Costs of detection bias in index–based population monitoring

摘要: Costs of detection bias in index�based population monitoring.� Managers wildlife populations commonly rely on indirect, count�based measures the making decisions regarding conservation, harvest, or control. The main appeal use such counts is their low material expense compared to methods that directly measure population. However, correct rests rarely�tested but often�assumed premise they proportionately reflect size, i.e., constitute a population index. This study investigates forest management for endangered Red�cockaded Woodpecker (Picoides borealis) and Wood Thrush (Hylocichla mustelina) at Piedmont National Wildlife Refuge central Georgia, U.S.A. Optimal decision policies a joint species objective were derived two alternative models dynamics. Policies simulated under scenarios of unbiasedness, consistent negative bias, habitat�dependent observed Thrush densities. Differences simulation outcomes between biased unbiased indicated the expected loss resource objectives (here, habitat birds) through decision�making based on biased counts. Given function used our analysis, losses were as great 11%, degree perhaps not trivial applications species management. Our analysis demonstrates costs uncertainty about relationship the population its observation can be measured units resource, which may offset apparent savings achieved by collecting uncorrected