Buy a copy of Sean Gallagher’s new book, Wild Steelhead: The Lure and Lore of a Pacific Northwest Icon
Chapter Eight: Around the Campfire with Bill McMillan
It is hard to pinpoint my first exposure to Bill McMillan. He was just someone you knew about. Maybe it was all those Salmon Trout Steelheader magazine articles he wrote during the 1970s. He certainly was an innovator in the steelhead fly-fishing world. But it was the 1982 reprint of Greased Line Fishing for Salmon by Jock Scott (1935), retitled Greased Line Fishing for Salmon [and Steelhead] by Frank Amato Publications that stands out. Bill wrote the introduction. It was fascinating reading. Although the original had been published nearly 50 years earlier, half a world away, Bill was instrumental in bridging the gap. He introduced us to the parallels in steelhead presentation. It instantly became one of the most important instructional manuals for the modern-day steelhead fly fisher. It is no wonder that when his book titled Dry Line Steelhead came out in 1987 my buddies and I rushed out to buy a copy and devoured it.
Sometime in the mid-1980s, I met Bill at an FFF Steelhead Committee meeting. By then his stature had changed; he had transformed himself from apostle of Arthur Wood to a preeminent—and immensely knowledgable—spokesman for the resource. I was taken aback by his soft-spoken demeanor. And something else: I remember his telling me that one of his good fishing companions was a gear angler. This man was not some stuffy elite, as I had wondered. He was a sensitive, contemplative, dedicated human being. The Common Man using his influence to make a difference in our world. I have nothing but high regard for Bill McMillan and his philosophy of resource stewardship. All of us in the steelhead community owe him a debt of gratitude, not only for his contributions as an innovative angler, but for his years of service as an advocate for the wild resource.
In 1954, young Bill’s father was a struggling tailor in the little town of Camas, Washington. To help make the family’s financial ends meet, once every two or three weeks, his mother and father loaded Bill and his brother into the car and drove across the Columbia River to downtown Portland, to the Meier & Frank department store’s Friday surprise sale. The youngsters made a beeline to the sporting goods, where Bill’s brother would pound the mitt of a Stan Musial baseball glove, and Bill would watch Audrey Joy tying flies. One day Bill wandered over to the book department, where he spied a prominently displayed new title. On the paper jacket was a fish leaping out of the water against a black background, under a headline that read, Steelhead to a Fly by Clark Van Fleet. He picked up a copy, sat down, and started reading.
Fast forward to the summer of 1961. Bill was 16. He had his driver’s license. Behind the wheel of an automatic, six-cylinder, four-door Ford nicknamed the Green Turd, he took corners on river roads fast and hard, challenged stock-car racers—and occasionally beating the older, experienced drivers. The mild-mannered, introspective teenager was a maniac behind the wheel. He raced over to the Wind River canyon twice and caught nothing. Then, on a third trip with a spin-fishing friend, near the Saint Martin Hot Spring outlet, Bill lifted his Wright & McGill tubular rod, came forward and threw the sticky silk fly line from an old pre-war stash across the current. The Red Ant—a Royal Coachman with no tail and a brown bucktail wing, tied on a stout No. 8 hook—hit the water. The line started moving and stopped; he was stuck. And then a silver-bodied fish came out of the water, again, and again, screaming upstream. Bill had made the trout fisherman’s mistake of holding coils of line in his hand. Sure enough—the knotted line went zip, zip zip through the guides and jammed in the second guide from the tip. Bill thought, oh, shit! The fish was leaping all over the place and Bill remembered the six-pound leader. The fish ran downstream, jumping in a big half-circle toward him. Bill slammed the butt of the rod into the water, waded out waist deep and managed to untangle the line. He finally got the knot undone and cranked in all the slack line. The fish was lying right at his feet and then went running off again. It had already jumped a dozen times. Bill begged his friend to go back to the car and get a net. His friend came back and tried to net the steelhead but the trout net was too small. He hit the taut line with the net and Bill thought, oh, my God! But by then the exhausted fish lay on its side. Bill dragged its head up on a rock in the middle of the river—a 24-inch, glistening wild male steelhead.
“It’s one of the few fish I’ve landed that I’ll never forget,” Bill said. “I remember every moment.”
Bill McMillan is one of a kind—a pivotal figure in the history of the Northwest’s rich steelhead legacy. Now our legacy is threatened. On a rainy November day, Tom Pero and I talked with Bill at his riverside home on the Skagit River, where he had lived since 1998, beside a glowing wood stove in his old rustic cabin where each day he wages a war of words for a regional return to steelhead abundance.
We’re sitting here on the banks of the Skagit River, once of the brightest stars in the steelhead constellation. This is where I first met a steelhead. It’s hard for me to see this star nearly extinguished.
Bill McMillan: There’s total closure now on the late steelhead catch-and-release fishery. The state and the co-managing tribes make a preseason assessment on what they’re going to get back in the way of adult steelhead returns, based largely on smolt-trap information down in Mount Vernon and the number of smolts that went out two and three years before. It is a very rough estimate because the trap effectively samples only a narrow swath of a big, broad river. In fact, they know the smolt trap rarely captures juvenile sockeye at all, despite large numbers outmigrating. They know it’s not a very effective location for good smolt sampling, but it has the advantage of long-term comparative information. So predicted adult returns are based on what may not be very good outgoing smolt data and on what predicted ocean survival conditions might be based on recent years of steelhead ocean survival. Out of this alchemy of predictions they determine how to manage the fisheries for the coming winter season.
