But no factor is more critical to salmon prosperity than ocean conditions, experts say, and the complex interaction between biologically distinct groups of salmon and changing ocean habitats has created challenges for resource managers.

At the same time a projected huge run of spring chinook salmon are entering the Columbia River, fishing on one of its major tributaries -- the Willamette River -- has been closed because of a low estimate of returning fish. And offshore salmon seasons are either shut down or sharply curtailed along the entire West Coast this spring and summer because of a projected historic low return of fish to the Sacramento River basin.

The common denominator in the good and bad runs is the ocean.

Bill Peterson, a fisheries biologist with NOAA who is based at Oregon State University's Hatfield Marine Science Center, says this year's salmon debacle can be traced back to unusual ocean conditions in 2005. A delay in the ocean upwelling caused ocean conditions "to collapse."

"The delayed upwelling off the Oregon coast meant that in the critical time when juvenile salmon were entering the ocean, there was nothing for them to eat – and most of them died," said Peterson, who is a courtesy professor in OSU's College of Oceanic and Atmospheric Sciences. "But you don't see the impact until two or three years later, when the fish should first begin returning as adults."

Wind-driven upwelling brings nutrients from deeper water to the surface and fuels phytoplankton blooms. Lipid-rich copepods and other zooplankton feed on the tiny plants, and in turn are consumed by anchovies, sardines, herring and other small fish that are staples in the diet of salmon and other fish. The delay in upwelling was caused by late arrival of seasonal winds, according to researchers at OSU, who published their findings in the Proceedings of the National Academy.

The delayed upwelling can explain why most fish runs are plummeting, yet fisheries managers are predicting a huge number of spring chinook bound for the Columbia River this year. Why? The answer, Peterson says, can be found by tracing where juveniles from different river systems go once they enter the ocean.

For the past 10 years, Peterson has participated in a research project funded by the Bonneville Power Administration that analyzes the distribution of juvenile salmon off the West Coast and uses genetic tracking to determine their river origin. Juvenile fish from many of Oregon's coastal rivers, along with those from the Willamette River and the Sacramento River, congregate just off the Oregon coast once they leave their river systems.

When the ocean collapse came in 2005, most of those fish starved.

"But Columbia River spring chinook don't stay off the Oregon coast," Peterson said. "In our 10 years of sampling, we've only caught a few Columbia River juveniles just off our coast, so it's obvious they go somewhere else. If you look this year at chinook salmon in Alaska, they're doing well. So it's possible that Columbia River juveniles head to the same place as Alaska juveniles."

Peterson speculates that perhaps young Columbia River salmon may migrate toward a unique ecosystem several hundred miles off the Northwest coast. In that deep, cold water, lipid-rich fishes known as myctophids, or "lantern-fish," provide a bountiful diet for a variety of marine life. These fishes are "very abundant" in the mesopelagic zone, he added, and could provide a rich forage base for young chinook salmon.

"It's just a theory at this point," he said. "We need to go out there and sample for juvenile salmon. But the situation this year underscores how fascinating research on salmon can be. We used to have a lot more genetic diversity in our salmon runs. They used to spawn at different times and hang out offshore at different times. We may be paying for the loss of that diversity."

Ocean conditions off Oregon in 2006 and 2007 were somewhat better for salmon survival, but still were less than ideal. The good news, Peterson says, is that the influence of La Niña over the winter has created what appear to be excellent ocean conditions thus far in 2008. But, he added, it's premature to celebrate.

"The system can't recover from a near-complete collapse in one year," Peterson warned. "There may not be enough adults in the streams to repopulate the runs. We need three or four years of good conditions before we can breathe a little easier."