Saving Tarboo Creek

by Scott Freeman

  〜

  The little salmon stream that my family is helping to restore is located on Washington’s Olympic Peninsula and is called Tarboo Creek. The name is pronounced like two beats on a Suquamish drum, with slightly more emphasis on the first—a gentle tar-boo. When people say tar-boo, they scare the fish.

  Tarboo Creek’s waters flow from north to south and empty into Tarboo Bay—a tiny fiefdom in the Pacific Ocean’s vast empire. Tarboo Bay runs into two long sand spits, then broadens into Dabob Bay, which opens into Hood Canal, a fjord that forms the southwest corner of Puget Sound.

  To the south, Hood Canal ends near the steps leading to the state capital building in Olympia, Washington. The canal and the rest of the sound’s western shore are sparsely settled, but its east coast is home to more than 2.5 million people. There, bays that are large enough to support ports attracted the European settlers who founded the cities of Bellingham, Everett, Seattle, and Tacoma.

  To the north and west, the sound connects to the Strait of Juan de Fuca—a conduit to the open Pacific that is an international boundary, separating Washington’s Olympic Peninsula and British Columbia’s Vancouver Island. If you put a kayak in the Strait of Juan de Fuca and paddled straight west, you would eventually run aground on Sakhalin Island, off Siberia’s southeast coast.

  Tarboo Bay is one of the highest quality estuaries left in Puget Sound. Unlike most of the region’s coast, the bay’s shoreline has a minimum of homes, breakwaters, docks, and jetties; the water quality is good enough to support natural spawning by Pacific oysters. The bay drains at low tide, creating an expanse of mudflat dotted with glaucous-winged gulls and Northwest crows. Tarboo Creek flows out toward the spits in a winding channel through the sand and muck, chasing the retreating salt water. But at high tide, the sea fills the bay to brimming. Ducks bob; salmon swim; seals hunt.

  In November, when the salmon follow a high tide into a creek channel deepened by the first rains, they enter one of the last tiny patches of low-elevation old-growth forest left in Puget Sound. I’ve seen aquatic biologists fall into a reverent silence there, marveling at the cool, even flows of clear water rippling over gravel beds and wending around sandbars, before stooping like eight-year-olds to turn over rocks and begin cataloging the scores of midge, caddis fly, and mayfly larvae that cling to the cobbles. The gradient is low and the channel wide, lessening the chances that storm-driven floods will scour the bottom and rip out salmon nests. There are jumbles of mossy downed logs and jangles of root wads, creating fish-filled pools and eddies. Bracket fungi peek out from the wet snags; they are releasing nutrients into the creek, at a slow drip, as they rot the downed trunks.

  Sword ferns cover the sandy deposits that line each bank. Above them, sprawling vine maples form a shrub layer; higher up, a smattering of small red alders holds sway. Over everything, craggy old bigleaf maple trees and towering Sitka spruce form a canopy. The maples are covered with moss and dripping with lichens; the spruce are massive—6 feet or more across at eye level, with trunks that rise five stories high at the first limb and then disappear in a tangle of needles and branches, perhaps topping out at 175 feet or more. The biggest Sitka of all has a root 3 feet in diameter that buttresses away from the base and bridges a tributary to the creek. You can sit on that root on a wet November day and watch chum salmon swim under your feet—headed up the tributary to spawn. We can only guess, but the tree may be more than 750 years old. It is a great-great grandmother, many times over.

  Walking the banks of the creek, you can find cougar, bobcat, or bear tracks, or at least the spindly impressions from a great blue heron. As you walk away from the stream into the uplands above the floodplain, you begin wending your way around pillars of reddish-brown western redcedar trunks. There a lacy canopy of drooping boughs casts enough shade to limit the undergrowth to a few scattered ferns.

