Breeding success and mate retention in birds: a meta-analysis

Abstract.

Several hypotheses about the adaptive value of divorce in birds predict a positive effect of breeding success on mate retention. Although some studies have provided direct support for this prediction, others have failed to demonstrate any significant influence of breeding success on mate retention. To date, no one has investigated the overall empirical evidence for such a relationship. We used a meta-analysis of published results to investigate whether the rate of divorce between consecutive breeding seasons differed between successful and unsuccessful breeders among monogamous bird species. The analysis was based on 38 studies involving 35 different species. The mean weighted effect size, measured as the Pearson product-moment correlation coefficient, was 0.112 and differed significantly from zero. Despite the relatively low mean effect size, this finding is extremely robust given the large fail-safe number, indicating that more than 1,333 additional nonsignificant studies would be necessary to reduce the effect size to a nonsignificant value. There was no evidence that year of publication had a significant influence on the results. Comparing between species, the difference in divorce rates between unsuccessful and successful pairs was negatively related to clutch size, even after controlling for phylogenetic dependence and the potential effect of adult survival rate. This indicates that the traditional measure of breeding success (failed versus fledged at least one young) may not be appropriate for species that typically lay large clutches. Therefore, the relatively moderate overall effect of breeding success on divorce might be considered as an underestimation of the true effect.