But a University of Washington researcher writes in this month's BioScience journal that the silt, minerals and nutrients that are unleashed have ecosystemwide effects, causing changes in rivers and lakes far from the nests.

From decreasing the amount of algae there is to eat to possibly influencing when aquatic insects emerge, spawning salmon can be extraordinary "environmental engineers," says Jonathan Moore, a UW graduate student in aquatic and fishery sciences.

Ignoring this role can cause missteps in managing salmon runs or attempting to rehabilitate habitat, he says. A major loss in the number of salmon, for example, doesn't just affect future generations of that fish alone.

"In streams with high densities of salmon, the disturbance from spawning impacts virtually all aspects of stream ecology," he says.

The female salmon, of course, isn't concerned about all that. She simply wants to lay her eggs in a nice, gravel-bottom bowl that's free of fine sediments that can smother them.

But consider the efforts of the grand dames of the salmon world, the female chinook � or king � salmon. The largest females are more than a yard long and tip the scales at 45 pounds or more. The biggest nests are nearly a foot and a half deep and extend up to 17 square yards �about the size of two parking-lot stalls. The rims around these craters can be the bane of boaters who, even with boats meant to navigate in shallow waters, can find their vessels grounded.

Smaller species of salmon that spawn at higher densities are capable of even more widespread tilling, according to Moore, who has for six summers worked with sockeye salmon through the UW's Alaska Salmon Program. Using counts of spawning sockeye for the last 50 years and previously measured nest sizes, Moore calculates that every summer the sockeye disturb at least 30 percent of the stream beds of two Alaskan streams he studied.

And in years when salmon populations are high, sockeye dig up entire stream beds more than once, being forced to superimpose new nests on top of old nests when the females run out of room.

Moore's work and the UW's Alaska Salmon Program is funded by the National Science Foundation, the Gordon and Betty Moore Foundation and Alaska salmon processors.

The rototilling effect probably happens wherever salmon are found in high densities, such as British Columbia, Alaska and even some individual streams in the Pacific Northwest, Moore says. Kennedy Creek, for example, is a small stream that flows into south Puget Sound and since 1968 has had chum salmon in high-enough densities that they have caused the amount of algae and stream insects to decline.

Scientists have long known of the habitat-changing activities such as dam building by beavers, but much of ecology has assumed that most other organisms simply react to the physical and chemical conditions around them. A handful of papers since 1999, including one by Moore in 2004, has focused on the disturbance created by spawning salmon and the ecosystem at large. In this most recent paper, Moore also outlines a conceptual framework he's developed that gives ecologists a way to formulate when the abundance of animals and their activities � be they spawning salmon or some other kind of animal � make them impossible to ignore as ecosystem engineers.

"One specific application to stream restoration in the Pacific Northwest, for example, deals with the practice of fertilizing salmon streams with the carcasses of dead salmon obtained from hatcheries," says Daniel Schindler, UW associate professor of aquatic and fishery sciences. "Although this does replace some of the nutrients that salmon returning from the sea would normally provide, it entirely ignores the fact that live salmon play a diverse suite of roles in streams, including helping to disperse those nutrients."

Related Links
University of Washington