  At first the woods look like they could go on and on, but they are actually just a 160-acre preserve. To the north, things get difficult from a salmon’s point of view. For much of its 7.5-mile length, Tarboo Creek was ditched or diverted and the surrounding wetlands drained. A series of badly installed culverts blocked fish passage to large portions of the main channel and tributaries. If the culverts that carry a stream under a road or driveway are too small or placed at the wrong grade, two things happen. After storms, large flows give them the look and feel of a fire hose, with so much water traveling so fast that salmon can’t fight their way up. At the downstream end, the torrent scours the surface below the lip of the pipe, eventually creating a waterfall and perching the culvert 3 feet or more above the rest of the stream. This is too high for fish to jump, even the athletic coho salmon. A poorly designed culvert can block migrating fish as effectively as a net.

  〜

  An old gentleman who grew up in a house near the creek said that as a boy he used to lie awake in his second floor bedroom, unable to sleep for the splashing of salmon spawning in the distance below. By the late 1990s some chum and coho were still nesting in Tarboo Creek, along with a smattering of sea-run cutthroat trout, but their numbers were a fraction of historical norms. And steelhead—rainbow trout that breed in freshwater but spend most of their life in the ocean—had been extirpated decades earlier. More than seventy years had passed since the stream had seen salmon runs like the old man described.

  In 2004 our family bought an 18-acre parcel that straddles the main channel of Tarboo Creek. When we did, we joined a community of individuals and organizations working to reforest abandoned pastures and degraded wetlands in the valley in hopes of restoring Tarboo Creek as a high-functioning salmon stream. Most of the individuals involved are private landowners in the watershed; the organizations include county, state, and federal government agencies, local Indian tribes, and private nonprofits like the Jefferson Land Trust. The entire project is coordinated by salmon biologist Peter Bahls, who directs the Northwest Watershed Institute (nwi).

  By the time we got involved, the project was already gaining momentum. The 160-acre tract that includes that patch of old growth near Tarboo Bay had recently been purchased by the State of Washington’s Department of Fish and Wildlife. The acquisition went through just before the tract was scheduled for clear-cutting by a private forest products company. The biggest Sitka spruces would’ve been spared—forest use regulations require loggers to maintain a 50-foot setback near fish-bearing streams—but the cedars on the bluff would’ve been felled for utility poles or decking and replaced with a plantation of fast-growing Douglas-fir cultivars. The cut was delayed at the last minute when Peter Bahls discovered an osprey nest and notified the Department of Natural Resources, which oversees logging. It’s illegal to cut forest near an active nest of a protected bird-of-prey species; the osprey discovery bought enough time for the Department of Fish and Wildlife to budget the money for the purchase.

  By 2004, Jim Yeakel and Joan Purdy, who run a farm and a cabinet-making business not far from the mouth of the creek, had worked with nwi to regrade and replant a rivulet that connected the main stem of the creek to a pond on their land—ideal habitat for juvenile salmon to grow in before returning to the ocean. Near the headwaters, Alan Iglitzin—a violist who directs chamber music concerts each summer at the Olympic Music Festival—had helped nwi remove a culvert that was blocking salmon passage and replace it with a wooden bridge. In between, the county highway department had replaced two other old culverts that were preventing salmon from moving upstream.

  So when we arrived, the mouth of the creek was protected and the worst of the blockades had been removed. The creek was reopened for business. Salmon could move into habitat that had been closed for more than fifty years.

  Our land is about halfway between the headwaters and the mouth of the stream, and was the next project on the list. The watercourse there looked like an open wound—it was easily the most badly degraded stretch of stream in the watershed. The creek had been channelized through our property in the early 1970s. Thirty-five years later
, the water was shooting through a steep-sided, arrow-straight ditch, excavating the bottom as it went. In the worst sections we could stand on the gravel streambed and not see out. The creek was 6 feet under. It was a sluiceway conveying sediment that would slowly choke the life out of Tarboo Bay.

  In what was left of the creek’s floodplain at our place, thistles and horsetail were chin high. Himalayan and Eurasian blackberry formed an archway over most of the creek channel, interwoven with tangled mats of reed canary grass. All three plants are nonnative and invasive, and are listed as noxious weeds in western Washington. The blackberry canes, in particular, are armed and dangerous. They brandish centimeter-thick stems studded with spikelike thorns and can grow 10 yards or more a year.

