No abstract

Bvtltinella are minute prosobranch gastropods inhabiting springs. The northern border of the distribution area of the genus goes through South Poland. We studied departures from Hardy-Weinberg equilibrium (HWE) within and the level of gene flow between 20 Central European populations of Bythinda. From nearly all populations 30 or more specimens were assayed for eight enzyme systems coded by nine loci (AAT, ALP, GPI, IDDH-I, IDDH-2, MDH, MDHP, PGDH, PGM). Six polymorphic loci (ALP, GPI, IDDH-I, IDDH-2, PGDH, PGM) were analyzed for departures from HWE by means of exact probability tests and F-statistics. Two populations (9 and 20) were in HWE for all the loci, heterozygote excess was never observed. Heterozygote deficiency was found for two loci in two populations (population 2: IDDH-I and PGM; population 13: IDDH-2 and PGM), and for one locus in the remaining 16 populations, which in 13 cases was IDDH-1. F-statistics and other data suggest that not inbreeding but other disturbing forces appear responsible for the observed deficiencies. We assume differential temporal selection on juveniles, the Wahlund effect, the negative fitness of heterozygote juveniles due to their higher food demands as limiting factor, and/or the lower growth rate of homozygotes leading to a longer life span as possible causes. The presented estimates of gene flow were computed from B (Wright's FST) and with the private allele technique; the former was found to reflect the reality better. Compared with similar estimates known from the literature for molluscs, the average value of B was high in our populations. Pairwise B values for each pair of populations were used to compute a minimum spanning tree projected on multidimensional scaling to visualize the pattern of interpopulation allozymic differences. Hierarchical F-statistics did not confirm the reality of the earlier distinguished species of the genus. 0 was, whereas N,,, was not, significantly correlated with geographic distance.

Morphology and anatomy of 20 Central European populations of the spring snail Bythinella Moquin-Tandon are described and illustrated. Detailed descriptions, drawings and SEM photographs, as well as measurements of shells, opercula, radulae and soft parts are presented, special attention being paid to the reproductive organs. Intrapopulation and interpopulation differences in 40 biometrical characters in males and 42 in females are analysed, for each sex separately, by means of descriptive statistics and ANOVA. Relations between variables are tested with Kendall's correlation coefficient. Biometrical characters vary widely within and slightly between the studied populations, whose variability ranges overlap. The length ratios: shell/radula, bursa/bursa duct and bursa/receptacle, as well as the length of bursa copulatrix and ctenidium display a statistically significant S-N clinal variation. The wide intrapopulation variation combined with the slight differences between the distinguished morphospecies in almost all the morphological characters examined indicates a limited diagnostic value of these characters in Bythinella.

The paper describes morphometric and allozymic differences among four European species of the family Viviparidae: Viviparus contectus (Millet, 1813), V. viviparus (Linnaeus, 1758), V. acerosus Bourguignat, 1862, and V. ater (Cristofori et Jan, 1832). Fourteen continuous biometrical characters were measured. Incremental discriminant‐function analysis, principal‐component analysis, and non‐metric multidimensional scaling were applied to analysis of the morphometric differences. All the techniques confirmed a similar picture: a slight morphometric differentiation, with the variability ranges of the species overlapping. On the other hand, the allozymic differentiation, studied at 12 loci, eight of them intra and/or interspecific polymorphic, is much better marked, the intraspecific Nei's distances among the four V. contectus populations ranging from 0.0014 to 0.0397, mean 0.0166, and interspecific distances ranging from 0.2306 (V. ater–V. acerosus) to 0.9888 (V. contectus 2 and V. viviparus), mean 0.6871. The allele frequencies and genetic distances (Nei's distance and Cavalli‐Sforza and Edwards’chord distance) were used to compute maximum likelihood, additive Fitch–Margoliash and ultrametric Fitch–Margoliash trees. All the trees presented a similar pattern, but the maximum‐likelihood and additive trees, based on Cavalli‐Sforza and Edwards’distance, seem to reflect the phylogeny best. The results are compared with the most parsimonious phylogenies inferred for radular, soft‐part morphology and anatomy, and opercular data from other papers by us, and the inferred phylogenies are also compared. Although the inferred molecular and morphological phylogenies are little different in topology, the amount of evolution along the corresponding branches (measured as the number of changes averaged over all reconstructions) is very different, the value of correlation coefficient between the two phylogenies being statistically insignificant. The occurrence of interspecific hybrids is discussed, and the isolation‐by‐competition mechanism is postulated. The probable origin of V. viviparus from a founder population extremely restricted in number is stressed. The possible history of the group is briefly discussed. The species is suggested to have originated in an unusual habitat of melt water at a glacier foreland that could have promoted genotypic differentiation and sympatric speciation.

No abstract