A Generalised Likelihood Framework for Partially Observed Capture-Recapture-Recovery Models

Ruth King and Rachel McCrea

Universities of St. Andrews and Kent

Summary

We provide a closed form modular-type likelihood for multi-state mark-recapture-recovery data when the state of an individual may be only partially observed. The generalised matrix form likelihood presented uniﬁes much disparate existing theory and provides a consistent and uniﬁed framework with many standard models as special cases, including, the Arnason-Schwarz model, observation error models and mixture models. The likelihood speciﬁcation paves the way for model discrimination, assessing absolute goodness-of-ﬁt, in addition to its use in model ﬁtting and parameter estimation.

Keywords:

Mark-recapture-recovery data; Closed form likelihood; Multi-state; Partially observed states; Suﬃcient statistics