Experimental test on public information use in the colonial Lesser Kestrel

The formation of colonies and the evolution of coloniality have been suggested to be a by-product of the use of public information to select breeding habitats. In this study we performed an experimental test to investigate whether the Lesser Kestrel (Falco naumanni) uses breeding success of conspecifics as a source of public information to select a breeding colony. We considered four potential cues based on reproductive performance of conspecifics in each colony: mean and variance of breeding success measured in both all pairs (TRS, total reproductive success), and only successful pairs (PRS, partial reproductive success). Both mean PRS and mean TRS could be predictive cues of the future reproductive output at a given site, because they varied among colonies and were autocorrelated from one year to the next. To dissociate any relationship between habitat quality and reproductive success we manipulated the mean brood sizes by transferring chicks among ten colonies. In five colonies, all the broods with at least two nestlings were reduced in one chick and in other five colonies all broods were enlarged by one chick. In the following year, we found that the number of adult immigrant kestrels settled at a colony was only explained by mean PRS in the previous year. We also examined the settlement of yearlings, which lack of information on breeding success in previous years. Unlike adult immigrant kestrels, yearlings tended to settle more frequently in those colonies with low mean PRS in the previous year, probably because of the lower pressure for occupation in those colonies. A multiple regression analysis showed that the growth of colonies was positively dependent on the mean PRS in the experimental year, and negatively on the number of predated nests. This study provides a solid support to the hypothesis that colonial species can use mean breeding success of successful pairs as a source of public information to make settlement decisions.This study was supported by Ministerio de Ciencia y Tecnología (project REN 2002–03295). RB and AM were supported by fellowships from the La Junta de Comunidades de Castilla-La Mancha.Peer reviewe