Variation with habitat in Cepaea nemoralis: the Cain & Sheppard diagram

In the 1950s Cain and Sheppard showed that in mature agricultural landscape in southern England samples of the polymorphic snail Cepaea nemoralis (L.) from woods had more unbanded, non-yellow individuals and fewer yellow bandeds than those from hedgerows and grassland. They considered mid-banded shells to be visually unbanded. This pattern was interpreted as resulting from selective predation acting on the phenotype rather than on specific genotypes. Evidence from this and 18 further data sets has been re-analysed. The association of yellow banded with open habitats and non-yellow, unbanded with woods is confirmed. The distribution of mid-banded is not consistent, and inclusion of other effectively unbanded phenotypes makes little difference to the result. The association therefore depends on the colour/banding system, but very little on other aspects of the phenotype. It is not incontrovertible evidence of selective predation; non-visual differences in fitness could also be involved. Further direct observation of the action of predators is needed to assess the argument for selective predation.