Ryukyu Islands (e.g., Tsukamoto, 2006; Mukai et al., 2009). Furthermore, the Kuroshio Current (along with the South China and North China Coastal Currents) is thought to have influenced the recolonization of the Ryukyus by the flathead mullet, Mugil cephalus, from southern refugia following Plio-Pleistocen sea level changes (Shen et al., 2011). For anchialine taxa, the strong northward currents of the Indo-West Pacific may have resulted in the colonization of the anchialine habitats of the Ryukyus from the southern Indo-West Pacific into the Ryukyus As mentioned above (see Introduction), climatic oscillations during the Pleistocene caused large fluctuations in sea levels globally, dramatically impacting the diversification and distribution of many species throughout the Indo-West Pacific (e.g., Barber et al., 2006; De Bruyn and Mather, 2007; Crandall et al., 2008; Fitzpatrick et al., 2011). For anchialine Carideans, these cyclical fluctuations in sea-level may have lead to multiple ?waves? of colonizations into the Ryukyus associated with periods of increased dispersal followed by subsequent divergence in isolation. Assuming larger/higher islands have been above sea level for longer periods of time then smaller, lower-lying islands, there appears to be a strong correlation between the distribution of anchialine lineages and island age. For example, one could hypothesize that, given their current elevation (> 100 m) and size (> 200 km2), the islands of Ishigaki, Miyako and Irabu may have been colonized by Halocaridinides at a time in the past when sea levels were higher then 119 present and many of the Ryukyu Islands were submerged. Then later, as sea levels receded to present day conditions or lower, as second (and maybe third) wave of Halocaridinides may have swept through the islands colonizing the once submerged islands of Taketomi and Tarama (elevation <30 m; area <15 km2). The more recent colonization of these smaller/younger islands is supported by the fact that the Taketomi/Tarama lineage possesses the signal of a recent strong bottleneck and subsequent expansion indicative of a recent colonization/founder event, as evident by the largest negative Tajima?s D and Fu?s FS values encountered in the study (Table 3). Furthermore, the only lineage that shows no evidence of a recent population expansion (Table 3) is lineage M. minutus endemic Minami-daito, which is thought to be considerably older then the islands of the Southern Ryukyus (Ota and Omura, 1992) suggesting it may have been colonized prior to the emergence and colonization of the southern islands of Ishigaki, Miyako and Tarama. For Antecaridina, one lineage has an impressive distributional range with haplotypes being shared among Christmas Island, Hawai?i, and three islands in the Ryukyus (Ishigaki, Tarama and Minami-daito) suggesting a recent colonization or ongoing dispersal/gene flow throughout the Pacific while the other lineage is restricted to the island of Ishigaki and may represent an older colonization. While the timing of specific colonization events cannot be determined with the current sampling, all three species appear to share a history of recurring periods of increased dispersal associated with sea-level fluctuations. 5.5.2 POPULATION STRUCTURE OF ANCHIALINE CARIDEANS IN THE RYUKYUS ARCHIPELAGO 120 Despite having similar ecologies and broadly sympatric distributions within the Ryukyu Archipelago, lineages Metabetaeus, Antecaridina and Halocaridinides exhibit radically different scales of geographic distributions, intraspecific diversity, and population structuring. While the similar phylogenetic patterns throughout the Indo-West Pacific suggest similar evolutionary histories for all three species, subtle differences in intrinsic life-history characteristics may have played a role in creating this phylogeographical discordance within the Ryukyus. For Caridean shrimp, differences in life history characteristics (i.e., egg size, larval stages, and larval habitat) and larval feeding mode (i.e., lecithotrophy versus planktotrophy) have been shown to strongly influence population structure and dispersal potential (Shokita, 1979, Page and Hughes, 2007; Craft et al., 2008; Russ et al., 2010). Very little is known about the life history of Metabetaeus minutus, but recent field and laboratory observations suggests that Metabetaeus typically produce 20-29 eggs that develop as planktotrophic larvae (Fujita, unpublished). As planktotrophic larvae are typically associated with wide dispersal abilities and low genetic differentiation, it is not suprising that populations of Metabetaeus from the Southern Ryukyu genetic lineage are homogenized across the three southern islands of Miyako, Tarama and Ishigaki, which span a distance of ~150 km (Fig. 3). This homogenization of M. minutus inhabiting anchialine habitats on islands, isolated by expanses of ocean of ~50 km, implies an absence of isolating barriers at this scale, thus leading to panmixia between populations on the three islands. This total lack of population structure at this scale is similar to what has been reported for other anchialine species like the neritiliid snail Neritilia cavernicola and alpheid shrimp Metabetaeus lohena, both of which possess planktotrophic larvae and 121 exhibit little to no genetic differentiation over similar ranges of ~200 km in the Philippines and Hawaiian Islands, respectively (Kano and Kase, 2004; Russ et al., 2010). Additionally, the two populations of M. minutus sampled from the Minami-daito lineage show no population structure across this island, again similar to populations of M. lohena occurring on any single Hawaiian Island. However, as evident from the deep (~10% COI p-distance) genetic divergence found between the two lineages separated by > 600 km, M. minutus appear to be isolated over greater geographic distances. Collectively, this suggests that Metabetaeus in general (minutus or lohena) may be a ?good? disperser in ?ecological? time when islands are in close (i.e., < 200 km) proximity. However, once islands get further apart (i.e., > 600 km), dispersal and colonization becomes an ?evolutionary sweepstake? event, with an apparent absence of gene flow following colonization. For