A SIGHTABILITY MODEL FOR BIGHORN SHEEP IN CANYON HABITATS

Visibility bias (failure to observe all animals) encountered during ae rial surveys produces biased estimates of population parameters. Facto rs affecting visibility during helicopter surveys of bighorn sheep (Ov is canadensis) have not been quantified. We measured visibility bias f or helicopter surveys of bighorn sheep in southwestern Idaho. Visibili ty was influenced (P < 0.05) by activity, habitat, sex composition of groups, light condition, position of sheep relative to the helicopter, and topographic position but not by group size (P = 0.781). Multivari ate regression indicated that activity (P < 0.001) and habitat (P < 0. 002) variables were the most important factors affecting visibility. A sightability model was developed to estimate bighorn population and c omposition parameters from data collected during helicopter surveys. W e conducted 12 surveys in southwestern Idaho. The estimated population observed during helicopter surveys ranged from 51.7 to 78.1% and aver aged 67.1% (CV = 10.6%). Confidence intervals for population estimates ranged from 16.4 to 22.9% and averaged 18.5% (CV = 16.0%) of the popu lation estimate. We recommend correcting survey data for visibility bi as to estimate bighorn sheep population parameters.