Sperm competition and its evolutionary effect on copula duration in the fly Scatophaga stercoraria

Abstract

The normal pattern of egg production in female dung flies consists of laying several successive egg batches of approximately equal numbers of eggs; although the number of eggs per batch is directly proportional to female size, the egg size remains constant. After a single insemination, a female can lay at least four batches without any drop in fertility. After a double mating in which the sperm from one male is labelled by exposure to 10 krad irradiation dose, the second male fertilizes 81·4 per cent of the eggs provided no further mating intervenes. After a double mating in which the second is interrupted, the competitiveness of the second mating can be shown to increase most steeply during the first half of copula. When females are mated immediately before laying (as in nature), matings labelled in various positions indicate that sperm from a given male can fertilize eggs in up to six successive batches, but that the percentage fertilization falls sharply as the number of batches (and hence number of subsequent matings) increases. With mated females, the last male to mate before oviposition fertilizes about 80 per cent of the succeeding batch irrespective of the number of previous matings. A model, which assumes that sperm from matings previous to the one before oviposition will compete for the remaining 20 per cent of the batch in the same proportionate relationship as they did for the previous batch, fits well with the experimental observations. Two patterns are therefore distinguishable for calculating the total egg gain of a male, depending on whether he mates with a virgin or non-virgin female. By assessing the chances of take-overs (where a second male ousts the male originally paired to a female) from field data, their effect is included in the model and the maximum, minimum, and probable average egg gains are calculated for males mating with females in different stages of their reproductive cycle. Egg gains with mated females could be increased by an increase in copula duration, but this would decrease the rate of capture of new females and also increase the probability of take-over. By using the model to assess the fertilization rate (eggs/min) achieved by males with varying copula durations in the present population and theoretical populations where the duration is both greater and smaller than at present, it can be predicted that selection would favour a value close to that observed. When the effect of virgin females is included, the fit is even better. Matings with post-oviposition females are much shorter than normal and the predicted optimum duration is very close to that observed.