An assessment of the accuracy of kangaroo surveys using fixed-wing aircraft

Abstract

The use of line-transect methodology, on foot or from a helicopter, is likely to return the most repeatable, least biased estimates of kangaroo density. However, the associated costs make both methods impractical for broad-scale surveys. For these, a fixed-wing aircraft remains the most cost-effective platform. Limitations of the standard fixed-wing method (200-m strip transects) are well known, but it continues to be used because it provides an index of trends, because there are now long runs of data (almost 20 years in some cases) collected in this standard form and an alternative method is lacking. In this study, four variations of fixed-wing surveys of kangaroos were investigated: two line-transect methods (involving different scanning techniques), the standard 200-m strip transect and a 100-m strip transect. Surveys using these methods were compared with helicopter line-transect surveys along the same flight lines in three areas (5000–7500 km²) in western Queensland. Both fixed-wing line-transect methods failed to produce consistently accurate estimates of density for all three species surveyed: red kangaroos (Macropus rufus), eastern grey kangaroos (M. giganteus) and common wallaroos (M. robustus). While generally more accurate than the uncorrected strip-transect counts, they were no less variable. However, the strip-transect counts still need to be corrected for bias for which this study offers revised estimates of correction factors for eastern grey kangaroos (3.7–10.2) and common wallaroos (3.8–4.1), and estimates for red kangaroos (1.7–2.7) that support currently used values. An attractive alternative is to survey in 100-m strip transects, which offer improved visibility (correction factors of 1.0–1.8 for red kangaroos, 2.1–3.6 for eastern grey kangaroos and 1.7–2.1 for common wallaroos) and are therefore likely to be more accurate and repeatable. However, these advantages need to be assessed in relation to continuing long runs of data using the standard 200-m strip transect. Correction factors for wallaroos are conservative as helicopter-based density estimates are known to be underestimates. Further work is needed to assess the generality of correction factors, both spatially and temporally.