Comparing Data‐Poor and Data‐Rich Stock Assessments to Generalise Guidance for Pacific Coral‐Reef Fisheries

Abstract

The growing use of data‐poor fisheries models provides unprecedented access to compelling stock metrics and management scenarios for coral‐reef fisheries. Yet, it remains unclear how well the assumptions surrounding data‐poor models derived from extensive cold‐water fisheries fit the life histories and ecology of tropical coral‐reef fishes. The present study compared the outcomes from length‐based spawning potential ratio (LBSPR) models versus observed trends from a novel meta‐analysis of nine fisheries‐independent (FI) and fisheries‐dependent (FD) data streams across Micronesia. Thirty target species that comprised over 70% of landings were examined across temporal and spatial fishing pressure gradients associated with FI and FD streams. Cluster analyses using normalised effect sizes from over 140 statistical tests grouped the stocks based on the magnitudes of decline in (proportional) biomass and size‐structure. Interestingly, many species with the greatest declines had modelled SPR and fishing mortality outcomes that did not trigger management based on established thresholds. These inconsistencies were attributed to several factors: variable compensatory density dependence across species, rapid growth over short time periods for small‐bodied species leading to hyper‐sensitive switch points in SPR metrics, and unusually long lifespans that did not always correlate with body sizes or growth rates. The complexity of life histories appeared to limit the use of snapshot data‐poor models for depicting relative stock status across species; however, data‐poor outcomes through time for each species better aligned with FI and FD trends. We generalise management guidance for a common and dominant group of tropical Pacific species.