Weighted Nest Survival Models

Abstract

Nest survival data possess two characteristics that are not fully exploited by traditional nest survival models: the data constitute a probability sample of the population and the length of time nests are observed provides information concerning age at the time of detection. Models that fail to acknowledge the probabilities with which nests are included in the sample are strictly conditioned on the observed, typically biased, data. In addition, if knowledge of the age at the time of detection is not utilized, the information contained in the data is not fully exploited. This paper presents a general model for the probability density function (pdf) of data observed in nest survival studies. The model incorporates both probabilities of inclusion and information concerning the age of nests at the time of detection. Maximum likelihood estimators of all parameters involved in the pdf, including survival rates, are based on the pdf of the observed data rather than the pdf of variables in the population. The nest survival model of Pollock and Cornelius (1988, Biometrics 44, 397-404) is a special case of the general model.