Multilocus allozyme heterozygote deficiencies in Crepidula onyx: geographic and temporal patterns among adult snails in Mission Bay, California

Although the phenomenon of multilocus heterozygote deficiencies in numerous taxonomically diverse molluscs has been studied for over 30 years it has resisted general explanation. Here we describe a case in Crepidula onyx Sowerby, a common protandrous slipper limpet on the Pacific coast of southern California. Spatial and temporal differentiation at 16 polymorphic allozyme loci were examined in samples of adult snails from three or four seasons and six localities in Mission Bay, San Diego. The populations were highly polymorphic (proportion of polymorphic loci P-L=0.65-0.69; mean individual heterozygosity H-O=0.13-0.29) and although there were slightly more changes in allele frequency between autumn and winter than between winter and spring, statistically significant spatial and temporal change or patterning was not detected. If the larvae are passive dispersers the lack of population structure is attributable to the fact that during any three-week period that they are in the plankton, the waters of Mission Bay are completely mixed by tidal currents. In marked contrast, statistically significant heterozygote deficiencies were found in all samples (locality and season) and subsamples (males and females). At each locality and season, deficiencies were as pronounced in males as in females. Furthermore, at each locality, there was no increase in level of multilocus deficiency between males in one season and females in a subsequent season. We conclude that the heterozygotes had been lost in the larval planktonic stage or at the time of metamorphosis and settlement. Several hypotheses to explain the heterozygote deficiencies in adults are discussed: inbreeding, self-fertilization, presence of null alleles, allele scoring bias, aneuploidy, parental imprinting, spatial and temporal Wahlund effects, age-effects, and direct natural selection on the marker loci. Although we can reject some of these hypotheses, others like inbreeding, require testing with markers with greater statistical power. Wahlund effects due to inadvertent mixing of two cryptic species and/or micro-spatial population genetic structuring due to reproductive asynchronies appear unlikely causes. Natural selection against heterozygotes in the larval phase of life may account for the observations, but selection does not appear to operate directly on any of the individual loci studied. An hypothesis is offered to explain the general selection against heterozygotes during the pelagic larval stage, and for the reverse during the sessile adult stage. Selection coefficients would, however, have to be unusually high and we conclude that we are unable satisfactorily to explain our observations.