Spatially coupled larval supply of marine predators and their prey alters the predictions of metapopulation models.

Oceanographic forces can strongly affect the movement of planktonic marine larvae, often producing predictable spatial patterns of larval delivery. In particular, recent empirical evidence suggests that in some coastal systems, certain locations consistently receive higher (or lower) larval supplies of both predators and their prey. As a consequence, rates of settlement and predation may be coupled spatially, a phenomenon I term the "coupled settlement effect." To investigate the metapopulation consequences of this phenomenon, I created discrete-time, patch-based analytical and simulation models with a common larval pool and uneven larval supply among patches. Using two complementary measures of subpopulation value as a basis of comparison, I found that models with and without the coupled settlement effect yielded strikingly different predictions. When prey and predator larval supplies were not coupled, patches supplied with a larger proportion of the larval pool made a greater contribution to the metapopulation. When settlement of prey and predator was strongly coupled, however, the opposite was true: subpopulations with lower rates of larval supply (above some minimum) were more essential to metapopulation persistence. These considerations could facilitate more effective selection of sites for protection in marine reserves.