Dive wiggles as a proxy of prey consumption in krill-feeding penguins

The dive profiles of pursuit-diving marine predators are often used to infer foraging behaviour, including potential indicators of prey consumption. ‘Wiggles’ are undulations in dive profiles that relate to foraging activity in a variety of marine predators. In penguins, wiggles are sometimes used as a proxy for prey consumption (e.g., catch per unit effort, CPUE), but this relationship remains poorly validated and likely varies with diet. We deployed animal-borne video cameras and depth recorders on chinstrap penguins (Pygoscelis antarcticus; n = 37) and identified over 17,000 euphausiid prey captures - mainly Antarctic krill (Euphausia superba) - during dives deeper than 3 m (n = 2458 dives). Using the video-observed prey captures as a reference, we tested how well various wiggle metrics derived from 1 Hz depth data predicted krill consumption by the penguins. Wiggle metrics generally showed a positive but noisy and highly variable relationship with the number of krill captured per dive, with association strength varying among metrics. While it is tempting to infer detailed foraging behaviours from dive wiggles (including ‘bottom distance’ generated by the R package diveMove), our results show: (1) notable rates of foraging – non-foraging dive misclassification; (2) only moderate agreement between CPUE estimated from wiggle counts and video observations; and (3) imprecise predictive models of actual prey consumption. While wiggle analyses offer some insight into prey consumption of krill-feeding penguins, our results suggest that alternative methods (e.g., acceleration-based indices) are needed to obtain more robust quantitative estimates of prey consumption.