Estimating abundance of a small population of Bryde's whales: a comparison between aerial surveys and boat‐based platforms of opportunity

Abstract

AbstractAccurate abundance estimates are essential for the development of effective conservation management strategies, yet they are difficult to produce for small populations that are elusive and sparsely distributed throughout their range. For such populations it is challenging to collect a representative dataset sufficient for robust estimation of detectability and abundance. Over a one‐year study, we used two methods to estimate abundance of a Nationally Critical, widely dispersed Bryde's whale population in the Hauraki Gulf, Aotearoa/New Zealand; (i) distance sampling from systematic line‐transect aerial surveys (n = 22 surveys, 9,944 km, total sightings 21–24 whales), and (ii) mark‐recapture (MR) using photo‐identification images collected from a platform‐of‐opportunity and small‐boat surveys (218 sampling occasions, 27 whales). From the aerial surveys, we estimated an average of 15 whales (95% CI = 6, 30; CV = 37%) at the sea‐surface at any time. For the boat‐based surveys, we developed a custom MR model to address seasonal and individual heterogeneity in capture probabilities and obtained an estimate of 72 distinct whales (95% CI = 38, 106; CV = 24%) in the population. These two approaches provide different perspectives on the abundance and dynamics of Bryde's whales. The aerial surveys estimate the average number of individuals present at any one time, whereas the MR model estimates the total number of animals that used the Gulf during the study. Although neither sampling method is optimal for estimating the abundance of this small, dispersed population, the use of two complementary approaches informs conservation managers about patterns of abundance and distribution over different temporal and spatial scales. It is common to have limited resources for marine research where model assumptions cannot be met. Here, we highlight pragmatic strategies showing how models can be customized to the population of interest to assist with monitoring species of conservation concern.