Do hatchery returns influence the prediction of wild returns?
McMillan: Yes, they then fine tune it for wild steelhead based on hatchery returns in early winter. It gives them a further idea of what the wild run may be—based on the same ocean conditions and the same year-class of the hatchery fish return. The managers base their fishery decisions on the early projection that includes returns to Marblemount hatchery and a test net fishery downstream by the tribal fishers. Given the more than decade-long trend of persistent decline, unless there is a significant indication of a shift from this pattern the river will likely continue to close to sport fishing early. Closure at the end of February or even end of January has become the norm. With few returns back to the hatchery and with the tribal test fishery poor, they just close it. The result in the winters of 2011-2012 and 2012-2013 have been closures at the end of January and in the previous year or two at the end of February. During the February closure years, the April catch-and-release fishery was retained from the bridge at Concrete on up to near Bacon Creek, and up the Sauk to Darrington. But with the January closures the April catch-and-release fishery has also been closed. Anyway, it has been a complex process of decisions, based on a kind of Ouija board of information. I have had no disagreement with the recent fishery outcome: closures. It has been apparent wild steelhead need protection. What has led to this is my problem with the Skagit River and Puget Sound management.
So the Washington Department of Fish and Wildlife, with the Sauk-Suiattle and Swinomish tribes, have eliminated our “taking” of wild steelhead by minimal catch-and-release mortality by keeping us off the river in springtime. What else are they doing to bring this population back?
McMillan: Although there have been some voluntary diminishment of the tribal net fishery, I had anticipated and hoped that maybe the tribal fishers would quit entirely at the end of January in recent years along with the closure of the sport fishery. But they hadn’t as of the 2011-2012 winter. This year will be wait and see. Tribal fishing has continued to occur for steelhead at least into February; now and then you’ll see tribal fishers out there in early March. But it has definitely decreased from what it was. Keep in mind the tribal fishers have treaty rights that go beyond our general perception of regulations. To their credit, they have voluntarily cut back their fishing time. My complaint is that we’ve got everything ass-backwards regarding steelhead fishery management.
McMillan: The reason we are in this desperate need to recover runs of Puget Sound steelhead—and most of the rest of the State of Washington—is because we have nearly eliminated early returning steelhead. Historically, wild winter-run steelhead here were dominated by December, January and February returns. If you go back to the tribal catch data from 1934 to 1959, prior to the hatchery programs for the 10 principal rivers where we’ve got data, from 85 to 98 percent of all the steelhead harvested in the tribal fisheries were between January and the end of February. Almost no steelhead were caught late. It’s possible that there were more later-returning fish than indicated by the tribal catch data; maybe the tribes didn’t target them because they weren’t as desirable to sell commercially at that later return date. Nevertheless, many of the later return steelhead are just as ocean fresh as the early return group on entry, and if they had been in large historic numbers there is no doubt the tribal fishers would have been fishing for them. Today a number of our Washington rivers may be at 50 to 75 percent of historic late-returning wild steelhead numbers. They’re actually not in that bad shape. But we are now largely missing the early entry winter fish. They used to be the dominate return group. They are the steelhead best adapted to Washington habitat conditions—and yet the very fish that steelhead management has largely eliminated.
Why are they best adapted?
McMillan: Because of their early return each winter. We don’t fully understand what made these early-return characteristics the most productive and abundant part of the run. But we do know some things. Some wild steelhead once spawned earlier than we have come to manage for over the past 30 years or more. Wild spawning often began in January, dependent on annual winter weather variables, the area of the watershed, and hydrological characteristics of their spawning destination. Some places they still do in Washington. In many areas, spawning was in full swing by February, much more than now, probably particularly so in lower-elevation, rain-driven tributary creeks. Still, the peak of spawning overall was probably March and April historically. But for over 30 years management has denied the presence of wild steelhead spawning prior to about March 15th.
Why is this important?
McMillan: Part of the story is suggested in the recent Skagit River steelhead study. Acoustic transmitters were inserted in 133 wild steelhead over a three-year period and tracked to their spawning destinations and back out to sea if they survived. Among the findings were that earlier-return steelhead stay in the system longer than late-return fish. Very few early return wild steelhead were captured to acoustic tag: zero in December, 2 in January, 11 in February. The remaining 120 steelhead were tagged from March through early May. Average time was nearly 50 days for steelhead tagged in February to reach Sauk River spawning destinations. The time diminished by month with steelhead tagged in April and May arriving at Sauk destinations in about 15 days. In other words, the earlier the entry, signficantly longer in-river time was required to reach the spawning grounds. There was a similar pattern for steelhead heading for the upper Skagit above Rockport. For mid-Skagit steelhead—from about Grandy Creek to Rockport—the time from tagging was more compressed into March and April. They got to spawning destinations in fewer than 10 days.
How long did the early fish stay in the main river?
McMillan: We don’t know. No December steelhead were tagged and only two were in January. But we can probably assume from the pattern that it was an even longer duration than for February steelhead. In the Sauk River basin there were also gender differences regarding spawning-destination arrival. Males tagged in February arrived 50 days earlier than the females. Other research has indicated the importance for males to begin staking out spawning territory at an early date to establish male dominance before female arrival. Unfortunately, there were no data to track this for February males at spawning destinations outside the Sauk.
The longer they are in the Skagit, obviously, the more vulnerable they are.
