Incorporating covariates in the analysis of capture re-encounter data

Methods of analysis of capture-re-encounter data involve fitting probability models to data collected from animals that have been previously given unique markings. Re-encounters can correspond to finding animals dead (recoveries), or capturing/resighting them alive (recaptures). For illustration in this paper we shall consider both the case of finding animals dead and also that of finding animals alive, at various times following marking. The model parameters are appropriate probabilities of annual survival, as well as probabilities of reporting/recapture of dead/live animals, respectively. An illustrative recovery data set is provided in Table 1, which is a subset of one of the data sets analysed in this paper. The full data set is available at www.tibs.org. The data correspond to birds being ringed throughout Britain, and later reported dead to the address on the rings.