AbstrAct.-Edge-of-range populations are often typified by patterns of low genetic diversity and high genetic differentiation relative to populations within the core of a species range. The "core-periphery hypothesis," also known as the "central-marginal hypothesis," predicts that these genetic patterns at the edge-of-range are a consequence of reduced population size and connectivity toward a species range periphery. It is unclear, however, how these expectations relate to high dispersal marine species that can conceivably maintain high abundance and high connectivity at their range edge. In the present study, we characterize the genetic patterns of two tropical echinoderm populations in the Kermadec Islands, the edge of their southwest Pacific range, and compare these genetic patterns to those from populations throughout their east Indian and Pacific ranges. We find that the populations of both Acanthaster planci (Linnaeus, 1758) and Tripneustes gratilla (Linnaeus, 1758) are represented by a single haplotype at the Kermadec Islands (based on mitochondrial cytochrome oxidase c subunit I). such low genetic diversity concurs with the expectations of the "core-periphery hypothesis." Furthermore, the haplotypic composition of both populations suggests they have been founded by a small number of colonists with little subsequent immigration. Thus, local reproduction and self-recruitment appear to maintain these populations despite the ecologically marginal conditions of the Kermadec Islands for these tropical species. Understanding rates of self-recruitment vs reliance on connectivity with populations outside of the Kermadec Islands has implications for the persistence of these populations and range stability of these echinoderm species.Population attributes are expected to differ according to range position. Differences in abundance, reproduction, and dispersal (connectivity) can often be observed across a species range, and over evolutionary timescales these population attributes are expected to lead to predictable neutral genetic patterns across their range. For

AbstrAct.-Marine animals inhabiting the Indian and Pacific oceans have some of the most extensive species ranges in the world, sometimes spanning over half the globe. These Indo-Pacific species present a challenge for study with both geographic scope and sampling density as limiting factors. Here, we augment and aggregate phylogeographic sampling of the iconic blue sea star, Linckia laevigata Linnaeus, 1758, and present one of the most geographically comprehensive genetic studies of any Indo-Pacific species to date, sequencing 392 base pairs of mitochondrial cOI from 791 individuals from 38 locations spanning over 14,000 km. We first use a permutation based multiple-regression approach to simultaneously evaluate the relative influence of historical and contemporary gene flow together with putative barriers to dispersal. We then use a discrete diffusion model of phylogeography to infer the historical migration and colonization routes most likely used by L. laevigata across the Indo-Pacific. We show that estimates of genetic structure have a stronger correlation to geographic distances than to "oceanographic" distances from a biophysical model of larval dispersal, reminding us that population genetic estimates of gene flow and genetic structure are often shaped by historical processes. While the diffusion model was equivocal about the location of the mitochondrial most recent common ancestor (MrcA), we show that gene flow has generally proceeded in a step-wise manner across the Indian and Pacific oceans. We do not find support for previously described barriers at the sunda shelf and within cenderwasih bay. rather, the strongest genetic disjunction is found to the east of cenderwasih bay along northern New Guinea. These results underscore the importance of comprehensive range-wide sampling in marine phylogeography.