Unplanned ecological engineering.

Fisheries can double the production of protein and revenue by abandoning current single-species management. This provocative prediction is the implication of the work in PNAS by Szuwalski et al. (1). Using the East China Sea as a case, they show how an indiscriminate fishery can support unexpectedly large catches by removing predators from the ecosystem. Such ecosystem engineering stands in stark contrast to reigning management paradigms that do not allow fishing down predators to increase the productivity of their prey. The theoretical support for such a feat of ecosystem engineering is well developed (2, 3). Trusting the Chinese catch statistics, Szuwalski et al. (1) provide empirical evidence that theory may be turned into practice. But their work is more than “just another fisheries paper;” it underscores highly controversial issues about the unavoidable trade-offs in managing fisheries and ecosystems. If we narrowly consider food security, maximizing fisheries catch from the ecosystem is a “no-brainer,” but from a conservation point of view, the loss of biodiversity in the East China Sea may seem like Aquacalypse come true (4). Can we really double fisheries’ production by turning the oceans into mega-scale mariculture operations? Is it what we want? Is It Really True? The surprisingly large catches in the East China Sea are reached by the ecological process of prey release. The removal of large predatory species from the ecosystem releases a substantial production from their prey, smaller species that are usually kept in check by predatory fish (Fig. 1). In this manner, the fishery can harvest what would otherwise have been eaten by the predators. This ecosystem cultivation parallels how humans turned from hunter-gathering to agriculture: By removing grazers from crops and isolating grazers from predators, the production of crops and meat for human consumption could be increased to levels beyond what natural ecosystems could support. It is difficult to perform controlled experiments on natural ecosystems. Consequently, Szuwalski et al. (1) resort to ecosystem models to explore the impacts of fishing. They use a “size spectrum” model that compensates for the lack of detailed information about the feeding relations among all species in the East China Sea, with a reliance on the rule that big fish eat smaller fish (Fig. 1). Themodel rests on balance between growth and mortality: Growth and reproduction of a larger predator are fueled by the death of its smaller prey. This process is a biologist’s version of mass conservation, and it is responsible for the observed prey release. Models are idealizations of reality, and our understanding of ecosystem processes is far from complete (5, 6). Even though prey release seems logical, not all models support it. For example, a highly influential paper did not see increased catches when all components of an ecosystemwere fished (7). By demonstrating correspondence between model simulations and the outcome of the “ecosystem experiment” in the East China Sea, Szuwalski et al. (1) lend credibility to models that replicate prey release. Combining this evidence with observations of trophic cascades, increases in smaller species when predatory fish are overfished (8, 9), makes prey release seem a likely process. Even though the oceans are severely altered by fishing, they are still wild and uncontrolled ecosystems. Things may therefore not go as planned: Similar to how an organic corn field may be invaded by weeds or insect Food Mortality Bi om as s