Abundance estimation of Antarctic minke whales via spatial modelling

Abstract

This paper describes our work on Antarctic minke whale abundance estimation using line‐transect survey data collected between 1986 and 2010 as part of the IWC’s IDCR/SOWER programme. While we focus on spatial modelling (producing smooth ‘maps’ of animal density), much of the detail is concerned with allowing for imperfect detection of whales on the trackline (g0 < 1). We present an account of operational survey details, along with full descriptions of the bespoke methodological minutiae that have to be considered when trying to account properly for survey protocols and Antarctic minke whale behaviour. This paper describes the reality of advanced distance sampling in a complicated setting. A complete mathematical background is given for the spatial models of both school density and school size; for the SOWER‐specific adaptation of Trackline Conditional Independence which we developed for dealing with g0; and for how all these parts of the overall process can be linked. We present some of our own estimates, alongside the IWC’s agreed consensus estimates, in part based on our results. The paper ends with comments on interpreting the estimates, the remaining unsolved problems when analysing these particular data, and lessons for future whale abundance surveys.