Repeatability of aerial surveys

In order to monitor population trends through space and time, aerial surveys need to provide density estimates that are a constant proportion of the true population density and are therefore repeatable. Previous work has identified factors that affect the visibility of kangaroos from fixed-wing aircraft and the relative contribution of these factors to the variation in counts. However, this work has relied on repeated surveys of a population of constant size. The development of accurate survey methods using helicopters has allowed the bias in fixed-wing surveys to be determined directly. Bias can therefore be estimated and assessed between areas and over time. Comparisons of kangaroo density estimated by these two aerial survey methods were made over 1991-97 near Blackall in central-western Queensland. Estimates of bias declined with increasing density for red kangaroos Macropus rufus, were relatively constant for eastern grey kangaroos M. giganteus and were highly variable for common wallaroos M. robustus. Nevertheless, medium-term population trends for the two kangaroo species were moderately well tracked. Currently, correction factors for fixed-wing surveys are applied at the scale of a 1 km² survey unit or a transect line of many units. These factors made some improvement to repeatability in this study, but the resultant density estimates remained biased. Correcting for bias on a regional scale may be more appropriate.