atyids, such as Antecaridina and Halocaridinides, egg size is thought to be an effective predictor of dispersal ability and population structure (Craft et al., 2008). For example, Caridina species possessing large geographic ranges and low levels of genetic structure generally have relatively small eggs (~0.4 mm), whereas those with large eggs (~1.6 mm) are more restricted in their distributions and exhibit higher levels of genetic structure (Page and Hughes, 2007). However, despite producing relatively small (~0.35 mm eggs; Fujita, personal observation) with planktotrophic development, Halocaridinides exhibits surprisingly strong population structure in the Ryukyus with strong differentiation found between islands separated by > 5 km of shallow ocean (i.e., Ishigaki vs. Taketomi). While local retention of larvae within nearshore coastal waters or the strong regional oceanic currents might explain the unexpected levels of genetic 122 structure (Weese et al., 2012), given their distributions the multiple lineages of Halocaridinides are more likely the result of independent colonizations from outside the archipelago (see above) that have remained genetically isolated due to subtle difference in reproductive traits, ecological factors, such as dietary differences, resource partitioning or competition. While Antecaridina ?lauensis? has historically been thought to have an "extremely disjunctive" distribution due to high dispersal abilities (Smith and Williams, 1981), the two lineages of Antecaridina in the Ryukyus represent two opposite ends of the dispersal spectrum. One lineage appears to have an impressive distributional range occurring on Christmas Island, the Hawaiian Islands, and at least three islands in the Ryukyus (Ishigaki, Tarama and Minami-daito), with minimal divergence found across this range. Given the correlations seen between reproductive traits (e.g., egg size and larval feeding mode) and dispersal ability, it could be predicted that this lineage would produce relatively small eggs developing as planktonic larvae. However, the second lineage is restricted to only the island of Ishigaki, suggesting this lineage may possess larger eggs; reducing dispersal abilities. Such intraspecific variation in egg size has been seen between cryptic lineages and species for Australian atyids (Paige et al., 2005; Cook et al., 2006; Paige and Hughes, 2007). In studying the Caridean shrimp of the Ryukyus, Shokita (1979) suggested that such an increase in egg size occurs as species transition from an amphidromous to a totally landlocked lifecycle; which may explain the difference in rage size of the two lineages. Alternatively, Antecaridina as a genus may produce medium size eggs undergoing planktonic development with moderate dispersal abilities, limiting long distance marine dispersal to evolutionary sweepstake events as 123 discussed above. In this context, the widespread Ryukyu/Hawai?i/Christmas Island lineage may be a recent dispersal event with insufficient time for divergence between populations. 5.5.3 DISTRIBUTION AND HIDDEN DIVERSITY OF ANCHIALINE SHRIMP OF THE INDO-WEST PACIFIC The anchialine fauna of the Indo-West Pacific is dominated by Caridean shrimp, with 11 species of shrimp being distributed throughout Pacific (de Grave and Sakihara, 2011). The majority of these anchialine Carideans, such as A. lauensis, M. minutus, Calliasmata pholidota Holthuis, 1973, Ligur uveae Borradaile, 1899 and Periclimenes pholeter Holthuis, 1973, are thought to have extremely widespread and disjunct distributions occurring throughout the Indo-West Pacific (see Maciolek, 1983). However, a number of shrimp species, such as Halocaridina rubra, M. lohena and Procaris hawaiana from the Hawaiian Archipelago (Maciolek, 1983), are characterized as being endemic to single islands or archipelagos in the Pacific with the open ocean hypothesized as a strong barrier to dispersal (Craft et al., 2008; Weese et al., 2012). Given the apparent multiple colonizations of the Ryukyus by Metabetaeus, Antecaridina and Halocaridinides and the presence of anchialine shrimp on isolated archipelagos such as the Hawaiian Islands, passive marine dispersal (i.g., Smith and Williams,1981), by either adult or larval stages, may be likely for these species as well and may be a common characteristic of anchialine Carideans in general. However, the varying levels of genetic structure and distributions revealed here for the multiple lineages of Metabetaeus, Antecaridina and Halocaridinides with in the Ryukyus, coupled with the restricted 124 distributions and populations structure of other anchialine Carideans such as H. rubra, M. lohena, Procaris hawaiana and C. rubella (Maciolek, 1983; Santos, 2006; Craft et al., 2008; Russ et al., 2011; Weese et al., 2012) suggest that the open ocean acts as a strong barrier to dispersal and long distance oceanic dispersal likely occurs on evolutionary, rather than ecological, timescales. Given the patterns that are beginning to be emerge for anchialine shrimp in the IWP, the diversity of taxa inhabiting the anchialine habitats of the Pacific may be vastly underestimated with many of the species previously described representing cryptic species complexes. In this study for example, both M. minutus and A. lauensis appear to represent possible species complexes comprised of at least, given the current sampling, two possible divergent species occurring in the Ryukyus. Furthermore, the population structure and genetic divergence found within H. trigonophthalma suggests a potential species complex of at least three Halocaridinides species occurring in the Ryukyu Islands. Additionally, cryptic species complexes have recently been revealed for the two anchialine atyids Caridina rubella (two potential species) on Miyako (Weese et al., 2012) and Halocaridina rubra (eight potential species) in the Hawaiian Archipelago (Santos 2006; Craft et al., 2008). In fact, of the six Pacific anchialine species studied from a genetic prospective to date, only M. lohena has failed to reveal a species complex (Russ et al., 2010). Given these results, additional genetic analyses of other widely distributed Pacific Carideans, such as C. pholidota, P. pholeter and L. uveae, are likely to reveal additional cryptic taxa. Given this, what has traditionally been thought to be 11 species of anchialine Carideans distributed throughout the Pacific (de Grave and Sakihara, 2011) may actually represent over 24 genetically distinct cryptic species, all more restricted in 125 their distributions then previously thought. While additional morphological studies will be required to resolve the taxonomic status of each complex, the restricted distribution of these ?potential species? should be taken into consideration when developing conservation and management strategies for anchialine Carideans and their habitats throughout the Indo-West Pacific. 