View Electronic Edition

Nice Boulders, but Where's the Fish?

This winter, three days after a rainstorm swelled local streams, a neighbor and I donned chest waders and scrambled up Mill Creek in search of spawning salmon. Mill is a small creek, narrow enough in places that you could lie with your head on one bank and your feet on the other. We clambered over mossy boulders beneath a canopy of big-leaf maple, bay, and fir, and covered six-tenths of a mile of sparkling riffles and cascades. It was a beautiful hike, marred only by the fact that we saw not one salmon, nor any evidence of their spawning.

We don't always get skunked on these expeditions. On my last time out, a fisheries biologist and I surprised a female chinook digging a nest in the gravel for her eggs. We paused for lunch on a nearby log, and watched in fascination when she resumed her task as we finished our sandwiches. A lucky crew might see six such nests, called "redds," on a half-day hike.

These spawner surveys have become an annual ritual for two dozen or so residents of the Mattole watershed on California's North Coast, where citizens have worked to revive faltering salmon runs since 1979. A spiderwork of nonprofit groups has arisen to learn more about the fish and their habitat, to communicate to our neighbors what we learn, and to do what we can to help the fish. Some projects have been volunteer efforts, with others funded by foundations, the state or federal government, or commercial salmon-fishing license fees.

My experiences surveying the spawning grounds made me wonder: after two decades of work, why is the sight of a single salmon so noteworthy? We've dragged boulders and logs into creeks to create better habitat for young fish, replaced culverts to make it easier for spawners to get upstream, decommissioned dirt roads that were fouling fishes' homes with sediment, and raised native salmon eggs in homemade redwood troughs.

We've pursued these projects for a variety of reasons. In part, we miss the taste of the fish and their value as a mainstay of local subsistence and economy. Fishing boats clad in peeling paint and "For Sale" signs, moored at half-empty docks, signal the decline of a livelihood that we hope to revive someday. In addition, we are moved by the power of the fish's journey from headwaters to ocean and back again. They've undertaken this odyssey for thousands of years, since their return fed the indigenous peoples of this region. What's more, Pacific salmon play a key role in the ecosystem, bringing a bounty of marine nutrients back to the landscapes that shelter their spawning grounds, and providing sustenance directly and indirectly to more than 100 species of birds, mammals, reptiles, and amphibians.

But results have been slow to materialize. Six salmon generations into our efforts, why don't the fish live up to the old-timers' yarns of streams so crowded with salmon that horses were too spooked to cross? The Mattole Salmon Group's best estimates put the returns of Mattole chinook in the mid-to-high hundreds, after having bottomed out at 100 ten years ago. There's no telling how low the runs to our 200,000-acre watershed might have gotten without our efforts. But we'd hoped that restoration would nudge the system back into balance so it could sustain itself without human management, thus working us out of a job. At this rate, our grandkids will still be setting logs in streams to build more habitat instead of setting nets to catch a share of the multitudes that return each year.

We are not alone in this state of limbo. Dozens of other salmon restoration projects began in the '80s, and by the '90s, they'd become epidemic throughout the Pacific coast. For the Sake of the Salmon, a Portland-based nonprofit, catalogs 410 watershed groups in California, Washington, and Oregon ( ?not counting salmonista brigades that are too small to show up on their radar screens, or that cross watershed lines. Despite all of these efforts, West Coast salmon runs started popping up on the federal threatened and endangered species lists in the early '90s. Twenty-six major runs remain there, and none has reproduced its way off the list.

So I started asking colleagues for examples of successful salmon restoration. The responses were instructive. People would brag about the superlative habitat their projects had created, or the number of miles of road where they'd reduced erosion problems. That's great, I assured them, but I mean the ultimate success: fish. Plenty o' fish. Streams writhing with horny salmon.

They hedged. Mumbled about ocean conditions?a shorthand for the impoverishment of the food chain caused by the 1997?98 El Niņo weather pattern, which shut off the upwelling that normally brings nutrient-rich water to the surface. Alluded darkly to drift netting, the destructive and wasteful practice that had its heyday in the North Pacific in the early '90s, ensnaring millions of immature salmon. Reminded me that numbers are a poor yardstick, when really our success is measured by the transformation of our culture so that we again care for salmon-like relatives.

Yeah, yeah, I replied, but I still want to know if anyone has brought a salmon run back to abundance, to the point where we can rest from all this restoration. For all my searching, I found surprisingly few examples.

Restoration workers may not reap the satisfaction of seeing fish return in large numbers, but we need to tune all of our senses to determine what the salmon need, and act in the belief that our best and humblest efforts will pay off. We need that humility in light of past mistakes, like pulling logs out of streams in the '70s to address imagined barriers to fish migration, only to pull the same logs back into streams in the '90s to create cover and pools. And a little luck will help, too, to shepherd recovering salmon runs past the threats of drought, flood, and fire.

Recovery will come in its own time, if it does at all, at a pace dictated by the magnitude and type of disturbances that caused salmon runs to dive in the first place. Along the way, we undertake certain chores, like creek clean-up, road maintenance, and spawner surveys, as a ritual of paying homage to salmon and inviting them back up our rivers and into our lives. We perform those acts because it is right, like telling the truth and not stealing your neighbor's bike from his garage. We perform them even though the ocean might be cruel to the young fish when they swim out to sea, for we are the custodians of the freshwater habitat and it is up to us to keep it fit for the fish when they return. And while we wait for the salmon to resume their former abundance, we do what we can out of the faith that they eventually will.

The successes I encountered showed me that I needed to shift my focus beyond the near term.

It's a tricky assignment, because we hu-mans are notoriously impervious to long-term thinking. As consumers of the quick fix, we resemble mosquitoes who watch with contempt as gardeners ploddingly plant, water, and weed their vegetable beds. The harvest is a few months and a couple of mosquito generations down the road. "Where are the results?" a mosquito might wonder, poking her proboscis into a gardener's capillary. "They'd be better off to be parasites like us." We hope the practices of restoration and habitat stewardship can elevate us from planetary bloodsuckers to worthy long-term members of the ecosystems.

Butte Creek

One place that has witnessed dramatic salmon recovery is Butte Creek, near Chico, California. Its threatened run of spring-run chinook enters the Sacramento River when snowmelt engorges Sierran streams. The fish then make their way through lower Butte Creek, a maze of rice fields and orchards where nature's plumbing has been redesigned for the convenience of tractors and irrigators. Above that, though, the habitat is in good shape. All summer long, the fish hold in deep natural pools, fasting until temperatures cool with the coming of autumn, when they lay their eggs and complete their lives.

By the early '90s, the run of spring chinook on Butte Creek was on the ropes. Surveys of these twenty- to forty-inch-long fish, made by snorkelers floating down the creek in August, showed that the run had fallen from the high thousands in 1960 to the low- to mid-hundreds.

State Fish and Game biologists realized that the chinook had no way to reach the good habitat higher in the creek. Several small irrigation dams sucked out much of the flow and blocked the fishes' migration to the more pristine parts of the stream.

With farmers' cooperation, fish ladders were improved, five dams were removed entirely, and alternate water sources were found for some irrigators. These changes put 40 cubic feet of water per second back in the creek during the spring chinook's upriver migration. Irrigation ditches were screened to keep juvenile salmon from swimming into rice paddies, thinking that that was the way to the Pacific. Happily, other improvements outside the watershed had al-ready begun: the operation of pumps in the Sacramento Delta downstream had been adjusted to reduce the number of young fish swept toward the pumps that send water to southern California. And a series of wetter years beginning in 1993 enabled the fish to migrate farther upstream and to survive more easily over the summer.

These factors have had an astounding effect. In the last five years, Butte Creek has seen one run of 7,500 spring chinook, and another three years later of 20,000. According to biologists, these salmon return primarily at age three. Researchers could thus compare each year's spawners with their parents' generation, and found that runs were typically increasing by two to three times over the previous cohort. One year saw a ten-fold increase.

The skeptic in me wondered whether some of these fish had simply strayed from the Feather River Hatchery on a nearby tributary of the Sacramento. But DNA testing has shown that the Butte fish are quite distinct from those hatchery stocks.

Butte Creek is an encouraging sign. It meant that if upstream habitat were intact, salmon could rely on it for recovery as long as we allow them to reach it. The efforts needed to replace the water diversions had been modest, and the gains spectacular.

Can Restorationists Restore?

But the problems of most watersheds are less tractable than Butte Creek's. Often?as in the Mattole?the habitat upstream is far from intact. I wondered whether any kindred watershed had managed to jump-start a faltering salmon run so successfully that they'd put aside their restoration work and let the salmon go it on their own.

