Here we describe several cases of within-season divorce shortly before the onset of breeding in a wild population of Blue Tits. In total, 23% of pairs divorced, and at least one of the partners re-mated with a new territory holder. Within-season divorce was associated with fitness costs in that pairs that re-mated had a lower fledging success. However, due to the small sample size, these results have to be viewed with caution, and we therefore refrained from testing classical hypotheses to explain the observed cases of between-season divorce (Choudhury 1995). In the following, we discuss causes of within-season divorce and potential mechanisms responsible for the reduced breeding success.
There is no published data on within-season divorce in Blue Tits; as such, we are not able to link our observations to the naturally occurring rate of this type of divorce. The between-season divorce rate is around 50% in our Blue Tit population (Valcu and Kempenaers 2008) and includes any change of partner between two breeding seasons irrespective of the timing of the partner change. It is possible that many of the reported cases of divorce occur shortly before breeding and that our results may reflect a natural rate of within-season divorce. Alternatively, the relatively high rate of within-season divorce in this study may have been due to the experimental disturbance at the original nestbox shortly before egg laying. All nestboxes in the control and experimental plots were temporarily removed from their original site and re-erected the following morning. This disturbance during the nest-building stage may have weakened the pair bond, leading several pair members to try and ensure breeding in the current season by re-mating with a new mate. Females that lost their nestbox in the experimental plots may have successfully out-competed other females throughout the entire study area. Indeed, in two cases, females from experimental plots re-mated with males from control plots (see Fig. 1), supporting the hypothesis that floating females may evict territorial females from their territory. This hypothesis is in line with studies documenting intense female–female competition and aggression for available nesting sites shortly before egg laying (Slagsvold and Lifjeld 1994) and is particularly prevalent in Blue Tits (Kempenaers 1995).
In our study, the breeding pairs showing within-season divorce had lower reproductive success than faithful pairs. These results are in contrast to previous findings in Black-capped Chickadees showing that females benefitted from within-season divorce by re-mating with a male of higher quality (Otter and Ratcliffe 1996; Ramsay et al. 2000). The reduced fledging success in our study was partly explained by total brood failures during the incubation or nestling phase, which were equally common in control (n = 3; two faithful, one divorced) and experimental (n = 2; two divorced) broods. More studies on the fitness consequences of within-season divorce are certainly needed since our results are based on a small sample size and, additionally, on the occurrence of complete brood failures, which may be due to chance events like the death of one pair member. Alternatively, the reduced breeding success may be due to the nest-site limitation experiment that altered breeding conditions in the whole study area. Increased competitive interactions around nestboxes may have continued during the incubation and nestling phase, assuming that evicted individuals remained within the study site. The extra investment in territorial defence may trade-off with brood care or survival, and there may be additional direct negative effects of the presence of floaters on reproductive performance. In one experimental box, we found dead nestlings with open head wounds that most likely originated from attacks by adult conspecifics. This type of injury had never been observed before in this study population. Alternatively, the rapid change of partner shortly before egg laying may have altered a male’s perceived confidence in paternity, and this may negatively affected paternal investment (Whittingham et al. 1992; Sheldon 2002) and may partly explain the complete brood loss via brood desertion of the male in the control and experimental plots.
In contrast to other studies on within-season divorce (Otter and Ratcliffe 1996; Ramsay et al. 2000), we were unable to demonstrate any fitness benefits of a rapid change of partner before breeding onset. However, under conditions of sudden nest-site limitation, evicted individuals may have only two alternatives: either initiate a fast change of breeding partner and thereby gain some reproductive success, or forego breeding for 1 year. Hence, within-season divorce may reflect a flexible strategy of short-lived species, such as the Blue Tit, aimed at maximizing the probability to reproduce in the current season. This concept of flexibility in mate choice is partly in line with a study on European Starlings, which demonstrated that females adopt a similar re-mating strategy once their partner had disappeared or died shortly before breeding (Pinxten et al. 1993). To assess the adaptive value of within-season divorce, it would be necessary to compare the residual reproductive success of individuals that switched their partner shortly before egg laying with individuals that forewent a breeding bout. The fitness consequences of divorce are complex, but our results suggest that divorcing shortly before the onset of breeding may incur higher fitness costs compared to divorce and re-mating during the non-breeding season, thereby highlighting the importance of the optimal timing for mate-switching.