Current shifts and kin aggregation explain genetic patchiness in fish recruits.

The scales of population structure in marine species depend on the degree to which larvae from different populations are mixed in the plankton. There is an intriguing trend in marine population genetic studies of significant genetic structure for larvae, recruits, or populations at fine scales that is unpatterned across space and changes through time. This "chaotic genetic patchiness" suggests that larval pools are not well mixed in the plankton. However, few studies have been able to distinguish among potential causes of spatial and temporal genetic heterogeneity: changes in larval migration patterns, changes in environmental selection, or stochasticity caused by "sweepstakes" reproductive success of spawners creating detectable family structure. Here we use microsatellite markers to show that significant allele frequency shifts occurred sporadically in space and time for cohorts of recruits of Paralabrax clathratus (kelp bass) collected once every two weeks over two years from five sites in the Santa Barbara Channel, California, USA. We found that the pattern of genetic differentiation among cohorts was explained by a combination of (1) family structure in some cohorts, evidenced by half and full siblings, and (2) an indication of changes in larval delivery. It is unlikely but possible that environmental selection also plays a role. Although sampling of potential source populations was incomplete, cohorts arriving during western current flows show most genetic similarity with a population sample collected in the west, and cohorts arriving during current flows from the southeast show similarity with population samples collected in the south and east. Despite the family structure apparent in some cohorts, these "sweepstakes" events occur on too fine a scale to create lasting year class genetic structure. The results corroborate oceanographic models of larval dispersal, which suggest that larval mixing in the plankton is less extensive than previously believed.