Habitat modification and gene flow in Saimiri oerstedii: Landscape genetics, intraspecific molecular systematics, and conservation

Habitat modification, when it results in population fragmentation, often results in the loss of genetic diversity due to reduced gene flow, inbreeding, and genetic drift. However, the severity of these effects depends on how diminished dispersal and gene flow become between patches of suitable habitat. An empirical understanding of how habitat change affects dispersal and gene flow within and among patches is essential to predict the effects of increased habitat modification and landscape change on population persistence and processes of divergence. Recent studies in landscape ecology suggest that our understanding of dispersal in a heterogeneous landscape will improve by explicitly considering the heterogeneity of matrix habitats, or unsuitable habitats between patches of suitable habitat. In this dissertation, I describe population genetic structure and dispersal patterns in the Central American Squirrel Monkey (Saimiri oerstedii, Primates: Cebidae), a New World primate threatened with extinction and living in a heterogeneous, human-modified landscape, using analyses that explicitly consider matrix heterogeneity.;I focus on the more endangered S. o. citrinellus, whose already restricted distribution in the Central Pacific region of Costa Rica has undergone considerable anthropogenic modification since the early 1900s. I collected non-invasive fecal samples from S. o. citrinellus across the Central Pacific region, obtaining full genotypes from 233 individuals. I also obtained 11 samples from S. o. oerstedii in the Southern Pacific region of Costa Rica from a collaborator, as well as fine-scale landscape data for the Central Pacific.;I analyzed the data using molecular systematics, population genetics, and landscape genetic techniques. In this dissertation, first I explore whether molecular genetic support exists for the subspecies distinction between S. o. citrinellus and S. o. oerstedii. Second, I describe population genetic structure and recent migration patterns within S. o. citrinellus using traditional population genetic methods and Bayesian models. I also compare population genetic structure among males versus females to test for sex-biased dispersal patterns in S. o. citrinellus . Then, using landscape genetic approaches, I describe the relationship between landscape heterogeneity and genetic structure in S. o. citrinellus , and inferred which matrix habitats are costly to dispersal. Finally, I offer explicit recommendations for the conservation management of S. oerstedii.;My results provide genetic support for S. o. citrinellus and S. o. oerstedii as separate taxa referred to as subspecies. Also, I found evidence of population genetic structure in S. o. citrinellus , with two genetically distinct populations and lower genetic diversity in the western population. I did not find genetic evidence for female-biased dispersal in S. o. citrinellus as expected. Instead, my results suggest that both sexes disperse, with males dispersing over longer distances. The landscape genetic analysis suggests that landscape heterogeneity is important in determining local population genetic structure in S. o. citrinellus in the Central Pacific region of Costa Rica. Specifically, oil palm plantations are moderate barriers to gene flow between populations, but not other matrix habitats. However, these inferences are specific to the composition and configuration of the Central Pacific landscape, and should not be generalized to all S. oerstedii populations.;This study generated important information for conservation management. Based on my results, I recommend that conservation managers house the two S. oerstedii subspecies separately in captive facilities, and only transfer, reintroduce, or translocate among groups of the same subspecies. However, transfers, reintroductions, or translocations of either males or females are both likely to be successful for S. o. citrinellus in the Central Pacific region, pending further behavioral study. I also recommend that, in order to augment dispersal to the isolated western population of S. o. citrinellus, conservation efforts should focus on building biological corridors through or around adjacent oil palm plantations. Also, managers should prioritize the maintenance of existing forest connectivity in the Central Pacific region.;The results also have important implications for future studies of evolutionary and ecological processes in heterogeneous landscapes. This study contributes to a growing body of research that finds differences in dispersal patterns among local primate populations of the same taxon. My results suggest that predictive models for variation in dispersal patterns should consider both variation among the environments of local populations within a species and temporal variation in local environments (e.g. recent habitat disturbance). Finally, this dissertation also supports the idea that matrix heterogeneity should be considered explicitly in studies of dispersal and gene flow, as opposed to assuming that all non-suitable habitats have a uniform effect on these processes. In the future, agent-based simulation approaches combined with ecological niche models and data on adaptive genetic diversity could expand upon this work to inform predictive models for population divergence and speciation under different climate and landscape change scenarios.