A Comparison of Populations of the Polymorphic Land Snail CEPAEA NEMORALIS (L.) Living in a Lowland District in France with Those in a Similar District in England.

AIN and SHEPPARD (1950, 1954) considered that natural selection, in particular visual selection exerted by predators, was an important agent influencing shell color and banding-morph frequencies in populations of Cepaea nemoralis in England. LAMOTTE ( 195 1 ) on the other hand, working with French populations, concluded that selection was of little significance in most colonies, and that random evolutionary effects exerted an overriding effect. Subsequent research has modified some of the early conclusions, but the differences in opinions are still trenchant (e.g., CAIN and SHEPPARD 1961; CAIN and CURREY 1963; LAMOTTE 1959, 1966). Work in England has tended to confirm CAIN and SHEPPARD’S conclusions. At least in lowland and wooded districts in England, further circumstantial evidence has been forthcoming both that visual selection is an important factor, and that there is little reason to believe that random effects have operated to an appreciable extent in most colonies (CURREY, ARNOLD and CARTER 1964; CARTER 1968; CAIN and CURREY 1963). LAMOTTE (1959, 1966) describes variation both of color and banding-morph frequencies with habitat in populations in some parts of France. He is more inclined, however, t o attribute the phenomenon to climatic selection (favoring yellow and unbanded snails in open habitats and pink and banded ones in woods land shady habitats) than to the effects of visual selection. He also considers that the weakness of the correlation of morph frequencies with habitat is mainly due to random evolutionary events. GUERRUCCI-HENRION (1 966) has found that the frequencies of the color morphs in populations in Brittany are correlated with the amount of precipitation, and suggested that drier climates favor the yellow morph. She concluded, however, that most of the intercolony variation in banding-morph frequencies in Breton populations was best explained by random effects. The methods employed by both groups of workers differed. LAMOTE sampled extensively and irrespective of habitat type, and considered color and banding-morph frequencies separately. CAIN and SHEPPARD chose well defined habitats, selected largely within a small radius around Oxford, and considered color and bandingmorph frequencies together in any one colony. Direct comparison and interpretation of results has been somewhat difficult,