Breeding- and nest-site choice is a behavioral strategy often used to counter negative interactions. Site choices prior to breeding prevents costs of predation and competition but has been neglected in the context of brood parasitism. For hosts of brood parasites, the earlier brood parasitism is prevented in the breeding cycle the lower the future costs. Suitable nest-sites for cavity-nesting common redstarts (Phoenicurus phoenicurus), a host of the common cuckoo (Cuculus canorus), are a limited resource, but their cavity-nesting strategy could potentially deter predators and brood parasites. We altered the entrance size of breeding cavities and investigated redstart nest site choice and its consequences to nest predation and brood parasitism risk, while accounting for potential interspecific competition for nest sites. We set-up paired nest-boxes and let redstarts choose between 7 cm and 5 cm entrance sizes. Additionally, we monitored occupancy rates in nest-boxes with 3 cm, 5 cm and 7 cm entrance sizes and recorded brood parasitism and predation events. We found that redstarts preferred to breed in 5 cm entrance size cavities, where brood parasitism was eliminated but nest predation rates were comparable to 7 cm entrance size cavities. Only in 3 cm cavities were both brood parasitism and predation rates reduced. In contrast to the other cavity-nesting species, redstart settlement was lowest in 3 cm entrance size cavities, potentially suggesting interspecific competition for small entrance size cavities. Nest site choice based on entrance size could be a front-line defense strategy that redstarts use to reduce brood parasitism.

Each set-up has its own cvs file.

For 2012-2013 the data from the general population and the choice experiment was compiled. Note that the column cavity size corresponds to the choice between the pair of nest-boxes (i.e. one 5 cm and one 7 cm nest box) used in the choice experiment. Given that one of the nest-boxes was never occupied, it was not included in the dataset to avoid confusion, as both nest-boxes had the same box ID.

Also, note that the 2019 data has more than one row per nest box as the entrance size is a time-dependent variable because additional covers were placed onto a subset of the nest-boxes at different times along the season. Moreover, the 2019 data has two columns for species: SpeciesCox and Species. SpeciesCox is the one used for the cox model, and all nest-boxes occupied by great tits were scored as empty. While for the Species column one can see which boxes were occupied by great tits.