For nearly as long as canneries have been processing salmon, researchers have been measuring, aging, and otherwise studying the complex journey of salmon from egg to death.

What they've learned is nothing short of astounding, but what remains unknown is even more so. Portions of a salmon's journey, like ocean migration and navigation, marine ecosystem dynamics, and the complete failure of returning fish in some years, are scarcely understood at all.

Even more disconcerting is the potential effect of global warming on wild salmon populations worldwide. Scientists know that a natural 25-year cycle of periodic ocean cooling—Pacific Decadal Oscillation—and heating greatly affects marine productivity. As the ocean warms, productivity increases, meaning more food for salmon. During a cooling cycle, less food is available, resulting in a decrease in salmon production. Harvest records throughout the Pacific Rim over the last century support this fact. Too-warm water, however, appears to inhibit normal nutrient circulation in the ocean, which each year brings nitrogen, along with other essential plant nutrients, to the surface. Throw into this mix a rising earth's temperature due to the effects of human industrialization, and no one knows what will happen if the North Pacific no longer cycles between warm and cool periods. Could the Pacific eventually become too warm, shutting off forever this vast, ancient conveyor belt of immense productivity? That's a possibility no one throughout the Pacific Rim dares to contemplate.

Salmon scientists face other challenges, one of them at the microscopic level. Genetic diversity, that great potpourri of variability that allows a species to quickly adapt to changing environmental conditions, may be at risk. Each year hatcheries located in Kodiak, Cook Inlet, Prince William Sound, and throughout southeast Alaska release about 1.6 billion salmon into the ocean. Sometimes hatchery fish spawn with wild fish, diluting the millennia-old genetic code of wild salmon. In addition to genetic mixing, no one knows what effect hatchery salmon may have on the oceanic food sources of wild salmon. Can the ocean support so many salmon, particularly in poor production years?

Salmon biological diversity also may be threatened by fishery managers themselves. Because Bristol Bay sockeye arrive all at once within the span of a few weeks, harvesting them while still allowing a sufficient number of spawners to escape is a terrific challenge for biologists, particularly where sockeyes from different watersheds mix together before ascending their individual natal streams. To control these mixed-population fisheries, biologists use a variety of tools—from emergency fishing closures to aerial, tower, and sonar counting of fish in rivers—to determine the greatest number of harvestable salmon in each watershed.

At the same time, biologists must allow adequate salmon escapement. But what is adequate? For decades now, biologists, concentrating their efforts almost entirely on producing the maximum sustainable harvests, have assumed that numbers are all that count. If enough salmon throughout the run escape fishermen’s nets and ascend the rivers, the reasoning goes, then spawners will fill all the tributaries and habitats capable of supporting salmon, thus maintaining their biodiversity. But is this true? Managers point to the record runs of past years as proof that their method works, but harvest numbers alone are hardly indicative of overall diversity in the bay’s hundreds, perhaps thousands, of individual tributaries. Who can say which creek, stream, or rivulet contains salmon with the specific genetic code that stands the best chance in a warming ocean or unusually cold winter, let alone whether those particular fish are spawning? No one knows because no one is watching individual spawning streams each year to make sure an adequate number of salmon return to each and every one.

Aside from biodiversity, are enough salmon escaping to perpetuate the runs? Some researchers are coming to believe that large, so-called excess, escapements are more important than previously assumed, because they build resiliency in salmon ecosystems by increasing their capacity for future productivity. Smaller, weaker populations, like those of the Nushagak River and Lake Clark, prosper in excess escapement years. At the same time, abundant fish carcasses provide essential nutrients for successive generations, as well as the myriad forms of aquatic and terrestrial life that both sustain and depend upon salmon.

These questions have prompted some scientists to suggest that instead of managing the fishery for maximum sustained yield, it may be time to replace that goal with a new one: minimum sustainable escapement.