Natural and sexual selection on male behaviour and morphology, and female choice in a wild field cricket population: spatial, temporal and analytical components.

We used multiple regression, path analysis and non-parametric selection surface visualisation to investigate natural and sexual selection and, in addition, cross-sectional female choice statistics to analyse female choice in a wild population of the field cricket Gryllus campestris L. in central Germany. Adults (167 males, 75 females) were individually marked and followed daily over the entire adult stage. Two morphological traits (pronotum width, body condition) and two behavioural traits (burrow occupation time, daily displacement) were measured for each male and used in selection analyses. Individuals mated multiply and male mating success was strongly right skewed with less than 6% of the population achieving 50% of the copulations. In males, analysis of natural selection in terms of lifespan revealed positive directional selection on burrow occupation time and stabilising selection on daily displacement. Analysis of sexual selection in term of mating success showed positive directional selection on pronotum width and lifespan. Path analysis confirmed the close association between natural and sexual selection and illustrated indirect effects of the behavioural variables on mating success via their effect on lifespan. Multiple regression analysis further indicated positive quadratic (disruptive) sexual selection on lifespan but the non-parametric cubic spline regression showed this to be an artefact of the quadratic approach. In fact, lifespan was under “threshold selection” i.e. it was not under selection below a threshold and under positive directional sexual selection above the threshold. A positive correlational selection gradient between lifespan and body condition revealed that a high body condition is advantageous among long-lived males. Female choice statistics showed that females chose large and heavy males in the beginning of the season only and that choosiness decreases with increasing distance to potential alternative mates. Our findings highlight the benefits of combining several analytical methods to uncover selection patterns and to avoid misinterpretations based on single methods. [ABSTRACT FROM AUTHOR]