Abstract
Precise estimation of bat mortality at wind farms is essentialfor conservation planning. Carcass detection probabilities,however, are often low and data sparse, leading to large un-certainties in fatality estimates. Precision can be improved bypooling information across multiple sites and years. Using post‐construction monitoring data from 20 wind farms in Iowa, USA,we evaluated complete pooling (CP), no pooling (NP), andpartial pooling (PP) approaches for estimating searcher effi-ciency, carcass persistence probability, and the fatality esti-mates obtained by integrating these detection‐probabilitycomponents. For searcher efficiency and persistence proba-bility, PP produced estimates that balanced the extremes of CPand NP, yielding narrower credible intervals than NP, especiallyat data‐limited sites, while preserving meaningful site‐leveldifferences. Fatality estimates for 3 bat species, differing incarcass abundance, were generally similar in the PP and NPmodels, but both were often different from CP. Partial poolingshowed gains in precision, especially for abundant speciesand years with less monitoring effort. A simulation study ofscenarios with different amounts of heterogeneity among sitesshowed that PP provided more accurate predictions than CPand NP in intermediate and high heterogeneity settings. Ourfindings highlight the value of PP as a flexible strategy forimproving detection probability estimation under ecologicaluncertainty and support its broader application in wildlifeimpact assessment.