Effect of surfacing patterns on abundance estimates of long-diving animals

Abstract

Population sizes of long-diving marine mammals can be estimated with minimal bias using line transect sampling techniques in conjunction with data on surfacing/diving, even though the detection probability on the trackline is considerably less than 1. Using a simulation, it was investigated whether population sizes can be accurately estimated when only the mean cue production rate, without information on surfacing/diving patterns, is available. In addition, a new model that incorporates information on surfacing/diving patterns into the estimation procedure was developed. The simulation showed that the new model can provide less biased abundance estimates, whereas neglecting surfacing/diving patterns can result in serious underestimates of abundance, in particular for longdiving animals.