EFFECT OF ECTOPARASITIC MITES ON SEXUAL SELECTION IN A SONORAN DESERT FRUIT FLY.

We conducted a field study and a laboratory experiment to test whether ectoparasitic mites, Macrocheles subbadius, generate parasite-mediated sexual selection in the Sonoran desert endemic fruit fly, Drosophila nigrospiracula. Male flies gather on the outer surfaces of necrotic saguaro cacti where they engage in male-male competitive interactions and vigorous female-directed courtship. At these sites, operational sex ratios were significantly skewed toward males. The degree to which mites were aggregated among flies varied across the 25 fly populations sampled. The degree of mite aggregation across fly populations was strongly positively related to the mean number of mites per fly (intensity of infestation). Both the intensity and prevalence of infestation (fraction of flies infested) increased with the age of the cactus necrosis. Infested flies of both sexes were significantly less likely to be found in copula than uninfested flies, and mean intensity of infestation was significantly more pronounced in noncopulating than in copulating flies. The effect of attached mites on copulatory success exhibited dose-dependency, and this effect was more stringent in males: males or females with more than two and four mites, respectively, were never found in copula. The magnitude of parasite-mediated sexual selection was estimated for 12 fly populations by calculating selection differentials for each sex separately. The relation between intensity of infestation and magnitude of parasite-mediated sexual selection was stronger in males but significant for both sexes. We also assayed copulatory success of field-caught males in the laboratory, both during infestation and after experimental removal of mites. Males infested with two mites copulated less frequently than uninfested individuals, and in mating trials after mites had been removed, previously infested males copulated as many times as flies with no history of infestation. These findings, and the lack of difference in the number of mite-induced scars on copulating and single individuals in nature, strongly suggest that the reduced copulatory success of infested flies is attributable to an effect of mites per se, rather than to a character correlated with parasitism or previous parasite infestation.