126 5.6 REFERENCES Anker, A. 2010. Metabetaeus Borradaile, 1899 revisited, with description of a new marine species from French Polynesia (Crustacea: Decapoda: Alpheidae). Zootaxa 2552: 37-54. Avise, J. C. 2004. Molecular Markers, Natural History, and Evolution. Sinauer Associates, Inc., Sunderland, Massachusetts, USA. Baldwin, G.B., D.W. Kyhos, J. Dvorak, and G.D. Carr. 1991. Chloroplast DNA evidence for a North American origin of the Hawaiian silversword alliance (Asteraceae). Proceedings of the National Academey of Science 88: 1840-1843. Barber, P. H., M. V. Erdmann, and S. R. Palumbi. 2006. Comparative phylogeography of three codistributed stromatopods: Origins and timing of regional lineage diversification in the coral triangle. Evolution 60: 1825-1839. Benzie, J. A. H. 1998. Genetic structure of marine organisms and SE Asian biogeography. Pp. 197-209 in Biogeography and Geological Evolution of SE Asia, R. Hall and J.D. Holloway, eds. Leiden, The Netherlands. Bracken, H. D., A. Toon, D. L. Felder, J. W. Martin, M. Finley, J. Rasmussen, F. Palero, and K. A. Crandall. 2009. The decapod tree of life: compiling the data and moving toward a consensus of decapod evolution. Arthropod Systematics and Phylogeny 67: 99-116. Briggs, J. C. 2000. Centrifugal speciation and centers of origin. Journal of Biogeography 27: 1183-1188. 127 Buhay, J. E. 2009. ?COI-Like? sequences are becoming problematic in molecular systematic and DNA barcoding studies. Journal of Crustacean Biology 29: 96- 110. Burridge, C.P., R. Melendez, and B.S. Dyer. 2006. Multiple origins of the Juan Fernandez Kelpfish fauna and evidence for frequent and unidirectional dispersal of Cirrhitoid fishes across the South Pacific. Systamatic Biology 55:566-578. Cai, Y. and S. Shokita. 2006. Atyid shrimps (Crustacea: Decapoda: Caridea) of the Ryukyu Islands, southern Japan, with descriptions of two new species. Journal of Naturla History 40: 2123-2172. Carpenter, K.E., P.H. Barber, E.D. Crandall, M.C.A. Ablan-Lagman, Ambariyanto, G.N. Mahardika, B.M. Manjaji-Matsumoto, M.A. Juinio-Menez, M.D. Santos, C.J. Starger, and A.H.A. Toha. 2012. Comparative phylogeography of the Coral Triangle and implication for marine management. Journal of Marine Biology 2011: doi:10.1155/2011/396982 Clement, M., D. Posada, and K. A. Crandall. 2000. TCS: a computer program to estimate gene genealogies. Molecular Ecology 9: 1657-1659. Cook, B. D, A. M. Baker, T. J. Page, S. C. Grant, J. H. Fawcett, D. A. Hurwood, and J. M. Hughes, 2006. Biogeographic history of an Australian freshwater shrimp, Paratya australiensis (Atyidae): the ole life history transition in phylogeographic diversification. Molecular Ecology 15: 1083-1093. Craft, J. D., A. D. Russ, M. N. Yamamoto, T. Y. Iwai, S. Hau, J. Kahiapo, C. T. Chong, S. Ziegler-Chong, C. Muir, Y. Fujita, D. A. Polhemus, R. A. Kinzie, and S. R. Santos. 2008. Islands under islands: The phylogeography and evolution of 128 Halocaridina rubra Holthuis, 1963 (Crustacean: Decapoda: Atyidae) in the Hawaiian archipelago. Limnology and Oceanography 53: 675-689. Crandall, K. A., A. R. Templeton, and C. F. Sing. 1994. Intraspecific phylogenetics: Problems and solutions, pp. 273?297. In Models in Phylogeny Reconstruction, R. W. Scotland, D. J. Siebert and D. M. Williams, eds. Clarendon Press, Oxford. Crandall, E.D., M.A. Frey, R.K. Grosberg, and P.H. Barber. 2008. Contrasting demographic history and phylogeographical patterns in two Indo-Pacific gastropods. Molecular Ecology 17: 611-626. de Bruyn, M., and P. B. Mather. 2007. Molecular signatures of Pleistocene sea-level changes that affected connectivity among freshwater shrimp in Indo-Australian waters. Molecular Ecology 16: 4295-4307. de Grave, S., and T.S. Sakihara. 2011. Further records of the anchialine shrimp, Periclimenes pholeter Holthius, 1973 (Crustacea, Decapoda, Palaemonidae). Zootaxa 2903:64-68. Edmondson, C. H. 1935. New and rare Polynesian Crustacea. Bernice P. Bishop Museum Occasional Papers 10: 3-40. Excoffier, L., G. Laval, and S. Schneider. 2005. Arlequin (version 3.0): An integrated software package for population genetics data analysis. Evolutionary Bioinformatics 1: 47-50. Fitzpatrick, J.M., D.B. Carlon, C. Lippe, and D.R. Robertson. 2011. The West Pacific diversity hotspot as a source or a sink for new species? Population genetic insights from the Indo-Pacific parrot fish Scarus rubroviolaceus. Molecular Ecology 20: 219-234. 129 Folmer, O., M. Black, W. Hoeh, R. Lutz, and R. Vrijenhoek. 1994. DNA primers for amplification of mitochondrial cytochrome c oxidase subunit I from diverse metazoan invertebrates. Molecular Marine Biology and Biotechnology 3: 294- 299. Fu, Y. 1997. Statistical tests of neutrality of mutations against population growth, hitchhiking and background selection. Genetics 147: 915-925. Fujino, T. and S. Shokita. 1975. Report on some atyid shrimps (Crustacea, Decapoda, caridea) from the Ryukyus Islands. Bulletin of Science and Engineering Division, University of the Ryukyus (Mathematics and Natural Science) 18: 93-113. Fujita, Y. 2007. Decapod crustaceans inhabiting springs on Miyako Islands, the Ryukyus, Japan. Bulletin of the Miyakojima City Museum 11: 89-110. Fujita, Y. and H. Sunagawa. 2008. Cavernicolous and terrestrial decapod crustaceans of Tarama-jima Island, the Ryukyu Islands, southwestern Japan. Bulletin of the Miyakojima City Museum 12: 53-80. Grimaldi, D. A., L.H. Throckmorton, T. Okada. 1990. A phylogenetic, revised classification of genera in the Drosophilidae (Diptera). Journal Bulletin of the American Museum of Natural History 197: 1-139. Guindon, S., and O. Gascuel. 