I found two watersheds that seemed to fit the bill. When locals turned their attention to Freshwater Creek near Eureka in the late 1970s, its coho salmon run averaged 100 to 200 fish per year. The watershed had enjoyed a respite from logging since the early '50s, and was recovering nicely from a huge 1964 flood that triggered slides and washed silt into almost every coastal stream. Humboldt Fish Action Council (a group of commercial fishermen, anglers, and civic activists) started a program to fertilize and incubate roe taken from adults captured from the local run. That way, they could protect the eggs from sediment that was suffocating stream-bottom redds.

Meanwhile, the California Conservation Corps and the fish council cut passageways through impenetrable logjams so coho could reach good habitat upstream, and built structures that created pools in the creek or improved their value to the growing fish. For the first decade, signs were encouraging. Numbers rose through the '80s, peaking at around 800 fish in 1988, and staying at 500 or more through the mid-'90s. In 1994, the fish council determined that the creek was fully seeded with coho and decided to see whether the run would sustain itself.

In theory, the coho should have thrived. The commercial coho fishery had been closed throughout the California coast since 1993, and instream work had increased the amount of available habitat by one-third.

But trouble was brewing uphill. Pacific Lumber Company, under new management since a 1986 junk bond takeover by Maxxam Corporation, began intensive logging on its redwood forests in the Freshwater basin in the late 1980s. More than 85 percent of one tributary watershed was cut in a decade; another feeder drainage was 60-percent clear-cut in four years. As a result, erosion increased, reducing eggs' chances of survival, and pools filled with sediment, thus displacing young coho from their rearing habitat.

Coho returns plummeted, and the 2000?01 run totaled just 160 fish, down to what it was two decades ago. The Freshwater experience reminds us that recovery is fragile, and can easily be reversed if damage to the watershed resumes.

Forty miles inland, the story on Horse Linto Creek started out in a similar vein. Logging had begun on this Trinity River tributary during the post-war building boom, and the creek suffered when the 1964 flood destroyed poorly placed roads. Even though the headwaters of many Horse Linto tributaries are protected in wilderness, the influx of silt from cutover and roaded areas had decimated fall-run chinook populations by the late '70s.

Fish populations needed an immediate boost. So in 1985, the Pacific Coast Feder-ation of Fishermen's Associations launched a project to raise chinook on Horse Linto Creek from native broodstock. They captured Horse Linto fish, waited until they ripened, took their eggs and fertilized them, then raised the young in troughs and ponds. In the spring, the juvenile fish were marked with a fin clip and tagged with a coded wire in their noses. Like branded livestock, they ventured onto the ocean pastures of the North Pacific, to return two to four years later.

The fishermen ran this program until 1994, trapping as many as a quarter of the run as the fish returned each November and December. Although they took just 15 to 25 percent of the adult females, the fish they raised comprised some 45 to 50 percent of the returning adults.

Meanwhile, the Forest Service began to rehabilitate the watershed. They stabilized a landslide that was bleeding silt into the creek, and installed 140 structures to improve spawning beds and create better rearing habitat. They decommissioned roads or reengineered them so they wouldn't impede the flow of water downhill and trigger landslides and gullies. Then, after a decade of raising salmon, the collaborators pulled the plug on the hatchbox program. That act set them apart from permanent hatcheries, whose fish runs are forever dependent on the hatchery, making the watershed obsolete from a salmon's perspective.

Seven years later, the run is going strong. Over the last five years, a stretch of creek that had yielded just a couple dozen redds each year has produced an average of seventy-six?a three-fold increase. But variable conditions make redd counts a shaky way to compare spawning populations from year to year. The more reliable measure of success is the migration of smolts to the ocean, which occurs during the low-water season and thus can be sampled more consistently. Since 1994, Horse Linto's graduating class of chinook smolts has risen steadily from 11,000 to more than 55,000. The run has outgrown its need for human assistance.

Watershed rehabilitation can't take all the credit for these increases?the timing was right, too. The kind of damage that occurred in Horse Linto Creek after the 1964 flood takes de-cades to heal. Sub-sequent logging took place slowly enough that it didn't reopen fresh wounds as soon as the old ones scabbed over. And all public land in the basin has been dedicated to forest protection since President Clinton's 1994 Forest Plan. Good fortune has also helped. Some 40 percent of the 46,000-acre basin burned so hot in 1999 that it left the forest a pincushion of blackened snags, setting the stage for dangerously rapid erosion. But snows blanketed the area and shielded the soil through the dicey first winter, before groundcover reestablished itself.

Measuring the Mattole experience against Horse Linto and Butte creeks, I can see why our efforts have been slow to pay off. Poorly designed logging and road building continued to destabilize Mattole hillsides well into the 1980s. When a stream is reshaped by excessive erosion, it can take decades or even centuries to recover, and no amount of artificial habitat can make up for that on a large scale. When riparian forests are cut or swept away, they can begin a lasting recovery only after the banks have stabilized, and then they will take decades longer to reach maturity. We can help initiate these processes, but they will proceed with their own rhythm.

Dead Salmon = Fresh Salmon

We need to be particularly patient because recent research shows that it takes dead salmon to make fresh salmon. Through stable-isotope studies of young coho that have never seen the ocean, we know that 25 to 40 percent of the nitrogen and carbon in their bodies comes from marine sources. Those oceanic building blocks arrived with previous generations of adult salmon, which swam upstream, spawned, died, and were eaten by scavenging invertebrates which then became part of the salmon's food chain. Young fish also feed directly on stray unburied eggs and carcasses. As a result, there's an inherent inertia to low salmon populations, because a meager spawning year provides enough carcasses to feed just so many young fish. To overcome that inertia, some salmon supporters scatter dead hatchery salmon along streams to fertilize them. In 1999, the Washington Dept. of Fish and Wildlife planted 120,000 salmon carcasses in twenty-three watersheds for this.

Biologists on the Keogh River on northern Vancouver Island have scored remarkable results with a related treatment: they seeded the twenty-mile-long stream with several hundred pounds of time-release nitrogen fertilizer pellets. Algae bloomed, insect populations skyrocketed, and juvenile fish are surviving at much higher rates than before. The British Columbian researchers suggest this technique could tide fish over until the return of adult spawners provides enough biomass to fertilize the ecosystem.

The carcass connection illustrates a complex knot of relationships that we are only beginning to comprehend. Another dimension includes beavers' importance to salmon populations: beaver ponds retain organic matter that enriches the food chain, and provide rearing pools for juvenile fish. Salmonologists speculate that intensive beaver trapping in the eighteenth- and nineteenth-century Northwest had already degraded salmon habitat by the time the first canneries opened in the 1860s, and that earlier runs may have exceeded even the most bountiful ones in the historical record.

The lessons of beaver resurfaced on Knowles Creek, a tributary of Oregon's Siuslaw River. In 1982, a debris torrent poured off a clear-cut, carrying huge trees from the downstream, uncut forest. When this stew of logs and mud reached the creek, it dammed the stream and created such a mess that the landowner was fined $5,000. The incident became a poster child of bad logging. But biologist Charley Dewberry found the next summer that young coho were most plentiful in the pool behind the logjam. The debris flow had created a mock beaver pond.

As we try to coax salmon runs back to abundance, our greatest teachers and allies are the salmon themselves. From time immemorial, they thrived in a landscape marked as much by disturbance as stability. Unlike the stereotype of salmon returning unerringly to their natal streams, salmon are innately resilient and opportunistic. After Mt. St. Helens erupted in 1980 and filled the Cowlitz River with ash, salmon swimming up the Columbia to the mouth of the Cowlitz smelled trouble and swam right on by, to other, undamaged tributaries. When glaciers receded from Puget Sound thousands of years ago, salmon strayed into previously ice-locked fjords and colonized newly exposed rivers. Indeed, those salmon runs helped to rebuild the soils of barren glacier-scraped landscapes with the nutrient bounty of their carcasses.

These adaptations require healthy habitat to help the fish withstand the disturbance du jour. Without big trees downstream, the Knowles Creek debris torrent would have been an unmitigated disaster. So it is up to us to protect areas that are still intact and to set the stage for watershed healing. That means addressing the sources of human-caused erosion ? such as failing roads and culverts ? knowing that we may not live to see the results translate into abundant salmon runs. And it calls for dogged vigilance to make sure those intact areas, which Dewberry calls anchor habitats, are not trashed. It calls for us to become attentive to what salmon are telling us ? by the places they congregate and spawn successfully, the places they avoid, the seasons when they thrive and those when they decline.

For restoration to succeed, we need to recreate the context for salmon to inhabit, to have the humility to heed what they are showing us, and to find the patience for the results to manifest. Then the salmon can work their magic?on their own populations, on the landscape, and on us.