McMillan: Yes. This long period of river time before spawning makes early return steelhead particularly vulnerable to harvest—and to potential loss from mortality related to repeated catch-and-release in an intensive sport fishery. Another interesting finding is that most of the earlier steelhead were bound for the mid-Skagit and Sauk with early arrival time apparently particularly important at these two locations. The latest-entry steelhead in May were bound for the upper Skagit and Sauk, although, with only four tagged fish, the conclusions aren’t definitive. The earliest arrival month at spawning locations was for the lower mid-Skagit in February. At one time, even earlier spawning may have occurred in this area, but we’re largely missing that component of the run to build evidence by tagging.
How long do Skagit steelhead typically spend on the redds?
McMillan: That’s another interesting thing—the earlier their entry, the longer the time the steelhead spend at their spawning destination. February entry steelhead spend an average of 45 to 50 days at spawning sites while May entry steelhead average fewer than 20 to 30 days. This, again, increases their vulnerability to harvest or catch-and-release mortality if those spawning destination areas are in active fishery areas.
And after spawning?
McMillan: Post-spawning time to exit the Skagit basin to salt water ranged from as little as one day to as many as 255 days. The average was roughly 15 days. That one 255-day outlier was a male. It resided in the Skagit basin from entry to exit for nine months. The average length of time from tagging to exit at Whidbey Island was about 70 days and male residency in river was about a month longer than for females. This is an important finding. During the three-year study, there were 77 females tagged but only 56 males. Female numerical dominance varies by steelhead population but is the overall norm. It is vital for males to fertilize more than one female—otherwise fertilization depends on resident males. This is also necessary to maximize genetic diversity within egg batches.
How do these latest findings compare with the historical picture?
McMillan: A big shift appears to be in where they spawn. The 1977-1981 Skagit steelhead study found that an average of 75 percent of wild winter steelhead spawning occurred in the tributary creeks of the Skagit basin—80 percent one year. Today it is common to have less than 25 percent of spawning occurring in tributary creeks. The steelhead that were once critical to fill these diverse small tributary habitats were likely early-run steelhead that are now largely missing. It is like an eight-cylinder automobile trying to run on only two to three cylinders. These small rain-driven tributaries were once probably the safest destinations to lay eggs to maximize survival to emergence time. But being rain-driven, some of the creeks go dry or nearly dry by mid-June to early July. Early entry and spawning is a probable advantage in order to get the emergent fry out of the gravel soon enough to migrate downstream to the main stem Skagit for primary rearing before the creeks go dry, or before habitat is greatly diminished in available summer flow.
So they’re like summer steelhead in that they have their niche, and fish coming later might not be able to reach and populate those marginal habitats.
McMillan: The other thing that is important to know about winter steelhead is that early returns can be dominated by males and that males almost invariably arrive on the spawning grounds earlier than the females they will spawn with; another interesting finding has been that in the latest portion of the steelhead spawning period there can be very few remaining male steelhead as was found on the Quillayute basin of the Olympic Peninsula as found in John’s published research work there [Bill’s son, fisheries biologist John McMillan]. It is thought that male steelhead may have virtually spawned themselves to depletion by the time of the latest arriving females. It subsequently leaves resident males to fill the void for fertilizing the eggs of late return female steelhead. Early arrival on the spawning grounds is important for males in order to maximize the number of females each male can mate with. They’ve got to be there early. By eliminating all those early males, as has occurred since the earlier historical period prior to large returns of hatchery steelhead and subsequent management to harvest hatchery steelhead, it has diminished wild steelhead productivity through elimination of early run-timing in general, and may particularly have eliminated the best spawning contribution by wild male steelhead that is now divided between hatchery and wild females and further diminished by loss of earliest entry time to the spawning ground destinations.
How have we altered this dynamic?
McMillan: Harvest, primarily, but not only harvest. We also have introduced interaction with hatchery fish to the early spawning of wild fish, particularly with males. We’ve continually assessed steelhead spawning times by female maturation. By the time the females dig the redds with eggs at mature development, the males have often been at full gonad development and ready to go much earlier than the females. As demonstrated in the Skagit steelhead study—and in other research—male steelhead can be on the spawning grounds for a long time. I suspect where we are getting a lot of damaging hatchery interaction is with wild males spawning with hatchery females that average earlier egg maturation and spawning than wild females. The males are wasting themselves before the wild females are around. With the numbers of wild males greatly reduced already, the big primary males in particular probably are likely depleting themselves of sperm before it counts—when more wild females are available at later dates. This creates a two-fold depletion in steelhead productivity in the wild: 1) spawning with hatchery females results in greatly reduced survival to adulthood of the resulting hatchery/wild progeny; and 2) many wild males may be depleted before many wild females are ready to mate and lay eggs resulting in even lower numbers of males in wild steelhead populations that are already typically top-heavy with females and are dependent on every male spawning multiple times to maximize the productivity of the wild population. This is no longer occurring. The other thing is that we haven’t managed for resident rainbows, which John’s work is showing to be very important to spawning success and productivity of the overall mixed wild steelhead-rainbow population. Our rivers no longer have many resident males of optimal spawning size because we have insufficient angling regulations to protect them. In most rivers, resident and estuarine fish of 12 to 20 inches—half-pounder saltwater life histories—are open to harvest. These fish are typically male. They spawn in wild steelhead populations otherwise low on adult males. This has been well recognized on Russia’s Kamchatka Peninsula for many years, dating back to the late 1960s and noted in California steelhead studies as early as the 1930s and 1940s. Yet management in most of Washington has yet to develop adequate protection of these important alternative male life histories that are so important to overall steelhead productivity.