2003. A simple, fast and accurate algorithm to estimate large phylogenies by maximum likelihood. Systematic Biology 52: 696-704. Haig, J. A., R. M. Connolly and J.M. Hughes. 2010. Little shrimp left on the shelf: the roles that sea-level change, ocean currents and continental shelf width play in the genetic connectivity of a sea-grass associated species. Journal of Biogeography 37: 1570-1583. 130 Harbaugh, D.T. and B.G. Baldwin. 2007. Phylogeny and Biogeography of he Sandalwoods (Santalum, Santalaceae): repeated dispersals throughout the Pacific. American Journal of Botany 94:1028-1040. He, L., A. Zhang, D. Weese, C. Zhu, C. Jiang, and Z. Qiao. 2010. Late Pleistocene population expansion of Scylla paramamosian along the coast of China: A population dynamic response to the Last Interglacial sea level highstand. Journal of Experimental Marine Biology and Ecology 385: 20-28. Hewitt, G. 2000. The genetic legacy of the Quaternary ice ages. Nature 405: 907-913. Hikida, T., and J. Motokawa. 1999. Phylogeographic relationships of the skinks of the genus Eumeces (Repitilia: Scincidae) in East Asia. Pp. 231-247 in Tropical Island Herpetofauna- Origin, Current Diversity, and Conservation, H. Ota, eds. Elsevier, Amsterdam. Holland, B.S. M.N. Dawson, G.L. Crow, D.K. Hofmann. 2004. Global phylogeography of Cassiopea (Scyphozoa: Rhizostomeae): molecular evidence for cryptic species and multiple invasions of the Hawaiian Islands. Marine Biology 145:1119-1128. Holthuis, L. B. 1963. On red coloured shrimps (Decapoda, Caridea) from tropical land- locked saltwater pools. Zoologische Mededelingen (Leiden) 38: 261-279. Holthuis, L. B. 1964. The Atyidae of Madagascar. M moires du Mus um national d'histoire naturelle 33: 1-48. Holthuis, L. B. 1973. Caridean shrimps found in land-locked saltwater pools at four Indo-West Pacific Localities (Sinia Peninsula, Funafuti Atoll, Maui and Hawai'i islands) with the description of one new genus and four new species. Zoologische Mededelingen (Leiden) 128: 1-48. 131 Humphreys, W.F., and S. Eberhard. 2001. Subterranean fauna of Christmas Island, Indian Ocean. Helictite 37: 59 -74. Kimura, M. 2000. Paleogeography of the Ryukyu Islands. Tropics 10: 5-24. Koba, M. 1980. Distribution and age of the marine terraces and their deposits in the reefcapped Ryukyu Islands, Japan. Quaternary Research 8: 189-208. Kojima, S., R. Segawa, and I. Hayashi. 1997. Genetic differentiation among populations of the Japenes turban shell Turbo (Batillus) cornutus corresponding to warm currents. Marine Ecology Progress Series 150:149-155. Komai, T., and Y. Fujita. 2005. A new stygiobiont species of Macrobrachium (Crustacea: Decapoda: Caridea: Palaemonidae) from an anchialine cave on Miyako Island,Ryukyu Islands. Zootaxa 1021: 13-27. Konishi, K., and K. Sudo. 1972. From Ryukyus to Taiwan. Kagaku 42: 221-230. Liggins, L., D. G. Chapple, C. H. Daugherty, and P.A. Ritchie. 2008. Origin and postcolonization evolution of the Chatham Islands skink (Oligosoma nigriplantare). Molecular Ecology 17: 3290-3305. Lin, S. M., C. A. Chen, and K. Y. Lue. 2002. Molecular phylogeny and biogeography of the grass lizards genus Takydromus (Reptilia: Lacertidae) of east Asia. Molecular Phylogenetics and Evolution 22: 276-288. Liu, M. Y., Y. X. Cai, and C. S. Tzeng. 2007. Molecular systematics of the freshwater prawn genus Macrobrachium Bate, 1868 (Crustacea: Decapoda: Palaemonidae) inferred from mt DNA sequences, with emphasis of East Asian Species. Zoological Studies 46: 272-289. 132 Long, Y., H. Wan, F. Yan, C. Xu, G. Lei, S. Li, and R. Wang. 2006. Glacial effects on sequence divergence of mitochondrial COII of Polyura eudamippus (Leipidoptera: Nymphalidae) in China. Biochemical Genetics 44: 361-377. Maciolek, J. A. 1983. Distribution and biology of Indo-Pacific insular hypogeal shrimp. Bulletin of Marine Science 33: 606-618. Maciolek, J. A. 1986. Environmental features and biota of anchialine pools on Cape Kinau, Maui, Hawai'i. Stygologia 2: 119-129. Maekawa, K., N. Lo, O. Kitade, T. Miura, and T. Matsumoto. 1999. Molecular phylogeny and geographic distribution of wood-feeding cochroaches in East Asian islands. Molecular Phylogenetics and Evolution 13: 360-376. Meijaard, E. 2003. Mammals of south-east Asian islands and their Late Pleistocene environments. Journal of Biogeography 30: 1245-1257. Mukai, T., S. Nakamura, M. Nishida. 2009. Genetic population structure of a reef goby, Bathygobius cocosensis, in the northwestern Pacific. Ichthyology Research 56:380-387. Murphy, N. P., and C. M. Austin. 2005. Phylogenetic relationships of the globally distributed firewater prawn genus Macrobrachium (Crustacea: Decapoda: Palaemonidae): biogeography, taxonomy and the convergent evolution of abbreviated larval development. Zoologica Scripta 34: 187-197. Mylroie, J. E., and J.R. Mylroie. 2011. Void development on carbonate coasts: creation of anchialine habitats. Hydrobiologia doi 10.1007/s10750-010-0542y. Myers, N., R. A. Mittermeier, C. G. Mittermeier, G. A. B. Da Fonseca, and J. Kent. 2000. Biodiversity hotspots for conservation priorities. Nature 403: 853?858. 133 Naruse, T., and H. Tamura. 2006. A first record of anchialine crab of the genus Orcovita Ng and Tomascik, 1994 (Decapoda: Brachyura: Varunidae) from Japan, with description of the species. Limnology 7: 147-151. Nattier, R., T. Robillard, L. Desutter-Grandcolas, A. Couloux, and P. Grandcolas. 2011. Older than New Caledonia emergence? A molecular phylogenetic study of the eneopterine crickets (Orthoptera: Grylloidea). Journal of Biogeography 38:2195- 2209. Ota, Y., A. Omura. 1992. Contrasting styles and rates of tectonic uplift on coral reef terraces in the Ryukyu and Daito Islands, Southwestern Japan. Quaternary International 15/16: 17-29. Page, T. J., K. Rintelen, and J. M. Hughes. 2007. An island in the stream: Australia?s place in the cosmopolitan world of Indo-West Pacific freshwater shrimp (Decapoda: Atyidae: Caridina). Molocular Phylogentics and Evolution 43: 645- 659. Page, T. J., and J. M. Hughes. 2007. Phylogeographic structure in Australian freshwater shrimp largely pre-dates the geological origins of its landscape. Heredity 98: 222- 231. Page, T. J., W. F. Humphreys, and J. M. Hughes. 