Habitat and the shell polymorphism of Cepaea nemoralis (L.): interrogating the Evolution Megalab database

Abstract

The Evolution Megalab database on the shell polymorphism of Cepaea nemoralis contains data on the majority of populations sampled over the last 100 years. Here, we use the database to examine geographical variation in the strength of the known relationship between morph frequencies and habitat. Unique pairs of samples from woodland and the nearest open habitat (hedges and grassland) are used to analyse the direction and magnitude of differences in morph frequencies. Shell colour shows stronger and more consistent variation with habitat than do banding patterns, yellow shells being at higher frequencies in open habitats than in woods. There is, however, marked geographical variation in this relationship, which is stronger in southern England and northern France than elsewhere. In some regions there is no evidence of any effect. Only in southern England are there significant associations between habitat and banding, with unbanded shells being at higher frequencies in woods than in open habitats. Within Great Britain a more refined analysis shows that the effects of habitat are evident mainly in areas of ancient countryside, in which many woods suitable for the species have existed for centuries. Less stable areas frequently exhibit ‘area effects’ in which variation does not relate to habitat. In a smaller set of pairs, such regional variation is not apparent in Germany. Earlier disputes about the role of drift and natural selection in determining patterns are in part a reflection of the region in which studies were carried out. The Megalab database does not cover all aspects of the polymorphism. In particular, band fusions are not recorded. The lack of evident effect in many regions does not demonstrate that there is none; more data and some knowledge of historical changes in habitats might reveal trends not apparent in this limited and conservative study.