Those resident males are contributing to the gene pool and population viability by spawning with the larger fish that have returned from the ocean.
McMillan: Yup, and they are particularly important as found in John’s work for being there for the late-spawning wild females. As he found in his work on the Olympic Peninsula, on the Quillayute system, by May through July, the only fish left to spawn with the female steelhead are resident males. They are a vital part of the mix. We have yet to provide protection for them in Washington. Larger sized male rainbow trout with the largest gonads and spawning capacity have in some instances been almost entirely wiped out. We’ve perpetually allowed harvest on them. Same thing for jack steelhead, predominately male, that typically come back between 15 and 20 inches long. I can remember when we had such fish return in the Washougal—they just virtually disappeared back in 1980s. Kurt Beardslee, and others of Wild Fish Conservancy, in snorkel survey work in the Tolt River, found that one of the most interesting factors in watching the recovery of the wild Tolt summer steelhead run has been the reappearance of fish that have estuarine life histories or are residents. These fish were largely absent when they started doing the snorkel surveys in the late 1980s and early 1990s, but they now commonly observe fish of 14 to 19 inches. Kurt would like to find funding to further study how prevalent these resident or estuarine life histories once were in Puget Sound rivers prior to their depletion through harvest, and what part they may be playing in steelhead recovery on the Tolt.
Based on your research, Bill, please place the fate of steelhead running the Skagit River in perspective—how many steelhead do you think were here historically?
McMillan: The earliest commercial fishing records began when Washington became a state in 1889. Steelhead catch in those records for Puget Sound—separated out from salmon—don’t occur until about 1892. For some of the rivers, including the Skagit, there is no steelhead data until 1895. It happens that 1895 was the commercial steelhead peak in Washington. Soon after there was a steep and sustained decline in the catch. Steelhead were no longer recorded in the commercial catch from about 1930 onward. There may have been a period of wild steelhead recovery suggested by sport-catch data that began in 1948 to sometime in the early 1960s. But since the late 1960s, runs here in Puget Sound have been on a steady and relatively continuous decline heading toward extinction. It’s also very clear from early commercial records that they began fishing commercially for steelhead by mid-November in the Skagit and elsewhere in Puget Sound. Steelhead management over the past 30 years has denied the historic recognition of these vitally important early wild steelhead.
The freshwater homes of the steelhead were also under assault from logging upriver and clearing of log jams downriver.
McMillan: Yup, although a lot of the habitat destruction—some of the worst—was actually inflicted before 1895. Fifty to 70 percent of the lower Skagit basin from Mount Vernon to the coastline was historically water. The incredibly rich floodplain and tidelands…. Ponds, tidal marsh, beaver dams and beaver ponds—one vast waterscape. Now it’s 10 percent water. Rearing areas, we lost vast amounts of rearing areas. Prior to 1877 there were two river-spanning log jams between the present I-5 bridge and the Highway 9 bridge. One of those jams was a mile long; the other a half-mile long. They had been in the river for so many years that the larger of the two jams had a 100-year-old forest growing on top of it! To open Euro-American settlement of the upper Skagit basin, they had to get those jams out. Before that, the only way people got upriver was to portage canoes over these long jams to get to the upper basin. They frequently hired tribal people. In 1877, some volunteers in the Mount Vernon area began to whittle away at those jams. Over a period of about two years, with only half a dozen men continually working away, they eventually eliminated—or sufficiently notched—those jams. This created the ability for steamboats to travel upstream with supplies for settlers, and allow commercial activities back and forth. A government-operated snag boat continued removing large wood out of the Skagit channel to further accommodate steamboat travel. By the time we get to 1895—the year of peak steelhead harvest—a very high percentage of former large wood in the Skagit between the mouth and Marblemount had been removed that once provided important fish habitat. The original lower river had a mile and a half of river-spanning log jams. No predator could get to those fish. No human fishermen could get to them. They were totally protected right there at the top end of tidewater. It had to have provided a vital sanctuary for steelhead and salmon, for both juveniles and adults.
What do you think the run was?
McMillan: The commercial steelhead harvest on the Skagit in 1895 was 26,000 steelhead based on an average of about eight pounds per steelhead.
McMillan: Killed. We also found from early record-keeping that during the same year, an equal number of steelhead was estimated harvested by the ranchers of the upper Stillaguamish River basin. We found other data that clearly indicate this would not have been limited to the Stillaguamish—all the settlers in upper river basins would have been equally compelled to harvest fish. The perplexing thing to me was how and why so few people took so many fish. Settlers in those upper basins was probably limited to maybe a dozen families in 1895. What would a dozen families do with so much perishable protein? I searched and searched the literature and finally came across the most illuminating passage. It came from what is today called Hanson Creek at the east end of Sedro-Woolley on the Skagit River. There’s been a habitat recovery project ongoing there in recent years; it’s where an old mental hospital used to stand. They had a garden that patients worked to grow their own food. It was called Benson Creek in the 1800s. Somewhere along the line it got shifted to Hanson Creek. Anyway, in 1906, An Illustrated History of Skagit and Snohomish Counties was published with interviews of some 1,200 early settlers. One of the interviews was particularly revealing. An early pioneer of Benson Creek said that fish appeared over a protracted spawning period. When the fish came into the creek, each salmon species by its season, they were so numerous that children, bears, wolves, and farm dogs were chasing them up and downstream. Everyone had a gay old time—critter and human, all kind of independent from each other, not paying much attention to each other, all focused on chasing the fish. It sounded like great fun. Each family in the area had its wagon and team and each was easily filled with 400 fish in about two hours of pitchforking or spearing the fish out of the creek. They used these hauls to fertilize their gardens and orchards and to feed livestock—mostly hogs. This was happening on all the Skagit creeks. This explains how a relatively small number of settlers could harvest as many steelhead as the commercial fishery. During the fall months they used salmon to fertilize and during the spring months they used steelhead. You have to remember that when they removed the old-growth forest, the soil left was highly acidic and unproductive for agriculture. They didn’t have any money or access to commercial fertilizer. For effective agriculture, the way you brought the land back to productivity was to add fish. Your first priority would have been to pick a piece of property that had a creek on it to provide your drinking and cooking water, but the secondary benefit was to provide abundant fish fertilizer which would grow food for you and to feed your animals.
