Validation of quantitative fatty acid signature analysis for estimating the diet composition of free-ranging killer whales.

Accurate diet estimates are necessary to assess trophic interactions and food web dynamics in ecosystems, particularly for apex predators like cetaceans, which can regulate entire food webs. Quantitative fatty acid analysis (QFASA) has been used to estimate the diets of marine predators in the last decade but has yet to be implemented on free-ranging cetaceans, from which typically only biopsy samples containing outer blubber are available, due to a lack of empirically determined calibration coefficients (CCs) that account for fatty acid (FA) metabolism. Here, we develop and validate QFASA for killer whales using full blubber from managed-care and free-ranging individuals. First, we compute full, inner, and outer blubber CCs from the FA signatures across the blubber layers of managed-care killer whales and their long-term diet items. We then run cross-validating simulations on the managed-care individuals to evaluate the accuracy of diet estimates by comparing full-depth and depth-specific estimates to true diets. Finally, we apply these approaches to subsistence-harvested killer whales from Greenland to test the utility of the method for free-ranging killer whales, particularly for the outer blubber. Accurate diet estimates for the managed-care killer whales were only achieved using killer whale-specific and blubber-layer-specific CCs. Modeled diets for the Greenlandic killer whales largely consisted of seals (75.9 ± 4.7%) and/or fish (20.4 ± 2.4%), mainly mackerel, which was consistent with stomach content data and limited literature on this population. Given the remote habitats and below surface feeding of most cetaceans, this newly developed cetacean-specific QFASA method, which can be applied to outer-layer biopsies, offers promise to provide a significant new understanding of diet dynamics of free-ranging odontocetes and perhaps other cetacean species throughout the world's oceans.