Incorporating uncertainty reveals resilience and efficiency differences between three neighboring marine food webs.

While general principles of ecosystem-based approaches to fisheries management have been articulated for over a century, their large-scale implementation remains a challenge. Operationalizing ecosystem-based management demands the ongoing development of indicators and ecological reference points, analogous to biological reference points of single species fisheries management. For such ecosystem-level indicators to be useful in applied contexts, we must appropriately account for uncertainty. Here, we focused on advancing the utility of one specific indicator, relative ascendancy, by incorporating parameter uncertainty. Relative ascendancy quantifies the efficiency of biomass flow through a food web, and efficiency is thought to be inversely related to overall system resilience. We constructed mass balanced food web models of three neighboring regions within the Northeast US continental shelf using the open-source Rpath framework. We incorporated parameter uncertainty into indicator calculations using the Ecosense simplified Bayesian synthesis routine, which allowed us to compare across model ensembles, rather than single parameterizations. Although our regions are adjacent and interconnected, and our models were parameterized using identical data sources, the inclusion of parameter uncertainty revealed meaningful differences in their ascendency properties. Upon further investigation, we linked overall system efficiency to whether its dominant trophic pathways are benthic, pelagic, or both. We present our uncertainty-informed analyses in the hopes it will advance the utility of the relative ascendancy indicator and support ecosystem-based management in the Northeast United States and beyond.