How many wild steelhead return to the Skagit River today?
McMillan: Well, in 2009 it was no more than 2,500—at most. That’s the official estimate. Historically, using catch records to reconstruct the likely strength of the run, we know that 26,000 steelhead were netted in the 1895 Skagit commercial fishery. Adding the estimate of equal steelhead numbers used by settlers, then that’s a harvest of 52,000. Using a basic rule of thumb of harvest, if you’re not diminishing a resource you need a 50 percent escapement—so potentially the Skagit steelhead run could have been 104,000 as my initial estimate back in 2006. However, there are differing mathematical approaches for better refining such estimates. One of these is called a Bayesian analysis into which one can enter ranges of uncertainties in the data. Computers can now run these data over and over again that can include not only the catch data in pounds and probable range of steelhead weights. But it also can factor in variables such as what the escapement range compared to harvest range might have been, based on the probability of whether a specific river’s settlement was more heavily weighted toward industrial or agricultural uses, as well as the pace of such development. The resulting range is that the size steelhead run would fall within all these given considerations. Nick Gayeski, Pat Trotter and I teamed to put our strengths together to develop a Puget Sound historic steelhead number estimate as well as those for four individual river basins that had well-defined catch data: Skagit, Nooksack, Stillaguamish and Snohomish. Nick is the mathematician, I was the primary historian, and Pat the primary editor with our resulting paper, Gayeski et al. 2011, “Historical abundance of Puget Sound steelhead, Oncorhynchus mykiss, estimated from catch record data.” It was published in the Canadian Journal of Fisheries and Aquatic Science. The Bayesian analysis resulted in a Skagit River steelhead run size in 1895 ranging from 70,000 to 149,000 with a mean of 105,600 steelhead—almost exactly our original 104,000 estimate. Using different approaches, we came to the same number. The past 10 years the Skagit run has averaged about 5,000 steelhead— four to five percent of the historic run size in 1895.
This is pre-dam?
McMillan: This is pre-dam. Dams weren’t put in on the Skagit until about 1922. That was when the Gorge dam blocked the upper river. The Baker dam was built on the Baker River in 1925.
Can it ever be brought back to any semblance of what it was?
McMillan: I think we can bring back to what we had in the 1950s.
Give us the Bill McMillan six-point program on how to bring back the Skagit.
McMillan: There can be no semblance of recovery until we bring back early-return steelhead. We’ve got management all wrong. We’re focusing harvest on early-return steelhead because we have to, in order to minimize hatchery steelhead interacting with wild steelhead.
So how do you have hatchery steelhead at all—how can you possibly accommodate hatchery steelhead into a Skagit restoration?
McMillan: At this point hatchery fish make wild recovery impossible. If you try to harvest all the hatchery fish to prevent swamping the wild population on the spawning grounds, you’re going to harvest the wild population right along with them and keep us perpetually where we are—with remnant, severely depressed returns of early winter fish. Until we can get that early-return life history back, there’s no potential to ever recover our steelhead populations—not in the Skagit, not throughout Puget Sound, not throughout western Washington. There should be far more steelhead than there are on the Queets River, for instance. There’s perfectly good evidence that the Queets had a particularly large steelhead population on the Northwest Coast. It’s more difficult to find accurate, long-term harvest records for other Olympic Peninsula streams, but there are such records for the Queets, including a particularly large harvest of steelhead in 1923 as indicated in cannery operation there at the time. The Queets was also once noted for very early return steelhead as found in the 1934-1959 tribal-catch data as well as the 1948-1959 sport-catch data. Over 95 percent of the Queets tribal catch was composed of December to February wild steelhead. All of this was long before the hatchery program began on the Queets in about 1980. We have to be able to recover those early-return wild steelhead. At this point, the strategy of providing protection for the late-return fish is not paying off. We’re within shooting distance of recovery back to 1950 levels in the Skagit basin if the hatchery steelhead releases cease. In the case of the Queets, where much of the habitat remains as it was in the 19th century in Olympic National Park, achieving wild steelhead run sizes back to those of the 1930s and ’40s may be achievable—again, by stopping hatchery plants.
In your judgment, is there enough wild seed stock to get the restoration going, or are they extinct?
McMillan: We haven’t completely eliminated them. They’re still there. I’ve got the modern tribal fishing records. Some wild steelhead are still caught in December in the Skagit basin and more in January and February. This was similarly the case for sport catch data when wild harvest still occurred in the Skagit basin until about 2002. The basic recovery building blocks remain.
So how do we do it—how do we bring back the icon of the Northwest?
