Surveying malleefowl breeding densities using an airborne thermal scanner

Abstract

When malleefowl, Leipoa ocellata (Megapodiidae), open their incubator-nests (mounds), relatively high temperatures are exposed, providing the possibility of sensing mounds remotely with an airborne thermal scanner. The feasibility of using this technique for surveying malleefowl populations was evaluated by conducting a test scan over four sites where the locations of active mounds were known, and by a groundbased study that examined the factors associated with the time and frequency of mound opening by the birds. In all, 26% of known active mounds were detected on 'quick-look prints' produced by the scanner. Detailed image analysis revealed further mounds and showed that all mounds detected were unambiguously indicated by maximum pixel temperature. The ground-based study showed that weather conditions in spring were poor predictors of mound-opening behaviour. However, the opening time of mounds was positively correlated the date, suggesting that scans would be most successful early in spring. In summer, the mound-opening behaviour of malleefowl differed markedly from that in spring; fewer mounds were opened on summer mornings and opening times were later and were strongly correlated to weather variables (but not with date). Using the ground-based data to model the probable success of scans under differing conditions, we estimate that scans covering 90 km2 (90 min duration) would detect up to 36% of active mounds on cloudy mornings in mid-October, compared with about 25% in mid-November and about 15% in summer. Repeated scans would substantially increase detection rates. We conclude that the technique is feasible, cost-effective and capable of vast coverage, although further development is required before broad-scale application.