The phylogeography of three species of African bovids, the hartebeest (Alcelaphus buselaphus), the topi (Damaliscus lunatus), and the wildebeest (Connochaetes taurinus), is inferred from sequence variation of 345 sequences at the control region (d-loop) of the mtDNA. The three species are closely related (tribe Alcelaphini) and share similar habitat requirements. Moreover, their former distribution extended over Africa, as a probable result of the expansion of open grassland on the continent during the last 2.5 Myr. A combination of population genetics (diversity and structure) and intraspecific phylogeny (tree topology and relative branch length) methods is used to substantiate scenarios of the species history. Population dynamics are inferred from the distribution of sequence pairwise differences within populations. In the three species, there is a significant structuring of the populations, as shown by analysis of molecular variance (AMOVA) pairwise and hierarchical differentiation estimations. In the wildebeest, a pattern of colonization from southern Africa toward east Africa is consistent with the asymmetric topology of the gene tree, showing a paraphyletic position of southern lineages, as well as their relatively longer branch lengths, and is supported by a progressive decline in population nucleotide diversity toward east Africa. The phylogenetic pattern found in the topi and the hartebeest differs from that of the wildebeest: lineages split into monophyletic clades, and no geographical trend is detected in population diversity. We suggest a scenario where these antelopes, previously with wide pan-African distributions, became extinct except in a few refugia. The hartebeest, and probably also the topi, survived in refugia north of the equator, in the east and the west, respectively, as well as one in the south. The southern refugium furthermore seems to have been the only place where the wildebeest has survived.

Intraspecific phylogeographic methods provide a means of examining the history of genetic exchange among populations. As part of a study of the history of Helix aspersa in the Western Mediterranean, we performed a phylogenetic analysis based on partial sequences of the mitochondrial large ribosomal subunit (16S) gene. Our samples include 31 H. a. aspersa populations from North Africa previously investigated for anatomical and biochemical characters. To clarify subspecific relationships, three individuals of the subspecies H. a. maxima were also studied. The molecular phylogeny inferred agrees largely with previous results, in splitting H. a. aspersa haplotypes into an eastern and a western group. H. a. maxima haplotypes form a third lineage arising before the H. a. aspersa groups. Divergence times estimated between the lineages suggest that dispersal during Pleistocene glaciation and vicariance events due to Pliocene geological changes in the western Mediterranean may both have played a significant part in the establishment of the present range of H. aspersa.

Reproductive traits of the land snail Helix mpersa Miiller were' investigated under artificial conditions from two samples, one collected from a population exposed to unpredictable human pressures in its natural environment, i.e. a recently created polders area with intensive agriculture, and the other from a snail farm in which animals were reared under constant conditions defined as 'optimal' for growth and reproduction. Results were compared with data collected from natural populations of the same region (Brittany) and from habitats spanning the environmental heterogeneity of the range of the species. A large part of the variation among populations could be explained by different phenotypic covariances between shell size, clutch size and egg size, but not by the number of clutches per snail. Thus, the higher egg production of snails from the polders was related to (i) a strong correlation between clutch size and shell size, shell size being in the upper limit of the overall range for the region concerned; (ii) an uncommonly low egg weight in comparison with the 'norm' of Helix aspersa, this trait seeming to be involved in a trade-off with clutch size. Second clutches were smaller than the first ones, but their eggs were significantly heavier. This difference may be linked to a size-dependent mortality of juveniles during winter which arises in all populations in which hibernation occurs as an adaptation to low temperatures. In addition to the selective regime usually involved for populations of helicid snails from Western Europe, several unpredictable mortality factors occurred in the polders area: herbicide and pesticide treatments (lethal for young snails), human predation (lethal for adults) and burning (lethal for ali snails). Life-history patterns of Helix aspma are discussed in relation to its ability to successfully colonize a large range of habitats modified by humans, to such an extent that it can become a pest. 0 ZOO0 The Linnean Society of London

Inbreeding depression was estimated from an outbreeding population of the freshwater snail Lymnaea peregra, on the basis of two successive generations of enforced selling and outcrossing, and 70 maternal lineages. Outcrossing was analyzed under two treatments, groups of two and five individuals. The fitness parameters measured included fecundity, growth, and survival. In the first generation, we contrasted three treatments (selfers vs. paired outcrossers and group outcrossers). Very similar results were obtained between the two outcrossing treatments. A strong self-fertilization depression (which includes parental fecundity and progeny fitness) was detected in the selling treatment (about 90%). In the second generation, there was again marked evidence for self-fertilization depression, with the highest contributions coming from parental fecundity and progeny hatching rate. Our results suggest that the decreased parental fecundity is a consequence of the mating system in the previous generation, although the role of partial self-incompatibility and the copulation behavior could not be ruled out. Hatching rate and early survival data are suggestive of purging of lethal mutations. Significant variation in fitness among selfing lineages was found for most fitness traits. Our experimental design also allowed to test for interactions among fitness loci. Only one trait of the nine studied behaved as expected under synergistic interactions. However we cannot rule out some purging during the experiment, which could have biased results towards linearity. Inbreeding depression was also inferred from the change of inbreeding level across generations in the same population. We obtained a value similar to the experimental estimate.

No abstract