Variation of shell shape in the clonal snail Melanoides tuberculata and its consequences for the interpretation of fossil series

Interpreting paleontological data is difficult because the genetic nature of observed morphological variation is generally unknown. Indeed, it is hardly possible to distinguish among several sources of morphological variation including phenotypic plasticity, sexual dimorphism, within-species genetic variation or differences among species. This can be addressed using fossil organisms with recent representatives. The freshwater snail Melanoides tuberculata ranks in this category. A fossil series of this and other species have been studied in the Turkana Basin (Kenya) and is presented as one of the best examples illustrating the punctuated pattern of evolution by the tenants of this theory. Melanoides tuberculata today occupies most of the tropics. We studied variation of shell shape in natural populations of this parthenogenetic snail using Raup's model of shell coiling. We considered different sources of variation on estimates of three relevant parameters of Raup's model: (1) variation in shell shape was detected among clones, and had both genetic and environmental bases; (2) sexual dimorphism, in those clones in which males occur, appeared as an additional source of shell variation; and (3) ecophenotypic variation was detected by comparing samples from different sites and years within two clones. We then tested the performance of discriminant function analyses, a classical tool in paleontological studies, using several datasets. Although the three sources of variation cited above contributed significantly to the observed morphological variance, they could not be detected without a priori knowledge of the biological entities studied. However, it was possible to distinguish between M. tuberculata and a related thiarid species using these analyses. Overall, this suggests that the tools classically used in paleontological studies are poorly efficient when distinguishing between important sources of within-species variation. Our study also gives some empirical bases to the doubts cast on the interpretation of the molluscan series of the Turkana Basin.