A Model with First-Year Variation for Ring-Recovery Data

Publisher Website
Google Scholar
Abstract
A probability model is proposed for ring-recovery data in which the survival probability for first-year birds is allowed to vary with calendar year. This model may be fitted to data by maximum likelihood without the customary arbitrary additional constraint when only young birds have been ringed. The approach is illustrated on simulated data and two real examples. Profile log-likelihoods reveal the possibility of substantial differences between alternative methods for estimating confidence intervals. More parsimonius versions of the model may be obtained, for example, through regressing the first-year survival probabilities on covariates if that is feasible. Additionally, the ring-reporting probability may also be made year-dependent, and any variation can be tested by likelihood ratio tests.

This publication has 2 references indexed in Scilit: