From the Cover: Predicting overfishing and extinction threats in multispecies fisheries

Threats to species from commercial fishing are rarely identified until species have suffered large population declines, by which time remedial actions can have severe economic consequences, such as closure of fisheries. Many of the species most threatened by fishing are caught in multispecies fisheries, which can remain profitable even as populations of some species collapse. Here we show for multispecies fisheries that the biological and socioeconomic conditions that would eventually cause species to be severely depleted or even driven extinct can be identified decades before those species experience high harvest rates or marked population declines. Because fishing effort imposes a common source of mortality on all species in a fishery, the long-term impact of a fishery on a species is predicted by measuring its loss rate relative to that of species that influence the fishery’s maximal effort. We tested our approach on eight Pacific tuna and billfish populations, four of which have been identified recently as in decline and threatened with overfishing. The severe depletion of all four populations could have been predicted in the 1950s, using our approach. Our results demonstrate that species threatened by human harvesting can be identified much earlier, providing time for adjustments in harvesting practices before consequences become severe and fishery closures or other socioeconomically disruptive interventions are required to protect species.Marine fisheries are an important global source of food and livelihoods (1–4), but there are concerns that current fishing practices threaten some marine species with severe depletion or eventual extinction (2–5). Many of the largest commercial fishing methods, such as trawling, longlining, and seining, unavoidably catch multiple species simultaneously (6–9). Multispecies fisheries pose a particular threat of extinction or severe depletion because fishing can remain profitable as long as some valuable species remain abundant, even while others collapse (6–11). In contrast, in a single-species fishery profits tend to fall as the target population declines, thereby removing the incentive to fish before extinction occurs (10). Multispecies fisheries pose a threat to two types of species or stocks (populations): (i) commercially valued species, called “weak stocks”, which are more vulnerable to overharvesting than are other commercially valuable species (6), and (ii) by-catch species, which are caught accidentally and create little economic incentive to cease fishing as their populations collapse because they have little or no commercial value (7–9).Failure to prevent collapse of weak stocks and by-catch species can impose substantial long-term environmental and economic costs. Slow-growing populations are most likely to collapse, but can take several decades to recover (5). Recovery often requires long-term fishery closures or reductions in effort, having substantial economic and social consequences (3, 5). Moreover, population declines caused by one fishery can diminish yields and profits in other commercial or artisanal fisheries that depend on the same species (e.g., ref. 12).Despite these costs, species threatened by fishing have rarely been identified until after their populations have declined substantially (2–5, 7, 8). Assessments of fishery impacts on species mostly focus on estimating current exploitation rates or past population trends (13–15), which identifies already declining species rather than predicting future declines. Data limitations have made empirical prediction of future threats from fishing challenging, particularly for weak stocks and by-catch species. Oceans are difficult to sample extensively, and few economic incentives exist to gather data on species other than the most commercially valued species (7, 8). Some predictive models (e.g., ref. 16) have been developed to forecast the impacts of some fisheries, but these are often data intensive. Some of the characteristics that make a population susceptible to overfishing are well known—for example, low population growth rates (3–11, 17, 18), high value and/or low fishing costs (10, 11, 17–19), and schooling behavior (18). Recently, some correlative approaches based on these characteristics have been developed for assessing likely relative threats to data-poor species (4, 20–22). However, predicting the severity of future threats in absolute terms with this type of approach can be challenging.Here, we present a mechanistic approach that uses readily available data to predict the potential of current fishing practices, if maintained, to eventually cause a population to be driven extinct or “overfished”, here defined as depletion below its maximum sustainable yield (MSY) abundance (N_MSY) (3). Our approach identifies combinations of biological and socioeconomic conditions that are likely to eventually lead to high mortality rates and population declines. As we show, these conditions can be identified long before either occurs.We test the predictive power of our approach on eight tuna and billfish populations of the Western and Central Pacific Ocean fisheries. High-seas tuna and billfish have elicited recent conservation concern due to significant population declines and range contractions found in many species (17, 23, 24). Three of the populations in our study, bigeye tuna (Thunnus obesus) and both the northern and the southern striped marlin (Tetrapturus audax) populations, have been recently identified as experiencing overfishing—meaning their exploitation rates have exceeded the MSY exploitation rate (F_MSY) (24–27). A fourth, blue marlin (Makaira nigricans), whose overfishing status has been subject to considerable uncertainty (28), has undergone a significant population decline and range contraction (13, 23, 28). We determine whether our approach could have predicted threats to these four populations, using data from as early as the 1950s, and assess the threats predicted by the latest available data to all populations.