2008. Shrimps down under: Evolutionary relationships of subterranean crustaceans from Western Australia (Decapoda: Atyidae: Stygiocaris). PLoS ONE 3: e1618. Santos, S. R. 2006. Patterns of genetic connectivity among anchialine habitats: a case study of the endemic Hawaiian shrimp Halocaridina rubra on the Island of Hawaii. Molecular Ecology 15: 2699-2718. 134 Santos, S. R., and D. A. Weese. 2012. Rocks and Clocks: Linking Geologic History and Rates of Genetic Differentiation in Anchialine Organisms. Hydrobiologia 677: 54-63. Shackleton, N. J. 2000. The 100,000-year ice-age cycle identified and found to lag temperature, carbon dioxide, and orbital eccentricity. Science 289: 1897-1902. Shen, K., B. W. Jamandre, C. Hsu, W. Tzeng, and J. Durand. 2011. Plio-Pleistocene sea level and temperature fluctuation in the northwestern Pacific promoted speciation in he globally-distributed flathead mullet Mugil cephalus. BMC Evolutionary Biology 11: 83. Shokita, S. 1979. The distribution and speciation of the island water shrimps and prawns from the Ryukyu Islands. Bulletin of the college of Science, University of the Ryukyus 28: 193-278. Smith, M. J., and W. D. Williams. 1981. The occurrence of Antecaridina lauensis (Edmondson) (Crustacea, Decapoda, Atyidae) in the Solomon Islands. Hydrobiologia 85: 49-58. Soeparno, Y. Nakamura, T. Shibuno, and K. Yamaoka. 2012. Relationship between pelagic larval duration and abundance of tropical fishes on terperate coasts of Japan. Journal of Fish Biology 80:346-357. Su, Y., Y. Chang, S. Lee, and I. Tso. 2007. Phylogeography of the giant wood spider (Nephila pilipes, Araneae) from Asian-Australian regions. Journal of Biogeography 34: 177-191. Swenson, U., J. Munzinger, and I.V. Bartish. 2007. Molecular phylogeny of Planchonella (Sapotaceae) and eight new species from New Caledonia. Taxon 56:329-354. 135 Tsukamoto, K. 2006. Oceanic biology: Spawning of eels near a seamount. Nature 439:929. Ujiie, Y., H. Ujiie, A. Taira, T. Nakamura, and K. Oguri. 2003. Spatial and temporal variability of surface water in the Kuroshio source region, Pacific Ocean, over the past 21,000 years: evidence from planktonic foraminiera. Marine Micropaleontology 49:335-364. Vandergast, A. G., R. G. Gillespie, and G.K. Roderick. 2004. Influence of volcanic activity on the population genetic structure of Hawaiian Tetragnatha spiders: fragmentation, rapid population growth and the potential for accelerated evolution. Molecular Ecology 13: 1729-1743. Weese, D. A., Y. Fujita, M. Hidaka, and S. R. Santos. 2012. The long and short of it: Genetic variation and population structure of the anchialine atyid shrimp Caridina rubella on Miyako-jima, Japan. Journal of Crustacean Biology 32: Whitelegge, T. 1897. The Crustacea. Memoirs of the Australian Museum. 3: 127-151. 136 Table 1. Measures of genetic diversity and tests of neutrality for populations of Metabetaeus minutus, Antecaridina lauensis and Halocaridinides trigonophthalma from the Ryukyu Archipelago, Japan. Metabetaeus minutus Antecaridina lauensis Halocaridinides trigonophthalma Island (code) Site Name (code) n nh ? h n nh ? h n nh ? h Taketomi (TK) Mi-na-ga (MIN) 16 14 0.028 0.983 Ishigaki (IS) Yoshino Cave (YOSH) 6 6 0.006 1.000 16 7 0.065 0.775 6 3 0.001 0.600 Tarama (TA) Shuga-ga (SHUG) 17 10 0.003 0.838 Futatsu-ga (FUT) 6 5 0.007 0.933 Fushato-ga (FUSH) 16 14 0.005 0.975 Miyako (MY) Kikya-ga (KIK) 9 8 0.005 0.972 Tomori Ama-ga (TAG) 8 4 0.002 0.643 Irabu (IR) Sabaoki Well (SW) 15 4 0.001 0.371 Minami-daito (MD) Jr. High Cave (JR) 6 5 0.006 0.933 15 9 0.004 0.876 137 Shintou Cave (SHIN) 10 7 0.004 0.876 Miyahira Cave (MIY) 6 6 0.006 1.000 Total 33 23 0.051 0.966 58 27 0.040 0.868 61 33 0.094 0.878 n, number of sampled individuals; nh, number of unique haplotypes, ?, nucleotide diversity; h, haplotype diversity. 138 Table 2. Additional specimens and sequences included in 16S-rDNA analyses. Taxonomic Group Location Accession Number Source Antecaridina sp. East Timor EF173754 Craft et al., 2008 Antecaridina lauensis Christmas Island EU123850 Page et al., 2008 East Timor EU123853 Page et al., 2008 Hawai?i? ------ Unpublished Ishigaki XXXXX This study Minami-daito XXXXX This study Caridina rubella Miyako-jima XXXXX This study Halocaridina rubra Hawai?i EF173734 Craft et al., 2008 Hawai?i EF173734 Craft et al., 2008 Halocaridinides trigonophthalma Taketomi XXXXX This study Taketomi XXXXX This study Miyako-jima XXXXX This study Metabetaeus minutus Minami-daito XXXXX This study 139 Ishigaki XXXXX This study Christmas Island?? XXXXX This study Metabetaeus mcphersonae Moorea?? XXXXX This study Metabetaeus lohena Hawai?i XXXXX This study Metabetaeus sp. New Caledonia FJ943435 Bracken et al., 2009 Betaeus harrimani FJ943434 Bracken et al., 2009 Macrobrachium japonicum Okinawa DQ194935 Liu et al., 2007 ?Sequence provided by Timothy Page, Griffith University, Queensland Australia ??Samples provided by Arthur Anker, Florida Museum of Natural History 140 Table 3. Indicators of demographic expansion events in lineages of of Metabetaeus minutus, Antecaridina lauensis and Halocaridinides trigonophthalma of The Ryukyu Archipelago, Japan. *P < 0.05, ** P < 0.001 Species Lineage n nh Fu?s Fs Tajima?s D Metabetaeus minutus Minami-daito 12 9 -2.650 0.412 Miyako/Tarama/Ishigaki 21 14 -7.328** -1.842* Antecaridina lauensis Ishigaki 10 4 -1.345* -1.667* Ishigaki /Tarama/Minami-daito 48 23 -19.819** -1.951** Halocaridinides trigonophthalma Ishigaki/Miyako/Irabu 29 9 -6.021** -2.021** Tarama/Taketomi 30 24 -22.994** -2.024** Taketomi 2 1 NA NA NA, not attempted. 141 Table 4. Pairwise ?ST (below diagonal) and Snn (above diagonal) estimates as measures of genetic differentiation between populations of Metabetaeus minutus, Antecaridina lauensis and Halocaridinides trigonophthalma of The Ryukyu Archipelago, Japan. *P < 0.05 Metabetaeus minutus JR MIY KIK FUT YOSH JR - 0.389 1.000* 1.000* 1.000* MIY 0.028 - 1.000* 1.000* 1.000* KIK 0.956* 0.951* - 0.392 0.244 FUT 0.947* 0.940* -0.048 - 0.325 YOSH 0.951* 0.945* -0.054 -0.061 - Antecaridina lauensis JR SHIN SHUG YOSH JR - 0.454 0.525 0.747* SHIN -0.031 - 0.408 0.707* SHUG -0.001 -0.039 - 0.727* YOSH 0.575* 0.597* 0.597* - Halocaridinides trigonophthalma TAG IR FUSH MIN YOSH 142 TAG - 0.541 1.000* 1.000* 0.468 IR -0.002 - 1.000* 1.000* 0.552 FUSH 0.982* 0.986* - 0.364 1.000* MIN 0.908* 0.929* 0.070* - 1.000* YOSH -0.013 -0.026 0.982* 0.901* - 143 Figure 1. Map of the Southern Ryukyus Archipelago, Japan, indicating the location of the Ryukyu Archipelago (A) and enlarged maps of the Yaeyama, Miyako and Daito island groups from which individuals of the genus Metabetaeus, Antecaridina, and Halocaridinides were sampled for this study (B). Individual sampling site are listed in Table 1 and geographical coordinates of sampling sites are available from the corresponding author upon request. 