This repository includes 90 simulated avian monitoring datasets, each consisting of data for 160 occupied sites, as well as the results of field PAM surveys for American woodcock (Scolopax minor), Wood Thrush (Hylocichla mustelina), and Eastern Whip-poor-will (Antrostomus vociferus). We used three occupancy modeling approaches (single-season static occupancy models, the Royle-Nichols abundance model, and zero-inflated beta-binomial models) to estimate occupancy in each dataset. We include model fitting results for each monitoring dataset under a variety of surveying intensities (i.e., subset to a varying number of visits to each site) and, for the simulated data, varying underlying occupancy rates (i.e., 20 %, 50 %, or 80 % occupancy for the simulated data). This repository also includes the scripts used to generate simulated data, fit models, and generate the figures for the main manuscript and the supplemental information. These data and scripts can be used to investigate the scenarios under which animal behavior violates the assumptions of occupancy models. It can also be used to understand how assumption violation impacts the accuracy of occupancy estimation under a variety of survey design and modeling choices.
Data and code from: Addressing widespread detection heterogeneity in avian occupancy modeling using passive acoustic surveys
Like
Liked
Date:
