Evidence of a reproductive barrier between two forms of the marine periwinkleLittorina fabalis(Gastropoda)

Abstract

Studies of allozyme variation may reveal unexpected patterns of genetic variation which challenge earlier conclusions of species delimitations based on morphological data. However, allozyme variation alone may not be sufficient to resolve this kind of problem. For example, populations of the marine intertidal snailLittorina fabalis(=Littorina mariae) from wave exposed parts and from protected parts of the same shores are distinguished by different alleles of arginine kinase (Ark) while indifferent, or very nearly so, in another 29 loci. Intermediate populations have large deficiencies of exposed/sheltered heterozygote classes ofArkand we have earlier suggested habitat-related selection in this locus as the explanation. In this study we estimated growth rate of individual snails of differentArk-genotypes in three different habitats (exposed, sheltered and intermediate). In all habitats the snails homozygous for alleles of ‘exposed’ type grew faster and matured at a larger size than did snails homozygous for alleles of ‘sheltered’ types. This relationship was indirectly confirmed in three additional sites of intermediate exposure where exposedArk-genotypes dominated among large (>8 mm) snails while the sheltered genotypes dominated among small (<5 mm) snails of truly sympatric samples. We furthermore found small differences in allele frequencies of two other loci (PgiandPgm-2) and in shell colour frequencies, comparing sympatric snails of exposed and shelteredArk-homozygotes. Although we found no signs of habitat-related selection among snails of differentArk-genotype, or selection against heterozygotes, we cannot reject selection inArk, as our experiments only covered one island, one season and grown-up snails. The coupling between allozyme and phenotypic characters in strictly sympatric samples of snails suggests the presence of two gene pools. Perhaps the large and small forms ofL. fabalisrepresent very closely related cryptic taxa. However, introgression between them seems a possible explanation for the striking similarities in the vast majority of morphological and allozyme characters.