For at least 50 years Alaska's Bristol Bay has been one of the most valuable fisheries in the U.S. On average, fishermen net about 25 million sockeye salmon annually in the bay's chilly waters. In 2009 the catch was worth more than $120 million.

Scientists at the University of Washington in Seattle's (U.W.) School of Aquatic and Fishery Sciences think they know why Bristol Bay is so productive year after year: Several hundred discrete populations of sockeye salmon inhabit the network of rivers and lakes that empty into the bay, and this tremendous population diversity buffers the entire fishery against the vicissitudes of the environment.

Diversity within the species creates what the scientists call a "portfolio effect," named for its resemblance to a diversified investment portfolio: Some salmon stocks do better under certain conditions, whereas others thrive under different constraints, but the fishery as a whole remains stable.

Scientists have long known that biological diversity stabilizes ecosystems. But as the U.W. scientists report in a paper on the portfolio effect published June 3 in Nature, population diversity usually gets far less attention than species diversity, even though populations may be going extinct at rates 1,000 times higher than those for species. "The measurement we use for assessing biodiversity is focused almost entirely on loss of species," says ecologist Daniel E. Schindler, lead author of the paper. (Scientific American is part of Nature Publishing Group.)

Schindler and his colleagues analyzed five decades of data from Bristol Bay, including counts, ages and timing of salmon returning to spawn in the area's lakes, rivers and tributaries. Each fish returns to its birthplace, and the diversity of these nurseries is what produces variance within the species. For example, male sockeye coming home to a deep lake grow humped backs to attract females, becoming so round they look like dinner plates. But male sockeye returning to a shallow stream would be easy targets for grizzly bears if they had humps protruding from the water. Instead, these fish are shaped like torpedoes.

Timing also contributes to diversity: Some Bristol Bay sockeye spend a year in freshwater before migrating to the ocean, whereas others stay for two years. Some remain at sea longer—and in different locations—than others do; some return to spawn earlier in the season. Not all of these life-history strategies pay off every year, Schindler says, but "there are enough winners to make up for the losers."

The portfolio effect is beneficial for fishermen, too. Schindler and his colleagues predict that without this adaptation Bristol Bay's fisheries would be closed once every two to three years because of poor fish returns, rather than once every 25 years. In places where population diversity is much lower—such as the Sacramento River in California and the once-mighty Columbia in Washington State—salmon fisheries have declined precipitously and go through frequent boom and bust cycles.

The scientists say that their paper, which is the first to quantify the stabilizing effects of population diversity, is a "game changer" because it suggests better ways of managing species everywhere. "We start thinking not of how many fish we have coming back, but of the diversity of fish we have coming back," says co-author Ray Hilborn, also at U.W.

One way to promote diversity is by aggressively protecting habitat, rather than focusing on the stocks that seem to be doing well at the moment. "Anyone who put all his money on the 'hot' stock, whether it was Enron or Florida real estate, has learned that lesson," Hilborn says.

The portfolio effect also suggests that managers should minimize the use of hatcheries, which produce a more homogeneous population, and that weak stocks should be protected from overharvesting. "Populations that are really weak right now can easily turn into the populations that are strong in the future," Schindler explains.

The Bristol Bay findings offer hope for species affected by the Gulf oil spill, according to Schindler and his colleagues. Even if oil wipes out some populations of fish and other animals, others within the same species—occupying different habitat niches or hatching at different times—are likely to survive. The most resilient species will be those that don't put all their eggs in one basket.

The portfolio effect is also expected to help species adapt to climate change. Although Bristol Bay is targeted for mining and hydropower, it is currently surrounded by mostly undisturbed habitat, with no dams or hatcheries. Protecting this salmon stronghold will be even more important in a warmer future, says Michael Webster of the Gordon and Betty Moore Foundation, a co-author of the paper along with the U.W. scientists. "Salmon are going to be like Californians, moving up the coast."