Improving detectability of dolphin signature whistles for capture-recapture analysis: an examination of array configuration using real-world data

Abstract

Abundance estimates of cetaceans are often acquired through capture-recapture analysis of photographically identified individuals. An alternative method, using capture-recapture of individually distinct signature whistles detected from acoustic underwater recording units, has recently been demonstrated. Here we investigate the effect of array configuration (1–3 hydrophones within 0.45 km2) and recording duty cycles (six variations of 33%, 50%, or 66% sampling periods) on the detection rate of dolphin signature whistles. Twenty-one signature whistle types were identified and used to create capture histories for each hydrophone and all potential array configurations. Open population models were used to estimate capture probabilities and precision for all data sets. The effect of different duty cycles on detectability were investigated by artificially applying six duty cycles to the continuously recorded data. Results demonstrate that location is more important than redundancy in small-scale arrays, even with detection distances as small as 750 m, and that duty cycling can increase survey durations without decreasing detectability. To acoustically sample intermittent signals of dispersed populations, it is more effective to space hydrophones further apart, in known high-use areas. This study provides insight into the application of capture-recapture to signature whistles, improving methods for long-term, noninvasive monitoring of elusive delphinids.