Male-killing in the Coccinellidae: testing the predictions

Abstract

Male-killing endosymbionts have been widely reported in the invertebrates and are highly prevalent in the Coccinellidae. The presence of male-killers can lead to extreme bias in host population sex ratios and may have important and far-reaching consequences for the life-history and evolution of their hosts. Male-killers may have direct and indirect effects on host fitness and reproductive behaviour, as well as affecting the host genome, either via strong selection pressure imposed by highly female-biased population sex ratios or by selective sweeps caused as a male-killer conferring an advantage to infected individuals spreads through a population. Criteria used to predict which species are liable to male-killer invasion, based on a variety of ecological factors, have been produced. In summary male-killers are predicted to occur in aphidophageous species, that lay eggs in clutches, show sibling egg consumption and are liable to neonatal larval mortality due to starvation. We assayed 30 species of Coccinellid for the presence of such male-killers to assess the predictive accuracy of the criteria. Male-killers were identified in 8 species in which they were predicted to occur and were absent from all 10 species predicted not to harbor them. Analysis of the remaining 12 species, where male-killers were predicted by the original criteria, but where they were not found, allowed us to identify areas where the criteria can be refined and improved. We conclude that whilst the original criteria give a reasonably accurate prediction, there are refinements and improvements, concerning details of host diet and life-history, which make them more robust, especially in the light of discoveries of male-killing suppressors and when incorporated give a better fit to our findings from field samples.