A non-excitable behavioural mutant, d4-662, was previously characterized as the fourth pawn locus mutant pwD in Paramecium tetraurelia. We now provide data demonstrating that d4-662 is in fact controlled by a pwB allele that has the unusual feature of complementing other pwB alleles in heterozygous F1 progeny. Neither the cytoplasm nor the nucleoplasm of d4-662 cured the mutational defects of pwB and in the reverse combination of d4-662 and pwB, the result was the same. On the other hand, pwA, another non-excitable mutant, was cured upon cross-injection with d4-662 and mutants carrying trichocyst non-discharge marker genes were also cured. This evidence suggests that d4-662 is a new mutant belonging to pwB, and would be better designated as pwB662. Extensive crossbreeding analyses, however, showed an unusual genetic relationship between d4-662 and pwB (pwB95 or pwB96). When d4-662 was crossed with pwB mutants, many progeny expressing wild-type phenotype or mixed clones of wild-type and pawn cells were obtained in the F1. Less than 12·5% expressed the pawn phenotype. The appearance of wild-type progeny in this F1 strongly suggests that an inter-allelic interaction between pwB662 and other pwB alleles may occur during development of the macronucleus.

AxinFu is a mutation in mice that causes fused tails and other developmental abnormalities as a result of insertion of an intracisternal-A particle (IAP), a murine retrotransposon, into intron 6. In a small percentage of offspring we found that the mutant allele reverts to wild-type through loss of the insertion with concomitant disappearance of the mutant phenotype. Investigation of a series of microsatellite loci in the proximal region of chromosome 17 revealed novel alleles which arise simultaneously with disappearance of IAP from AxinFu. These novel microsatellite variants are distinct from the parental alleles and those so far discovered are organized into two haplotypes. Both haplotypes demonstrate stable Mendelian inheritance. Results show that these rearrangements, which are involved in the production of the new haplotypes, exceed millions of base pairs.

Chromosomal inversions largely inhibit recombination and may be associated with selective forces, such as hitch-hiking effects: the effect of positive selection on linked loci. A West African population of Drosophila melanogaster showed a high frequency (0·61) of the In(2L)t inversion. Departure from neutrality statistically associated with the inversion polymorphism was previously recorded at Su(H), a locus distant from the proximal breakpoint of the inversion. These results were consistent with hitch-hiking effects with recombination. The present sequence polymorphism survey involves a 1 kb fragment of the Vha68-1 locus located closer to the proximal breakpoint of the inversion. It shows a significant deficit of polymorphism with respect to divergence when compared with other loci studied in the same population, thus suggesting selective effects. Only 11 polymorphic sites are present in a sample of 20 chromosomes and these sites present a significant excess of rare-frequency variants. The major haplotype shows an unexpectedly high frequency. Our estimate of the background selection effect is not sufficient to account for the observed reduction of polymorphism. Intraspecific variation is structured between inverted and standard chromosomes; there are no shared polymorphisms but also no fixed differences between them. This pattern, together with that found on other loci previously studied near this inversion breakpoint, suggests hitch-hiking effects enhanced by the inversion.

Genetic variability for growth was analysed in three populations of Ostrea edulis, selected for resistance to the protozoan parasite Bonamia ostreae. This study was undertaken first to determine the potential for selection for growth in populations that have never been selected for this character, and second to estimate heterosis versus inbreeding depression. Growth was monitored in culture for 10 months. The selected populations (namely S85-G3, S89I-G2 and S89W-G2), their crossbred population and a control population were composed of full-sib families whose parents were already genotyped using five microsatellite markers. This genotyping allowed the estimation of genetic relatedness among pairs of parents. The parents' relatedness was then correlated with the growth performance of their offspring within each of the three populations, and inbreeding depression was estimated. The population effect for growth was highly significant, with the crossbred population having the highest growth rate, followed by S89I-G2 and S89W-G2, S85-G3 and the control population. The within-populations family effect was also highly significant, indicating, as well as the high value for heritability at the family level (between 0·57 and 0·92), that a potential for a further selection for growth still exists within the three populations. Estimates of inbreeding depression (relative to the mean, for complete inbreeding) were high (1 for S891-G2, 0·44 for S89W-G2 and between 0·02 and 0·43 for S85-G3), which correlates with the apparent heterosis for growth observed in the crossbred population. These results are discussed in the context of the future management of the selected populations.

Group selection has historically been an important and controversial subject in evolutionary biology. There is now a compelling body of evidence, both theoretical and experimental, that group selection not only can be effective, but can be effective in situations when individual selection is not. However, experiments in which true population-level traits have been shown to evolve in response to group selection are currently limited to two species of flour beetle in the genus Tribolium and RNA viruses. Here we report the results of an experiment wherein we imposed group selection via differential extinction for increased and decreased population size at 6-week intervals, a true population-level trait, in the poeciliid fish Heterandria formosa. In contrast to most other group selection experiments, we observed no evolutionary response after six rounds of group selection in either the up- or down-selected lines. Populational heritability for population size was low, if not actually negative. Our results suggest that group selection via differential extinction may be effective only if population sizes are very small and/or migration rates are low.

Genetic differences among populations exposed to selection form barriers against genetic exchange by mortality among hybrids. The strength of such a selection barrier, with which one (recipient) population reacts against immigration from another (donor) population, may be measured as the cumulative mean fitness of hybrids and their descendants relative to the fitness of the recipient population. Previous work analysed a case of weak selection with pairwise epistatic interactions by assuming small genetic distance between two populations in contact. The present study allows large genetic difference between the donor and recipient populations and considers weak multilocus selection with arbitrary epistatic interactions between two or more linked loci. An approximate analytical expression for the barrier strength is obtained as an expansion in which the strength of selection plays the role of a small parameter. It is shown that allele frequencies and gametic linkage disequilibria contribute in different ways to the strength of the selection barrier.

In a previous contribution, we implemented a finite locus model (FLM) for estimating additive and dominance genetic variances via a Bayesian method and a single-site Gibbs sampler. We observed a dependency of dominance variance estimates on locus number in the analysis FLM. Here, we extended the FLM to include two-locus epistasis, and implemented the analysis with two genotype samplers (Gibbs and descent graph) and three different priors for genetic effects (uniform and variable across loci, uniform and constant across loci, and normal). Phenotypic data were simulated for two pedigrees with 6300 and 12300 individuals in closed populations, using several different, non-additive genetic models. Replications of these data were analysed with FLMs differing in the number of loci. Simulation results indicate that the dependency of non-additive genetic variance estimates on locus number persisted in all implementation strategies we investigated. However, this dependency was considerably diminished with normal priors for genetic effects as compared with uniform priors (constant or variable across loci). Descent graph sampling of genotypes modestly improved variance components estimation compared with Gibbs sampling. Moreover, a larger pedigree produced considerably better variance components estimation, suggesting this dependency might originate from data insufficiency. As the FLM represents an appealing alternative to the infinitesimal model for genetic parameter estimation and for inclusion of polygenic background variation in QTL mapping analyses, further improvements are warranted and might be achieved via improvement of the sampler or treatment of the number of loci as an unknown.

Sponsored by: The Genetical Society, Promega (UK) Ltd and B&K Universal Group Ltd