Correcting Bias in Aerial Survey Population Estimates of Feral Livestock in Northern Australia Using the Double-Count Technique

Abstract

(1) The suitability of the double-count technique for estimating and correcting negative bias in population estimates, of feral livestock derived by multispecies aerial survey was examined. (2) Visibility bias was estimated for buffalo, cattle, horses and donkeys, in open and woodland habitats, from double counts of groups obtained during a large-scale aerial survey in the Northern Territory. Correction factors were derived to adjust observed number of groups to absolute estimates (corrected number of groups x mean group size). (3) The degree of bias was greater between habitats than between species, but was unrelated to the size of groups. Despite surveying multispecies assemblages in two habitats, the counts by observers were consistent and similar. (4) The validity of the double-count technique for correcting aerial survey bias in population estimates of donkeys and buffalo in woodland was tested by removal experiments. Population estimates derived by the index-manipulation-index method were similar, and showed that the double-count technique was sufficiently accurate. (5) The double-count technique is a suitable and cost-effective method of improving the accuracy-of population estimates of feral livestock from multispecies aerial survey.