Patterns of reproductive success determined by DNA fingerprinting in a communally breeding oceanic bird

Abstract

Using DNA fingerprinting we estimated the reproductive success of 49 adult birds belonging to 16 breeding groups of the sexually monomorphic brown skua (Catharacta lonnbergi) from the Chatham Islands (New Zealand). This population has a variable mating system, breeding in both monogamous and polyandrous groups. The parentage of 45 chicks produced over three breeding seasons was unequivocally determined using the multilocus probes 33.15 and 33.6. We found no evidence of either extra-pair or extra-group fertilization and there was no evidence to suggest egg dumping by females in any breeding group. Consequently, in the case of pairs, parentage of all chicks was assigned to the resident adult birds. In addition, band sharing analysis indicated that members of communal groups were not close relatives. In the 10 communally breeding groups examined, multiple paternity within a clutch was recorded on two of the 12 occasions in which two chicks were reared. Analysis of the parentage of offspring belonging to different groups, from different years, demonstrated that the number of chicks produced by some adult males varied considerably between seasons. In contrast, the reproductive success of other individuals was constant; for example, one male produced two chicks in each of the three seasons it was studied, while other males in communal groups did not produce any chicks during the course of this study. Fitness is a lifetime parameter, and any assessment of it requires studies over at least the average lifetime of an individual. The findings presented in this study suggest that, for brown skuas, there are significant differences in the reproductive success of some adult males in different breeding seasons. These results indicate that estimates of reproductive fitness based on only a single breeding season's data can be seriously inaccurate. Should temporal changes in paternity (and/or maternity) be shown to be common phenomena in other species, then such results would have major implications for the interpretation of parentage studies.