The scientists here will never complete their research, because there is always more to learn about how this ecosystem works. Hauser, for instance, through surveying two creeks and the lakeshore spawning site between them, day after day for 10 years, has discovered that a single male sockeye that spawned with two females in 2004 had about 34 offspring that returned in 2008-'09 – about 10 percent of the run. That's amazing, he said, because more than half the spawning sockeye have no offspring that make it back. He doesn't know why that particular male's spawning was so successful, but thinks more research might provide an answer.

Hauser also hopes to answer a question about the salmon's anti-bear strategies: "No one really knows why" the salmon generally gather at the mouths of creeks and then rush up all at once to their spawning sites, he said, but it's probably because, by running with a crowd, each one has a slightly better chance of surviving the gantlet of bears long enough to spawn. Some of the salmon even go back and forth, entering and leaving the creeks repeatedly at certain times of day, avoiding the bears' most active feeding times around sunrise and sunset. Hauser wonders whether this is "risk-sensitive breeding" – are these particular fish just better at dodging bears and thus more successful in having offspring?

Connectivity is one of the main questions. If salmon are wiped out in one creek, how long does it take for other salmon to re-colonize it? And where do they come from? The answer matters to more than salmon; already, climate change is causing many species to re-locate in response to the upheaval in conditions.

"We have a really lame ability to predict the future," especially how ecosystems will evolve with climate change, Schindler said. "But what we can do is spread the risk around," by maintaining the complexity and diversity as "an investment strategy." There's already evidence that colonizing and re-colonizing can happen very quickly if the ecosystem is still complex and diverse. "Let's keep our options open for animals and species we value."

In his "continuous game" of attempting to keep all this research going, Schindler sat at the kitchen table in the main cabin one morning, sipping coffee made with lake water, surrounded by shelves of high-energy food brought in by boat – M&Ms, peanut butter, blocks of cheese – and fired up his computer. He was writing the annual progress report to the National Science Foundation, which provides roughly 50 percent of the program's current budget. The team needs at least a half-million dollars per year just for the core activities, not including their university salaries. "We lose sleep over funding this program," Schindler said.

Salmon canneries initially funded the research, hoping to determine how to maintain healthy runs, but since cheaper farmed salmon has flooded the market, their business has declined; they now provide just 10 to 15 percent of the program's budget. The rest is raised in bits and pieces from many sources, including the U.S. Fish and Wildlife Service and commercial fishermen who want to help ensure stable runs. The largest support in recent years has come from the Gordon and Betty Moore Foundation, based in Palo Alto, Calif. – Gordon Moore, the founder of Intel, is an avid fisherman – but the foundation is shifting away from supporting basic research.

"Our program's strength is long-term data, and that's the hardest thing to get funding for," Schindler said. The University of Washington itself doesn't provide much funding for this program; even though universities build their reputations on research, they usually just house it and skim off a percentage of the grants for overhead. "The longevity of the program depends on the energy and passions of a couple of faculty," Schindler said, citing Tom Quinn and Ray Hilborn, another veteran in UW's School of Aquatic and Fishery Sciences. "It's not the institutions."

Schindler himself puts in the most time at the camp, staying into autumn after the others leave, sometimes walking the creeks alone with a shotgun because the bears get frantic before hibernation. He and Armstrong even came here last winter, driving snowmobiles across the frozen lake to sample the chemistry of snow that would end up in streams in the summer. Tim Cline, one of the grad students, told me that Schindler can count salmon simply by walking up creeks to see what's there; at a glance, he can estimate hundreds accurately and even do the gender breakdown.

In many respects, Schindler is carrying on his father's legacy. David Schindler was in constant hot water in Canada due to his outspoken advocacy for the science on acid rain and phosphate detergents. More recently, he's been raising hell about the oil sands mining in Alberta. Says Daniel, "He's influential because he's not scared of anyone."

About 10 years ago, on the bank of Washington's Snohomish River, surrounded by roads and farms and cities, I met a guy who put on a snorkeling mask and swam with the remnants of the salmon run there – just for the love of it. I put it on my bucket list: Someday, somewhere, before I die, swim with salmon.

One afternoon here, on Sam Creek, I got my chance. First, Armstrong donned his snorkeling gear and a dry suit and slipped into the cold, clear water to take close-up photos of the vivid red sockeye in a pool where the creek meets Lake Nerka. Hundreds of salmon hovered side-by-side, facing upstream, getting ready to make their run, and they tolerated Armstrong slowly easing into their formation.

Then we walked up the creek through a litter of salmon carcasses, past fresh bear and gull prints in the sand, finding smaller pools where more sockeye hovered. I wrestled into the rubber suit, adjusted the facemask, and lay down in a pool. Almost instantly, it seemed, I was surrounded by the big reds.

As they undulated to keep themselves in formation, some brushed against me with their bodies, and some swiped their tails against exposed portions of my face. A few even wriggled under me, one by one forcing their way between my chest and the sandy bottom.

Shifting slightly upstream, I turned to look at them head-on. Dozens faced me, just a few inches away, their jagged teeth exposed through gaping, elongated jaws – another physical change that occurs as they spawn, as if they're reverting to a completely primitive form that matches the landscape. Their gills flexed water in and out, extracting oxygen, and their eyes, eerie golden circles, gazed at me implacably, as if nothing else mattered except their instinct to spawn. I reached out, touched one, and then another, and another.

---

Ray Ring, writer of this story, is an HCN senior editor based in Bozeman, Montana.

Jonny Armstrong, the photographer, was born and raised in Ashland, Oregon, and just finished his University of Washington doctorate, working with Daniel Schindler. He's now a 2013 Smith Postdoctoral Research Fellow based at the University of Wyoming, working to evaluate the effects of resource development on wide-ranging consumers in Alaska watersheds.

This coverage is supported by contributors to the High Country News Enterprise Journalism Fund.