The degree of genetic connectivity among populations in a metapopulation has direct consequences for species evolution, development of disease resistance, and capacity of a metapopulation to adapt to climate change. This study used a metapopulation model that integrates population dynamics, dispersal, and genetics within an individual-based model framework to examine the mechanisms and dynamics of genetic connectivity within a metapopulation. The model was parameterized to simulate four populations of oysters (Crassostrea virginica) from Delaware Bay on the mid-Atlantic coast of the United States. Differences among the four populations include a strong spatial gradient in mortality, a spatial gradient in growth rates, and uneven population abundances. Simulations demonstrated a large difference in the magnitude of neutral allele transfer with changes in population abundance and mortality (on average between 14 and 25% depending on source population), whereas changes in larval dispersal were not effective in altering genetic connectivity (on average between 1 and 8%). Simulations also demonstrated large temporal changes in metapopulation genetic connectivity including shifts in genetic sources and sinks occurring between two regimes, the 1970s and 2000s. Although larval dispersal in a sessile marine population is the mechanism for gene transfer among populations, these simulations demonstrate the importance of local dynamics and characteristics of the adult component of the populations in the flow of neutral alleles within a metapopulation. In particular, differential adult mortality rates among populations exert a controlling influence on dispersal of alleles, an outcome of latent consequence for management of marine populations.