Correction factors for prey size estimation from PenguCams

Abstract

The use of animal-borne cameras enables scientists to observe behaviours and interactions that have until now, gone unseen or rarely documented. Researchers can now analyse prey preferences and predator-prey interactions with a new level of detail. New technology allows researchers to analyse prey features before they are captured, adding a new dimension to existing prey analysis techniques, which have primarily relied on examining partially or fully digested prey through stomach flushing. To determine prey size, the video footage captured needs a correction factor (pixel:mm ratio) that allows researchers to measure prey dimensions using image measuring software and convert the pixels to actual measurements. This in turn will help estimating the prey energy content. This method requires a reference object with known dimensions (such as beak measurements) to ground truth your distance. Using PenguCams we determined the correction factor by measuring a 2 cm section of 1 mm grid paper from video footage taken at known distances (10, 20, 30, 40, 50, 60 cm) in different salinities ranging from air and fresh water, up to 35 psu in 5 psu increments while controlling for temperature and pressure. We found no significant difference between correction factors of water at different salinities. However, due to their considerable differences in refraction index, correction factors contrast between water and air. Linear equations modelled from correction factors at tested distances help predict correction factors between tested distances and, therefore, enable a wider application of this research. We provide examples from PenguCam footage taken of Humboldt (Spheniscus humboldti), Tawaki (Eudyptes pachyrhynchus) and King (Aptenodytes patagonicus) penguins to illustrate the use of identified correction factors. This study provides a tool for researchers to further enhance their understanding of predator-prey interactions.