  The previous landowners had also cut all of the 3-to-4-foot-diameter western redcedars that were growing on the hillside above the ditched creek. The cut left a smattering of spindly young cedars—whips in logger parlance—along with a stand of bigleaf maples. Foresters call this approach high-grading, or thinning from above. It’s a way to cash out quickly. If you thin from below, you remove smaller trees that are competing for light and water so that the bigger trees can get even larger over time, and more valuable. Thinning from below is a way to manage for longer-term, sustained yields.

  Fortunately for us, the owners had left the large bigleaf maples—apparently not realizing that they can contain patches of compression wood, which grows in response to the weight of leaning stems or branches. In maples, compression wood can be figured with dazzling swirls and streaks that are prized by luthiers for making violin and guitar backs. People in the business claim that a large maple with extensive figuring can be worth $5,000 or more on the stump. It’s common for maple poachers to patrol back roads on the peninsula, looking for big specimens. If the coast is clear they’ll drop the tree, find sections with figured wood, and cut them out, like ivory.

  The 18-acre parcel we bought was actually part of an 83-acre farm the previous owners had purchased. After logging it—clear-cutting 40 acres and high-grading the larger redcedar off our place and the big Douglas-fir off the farm next door—they subdivided it into three parcels and put them up for sale, all within a year of their original purchase. In this, they were following a long tradition. When Lewis and Clark returned to St. Louis from the Pacific coast, they turned in their notes describing the natural wonders the Corps of Discovery had seen—an Eden of rolling prairies brimming with wildlife, pristine rivers and lakes, and cathedral-like forests that stretched for hundreds of miles. Then Meriwether Lewis began speculating in land and the fur trade. He wanted to cash in.

  At our place, the previous owners had even dug two ponds and bulldozed a house site and circle drive, thinking they would sell to wealthy retirees. Instead the place sat, attracting almost no interest. It had been for sale for four years when we came along.

  〜

  An ecologist I know advises new landowners to wait five years before starting restoration projects. He wants them to walk their land—to watch it through the seasons and over the years. The idea is to learn a place before starting to change things. This is sound advice, but we couldn’t take it. The Northwest Watershed Institute had a grant to dig a new, meandering channel for the stream, and all the required permits were in hand. The idea was to make crooked that which was straight.

  Stream restoration permits usually start August 1 and expire at the end of September. The interval is timed to coincide with the late-summer drought that is typical of the region, when heavy rains are unlikely to erode newly dug channels. The regulatory agencies also want you to wait until the salmon fry of the year are big enough to handle the disturbance. Then the challenge is to get the work done and the machines out of the water well before the rains begin and the adults return to breed.

  In a floodplain, the purpose of a ditch is to collect surface water, drop it down, and get it to run off as fast as possible. A well-engineered, well-maintained ditch will dry the soils enough to make pasturing or plowing possible. But in the Tarboo watershed, the channelization and pasturing and plowing weren’t successful. The last dairy farms in the valley closed down in the 1960s—victims of poor agricultural soils, steep slopes, and competition from large factory farms. The ditching at our place was a last-gasp effort that didn’t change the outcome. A century of trying to grow livestock, instead of big trees and big fish, had ended in failure.

  The purpose of a stream restoration is to undo everything that channelization does: to bring the water back up to the surface, slow it down, and wet the soils. Once the stream is back at ground level and meandering its way seaward, the groundwater level will rise and soils will resaturate, supporting the growth of wetland plants. A ditch is straight, narrow, and predictable; a stream is curvaceous, broad, and fickle. Other than deepening, a straightened channel may not change appreciably in half a century; a free-flowing river changes almost constantly.

  To introduce complexity in a restored salmon stream, you have to create quiet, backwatered pools where the current is slow; rapid, noisy riffles where adult females like to lay their eggs; and everything in between. The pools are where the salmon fry will congregate, burning up a minimum of energy to maintain their place in the current and grazing in peace on organic debris and invertebrates, with hidey holes nearby in case a kingfisher or great blue heron visits. Riffles are preferred spawning sites because the current cascades over gravel, creating white water that oxygenates the eggs stashed below.

  Restorationists also want to install just as many logs as possible. The best salmon streams in the Pacific Northwest run through dense stands of towering old trees and are pierced and prodded and poked with windfall timber. The moss-draped logs create pools where both young and old fish can hide from predators; the decaying wood drizzles nutrients used by aquatic plants and insects, possibly ending up in young salmon.

  Peter Bahls and civil engineer Tom Smayda studied intact stretches of Tarboo Creek to figure out the size of the stream’s normal meanders, and they used recent weather and water flow records to model storm-driven surges. The research let them work out how deep and wide to make the channel—what its new profile would look like. They were designing for high water and for flooding, calculating how often the water would overtop the banks and where it would go when it did.

  The result was sheaves of blueprints. The ditch through our place ran about 1,000 feet; the remeandered stream would run 1,500. And it would have features: new meanders, logs, root wads, pools, and riffles and mini-rapids lined with imported gravel. It was all laid out, in blue and white.

  Susan’s uncle Luna Leopold had spent a career studying how streams work and was widely considered to be a world authority on river hydrology. He critiqued the initial drawings on what turned out to be his deathbed. When he and the state permitting agencies had signed off on Tom’s revisions, we had a project.

  〜

  The remeandering crew was made up of our family—including two sons, a sister, her boyfriend, a niece, Susan’s father, and her aunt Estella Leopold—along with Peter Bahls and Sean Gallagher from nwi. But the main man was Bob Harrison.

  For nine months of the year, Bob makes his living digging basements for suburban tract homes. But for three months of the year, he remeanders salmon streams. He smokes two packs of cigarettes a day, has a belly like a Buddha, and is the go-to guy for salmon stream restoration in northern Puget Sound. He’s a fisherman, so he knows how good streams look and act, and he understands his role: “We’re just trying to give the creek good bones,” he likes to say. “Then it’ll make itself over time.”

  Bob masters his machine. Most excavator operators can dig; Bob can sculpt. After watching him work his big orange Hitachi for a couple of hours, Susan turned to me and said, “If I could just put a brush in that thing’s claw, he could probably paint pictures with it.”

  For two weeks, Peter Bahls staked out the sinewy new channel and Bob dug and scraped and contoured. Peter surveyed to check depths and widths;
Bob adjusted both to match Tom’s drawings. Then he brought in long logs—damaged goods scavenged and trucked in from nearby clear-cuts—to place in the new channel. When a particularly big cedar log dropped into place, Bob would snatch the cigarette from his mouth and whoop like a cowboy. “Hot damn!” he’d holler. “The fish’ll love that.”

  Once the wood was in place and the excavator moved on, the field hands moved in. Sean led; we followed. As we worked and got to know Sean, we learned that he lived in the valley and had been working with nwi for three years, helping with restoration projects, doing spawning surveys to count and map salmon nests during the run, and sampling the creek throughout the spring and summer to count salmon fry. We also learned that he is of Irish-Italian-Inupiat (Eskimo) extraction. His maternal grandmother was a native of King Island, Alaska, a now-uninhabited rock in the Bering Straits. In 1778, King Island was home to two hundred people who looked out of their walrus-skin residences to see Captain James Cook sail by. Cook mapped the island and gave it its English name. Much earlier still, King Island had been part of the Bering land bridge. Sean’s ancestors may have been living on that rock since people first walked into North America from northeast Asia, more than fifteen thousand years ago.

  When Bob was done, the banks of what would become a meander were raw and the bottom of the channel dry. Our first job was to spread seed from rye and other annual grasses above and just below the future water line and add a layer of hay. By covering the bare soil, the hay would lessen the impact of rain and prevent little rivulets from forming and carrying silt into the channel. It would also mulch the grass seed, keeping newly sprouted seedlings from drying out and dying later that summer.

  Then we rolled out and pegged down a net of heavy jute twine called coir (coy-er). This was a key step: the coir would prevent the new stream from cutting into the raw banks, preventing erosion until the new vegetation could root and hold the soil. In time, the coir would decay and disappear.