What constitutes a nesting attempt? Variation in criteria causes bias and hinders comparisons across studies

Because reliable estimates of nesting success are very important to avian studies, the defnition of a “successful nest” and the use of different analytical methods to estimate success have received much attention. By contrast, variation in the criteria used to determine whether an occupied site that did not produce offspring contained a nesting attempt is a source of bias that has been largely ignored. This problem is especially severe in studies that deal with species whose nest contents are relatively inaccessible because observers cannot determine whether or not an egg was laid for a large proportion of occupied sites. Burrowing Owls (Athene cunicularia) often lay their eggs ≥3 m below ground, so past Burrowing Owl studies have used a variety of criteria to determine whether a nesting attempt was initiated. We searched the literature to document the extent of that variation and examined how that variation influenced estimates of daily nest survival. We found 13 different sets of criteria used by previous authors and applied each criterion to our data set of 1,300 occupied burrows. We found significant variation in estimates of daily nest survival depending on the criteria used. Moreover, differences in daily nest survival among populations were apparent using some sets of criteria but not others. These inconsistencies may lead to incorrect conclusions and invalidate comparisons of the productivity and relative site quality among populations. We encourage future authors working on cavity-, canopy-, or burrow-nesting birds to provide specific details on the criteria they used to identify a nesting attempt.