Population Status of Exploited Marine Fish Populations

Abstract

Estimating the population status and trajectory of marine fishes is hampered by limited data and the assumptions needed to augment these data gaps can have broad implications. Poor management decisions based on poor inference can have far reaching socio-economic and ecosystem consequences. Here, I examine critical assumptions commonly employed in fisheries models and how they affect our beliefs on the population status of exploited marine fisheries. The assumption that catch rates are proportional to abundance is a critical assumption that can influence estimates of stock status. In chapter 3, this assumption is evaluated via a state-space modelling framework to obtain better estimates of the stock status of cusk (Brosme brosme) by incorporating multiple population time series in the estimation of population biomass. By relaxing the assumption that catch rates were directly proportional to population biomass, I found that the research trawl survey was likely hyperdepleted, thus exaggerating estimates of population decline. Indeed, previous estimates of population decline based solely on the the research trawl survey were 93% compared to my estimate of 64% once non-proportionality was incorporated in the modelling structure. In chapter 4, I model the spatial distribution and population trajectory of different size classes of thorny skate (Amblyraja radiata) to test for density dependent habitat selection (DDHS). I found large declines in both abundance and distribution for all size classes with the greatest declines observed in large juvenile and adult size classes. I found strong evidence for DDHS for the large juvenile and adult size classes, however, small juveniles exhibited limited changes in distribution associated with changes in abundance. In chapter 5, I examine how biological reference points used by fisheries to monitor harvest sustainability compare to those developed by conservation organizations, such as the IUCN Red List, to estimate extinction risk. I show that the two approaches frequently arrive at similar categorizations of stock status and differences are primarily a result of decline thresholds used to trigger conservation action. Conservation and fisheries metrics aligned well (70.5% to 80.7% for riskier and more conservative reference points respectively) despite their mathematical disconnect in decline thresholds. My analyses suggest conservation and fisheries scientists will agree on the status of exploited marine fishes in most cases, leaving only the question of appropriate management responses for populations of mutual concern still unresolved.