Numerical Estimation of Survival Rates from Band-Recovery and Biotelemetry Data
Gary C. White · Journal of Wildlife Management · 1983
The estimation of survival rates from tagging or banding data has been well developed by Brownie et al. However, problems occur when sparse data sets result in undefined estimates, when survival estimates exceed unity, when a hypothesis about the data cannot be tested by any of the available models, andwhen constraints on model estimators are desired. This paper presents a general analysis method whereby of the models Brownie et al. and many other methods described in the literature are merely special cases. Models are specified algebraically as cell probabilities consisting of functions of the survival rates and other parameters to be estimated. These algebraic expressions and the observed cell values are input to the computer program SURVIV to provide maximum-likelihood estimates of the unknown parameters and perform hypothesis tests on the data. The generality of the model specification also allows estimation of survival rates form biotelemetry data. 3 tables, 1 figure.