Contributions to the rapid evolutionary loss of a sexual signal: Islands, parasites, and mating behavior in the Polynesian field cricket Teleogryllus oceanicus

Sexual signals are intended to attract members of the opposite sex, but frequently attract the attention of unintended receivers like natural enemies. In the Hawaiian Islands, the field cricket Teleogryllus oceanicus co-occurs with a deadly natural enemy, an acoustically orienting parasitoid fly, found nowhere else in their range. I show that selection to avoid the fly favored a wing mutation, flatwing, in one Hawaiian population, that eliminates males' ability to produce song, their sexual signal, by eliminating the calling structures on the wings. This is one of very few documented rapid changes in sexually selected traits. The mutation spread to >90% of males on Kauai in <20 generations because it protects males from the parasitoid fly. Breeding experiments show that the flatwing mutation is due to a change in a single sex-linked locus. Flatwing males appear to behave as satellites to the remaining calling males, settling in close proximity to callers and attempting to mate with females attracted to them. They show enhanced phonotaxis behavior, moving toward song more quickly and settling in closer proximity to the source of broadcast calling song than do normal-wing males. Satellite behavior existed as a behavioral option prior to the mutation in wing morphology, and appears to be used by males, regardless of wing morphology, when mating success has been low. This suggests that preexisting behavioral variation facilitated the rapid loss of singing ability on Kauai.;For a sexual signal to change, the change must be favorable for the signaler, but must also be accommodated by the receiver's perception and preferences. Despite conventional wisdom that females require males to produce a courtship song before mating, I show that females from ancestral, unparasitized Australian and Pacific Island populations as well as parasitized Hawaiian ones sometimes mate with silent flatwing males. Lab experiments support the hypothesis that island females have relaxed preferences relative to mainland Australian crickets, suggesting that the non-adaptive process of island colonization favored females with relaxed mating requirements (Kaneshiro's effect). Taken together, these results support the idea that variation in both male mating behavior and female preferences facilitated the rapid evolutionary loss of an important sexual signal.