Analyzing patterns in population dynamics using repeated population surveys with three types of detection data

摘要: Abstract To facilitate the use of population counts as an index change, we describe a generalization distance sampling methodology to analyze, in addition observer, two other ways estimate imperfect detection probability: multiple observers and time-to-detection, flexible manner, meaning that not all sites or years need have information be surveyed same way every year. We also account for effect partially-observed individual covariates, group size on probability. Finally, separate probability availability from itself. perform thorough, illustrated assessment pros cons this framework with simulations real case studies. First, compare simple linear models, illustrating magnitude bias caused by detection. Second, standard sampling, variation However, was weakly identifiable, ability it probability, therefore debias trend estimate, depended data configuration. Combining time-to-detection solved weak identifiability applied study. recommend using both type analysis showcase, regression count against time. Discrepancies between results complex analyses can help identify sources former loss precision latter within logistical constraints local wildlife management schemes.