Are nest sites actively chosen? Testing a common assumption for three non-resource limited birds

Many widely-accepted ecological concepts are simplified assumptions about complex situations that remain largely untested. One example is the assumption that nest-building species choose nest sites actively when they are not resource limited. This assumption has seen little direct empirical testing: most studies on nest-site selection simply assume that sites are chosen actively (and seek explanations for such behaviour) without considering that sites may be selected randomly.We used 15 years of data from a nestbox scheme in the UK to test the assumption of active nest-site choice in three cavity-nesting bird species that differ in breeding and migratory strategy: blue tit (Cyanistes caeruleus), great tit (Parus major) and pied flycatcher (Ficedula hypoleuca). Nest-site selection was non-random (implying active nest-site choice) for blue and great tits, but not for pied flycatchers. We also considered the relative importance of year-specific and site-specific factors in determining occupation of nest sites. Site-specific factors were more important than year-specific factors for the tit species, while the reverse was true for pied flycatchers. Our results show that nest-site selection, in birds at least, is not always the result of active choice, such that choice should not be assumed automatically in studies of nesting behaviour. We use this example to highlight the need to test key ecological assumptions empirically, and the importance of doing so across taxa rather than for single ‘‘model’’ species