Vocal identification of individual African Wood Owls Strix woodfordii: A technique to monitor long-term adult turnover and residency

The hoot-calls of Wood Owls Strix woodfordii, recorded in the Kruger National Park, South Africa, from 1986 to 1998, are sufficiently different to distinguish individuals. Thirteen hoot-call variables, measured from recordings over this period, show that male and female Wood Owls have such temporal stability that these calls can be used reliably as a long-term census technique. MANOVA, based on ordinations from principal component analysis, was used to identify individuals statistically between sampling periods. A forward stepwise discriminant function analysis achieved 100% classification success of individual male (n = 3) and female (n = 4) owls from a single sampling period. For all the individuals recorded over the whole study period we achieved a classification success of 80.9% (n = 9) for male hoot calls and of 96.3% (n = 13) for female calls. We found mean occupation periods of 82.25 and 65 months and annual turnover rates of 19.3% and 13.65% for males and females, respectively. Our survey, using vocalizations, is unique since we use data collected over a 12-year period to derive estimates of population turnover in Wood Owls, and consider what questions can be addressed in similar studies.