True polyandry and pseudopolyandry: why does a monandrous fly remate?

BACKGROUND: The rate of female remating can have important impacts on a species, from affecting conflict and cooperation within families, to population viability and gene flow. However, determining the level of polyandry in a species can be difficult, with information on the mating system of many species being based on a single experiment, or completely absent. Here we investigate the mating system of the fruit fly Drosophila subobscura. Reports from England, Spain and Canada suggest D. subobscura is entirely monandrous, with no females remating. However, work in Greece suggests that 23% of females remate. We examine the willingness of female D. subobscura to remate in the laboratory in a range of conditions, using flies from both Greece and England. We make a distinction between pseudopolyandry, where a female remates after an ineffective first mating that is incapable of fertilising her eggs, and true polyandry, where a female remates even though she has received suitable sperm from a previous mating. RESULTS: We find a low rate of true polyandry by females (4%), with no difference between populations. The rate of true polyandry is affected by temperature, but not starvation. Pseudopolyandry is three times as common as true polyandry, and most females showing pseudopolyandry mated at their first opportunity after their first failed mating. However, despite the lack of differences in polyandry between the populations, we do find differences in the way males respond to exposure to other males prior to mating. In line with previous work, English flies responded to one or more rivals by increasing their copulation duration, a response previously thought to be driven by sperm competition. Greek males only show increased copulation duration when exposed to four or more rival males. This suggests that the response to rivals in D. subobscura is not related to sperm competition, because sperm competition is rare, and there is no correlation of response to rivals and mating system across the populations. CONCLUSIONS: These results illustrate the difficulties in determining the mating system of a species, even one that is well known and an excellent laboratory species, with results being highly dependent on the conditions used to assay the behaviour, and the population used.