Assessing visibility bias associated with helicopter counts of feral pigs in Australia's semi-arid rangelands

Abstract

Aerial survey is a rapid technique for obtaining information about the abundance and distribution of feral pigs. To obtain estimates of actual pig abundance from such surveys, visibility bias associated with counts of pig groups must be estimated and accounted for. A double-count survey technique was used to estimate habitat-specific visibility bias associated with counts of pigs obtained from standardised helicopter surveys conducted in the semi-arid rangelands. Visibility bias differed significantly between riverine black-soil floodplains and non-riverine habitats, suggesting that at least two visibility-bias correction factors (3.61 and 1.47,respectively) were needed to derive absolute estimates of abundance from observed counts. The density of pigs was reduced experimentally on six sites by helicopter shooting or trapping to (1) evaluate the effect of helicopter shooting on visibility bias during subsequent helicopter surveys, and (2) provide further information on visibility bias associated with black-soil floodplain habitat. The experiments indicated no effect of helicopter shooting on visibility bias. However, a significant positive relationship between percentage cover of black-soil floodplain and visibility bias on study sites used for removal experiments suggests that estimates of visibility bias for this habitat derived from double-count surveys may be underestimates.