A hierarchical nest survival model integrating incomplete temporally varying covariates.

摘要: Nest success is a critical determinant of the dynamics avian populations, and nest survival modeling has played key role in advancing ecology management. Beginning with development daily models, proceeding through subsequent extensions, capacity for effects hypothesized factors on expanded greatly. We extend models further by introducing an approach to deal incompletely observed, temporally varying covariates using hierarchical model. Hierarchical offers way separate process observational components demographic obtain estimates parameters primary interest, evaluate structural ecological management interest. built model analyze data from reintroduced whooping cranes (Grus americana) Eastern Migratory Population. This reintroduction effort been beset poor reproduction, apparently due primarily abandonment breeding birds. used assess support hypothesis that caused harassment biting insects. obtained indices blood-feeding insect populations based spatially interpolated counts insects captured carbon dioxide traps. However, trapping was not conducted daily, so we had incomplete information variable covariate therefore supplemented our parallel estimating values missing covariates. Bayesian selection identify best predictors survival. Our results suggest black fly Simulium annulus may be negatively affecting cranes, decreasing as abundance S. increases. The framework have developed will applied future larger set biting-insect other hypotheses nesting failure this population; resulting inferences ongoing efforts manage population via adaptive approach. Wider application promise varying, but imperfectly observed survival, including possibility collected incubating adults.