Population growth rate and genetic variability of small and large populations of Red flour beetle (Tribolium castaneum) following multigenerational exposure to copper.

We reared large (1000 individuals) and small (20 individuals) populations of Tribolium castaneum on diet contaminated with copper in order to determine if the size of a population affects its ability to adapt to adverse environmental conditions. After 10 generations, we used microsatellite markers to estimate and subsequently compare the genetic variability of the copper-treated populations with that of the control populations, which were reared on uncontaminated medium. Additionally, we conducted a full cross-factorial experiment which evaluated the effects of 10 generations of "pre-exposure" to copper on a population's fitness in control and copper-contaminated environments. In order to distinguish results potentially arising from genetic adaptation from those due to non-genetic effects associated to parental exposure to copper, we subjected also F11 generation, originating from parents not exposed to copper, to the same cross-factorial experiment. The effects of long-term exposure to copper depended on population size: the growth rates of small populations that were pre-exposed to copper were inhibited compared to those of small populations reared in uncontaminated environments. Large Cu-exposed populations had a higher growth rate in the F10 generation compared to the control groups, while the growth rate of the F11 generation was unaffected by copper exposure history. The only factor that had a significant effect on genetic variability was population size, but this was to be expected given the large difference in the number of individuals between large and small populations. Neither copper contamination nor its interaction with population size affected the number of microsatellite alleles retained in the F10 generation.

Sexual selection and population divergence II. Divergence in different sexual traits and signal modalities in field crickets (Teleogryllus oceanicus).

Sexual selection can target many different types of traits. However, the relative influence of different sexually selected traits during evolutionary divergence is poorly understood. We used the field cricket Teleogryllus oceanicus to quantify and compare how five traits from each of three sexual signal modalities and components diverge among allopatric populations: male advertisement song, cuticular hydrocarbon (CHC) profiles and forewing morphology. Population divergence was unexpectedly consistent: we estimated the among-population (genetic) variance-covariance matrix, D, for all 15 traits, and Dmax explained nearly two-thirds of its variation. CHC and wing traits were most tightly integrated, whereas song varied more independently. We modeled the dependence of among-population trait divergence on genetic distance estimated from neutral markers to test for signatures of selection versus neutral divergence. For all three sexual trait types, phenotypic variation among populations was largely explained by a neutral model of divergence. Our findings illustrate how phenotypic integration across different types of sexual traits might impose constraints on the evolution of mating isolation and divergence via sexual selection.

Sexual selection and population divergence I: The influence of socially flexible cuticular hydrocarbon expression in male field crickets (Teleogryllus oceanicus).

Debates about how coevolution of sexual traits and preferences might promote evolutionary diversification have permeated speciation research for over a century. Recent work demonstrates that the expression of such traits can be sensitive to variation in the social environment. Here, we examined social flexibility in a sexually selected male trait-cuticular hydrocarbon (CHC) profiles-in the field cricket Teleogryllus oceanicus and tested whether population genetic divergence predicts the extent or direction of social flexibility in allopatric populations. We manipulated male crickets' social environments during rearing and then characterized CHC profiles. CHC signatures varied considerably across populations and also in response to the social environment, but our prediction that increased social flexibility would be selected in more recently founded populations exposed to fluctuating demographic environments was unsupported. Furthermore, models examining the influence of drift and selection failed to support a role of sexual selection in driving population divergence in CHC profiles. Variation in social environments might alter the dynamics of sexual selection, but our results align with theoretical predictions that the role social flexibility plays in modulating evolutionary divergence depends critically on whether responses to variation in the social environment are homogeneous across populations, or whether gene by social environment interactions occur.