McMillan: Number one, eliminate hatchery steelhead plants. If there’s any legitimate time and place to target wild steelhead harvest by tribal fishers, it’s the late run that is now the strongest wild return period and far closer to historic numbers than early return wild steelhead. Again, we’ve got it all ass-backwards by targeting the tribal fishery on the early steelhead that are now the weakest return component of the run. We should have catch-and-release fisheries for sportsman early in the season to give those fish maximum protection on the spawning run and then get our lines out of the water where the fish are getting close to their spawning destinations. The lower river may still provide some opportunity for a later catch-and-release sport fishery if it is tailored in a way to avoid concentrations of steelhead holding near their spawning destinations. Many of the late-return steelhead are migrating quickly through the lower river toward upper destinations as found in the Skagit acoustic tagging. Meantime, we have to somehow provide the treaty rights of tribal people on the lower river to harvest some proportion of those fresh fish coming through in the late run. The late run is the only remaining component that can possibly provide some limited harvest opportunity in a management plan driven by wild recovery. Then, as the recovery progresses and the runs grow in strength, expand out with an eventual goal of having relatively even distribution of fishing time over the full length of the run, with no concentrated harvest of any particular segment of the run. A plan for eventual low to moderate tribal harvest throughout the return period is the only way to sustain a wild steelhead future that can viably include harvest.
Give us the highlights of your big study of the Skagit.
McMillan: My assigned task in the Skagit study was to compare historic to present wild steelhead trends of other Pacific Coast rivers in Oregon, Washington, and southern British Columbia to that of the Skagit. I set out to see if there was or was not a pattern of hatchery impacts in these comparisons. It’s most interesting to compare the long-term trend of Skagit wild steelhead diminishment as an apparent response to fishery management that became dominated by a continual increase in hatchery steelhead smolt releases to other rivers. Many of these rivers have had a similar management focus. My job was to compare the Skagit to those few rivers on the coast where hatchery steelhead programs were never implemented or where hatchery programs have ceased. Record-keeping of the steelhead sport fishery began in 1948 in Washington when anglers had to begin recording their catches on punch-cards for subsequent annual catch tabulation by stream. In nearly the same period of time on the North Fork Umpqua River in Oregon—where winter-run steelhead have had no long-term winter steelhead hatchery program—the counts of steelhead past Winchester dam since 1946 reveal that while periodic upward and downward returns have occurred like a roller-coaster over time, the wild steelhead population has remained virtually stable for over 65 years with no declining trend. [motions with hands]
Ups and downs, ups and downs—pretty stable in the long run.
McMillan: No deep valleys. You see cycles of abundance reflected in normal changing freshwater and ocean conditions. Those same conditions affected runs returning to the Skagit, except abundance over time goes continually downward as shown by punch-card catch data and tribal catch data since the latter 1960s. [extended hand points downward]
You’ve still got the cyclical peaks and valleys, but overall it’s like an escalator going down.
McMillan: Furthermore, the North Umpqua has long provided wild steelhead harvest opportunity. It’s only within the last three years that many anglers have insisted on increased emphasis of catch-and-release with considerably reduced harvest. However, in the case of the North Umpqua it is apparent that it has sustained remarkably even, but moderate, harvest opportunity without resulting in wild diminishment. This is likely due to its comparative historical absence of hatchery winter-run steelhead. The North Umpqua has a consistent wild return past Winchester dam of 6,000 to 10,000 wild winter-run steelhead. It’s totally sustainable even with an annual harvest of 1,000 to 2,000 fish on average all the way along. [follows printed graph with finger] That’s the ideal target for a fishery, particularly if harvest is a necessary part of the management, as is the case in Puget Sound due to tribal treaty rights agreed to in 1855.
That’s the ideal management situation. You have a stable wild resource and if some want to harvest a small portion, they can. They have that opportunity. The resource isn’t going to suffer.
McMillan: It shows what a healthy river is. One can certainly build a case that steelhead don’t really need catch-and-release regulations on the North Umpqua, unless the goal of “maximum sustained recreation” is the driver built on the premise that catch-and-release mortality remains low enough to be less than outright harvest. My basic argument is that in the case of the Skagit and the other streams in the Boldt Case area of Washington, we don’t have that option of narrowing management here to “maximum sustained recreation.” The tribes have every right to expect their 1855 peace agreement that included the right to harvest fish in a manner that does not compromise the ability for steelhead and salmon to healthily sustain themselves. The management route to sustainably providing that is demonstrated by the North Umpqua—no hatchery plants and moderate harvest on winter steelhead that does not exceed about 20 to 25 percent of the run size. Moderation is the key to sustainability and which, of course, was the original tribal practice.
If somebody’s going to harvest them—it’s going to be the tribes.
McMillan: Yup, we’ve got to provide some sort of carrot to get the tribes on the bandwagon for recovery of wild fish. We have to be able to prove that wild fish can provide modest, steady harvest. But you cannot expect pie in the sky. Current management is based on expectations of mixed hatchery-wild return harvests of the 1960s, which isn’t going to happen. And which can’t ever again happen if steelhead are to be sustainable—and today they are not sustainable as managed in the Skagit basin and the rest of Puget Sound.
Yeah, it’s kind of like saying we should go back to the 1970s for our harvest of timber or back to the Westport days of big salmon limits. That’s not going to happen.
