AN ALTERNATIVE AVIAN POPULATION ESTIMATE FOR OVERDISPERSED POPULATIONS FOR USE IN MARK RECAPTURE STUDIES OF PESTICIDE EFFECTS

Statistical models that have been commonly applied to mark-recapture/r esighting data, such as the Jolly-Seber estimate, assume that capture and recapture/resighting probabilities are constant for all birds. Thi s assumption is likely false. An alternative approach that is based on the beta-binomial distribution that allows for heterogeneous capture/ recapture is presented. Use of the approach is illustrated with data o btained in a field study of NEMACUR(R) 15G applications. A total of 24 ,000 simulations indicated that the power of the beta-binomial in dete cting a 10% reduction in survival varies from 56.2 to 99% in resightin g data and is more than 84% in recapture data. Population estimates we re obtained using the maximum likelihood method. Under conditions of h eterogeneous probabilities, more precise estimates of population size were produced with this approach than with the Jolly-Seber method.