Candidate detection function models used to estimate population densities of NPSA landbirds (2011 and 2018)
Judge SW, Camp RJ, Vaivai V, Hart PJ. 2022. Candidate detection function models used to estimate population densities of NPSA landbirds (2011 and 2018). Pages 5. In Status of landbirds in the National Park of American Samoa
Model parameters and model-selection results for forest birds of the 2011 and 2018 National Park of American Samoa survey. Within each species analysis, models were sorted by differences in second-order Akaike’s information criterion corrected for small sample size (∆AICc) between each candidate model and the model with the lowest AICc value. Key models examined included half-normal (HN) and hazard-rate (HR) key detection functions and with cosine (COS), Hermite polynomial (Hpoly), and simple polynomial (Spoly) series adjustments. Covariates were incorporated with the most parsimonious key model to improve model precision. Covariates included the categorical variables cloud cover (Cloud), amount of rain (Rain), Beaufort wind scale (Wind), Beaufort gust scale (Gust), observer (Obs), survey year (Yr), and survey panel (Panel). Time of day (MinSS) was evaluated as a continuous variable. The number of estimated parameters (Num params), and negative log-likelihood (-LogL) are presented. The Akaike model weight (AICc weights) is the likelihood that each model is the best of the converged models evaluated for each species.