McMillan: My theory is the reason that we can never go back is that we have developed predator populations that specialize in eating the hatchery smolts that go out. It took predators about a decade to build breeding colonies around the golden opportunity for preying on massive numbers of hatchery fish flooding the estuaries at once. Hatchery smolts don’t have a clue of how to avoid predators in their year of hatchery pond rearing. Avian predators, seals, dogfish, sea lions and everything else—a feeding bonanza like they’d never experienced before. It provided the easiest way to make a living they ever saw. And once those predator levels built up, we can never go back to what the productivity level was like in the 1960s. And wild fish have to outmigrate as smolts through this same increased horde of predators with increased losses. Our reliance on hatcheries has created this whole disruption of the natural ecosystem. It’s completely out of whack. If we get rid of the hatchery fish, it will not take that many generations to get the wild fish back—if we provide that early-return component the opportunity to exploit what appear to be their advantages. In looking at the tribal harvest data, the Skagit still has enough early steelhead to build recovery on that return in December, January and February. Even my own fishing data indicates that in January is when one can well expect to begin encountering wild fish in predictable numbers. The tribal harvest data indicated that considerable numbers of January steelhead are caught. This is what Skagit recovery has to build on.
I think they can come back. I do.
McMillan: The Skagit basin is one of the more hopeful places in Washington due to quite extensive habitat for steelhead remaining and the relatively low human population in Skagit County. Nevertheless, it’s discouraging to look at places like Finney Creek in the lower valley. The forest above looks like it was detonated. The stream channel is now so broad in width that in places it’s the width of the Sauk River. It may take a century for its degraded channel to recover if we were to once make the decision to effectively alter forest practices to allow channel recovery to occur in that time. But it’s going to be a long process. It’s not going to be as productive as it historically was for a long time. But wild steelhead, in the absence of hatchery fish, will make better use of what remains in Finney Creek and other degraded creeks than they have in the past.
Yet there are huge areas of intact habitat on the Sauk, the Suiattle, the Cascade and all the other upper tributary streams.
McMillan: Yes, indeed. One of the big surprises in the recent Skagit study is that most of the hatchery interaction is taking place in all these little tributaries throughout the basin—that is where the hatchery fish are predominantly going to interact with the wild fish. And it is showing up in the genetics. There were the 1977-1981 studies done on the Skagit that tie in, one a genetic study and one a spawning study. The spawning study indicated that at that time—when they were doing spawning surveys in 1978, 1979 and 1980—from 65 to 80 percent of all Skagit River steelhead spawned in the small tributaries. Now it’s just the reverse. Most spawn in the main stems. This suggests to me that this shift is another reason why we are now missing early steelhead. The early steelhead must have historically spawned early in these tributaries. But since we’ve reduced their numbers these small streams are no longer as productive. It’s also where the genetics tell us that the hatchery and wild interactions have taken place. These small streams are vital for spawning. But they tend to diminish in the summer. Thereafter the Skagit itself provides vast areas of rearing space for creek origin juveniles. As the streams dried up, the juveniles had to migrate out into the Skagit. In 1982, the researchers couldn’t figure out how it was possible for the small tributaries to sustain that level of continued returns of numbers of adult fish, given the small areas for rearing. They theorized that the main stem Skagit provided the rearing while the tributaries supplied the spawning habitat, but even at that time the main Skagit was considered to be greatly underutilized rearing steelhead. This is where I base my estimate that the wild Skagit steelhead future has much room for recovery success.
How about the dam fluctuations—do you think that has a influence on the river for spawning?
McMillan: Not so much on steelhead spawning, but the fluctuations do greatly impact the chum and pink salmon below the entry of the Baker River due to operation of the Baker dams. The flows are daily turned on and off into the Skagit with common dewatering of salmon redds downstream—so the steelhead are affected in the diminishment of salmon that provide nutrients. Young steelhead are ultimately dependent on chum and pinks, primarily for nutrients to sustain them through the winter. After juvenile steelhead migrations out of the Skagit tributaries, what are they going to feed on? The river doesn’t have the aquatic insect populations of, say, Oregon’s Deschutes. What it used to have, historically, was tons and tons of pink and chum salmon dying and providing egg and carcass nutrients. So the dam fluctuations are damaging in that way. Remember the Baker River salmon kill back in 2000-2001, when we got front-page coverage in a number of major newspapers on both coasts?
Has this situation improved?
McMillan: At the time, Puget Sound Energy was fighting their federal relicensing agreement, which they had to have in place by 2008. They had been resisting doing anything about modifying or providing minimum flows to the Skagit. They were saying that their only obligation was to the Baker River. What they were providing the Baker River was trucking the adult sockeye and chinook upstream of the dams where there was still productive habitat. Trucking of steelhead, however, was not consistently done. They claimed that dewatering the lower Baker was doing no harm. That’s where they continued to make their stand until 2000-2001 when the salmon kill hit the front page of The New York Times and the third page in the Wall Street Journal. Their stocks took a big hit. That was a big wakeup call and subsequently became more negotiable on flow agreements. The agreements on minimum flows coming out of the Baker are predicated on their finishing installation of two more generators at the powerhouse which will include an ability to provide more gradual up and down ramping rates as well as not to go below a minimum flow standard. Before they had no minimum flow standards. The relicensing agreement indicates that the Baker is not to go below about 1,200 cfs discharge into the Skagit sometime in 2013. Construction presently continues as the deadline fast approaches to get those two generators operational and minimum flows met. That will help, particularly if they operate their generators the same way that Seattle City Light does on the upper Skagit. As power companies go, Seattle City Light is the best case example of any dam managers that I know for making progressive changes in the way they operate dams for fish. Chinook and chum and pink salmon have all dramatically benefited between Rockport and Newhalem in the area of primary Seattle dam influence. While ESA-listed chinook salmon have stablized in the upper Skagit, influenced by dam operations, everywhere else in the Skagit basin they continue to decline at about six percent per decade. In fact, at the time when Dave Pflug and Ed Conner [Seattle City Light biologists] had their upper Skagit paper published in the mid 2000s, chinook were actually increasing slightly in that section of river. Both pink and chum salmon have dramatically increased in that section. Prior to the 1970s, Seattle City Light was still doing some of the same things as Puget Sound Energy was on the Baker—dewatering, leaving flows off all weekends, and generally making fish production ineffective.
