A new approach to the "apparent survival" problem: estimating true survival rates from mark-recapture studies.

摘要: Survival estimates generated from live capture-mark-recapture studies may be negatively biased due to the permanent emigration of marked individuals study area. In absence a robust analytical solution, researchers typically sidestep this problem by simply reporting using term "apparent survival." Here, we present hierarchical Bayesian multistate model designed estimate true survival accounting for predicted rates emigration. Initially use dispersal kernels generate spatial projections probability around each capture location. From these projections, individual and resulting values bias-adjusted histories. When tested simulated data sets featuring variable detection probabilities, rates, patterns, consistently eliminated negative biases shown apparent standard models. applied case concerning juvenile in endangered Cape Sable Seaside Sparrow (Ammodramus maritimus mirabilis), increased more than twofold above estimates. Our approach is applicable any design should particularly valuable organisms with dispersive life stages.