Harbour porpoise behavior and density in tidal rapids.

Speaker: Jamie MacAulay (University of St.Andrews)

Abstract

Density estimation in the complex and dynamic 3D environments of tidal rapid sites is challenging due to both large variations in animal density and substantial changes in noise, bathymetry and animal vertical distribution over small spatial scales. This results in a highly variable and constantly changing probability of detection. Surveys were conducted with a combination of different drifting PAM units at a tidal rapid site in Scotland. These consisted of both multi-channel arrays and simpler single channel acoustic recorders. The arrays can localise the positions of vocalising harbour porpoises and thus provide detailed information on underwater behaviour and movement. However the localisation range of the arrays is far less than the maximum detection range and thus tabulated distances cannot be used to construct a probability of detection function. Here we propose an alternate framework for calculation of density in tidal streams based on distance sampling: first the detailed behavioral information collected in the tidal rapids and beam profiles from captive animals are used to simulate a 2D probability of detection with respect to depth and range. The total effective detection volume is then calculated based on the calculated  probability of detection, noise and bathymetry. Density in different areas of the tidal stream is then modeled using both localised and non-localised clicks at different depths allowing for a 3D estimate of relative click density. The density of both buzz (foraging) clicks and echolocation clicks can be calculated which allows for a metric of foraging rates in different parts of the tidal rapids.