2021 journal article
Estimating discard mortality using meta-analysis and fishery-dependent sampling
FISHERIES RESEARCH, 240.
Estimates of discard mortality are difficult to obtain. Meta-analysis or life-history-based approaches to estimate discard mortality could provide informed estimates when direct empirical estimates are not available. We used data from published literature across a variety of fish species to determine if hooking condition (good vs. poor) and species-specific values for the Brody growth coefficient (K: a measure of fish physiology) were meaningful factors influencing discard mortality in hook and line fisheries. We then examined whether a two-step approach, combining condition- and physiology-specific estimates of discard mortality with data on proportion-by-hooking-condition hooking information for a fishery, could result in an estimate of discard mortality for dolphinfish Coryphaena hippurus comparable to an empirical estimate. A model with hooking condition, K and their interaction best fitted the published discard mortality data. K was an important negative covariate of discard mortality for good hooking condition, with higher K species experiencing greater rates of survival. In contrast, species in poor condition had similarly low rates of survival across a range of K values. Results suggests that hooking condition is the dominant source of mortality when fish are hooked in vital areas but that physiology should also be taken into account when estimating discard mortality for good condition fish. For the recreational dolphinfish fishery in the southeastern US, we estimated a median proportional discard mortality rate of 0.12 (95 % credible set: 0.07, 0.17) when combining the meta-analysis and field-collected proportion-by-condition data. This estimate was lower than the empirical estimate of dolphinfish discard mortality but the credible sets overlapped (median: 0.25; 95 % credible set: 0.05, 0.39). Estimates of discard mortality from our meta-analytic approach may be applicable to fisheries where empirical estimates of discard mortality are not available and hooking injuries are the dominant source of mortality.