Dispersal, inbreeding and fitness in natural populations

Abstract: To evaluate how inbreeding affects the viability of small populations, it is crucial to determine what constitutes a population (i.e., the number of breeders in local patches and the degree of dispersal), and also to estimate the reproductive consequences of philopatry and emigration. In this thesis, I concentrate on dispersal, inbreeding and fitness in natural populations. I consider dispersal rates, individual and population parameters predicting dispersal, and reproductive costs and benefits of dispersal. Furthermore, I evaluate how reproductive effects of inbreeding can be quantified by means of genetic markers. I base my arguments on empirical data gathered from newly founded nonequilibrium populations of great reed warblers (Acrocephalus arundinaceus) in Sweden. My results show that it is important to consider scale when evaluating possible predictors of dispersal. This was evident both from a capture-recapture analysis, which revealed that dispersal varied between years on the larger scale and between hatching sites on the smaller scale, and from a molecular analysis, which revealed a sex bias in dispersal on the larger scale that was lacking on the smaller scale. The capture-recapture study also revealed that dispersal was partly heritable in great reed warblers. Occasionally, closely related individuals mated. When this happened, the reproductive success decreased as a result of poor egg-hatching success. Also, there were fitness differences between philopatric and immigrating birds. In males, philopatric individuals had higher life-time mating success than immigrants. In females, on the other hand, short-distance dispersing immigrants recruited a higher proportion of their offspring than both philopatric individuals and long-distance immigrants. This suggests that there exists an optimal dispersal distance in female great reed warblers. Finally, I found that surviving great reed warblers had significantly higher microsatellite heterozygosity than their non-surviving siblings. This result led to a major conclusion in my thesis: that heterozygosity-fitness correlations might occur in nonequilibrium populations for reasons that are neither related to selection at the markers per se, nor to variation in the inbreeding coefficient. Instead, and because of high levels of linkage disequilibria in the population, the association between heterozygosity and survival might have arisen from selection at fitness loci that were linked to the genetic markers.