ESTIMATING AND DEALING WITH DETECTABILITY IN OCCUPANCY SURVEYS FOR FOREST OWLS AND ARBOREAL MARSUPIALS

摘要: Surveys that record the presence or absence of fauna are used widely in wildlife management and research. A false occurs when an observer fails to a resident species. There is growing appreciation importance absences surveys its influence on impact assessment, monitoring, habitat analyses, population modeling. Very few studies explicitly quantify rate these errors. Quantifying provides basis for estimating survey effort necessary assert species absent with pre-specified degree confidence allows uncertainty arising from be incorporated inference. We estimated 2 forest owl 4 arboreal marsupial based 8 repeat visits 50 locations south-eastern Australia. obtained estimates using generalized zero-inflated binomial model. presented detectability curves each convey number required achieve specified level will detected. The observation error rates we calculated were substantial but varied between For least detectable species, powerful (Ninox strenua), our standard returned 87% visits. However, more sugar glider (Petaurus breviceps) 45% rate. predict approximately 18 would 90% sure detecting owls 5 provide gliders. fitted hierarchical logistic regression models data describe variation detection explained by environmental variables. found temperature, rainfall, quality influenced most Consideration could result important changes resource conservation planning.