Sensitivity analysis identifies high influence sites for estimates of penguin krill consumption on the Antarctic Peninsula

Abstract

AbstractKrill consumption by natural predators represents a critical link between surveys and models of standing krill biomass and the design of a sustainable krill fishery for the Scotia Sea. Antarctic krill (Euphausia superba) is a significant component of diet for penguins breeding in this region and, consequently, uncertainties regarding penguin population abundances contribute to uncertainties in krill predation estimates. We use a comprehensive database of Antarctic penguin abundances to identify 14 breeding colonies that contribute most significantly to uncertainty regarding the total number of pygoscelid penguins breeding in this region. We find that a high quality survey of Zavodovski Island alone would decrease uncertainty in total population by 24.8%, whereas high quality surveys of all 14 “high-influence” locations would decrease uncertainty by almost 72%. Updated population estimates at these sites should be considered top priority for future fieldwork in the region. Our results are based on a robust quantitative method for assessing data priorities in estimating krill consumption that is easily extended to other groups of krill predators.