by Stephen J. Hall

Reprinted with permission from Trends in Ecology & Evolution, 1998, Vol. 13, No. 8, pp. 297-298

One measure that often meets great resistance from fishermen, but is beloved by conservationists, is the establishment of marine protected or "no take" areas (MPAs). However, although MPAs might serve a number of conservation functions [2], until recently, advocates of the approach could point to few convincing data and little underlying theory that would support the case for establishing them to enhance fish stocks.

Assessing the status and trends in the abundance of fish stocks is subject to considerable uncertainty and is now recognized as a major problem for both assessment biologists and fisheries managers. With the recent collapse of northern cod (Gadus morhua), this problem is now evident even for stocks that were traditionally considered to exhibit fairly predictable population dynamics [3]. Finding ways to deal with this uncertainty is uppermost in the minds of many fisheries biologists.

One approach is simply to try harder by broadening the information base upon which fisheries management advice is founded. Developing multispecies models that include interactions between species, rather than managing stocks as separate entities [4], or incorporating an understanding of recruitment processes from fisheries' oceanography studies [5], could well lead to improved management practices. However, such efforts carry a substantial price tag. There are many who feel that, in many cases, the benefits could be marginal relative to the expense [6].

An alternative approach is to seek better ways to deal with the fact that the world will always be uncertain and there are limits to what we can know [7]. The current strategy, however, is to ignore the uncertainty. Having made every effort to obtain the best possible estimate for a stock's current abundance and future trends, catch quotas are set with the best guess. This approach is understandable, because any admission of doubt on the part of fisheries managers or scientists opens a chink in the armor, allowing the fishing industry to argue for quotas at the upper limit of the confidence boundary. The industry can argue that there is "no proof" that a lower level of catch is necessary.

If high catch levels one year led to lower stock estimates and enforceable lower catch quotas the next year, a quota that was too high in a given year might not be a problem. Unfortunately, however, fisheries are not usually sufficiently self-correcting. A high catch this year will not necessarily lead to much lower estimates of stock abundance in subsequent years. Even if it did, the structure of most fisheries is such that annual adjustment of quotas does not seem to work very well, especially when the adjustment has to be large and downward. Besides, lowering high quotas is difficult to achieve politically when fishermen have loans to pay on expensive boats and want to maintain a decent standard of living.

Uncertainty, therefore, is a major problem for both assessment and management. In other uncertain situations, the normal response would be to hedge one's bets in some way and adopt a range of activities to reduce the risk of catastrophe. Such bet-hedging involves a cost in terms of reduced expected benefits, but it is outweighed by the reduced risk - in this case, of collapse or severe decline caused by overfishing. Could closing areas to fishing be a suitable bet-hedging tactic? The argument and analysis put forward by Lauck et al. [8] suggest that it can.

Fish stocks that inhabit large areas of ocean floor are the staple fishery resource for the trawl fleets of most industrialized nations. A reasonable management goal for these resources would be that the stock should remain at >60% of the carrying capacity over a given time horizon, say 20 years. Maintaining such levels would put the stock in the region of the optimal sustainable yield that many fisheries aspire to. Lauck et al. [8] consider how the establishment of an MPA might help achieve this goal using a mathematical model that incorporates uncertainty in harvest rates (box 1).

Figure 1

The model results are striking (figure 1), indicating that even when the levels of uncertainty in harvest are moderate (coefficient of variation <50%), the chances of achieving the management goal drop dramatically once the fraction of the total area available for harvesting becomes greater than about 30%. Perhaps more significantly, when the levels of uncertainty are larger, even allowing fishing access to only very small proportions of the area still gives a very high probability of failure.

These results show that high levels of uncertainty require a considerable degree of management control to assure a sustainable future for stocks, regardless of whether it is through MPAs or simply controlling effort. Interestingly, however, the imposition of an MPA might not lead to reduced catches. On the contrary, when the authors compared the effect of lowering the allowable catch level to achieve the management goal, rather than imposing an MPA, they found that up to 50% lower catches were obtained for the same level of protection. It would appear, therefore, that MPAs could make sense for both fishermen and conservationists.

The model that was used to draw these conclusions is extremely simple and can be challenged on many grounds. It is important to recognize, however, that although such strategic models do not reflect reality, they alert us to important underlying principles. In this case, they show us that MPAs can be an important bet-hedging strategy in an uncertain world, and could act as an effective insurance policy that would not only protect the long-term future of stocks, but also yield higher average catches. On this basis alone, they deserve careful consideration by the fisheries management establishment, particularly because in many respects this model is no simpler than those that have underpinned fisheries for the past 30 years.

Many others have written about the potential benefits of MPAs, and there are a variety of reasons why one might want to establish them [10, 11]. Not least is the ethical argument that there should be some marine areas where human influence is minimized - an argument that requires no science. If, however, MPAs are to be established on the grounds that they will support fisheries, it is most likely to happen if a convincing case can be made that benefits to fishermen will also follow. It would be foolish to suggest that the analysis described here is sufficient to justify the establishment of MPAs to protect fisheries. Indeed, there can be no substitute for careful ecological research to ensure that key criteria for reserve establishment are met. Reserves must, for example, act as source population for fished areas and should protect spawning grounds and other areas critical to the population. These criteria constitute an important research agenda for marine ecologists in the coming decade. Thankfully, however, the theoretical foundations that justify the effort are finally being laid.

Send us your comments and ideas for future articles.

Endlinks