Studies onCepaeaII. Area effects and visual selection inCepaea nemoralis(L.) andCepaea hortensis

Abstract

The snailsCepaea nemoralisandC. hortensisshow an extensive and stable polymorphism involving the colour and banding pattern of the shell. The surveys on the Berkshire Downs, the Purbeck Hills and part of lowland Somerset reported here show, together with previous ones, that both species can occur in lowland and chalk upland localities, but the distribution of the two species in any one locality is different. This difference may be related to topography as on the Berkshire Downs and the Purbecks, or to habitat as in south-east Somerset. There are no regional differences in theC. nemoralispolymorphism, most morphs so far recognized being present in every locality. The colour polymorphism inC. hortensisis much more extensive in the southern localities than in those from central England and there are alo regional differences in the banding polymorphism. Two types of morph frequency variation occur in both species. The first described was variation with habitat, the second, variation related to geographical position. The surveys show that habitat variation occurs in both lowland and upland areas, although it does not necessarily occur in either. Thus both species show morph frequency variation with habitat round Oxford and on the eastern Berkshire Downs. OnlyC. hortensisshows such variation on the Purbeck Hills althoughC. nemoralisis present there. The evidence, although not conclusive, indicates that neither species shows habitat variation in lowland south-east Somerset, but there is good evidence that neither species shows such variation on the western Berkshire Downs, the Marlborough Downs or Salisbury Plain. The morphs showing frequency variation with habitat are not always the same in either species. Round Oxford, and on the eastern Berkshire Downs, brown, pink and bandedC. nemoraliswith at least the upper two bands missing, are at a high frequency in woods, and yellows and bandeds are at a high frequency in other habitats.C. hortensis, on the other hand, shows variation in the frequency of fused bands and yellow effectively unbandeds between habitats in these two localities. On the central Berkshire Downs, spread banded show habitat variation together with other morphs inC. nemoralis. On the Purbeck Hills where pinks, browns and unbandeds are common in C. hortensis the habitat variation is very similar to that ofC. nemoralisround Oxford. Habitat variation is not restricted to strikingly different morphs. Different types of pinkC. nemoralisshow such variation on the Berkshire Downs, and fusions of the upper or more bands inC. hortensisalso show variation there. All the variation of morph frequency described above is best explained on the hypothesis of visual selection. Morph frequency variations related to geography give rise to large areas of stable morph frequency irrespective of habitat, the so-called area effects. The surveys to date indicate that area effects are restricted in both species to upland localities. The Berkshire Downs survey ofC. nemoralisshows that area effects and habitat variation are not mutually exclusive. In different localities, related on an east/west axis, most morphs may show habitat variation, some show habitat variation and some area effects, or most show area effects. Area effects may occur in both species in the same locality as on the western Berkshire Downs, where the limits of the various area effects in the two species are semi-coincidental, or they may only occur in one species,C. nemoralis, on the Purbeck Hills. There is no experimental evidence as to the cause of area effects but there is some evidence from the present surveys which supports Cain & Currey’s (1963a, c) suggestion that they are maintained by selection.