Distance sampling to determine kaka (Nestor meridionalis septentrionalis) density within Waipapa Ecological Area, Pureora

Reliable information about population density and trends is essential for making valid inferences regarding conservation management. The suitability of point counts using distance sampling was examined as a means of monitoring a population of kaka (Nestor meridionalis septentrionalis), a large forest-dwelling parrot, inhabiting the Waipapa Ecological Area in the central North Island of New Zealand. Counts were conducted on 13 occasions between 2000 and 2007. The sampling design was tailored to maximise the detectability of kaka and to minimise violations of the four most important assumptions of distance sampling. Location errors and subsequent distance estimation errors were most likely to bias density estimates despite our attempts to minimise failures of this assumption. Densities estimated from counts conducted in October were similar between 2000 and 2007 (approximately 0.5 kaka ha^-1) with no evidence of either a positive or negative trend. Densities derived from counts in February or March were more erratic and seemed to reflect variation in the frequency and success of the preceding breeding season. Given the frequency of kaka breeding, the pest control regime during the study period, and our attempts to minimise violations of distance sampling assumptions, we are confident that the reported trends in density are realistic. Although distance sampling was found to give reliable density estimates of kaka at Waipapa, this may not be the case at other sites, particularly where kaka density is low, location error is high, forest structure or topography are more complex, or surveys of kaka are made part of more extensive multi-species surveys.