Estimating Demographic Parameters from Complex Data Sets: A Comparison of Bayesian Hierarchical and Maximum-Likelihood Methods for Estimating Survival Probabilities of Tawny Owls, Strix aluco in Finland

摘要: We compared a method of moments approach using estimates from maximum likelihood framework, ultrastructural models within and hierarchical estimated Markov chain Monte Carlo Bayesian framework for estimating survival recapture probabilities their variance components large, complex 20 year data set consisting both live recaptures recoveries. Estimates mean age-specific four age classes (young, second-year, third-year adult) were similar with all approaches, but the year-specific parameters some recovery on boundaries, leading to overestimates individual adult hence components. All approaches coefficients relationships between winter temperature probabilities, appeared exaggerate variation in relation prey abundance. Annual sensitive choice structure; modelling difference adults as random effects better patterns annual than treating independent. Our comparisons suggest that may be more likely produce reliable methods, even large sets, especially if there are many considerable sample sizes.