Correcting Bias in Survival Probabilities for Partially Monitored Populations via Integrated Models

Sammanfattning

We provide an integrated capture-recapture-recovery framework for partially monitored populations. In these studies, live resightings are only observable at a set of monitored locations, so that if an individual leaves these specific locations, they become unavailable for capture. Additional ring-recovery data reduce the corresponding bias obtained in the survival probability estimates from capture-recapture data due to the confounding with colony dispersal. We derive an explicit efficient likelihood expression for the integrated capture-recapture-recovery data, and state the associated sufficient statistics. We demonstrate the significant improvements in the estimation of the survival probabilities using the integrated approach for a colony of guillemots (Uria aalge), where we additionally specify a hierarchical approach to deal with low sample size over the early period of the study.Supplementary materials accompanying this paper appear online.

Nyckelord

Bias; Capture-recapture-recovery data; Hierarchical model; Partial monitoring

Publicerad i

Journal of Agricultural, Biological, and Environmental Statistics
2021
Utgivare: SPRINGER

SLU författare

• Hentati Sundberg, Jonas

  • Institutionen för akvatiska resurser (SLU Aqua), Sveriges lantbruksuniversitet
    

UKÄ forskningsämne

Sannolikhetsteori och statistik
Ekologi

Publikationens identifierare

• DOI: https://doi.org/10.1007/s13253-020-00423-1

Permanent länk till denna sida (URI)

https://res.slu.se/id/publ/110357