Utility of indicator thresholds across spatial gradients for applications to ecosystem‐based fisheries management Ecosphere

Ecosystem‐based fisheries management (EBFM) is increasingly recognized as an effective management strategy, especially in systems with dynamic components that have integrated responses to interacting drivers. Management targets therefore need to incorporate these complex dynamics, which may be accomplished by establishing targets based on the indicators. Here, we leverage existing data to evaluate the use of indicators in EBFM for nearshore ecosystems in Hawai‘i. We screened for potential nonlinearities in the spatial relationships between ecosystem indicators and environmental drivers over a 10‐year period, and assessed commonalities in identified indicator–driver thresholds. To understand how the ecosystem indicators relate to commonly used metrics in single‐species assessments, we also tested whether these relationships were consistent across species, species complexes, and functional groups. Gradient forests and generalized additive mixed models revealed detectable threshold relationships across five environmental drivers and six ecosystem indicators. Several indicators showed considerable overlap in threshold locations with common drivers, especially for rugosity (habitat complexity) and wave forcing. While some indicators comprising nested species, species complexes, and functional groups had comparable thresholds with a common driver, this was not always the case, highlighting the need for careful consideration when selecting indicators for management applications. We demonstrate the use of these thresholds in environmental drivers by evaluating them along a gradient of fish biomass. The use of multi‐model inference to incorporate multiple statistical methods to define the relationship between an ecosystem indicator and driver offers a quantitative basis for ecosystem management. Our results here provide a rubric by which to test for the existence of thresholds in indicator–driver relationships across spatial gradients and additionally identify clear thresholds for which to consider when managing fisheries. Grant no. NA20NMF4410262