Long-term survival rates in colour-ringed shorebirds — practical considerations in the application of mark-recapture models.

Capsule: Serious violations of the model assumptions were seen. Aims To illustrate the manner in which violations of the mark-recapture model assumptions may occur. Methods Attempts were made to model annual and seasonal survival of several shorebird species at Teesmouth, northeast England. The database contained 25 years of colour-ring resightings and several suites of models were tested. Results There was significant heterogeneity in survival and recapture probabilities for all five species of shorebird that had been colour-marked. Various attempts and blocking, stratifying and constraining our survival models, while in many cases providing more parsimonious solutions, had no effect on these heterogeneities. Conclusions Our problems stem from both the methodology used and lack of information about the status of individuals (e.g. territoriality and variability in the level of site fidelity). We suggest ways in which data more appropriate to survival analyses might be collected in future. [ABSTRACT FROM AUTHOR]