144 Figure 2. Phylogenetic relationships inferred for selected species of endemic Caridean shrimp of the Pacific Basin. (A) Map of sampling localities from the Pacific Basin and the Southern Ryukyus. Maximum likelihood (ML) trees for the Alpheid dataset (-lnL = -3,493.34; B) and the Atyid data set (-lnL = -4,958.78; C). Values above vertical lines represent bootstrap support as percentages of 100 re-samplings for ML analyses. CI- Christmas Island, ET- East Timor, HI- Hawai?i, IS- Ishigaki, MD- Minami-daito, MO- Moorea, MY- Myiako, NC- New Caledonia, TA- Tarama, TK- Taketomi. 145 Figure 3. Networks depicting relationships among cytochrome c oxidase subunit I (COI) haplotypes recovered for three genera of endemic anchialine shrimp sampled from the Ryukyus Archipelago, Japan. Map of sampled islands; Southern Ryukyus (A; TK-Taketomi [orange], IS- Ishigaki [yellow], TA- Tarama [green], IR- Irabu [red], and MY- Miyako [black]) and Minami- 146 daito (B; MD- Minami-daito [blue]). Geographic distribution of Metabetaeus (C), Antecaridina (D) and Halocaridinides lineages (E). For each network, black dots represent unsampled (missing) haplotypes while rectangles represent the haplotype with the highest outgroup probability according to the analysis. The size of circles and rectangles are proportional to the frequency at which a specific haplotype was recovered. Despite variable lengths, each branch implies a single mutational difference between haplotypes. Color codes for each lineage match island colors in panel A and B. Dashed lines separate lineages. 147 CHAPTER 6. Conclusions 6.1 GENERAL CONCLUSIONS We are currently witnessing the greatest loss of species since the extinction of the dinosaurs 65 million years ago (Warrick, 1998), with global climate change, the introduction of invasive species, and habitat destruction being significant threats to biodiversity around the world. Given this, there is now an urgency to accelerate research into unique island environments and their biota before they are lost forever. In this context, one of most threatened ecosystems on numerous islands may be the anchialine habitat. Historically, anchialine environments have experienced significant negative impacts from anthropogenic activities such as urbanization, groundwater extraction and contamination, and the introduction of invasive species. Thus, in order to manage these habitats and their fauna, it is important to develop an understanding of the biodiversity, ecology, and evolution of organisms from the anchialine niche since such knowledge will play a critical role in establishing conservation strategies for these environments around the world. Subterranean environments, such as the anchialine niche, have long fascinated biologists (e.g., Darwin?s ?wrecks of ancient life?; 1859) and much of this fascination has centered on understanding the geographic distributions through space and time (i.e., biogeography) of their inhabitants (Porter, 2007). Analyzing the population genetic and geographic structures of sympatric taxa can provide valuable information on the processes involved in generation patterns of diversity (Avise, 2000). If these processes are dominated by intrinsic factors (i.e. biological, ecological, physiological or behavioral), discordant patterns of population structure are expected to result, whereas extrinsic factors (i.e. physical, geological, environmental) are likely to result in similar patterns of population structure even among diverse taxa (McMillen-Jackson and Bert, 148 2003; Page and Hughes 2007). The goal of this dissertation was to examine the evolutionary history and phylogeography of organisms from the anchialine niche to gain a better understanding of the ecological and evolutionary forces influencing current biogeographic and diversity patterns in these unique habitats. Specifically, through the use of molecular markers (i.e., mitochondrial cytochrome oxidase subunit I [COI] and large subunit ribosomal [16S- rDNA] genes), I investigated the population structure, phylogeography and evolution of five Caridean shrimp (Halocaridina rubra, Caridina rubella, Antecaridina lauensis, Metabetaeus minutus and Halocaridinides trigonophthalma) endemic to the anchialine ecosystem of the Pacific Basin to gain a better understanding of the combination of factors involved in creating current distribution patterns, including dispersal ability, potential vicariant events and rates of evolution. Furthermore, using H. rubra as an example, I demonstrate how such genetic information can be utilized to identify populations and habitats warranting conservation management. Given the result presented here, it appears that many anchialine Carideans are much more restricted in their distributions then previously thought (e.g. Smith and Williams, 1981; Kano and Kase, 2004). In all cases, what has been previously thought to be one species distributed across an island (e.g., C. rubella), archipelago (e.g., H. rubra) or ocean (e.g., M. minutus, A. lauensis or H. trigonophthalma) actually consisted of multiple ?cryptic species? (see below) restricted to a particular aquifer, island or island group. Following the reasoning presented above, these concordant patterns seen across multiple taxa suggest that extrinsic factors, such as geologic history and oceanographic conditions, have a large impact on the population structure and evolution of anchialine organism. For example, in Chapter 4, it was shown that divergence time estimates between populations, genetic groups and/or lineages across the islands of Maui and 149 Hawai?i were highly consistent with the