Evaluating helicopter-based surveys for estimating densities of Himalayan thar

Abstract

Himalayan thar (Hemitragus jemlahicus Smith) inhabit a large area of New Zealand’s southern alps. Since 1993, a control plan for thar has required estimation of thar density within management units (MUs), and across their range. We evaluated helicopter-based survey as an alternative to current ground-based surveys to provide these estimates. Both survey methods sample thar density within predetermined survey-catchments located in MUs. Helicopter- and ground-based surveys undertaken in four survey-catchments (three located in MU4, one located in MU3) contrasted the performance of the two methods. Helicopter-based surveys used a double-count method to estimate the sightability of thar groups. Sightability varied from 0.53 for single thar, to 1.0 for groups of three or more. Correction factors accounting for sightability were applied to helicopter-based counts to estimate thar density within survey-catchments, which were contrasted with estimates from ground-based surveys. Density estimates for three of the survey-catchments were within 25%, and displayed no systematic bias, indicating reasonable convergence of the two methods. The effect of sampling rate on precision of density estimates was explored using helicopter-based survey results for survey-catchments sampled within MU4. Confidence intervals for the density estimate stabilised when 10–15 survey-catchments were sampled, equivalent to a sampling rate of 13–20% of the MU. Compared with ground-based surveys, helicopter-based surveys were 30% cheaper, less logistically constrained, and could be applied randomly across MUs. This suggests that they may be a better option to meet the information requirements for implementation of the thar control plan, at least in MUs where habitat is sufficiently open.