How are steelhead doing on the Skagit above Sauk?
McMillan: Until recent years, steelhead had not responded as well as salmon. But, then again, steelhead were not as immediately affected by the changes in how flows are operated on the upper Skagit. Their redds are made in the spring, when flows remain predictably high with snow run-off throughout the summer when fry emergence occurs. The biologists do chinook spawning surveys there by jet boat in August and September. They mark each of the redds and they take a GPS reading of where each of those redds is. Those closest to the bank are the ones that they are particularly concerned about—the location of the redds sets the limit of how low the flow is allowed to get on dam release. It’s not just an arbitrary measure. It’s their operational design not to dewater more than something like about two percent of those chinook redds. Essentially, it’s probably even better than natural conditions, and chinook have responded pretty well. Their numbers have stabilized. And pinks and chum have really benefited. One of the reasons why steelhead have not demonstrated marked increases in numbers is that they spawn in spring with very little likelihood of dewatering on through to July and even well into August. In recent years, steelhead numbers finally have been increasing in the upper Skagit basin. In fact, 2013 is looking to be an especially high count and this gradually increasing trend in steelhead redd counts has been occurring since about 2004 or so. The reasons are not yet clearly understood. The large floods in 2003 and 2006 had great impacts on the Sauk basin. I theorize that in part the impact of those floods on the spawning grounds, which became covered with two to three feet of sediment in many areas, resulted in larger than normal numbers of wild steelhead straying from the Sauk into the upper Skagit, which remained free of the flood effects. I suggest this due to knowing the response by Toutle River steelhead after Mount St. Helens erupted in 1980, which resulted in much steelhead straying for the first several years due to all the sediment in the Toutle basin. Perhaps that added straying by wild Sauk steelhead helped to jumpstart productivity in the uppermost Skagit main stem. However, this does not explain the still- increasing redd counts in the upper Skagit with the Sauk basin now largely clear of those sediment deposits that have since washed downstream and out of the stream channels. Another contributive explanation may be several combined factors. Hatchery steelhead plants were cut by more than half in the Skagit basin in 2008; this may help to explain the most recent upward bump in the upper Skagit from 2010 to now. Plants of steelhead smolts were also eliminated from Barnaby Slough in the upper Skagit. And in 2002 catch and release of wild steelhead was finally made mandatory throughout the year in the Skagit basin. Harvest of wild steelhead in the already depleted early return time period of December to February had been for some reason previously allowed in the sport fishery. Earlier initiation of spawning has been occurring in the upper Skagit—this may be an indicator of some positive benefits of implementing catch and release of those early wild fish. Late is better than never.
And they are not so much main-stem spawners.
McMillan: It’s a mix of both main-stem and tributary spawning in the upper Skagit. It is becoming increasingly clear that our steelhead problems in the Skagit may be more hatchery driven than for any other single reason.
But the hatchery bureaucracy is going to be tough to dismantle. The public—even anglers who should know better and have watched the runs disappear—think that more smolts planted results in more steelhead coming back.
McMillan: One of the positive development is that the tribal fishery managers and Seattle City Light biologists now acknowledge this reverse trend between hatchery smolt releases and steelhead returns. The more smolts that are planted the fewer steelhead—both hatchery and wild—come back. In the early to mid-1990s, Seattle City Light, as part of their dam relicensing and mitigation, agreed to provide more money for hatchery steelhead. So hatchery smolt releases increased dramatically from an average of about 225,000 to more than 500,000 annually.
But fewer adults returned….
McMillan: At a presentation about the steelhead hatchery program on the Skagit given by the Hatchery Scientific Review Group [independent team of biologist working on hatchery reform policies for Puget Sound and coastal Washington], I questioned the HSRG people about how they could scientifically justify their approval of the proposed Grandy Creek hatchery rearing pond plans in 2006—based on the evidence I and others had provided in the Grandy Creek hatchery pond review process, demonstrating that the more hatchery smolts planted, the fewer the returning adults. The HSRG folks weren’t very happy about my bringing this to their attention. Nevertheless, Seattle City Light and the Skagit tribes decided this wasn’t making any sense to continue escalating hatchery steelhead smolt plants. The Washington Department of Fish and Wildlife didn’t agree. They wanted to keep the mitigation money flowing for the hatchery program. But it was two against one for Skagit River management decisions. The tribes and Seattle City Light stayed firm. Subsequently, in 2008, they cut the hatchery plants by more than half. The hatchery steelhead are now back to around 220,000 smolts annually.
Do you think this will make a difference?
McMillan: I am hoping that we will see a slight increase in wild spawners over time, although I don’t expect any dramatic increases until we completely eliminate the hatchery steelhead problem in the Skagit basin. And it may yet happen.
Wow. This has been a real eye-opener, Bill. It’s refreshing to hear someone finally talk about practical solutions to saving wild steelhead. Maybe the old river has a future after all.