Genetic mark-recapture provides insights into bee movement and plant reproductive success

Genetic data offer a means of inferring the contemporary and historical movement of organisms, in study systems where direct observation is infeasible. However, the use of genetic markers as a proxy for the direct observation of movement presents its own challenges: the observed quantities (genotypes) are fundamentally stochastic. In many cases, movement can only be inferred from molecular markers by exploiting familial relationships among organisms. From a statistical perspective, this poses a unique challenge that requires linking ecological or behavioral hypotheses to an inherently noisy and constrained observation process. This thesis develops and applies statistical models to answer basic questions about movement in bees – a group of organisms that have tremendous ecological and commercial importance but are too small and too motile to track directly – by using molecular markers and exploiting family relationships among bees and among the plants they pollinate. Thematically, this thesis is organized into five chapters split across three sections. The first section (chapters 1 and 2) concerns bee foraging movements in times of food scarcity, and employs as a study system a common species of bumble bee in the Californian chaparral. The second section (chapters 3 and 4) concerns the spatial context of plant reproductive success, and uses as a study system a widely distributed tropical understory tree that is pollinated by a functionally diverse bee community. The fifth chapter concerns constraints on dispersal movements, and develops a statistical methodology for inferring how environmental heterogeneity influences migration rates, given patterns of extant genetic variation.