The impact of vagrants on apparent survival estimation in a population of Common Crossbills (Loxia curvirostra)

Abstract

The impact of nomadism on animal population dynamics (e.g. survival) is poorly understood. This factor is not inconsequential because our current view on animal survival rates could suffer from biases if it is ignored. The degree of nomadism in Common Crossbills (Loxia curvirostra) is reported to differ population-specifically, and thus they constitute a good model for the study of survival in nomadic species. Southern Europe hosts resident birds, probably also local vagrant birds and non-local, vagrant, northern European Crossbills. The impact of vagrants on the estimation of apparent survival rates has never been determined in areas where Crossbills have been reported as resident. We hypothesise that transients (birds for which survival from year t to year t + 1 is zero) will be detectable if a majority of vagrants remain at a particular site for just a few months. Alternatively, if vagrants remain for longer, transients will be absent by definition, so apparent survival estimation will be lower than would be if transients did not exist. As wing length in vagrants is commonly longer than in residents, we can expect a negative effect of wing length on survival, as long-winged (vagrant) Crossbills emigrate from the area after a few years, thus demonstrating lower apparent survival rates than resident, local birds on a local scale. Alternatively, if vagrants have a negligible impact on the estimation of local apparent survival rates (i.e. if residents clearly outnumber vagrants), we should not detect transients nor find any effect of wing length on survival. To test this hypothesis, we used data collected over a period of 16 years at a site located in a Scots pine (Pinus sylvestris) forest in northern Iberia. Capture–recapture data were analysed with Cormack–Jolly–Seber (CJS) models. Although transients were not detected, the CJS models showed that survival was negatively affected by wing length. Our results support the hypothesis that vagrants occurred in the area for more than 1 year before subsequently disappearing, and that their presence has a strong impact on local survival estimation. Accordingly, if the presence of vagrants is not considered, this can lead to the underestimation of local survival rates of resident Crossbill populations.