Dispersal data are inevitably biased toward short-distance events, often highly so. We illustrate this problem using our long-term study of Acorn Woodpeckers (Melanerpes formicivorus) in central coastal California. Estimating the proportion of birds disappearing from the study area and correcting for detectability within the maximum observable distance are the first steps toward achieving a realistic estimate of dispersal distributions. Unfortunately, there is generally no objective way to determine the fates of birds not accounted for by these procedures, much less estimating the distances they may have moved. Estimated mean and root-mean-square dispersal distances range from 0.22-2.90 km for males and 0.53-9.57 km for females depending on what assumptions and corrections are made. Three field methods used to help correct for bias beyond the limits of normal study areas include surveying alternative study sites, expanding the study site (super study sites), and radio-tracking dispersers within a population. All of these methods have their limitations or can only be used in special cases. New technologies may help alleviate this problem in the near future. Until then, we urge caution in interpreting observed dispersal data from all but the most isolated of avian populations.
Additional publication details
Natal dispersal in the cooperatively breeding Acorn Woodpecker