geologic age of the basalt basins of their anchialine habitats of H. rubra from the Hawaiian Archipelago. Similarly, the apparent repeated colonizations of the Ryukyus by multiple lineages of M. minutus, A. lauensis or H. trigonophthalma, given their population structure and distributions across the islands, appear to be correlated with the geologic history and emergence of the islands that they inhabit (Chapter 5). Other extrinsic factors influencing the population structure of anchialine organisms appear to include regional oceanography (i.e., circulation patterns, oceanographic currents and thermal/salinity gradients) and selection pressures (i.e., habitat availability) (Chapter 4). If these ?regional? factors (Page and Hughes, 2007) were the sole processes influencing geographic structuring in these environments, the anchialine taxa inhabiting a particular region (i.e., the Hawaiian or Ryukyu Archipelago) would exhibit similar patterns of geographic structuring. However, despite overall patterns of restricted dispersal, varying patterns of genetic differentiation and population structure was uncovered between sympatric taxa, suggesting the influence of intrinsic processes as well. For instance, while H. rubra exhibits extremely low levels of gene flow between populations over limited geographic scales (Santos, 2006; Craft et al., 2008) a later study of M. lohena, which produces planktotrophic larvae, revealed little to no population structure among populations sampled from the same anchialine habitats (Russ et al., 2011). Furthermore, if extrinsic barriers to gene flow were the sole factors influencing anchialine diversity, one would have expected similar phylogeographic patterns between C. rubella, M. minutus, A. lauensis and H. trigonophthalma distributed across Ryukyus (Chapters 4- 5). However, the phylogenetic relationships and geographic distributions of these species suggest that each responded uniquely to the dynamic geologic and oceanographic histories of the Indo-West Pacific. These unique responses are likely due to slight, but significant, differences in 150 intrinsic life-history characteristics or ecological factors, such as dietary differences, other resource partitioning and/or competition. For example, small differences in life history characteristics (i.e., egg size, larval stages, and larval habitat) and larval feeding mode (i.e., lecithotrophy versus planktotrophy) have been shown to strongly influence both population structure and dispersal potential of Carideans (Shokita, 1979, Page and Hughes, 2007). Collectively, these studies demonstrate how the ecology, population structure and evolutionary history of organisms endemic to the anchialine niche result from a complex interaction between extrinsic obstacles to gene flow (i.e., geologic history and oceanography) and intrinsic limits to dispersal (i.e., life history traits and developmental mode). Understanding these complex interactions is important since such information is vital for developing management plans to preserve the species and, by extension, the anchialine ecosystem in the Pacific Basin. These studies also suggest that the diversity of taxa inhabiting the anchialine habitats of the Pacific may be vastly underestimated. The anchialine fauna of the Indo-West Pacific is dominated by Caridean shrimp, with 11 species of shrimp thought to be distributed throughout Pacific (de Grave and Sakihara, 2011). However, in Chapter 4, it was shown that C. rubella on the island of Miyako-jima actually consists of two morphologically and genetically distinct species. Furthermore, M. minutus, A. lauensis and H. trigonophthalma of the Ryukyus were all found to represent possible species complexes, each comprised of at least two possible divergent species, all more restricted then previously thought (Chapter 5). Additionally, H. rubra of the Hawaiian Archipelago has also recently been revealed to be a complex of at least eight potential species (Santos 2006; Craft et al., 2008). In fact, of the six anchialine species from the Pacific Basin studied from a genetic prospective to date, only M. lohena has failed to reveal a species 151 complex (Russ et al., 2010). Given these results, genetic analyses of additional Carideans inhabiting the anchialine niche are likely to reveal additional cryptic taxa. While further morphological studies are needed to resolve the taxonomic status of each complex and its potential species, the restricted distribution of these ?potential species? need to be taken into consideration when developing conservation and management strategies for these Carideans and the anchialine habitats of the Indo-Pacific. 6.2 FUTURE PERSPECTIVES While these phylogeographic and population genetic studies have contributed to our understanding of the evolutionary and biogeographic history of anchialine Carideans, many questions sill remain. As with most population genetic studies, additional sampling localities need to be visited. For example, further sampling of M. minutus, A. lauensis and H. trigonophthalma from outside the Ryukyus would better elucidate the evolutionary history of these shrimp in the Pacific Basin. Furthermore, with additional samples, a better understanding of thier phylogenetic relationships and the identification of a few geologic calibration points, it may be possible to utilize the molecular divergence between populations from the different island groups to infer the geologic history and formation of the Southern Ryukyus (the original goal of my NSF proposal) for which no hypothesis has received universal acceptance (Muraji et al., 2008). Lastly, as mentioned above, cryptic species complexes were revealed for each of these three ?species?. With increased sampling from areas such as New Caledonia, Easter Island, Christmas Island, the Philippines, Palau, Guam and other locations where these ?species? have been reported, additional morphological studies will be needed to resolve the taxonomic status of each complex and its potential species. 