Accounting for connectivity is essential in marine spatial planning and the proper design and management of marine protected areas, given that their effectiveness depends on the patterns of dispersal and colonization between protected and non-protected areas. The genetic structure of populations is commonly used to infer connectivity among distant populations. Here, we explore how population genetic structure is affected by pre- and settlement limitations with a spatially explicit coupled metapopulation-gene flow model that simulates the effect of demographic fluctuations on the allele frequencies of a set of populations. We show that in closed populations, regardless of population growth rate, the maintenance of genetic diversity at saturating initial population density increases with species life expectancy as a result of density-dependent recruitment control. Correlatively, at low initial population density, the time at which a population begins to lose its genetic diversity is driven larval and post-settlement mortality (comprised in the recruitment success parameter)—the larger the recruitment success, the stronger the genetic drift. Different spatial structures of connectivity established for soft bottom benthic invertebrates in the Gulf of Lions (NW Mediterranean, France) lead to very different results in the spatial patterns of genetic structuration of the metapopulation, with high genetic drift in sites where the local retention rate was larger than 2%. The effect of recruitment failure and the loss of key source populations on heterozygosity confirm that transient demographic fluctuations help maintain genetic diversity in a metapopulation. This study highlights the role of intraspecific settlement limitations due to lack of space when the effective number of breeders approaches saturating capacity, causing a strong reduction in effective reproduction. The present model allows to: (i) disentangle the relative contribution of local demography and environmental connectivity in shaping seascape genetics, and (ii) perform in silico evaluations of different scenarios for marine spatial planning.