Sparsely furnished with only a desk, chair and computer, Dr. James Anderson's office in the Fisheries Science Department at the University of Washington reflects the fastidious mind of a seasoned scientist. Graying and soft-spoken, eager to discuss the intricacies of his work, Anderson seems at home in this building full of ambitious and hyper-intelligent biology nerds. He specializes in a branch of science known as fish-passage modeling -- complex mathematical calculations that estimate the effect of a range of possible river and ocean conditions on the health of migrating salmon. For three decades now, Anderson has focused on salmon passage through the federal dams on the Columbia River and its largest tributary, the Snake.

Less than a century ago, some 15 million salmon returned to their home waters in the Columbia Basin every year, some to creeks over a thousand miles from the ocean and more than a mile above sea level. Now, they're extinct in a third of their former range and severely impaired where they remain, with 13 of 16 stocks listed as threatened or endangered. Though mining, logging, farming and urban development have taken their toll, many fisheries scientists say the main culprit is the 11 dams blocking the mainstem Columbia and Snake. 

Anderson's models, however, have consistently shown something different: that dams are not nearly as hard on salmon as their detractors have made them out to be. And he says that helping salmon survive isn't as simple as diverting water from electricity-producing turbines and returning it to the fish.  

  Anderson's work has had a lot of influence on the great salmon debate in the Columbia Basin. Rich Zabel, who earned his Ph.D. under Anderson, took a version of his mentor's model to his job at the National Atmospheric and Oceanic Administration, the agency responsible for the recovery of threatened and endangered salmon. Over the years, NOAA has refined the model to account for a flood of new data, but its prognostications haven't deviated too far from Anderson's, more than a decade ago. And last month, when federal Judge James Redden convened his court to hear about NOAA's latest salmon recovery plan, the model was part of the science behind a familiar federal argument: Salmon actually don't need much more water than federal hydropower producers are willing to give them.

That's a controversial idea, to say the least. But Anderson says conclusions to the contrary deserve scrutiny, too. He believes that, on occasion, state and tribal fisheries managers and salmon advocates have bent data to fit their fish-friendly conclusions. "Once, early on, I pointed out the problems with a model salmon advocates were using," he recalls. "It showed salmon survival increasing to beyond 100 percent, commensurate with an increase in water flow. Well, that's just not possible, and I said so. At lunch the people from their side wouldn't even talk to me."

Not that Anderson minds a scientific quarrel. "If I find an argument that's horrid, I'll tell people that I think it's bunk. And I got really irritated along the way when I thought people were presenting opinions and hypotheses that were drawn by what they wanted the final result to be."

Anderson may believe that a kind of sentimentality about salmon has compromised science along the Columbia, but his critics hint that his own work has been tainted by money. The Bonneville Power Administration, the New Deal-spawned federal agency that markets and sells electricity produced by the Columbia Basin's dams, funds both Anderson's work as a university professor and as a private consultant. Armed with the results of his own modeling work, Anderson has testified in Congress on behalf of the BPA as well as for corporations like aluminum giant Alcoa, which enjoys heavily subsidized power contracts with the BPA. His testimony has reliably bolstered the BPA's claim that it's already doing everything it reasonably can for salmon. He's said that increased flows won't help salmon negotiate the dams, spoken out against a plan to breach four dams on the Lower Snake River, and approved of new irrigation schemes for agribusiness.  

Anderson's industry-friendly conclusions may seem dubious in light of who's paying for them. But funding arrangements like this aren't all that unusual in the realm of Columbia Basin salmon research. Several laws and treaties oblige the BPA to foot the bill for salmon recovery and science, and it also funds projects at its own discretion.

By its own reckoning, the BPA has spent somewhere between $8 and $9 billion over the past 30 years to bring salmon back to a semblance of their former abundance. The agency finances the vast majority of salmon research in the Columbia River Basin. It pays for graduate salmon study at major regional universities. It commissions research from private consultants. It pays tribal and state agencies to conduct their own biological inquiries. It gives generously, in some years, to the U.S. Fish and Wildlife Service. It has underwritten the Portland office of NOAA-Fisheries, the very agency charged with determining whether or not the BPA needs to make dramatic changes to the hydropower system to ensure salmon survival.

The investment, however, has not paid off. Most listed salmon stocks have continued to decline, and none are headed toward recovery. And the program's ineffectiveness, some scientists say, is due in part to the BPA's influence over how that money is spent.