Dispersal and social attraction affect colony selection and dynamics of lesser kestrels

We studied the mechanisms that regulate colony dynamics in a Spanish population of Lesser Kestrels, using eight years of data from banded individuals in 494 colony-years. Colony growth was positively related to breeding success at the colony the year before. However, individuals of all dispersal statuses, i.e., adult and first-breeding philopatric and immigrant birds, significantly contributed to changes in colony size, indicating an important effect of dispersal on colony dynamics via colony quality. Given that there is strong evidence that Lesser Kestrels base their settlement decisions on conspecifics, we tested whether immigrants used the number of previously settled residents in year t (social or conspecific attraction hypothesis) and/or the breeding performance of conspecifics in year t - 1 (performance-based attraction hypothesis) to select their breeding colony. Breeding success of colonies varied both in space and time and was autocorrelated from one year to the next. Moreover, lifetime reproductive success of Lesser Kestrels was positively associated with colony size, and individuals can predict final colony size early in the breeding season, so assumptions of both hypotheses were fulfilled. Our results support the social attraction hypothesis, since immigration was positively related to the number of philopatric adults, but not to conspecific breeding success the year before. Given that departure decisions of adults were based on personal information about breeding success and colony size is related to fitness prospects, previously settled individuals provide easy and reliable information about colony quality, and social attraction could be seen as a particular case of public information in Lesser Kestrels. Consistently, absolute numbers of both philopatric adults and immigrants increased with colony size the year before, although immigrants increased only up to a threshold beyond which this trend disappeared. Therefore, immigrants seem to be prevented from settling in the largest colonies, which could explain why all individuals do not concentrate in a few big colonies, but some settle in suboptimal colonies or colonize unoccupied sites. This opposing effect of conspecifics, together with the low levels of temporal autocorrelation in colony quality between time lags greater than or equal to2 yr, could promote colony size variability and facultative coloniality in this species.