152 For Halocaridina, the proverbial ?white whale? appears to be a well-resolved phylogeny between the eight lineages. It is hypothesized that colonization patterns of Halocaridina across the Hawaiian Islands would follow the ?Progression Rule?, in which populations from older islands act as founders for ones on younger islands (Funk & Wagner 1995). However, despite a number of attempts with a wide range of molecular markers including microsatellites, exon- priming intron-crossing markers, amplified fragment length polymorphism (AFLP) and a number of nuclear loci, the relationships between the eight Halocaridina lineages are largely unresolved (Craft et al. 2008). Hopefully, through the utilization of next-generation sequencing technologies, appropriate markers can be identified to fully resolve these relationships which might provide new insights on dispersal and colonization in the archipelago (Craft et al. 2008). Leaving the anchialine system, I believe it would be interesting to investigate the population structure and phylogeographic history of some of the freshwater atyids inhabiting the streams of the Hawaiian and Ryukyu Archipelagos since such studies may shed new light on both the origin and causes of species diversity both within an archipelago and on individual islands (Emmerson, 2002). Although the more than 21 species of freshwater shrimp (Cai and Shokita, 2006) inhabiting the river and streams of the Ryukyu Islands have been relatively well studied in recent decades (Cai and Shokita, 2006), none have been examined from a genetic prospective. Such studies would offer the opportunity to (i) investigate the colonization history, evolution and origin of the Ryukyu shrimp fauna (e.g., Maekawa et al., 1999), (ii) test hypotheses regarding the effects of life history traits on the dispersal abilities and colonization of the ?inlandwater shrimps? of the Ryukyus put forth by Shokita (1979), (iii) examine the biological impacts of the Takara and Keroma Gaps on the freshwater diversity of the Ryukyus (e.g. Muraji et al., 2008). Furthermore, comparative phylogeographic studies between 153 freshwater and anchialine Carideans may provide further insight into the ecological and evolutionary forces acting on anchialine taxa in general. Lastly, recent advances in next-generation sequencing technologies have transformed the way we think about many aspects of molecular biology and evolution (Tautz et al., 2010). Utilizing these new techniques, a number studies regarding anchialine crustaceans could be envisioned, including, but not limited to: studying the genetics of phenotypic differences between subterranean and surface taxa through transcriptional profiling, examining genomic signatures of adaption to life in the anchialine environment and identifying single nucleotide polymorphisms (SNPs) to better understand diversity between closely related populations and species. 154 6.3 REFERENCES Avise, J.C. 2000. Phylogeography. Harvard Press, MA. Cai, Y. and S. Shokita. 2006. Atyid shrimps (Crusatcea: Decapoda; Caridea) of the Ryukyu Islands, southern Japan, with descriptions of two new species. Journal of Natural History 40:2123-2172. Craft, J. D., A. D. Russ, M. N. Yamamoto, T. Y. Iwai, S. Hau, J. Kahiapo, C. T. Chong, S. Ziegler-Chong, C. Muir, Y. Fujita, D. A. Polhemus, R. A. Kinzie, and S. R. Santos. 2008. Islands under islands: The phylogeography and evolution of Halocaridina rubra Holthuis, 1963 (Crustacean: Decapoda: Atyidae) in the Hawaiian archipelago. Limnology and Oceanography 53:675-689. Culver, D.C., T. Pipan, K. Schneider. 2009. Vicariance, dispersal and scale in aquatic subterranean fauna of karst regions. Freshwater Biology 54:918-929. Darwin, C. 1859. On the origin of species by means of natural selection, or the preservation of favored races in the struggle of life. John Murray, London. de Grave, S., and T.S. Sakihara. 2011. Further records of the anchialine shrimp, Periclimenes pholeter Holthius, 1973 (Crustacea, Decapoda, Palaemonidae). Zootaxa 2903:64-68. Emerson, B.C. 2002. Evolution on oceanic islands: molecular phylogenetic approaches to understanding pattern and process. Molecular Ecology 11: 951-966. Juan, C., M.T. Guzik, D, Jaume and S.J.B. Cooper. 2010. Evolution in caves: Darwin?s ?wreck of ancient life? in the molecular era. Molecular Ecology 19: 3865-3880. Maekawa, K., N. Lo., O. Kitade, T. Miura, and T. Matsumoto. 1999. Molecular phylogeny and geographic distribution of wood-feeding cockroaches in East Asian islands. Molecular Phylogenetics and Evolution 13:160-376. 155 McMillen-Jackson, A.L. and T.M. Bert. 2003. Disparate patterns of population genetic structure and population history in two sympatric penaeid shrim species (Farfantepenaeus aztecus and Liptopenaeus setiferus) in the eastern United States. Molecular Ecology 12:2895- 2905. Muraji, M., N. Arakaki, S. Ohno, and Y. Hirai. 2008. Genetic variation of the green chafer, Anomala albopilosa (Hope) (Coleoptera: Scarabaeidae), in the Ryukyu Islands of Japan detected by mitochondrial DNA sequences. Applied Entomology and Zoology 43:299- 306. Page, T.J. and J.M. Hughes. 2007. Radically different scales of phylogeographic structuring within cryptic species of freshwater shrimp (Atyidae: Caridina). Limnology and Oceanography 53:1055-1066. Porter, M.L. 2007. Subterranean biogeography: What have we learned from molecular techniques? Journal of Cave and Karst Studies 69:179-186. Russ, A.R., S.R. Santos, and C. Muir. 2010. Genetic population structure of an anchialine shrimp, Metabetaeus lohena (Crustacea: Alpheidae), in the Hawaiian Islands. Revista de Biologia Tropical 58:159-170. Santos, S. R. 2006. Patterns of genetic connectivity among anchialine habitats: a case study of the endemic Hawaiian shrimp Halocaridina rubra on the Island of Hawaii. Molecular Ecology 15: 2699-2718. Smith, M.J. and W.D. Williams. 1981. The occurrence of Antecaridina lauensis (Edmondson)(Crustacea, Decapoda, Atyidae) in the Solomon Islands: An intriguing biogeographical problem. Hydrobiologia 85:49-58. 156 Shokita, S. 1979. The distribution and speciation of the island water shrimps and prawns from the Ryukyu Islands. Bulletin of the college of Science, University of the Ryukyus 28:193-278. Tautz, D., H. Ellegren, and D. Weigel. 2010. Next generation molecular ecology. Molecular Ecology 19:1-3. Warrick, J. 1998. Mass Extinction Underway, Majority of Biologist Say, in Washington Post: Washington, DC. USA.