Most studies of reinforcement have focused on the evolution of either female choice or male mating cues, following the long‐held view in sexual selection theory that mating mastakes are typically more costly for females than for males. However, factors such as conspecific sperm precedence can buffer females against the cost of mating mistakes, suggesting that in some hybrid zones mating mistakes may be more costly for males than for females. Thus, the historical bias in reinforcement research may underestimate its frequency. In this study, we present evidence that reinforcement has driven the evolution of male choice in a hybrid zone between teh highly promiscuous lealf beetles chyrsochus cobaltinus and C. auratus, the hybrids of which have extremely low fitness. In addition, there is evidence for male choice in these beetles and that male mating mistakes may be costly, due to reduced opportunities to mate with conspecific females. The present study combines laboratory and field methods to quantify the strenght of sexual isolation, test the hypothesis of reproductive character displacement, and assess the link between relative abundance and the strenght of selection against hybridization. We document that, while sexual isolation is weak, it is sufficient to produce positive assortative mating. In addtion, reproductive character displacement was only detected in the relatively rare species. The strong postzygotic barriers in this system are sufficient to generate the bimodality that characterizes this hybrid zone, but the weak sexual isolation is not, calling into question whether strong prezygotic isolation is necessary for the maintenance of bimodality. Growing evidence that the cost of mating mistakes is sufficient to shape the evolution of male mate choice suggests that the reinforecement of male mate choice may prove to be a widespread occurrence.

No abstract

No abstract

We investigated patterns of cytonuclear disequilibrium between nuclear allozyme loci and partial mitochondrial COI and COII restriction fragment length polymorphism patterns within a population of hybridizing chrysomelid beetles and assessed to what degree the genotype frequencies of F1 hybrids were consistent with patterns of mate choice or endosymbiont infection. We document that in this population, > or = 50% of the heterospecific pairs at a given time are composed of Chrysochus auratus females and Chrysochus cobaltinus males, suggesting that at least half of the F1 hybrids should possess the C. auratus mitochondrial genotype. However, we found that the majority (89%) of F1 hybrids possessed C. cobaltinus mtDNA (P < 0.001). The lack of evidence for Wolbachia infection in these highly promiscuous beetles, coupled with the fact that F1 hybrids of both cross types do exist, indicates that endosymbionts are an unlikely explanation for the discrepancy between cytonuclear genotype frequencies and behavior. We argue that cytonuclear disequilibrium at this focal Chrysochus hybrid site is likely due to a strong directional bias in postmating prezygotic barriers in this system. The results presented here underscore the importance of combining both field and molecular data in studies of cytonuclear disequilibrium and point to the dangers inherent in attributing patterns of cytonuclear disequilibrium to assortative mating.