U.S. Dept Commerce/NOAA/NMFS/NWFSC/Publications
NOAA-NWFSC Tech Memo-19: Status Review for Klamath Mountains Province Steelhead


Population Trends

Coastwide Overview of Steelhead Abundance
Three substantial reviews of North American steelhead abundance have been undertaken: Sheppard (1972), Light (1987), and Cooper and Johnson (1992).

Sheppard (1972) reviewed historical commercial catch records from the 1890s through the 1960s. Total U.S. commercial steelhead catch declined sevenfold from an average of 2,700,000 kg in the 1890s to an average of 370,000 kg in the 1960s. Sheppard attributed most of this decline to restrictions on the fisheries rather than decline in abundance. For the period from 1945 to 1962, however, the Oregon coastal fishery was primarily an Indian gill-net fishery with relatively stable effort, so statistics for that fishery provide an index of abundance. This fishery declined from an average of 38,100 kg in 1945-49 to an average of 1,500 kg in 1958-62. The fishery was discontinued in 1962 due to declining stocks.

Sheppard (1972) also reviewed trends in sport catch of steelhead. Steelhead sport fishing statistics were first formally collected after World War II, when Washington and Oregon instituted punchcard systems in 1948 and 1952, respectively. California began using questionnaires to estimate steelhead catch in 1953 but discontinued regular reporting after 1956. In both Washington and Oregon, the number of anglers and total steelhead sport catch increased roughly twofold from 1953 to 1969, the last year included in Sheppard's study.

Finally, Sheppard (1972) provided rough estimates of total regional average adult steelhead runs in the early 1970s: California, 400,000; Oregon, 357,200; Washington, 606,400; Idaho, 42,500; British Columbia, 112,000; total, 1,528,000 (footnote 8). (This estimate for California can be compared with an estimate of 600,000 in the early 1960s; CDFG 1965.) These estimates were based on an expansion of total sport and commercial catch, assuming a 50% catch rate. Sheppard's overall assessment was that North American steelhead abundance remained relatively constant from the 1890s through the 1960s, although there had been significant replacement of natural production with hatchery production in California, Oregon, and Washington.

Light (1987) attempted to estimate total average run size for the mid-1980s based on sport harvest data, dam counts, and other resource agency information (Table 11). His coastwide total of 1.6 million was similar to Sheppard's estimate 15 years earlier.

Cooper and Johnson (1992) focussed on recent regional trends in steelhead abundance, using catch and hatchery returns as indices. They did not attempt an overall abundance estimate; however, they noted a recent (1985 to 1991) decline in steelhead returns (both hatchery and wild) in British Columbia, Washington, and Oregon. Cooper and Johnson suggested common factors that might be responsible for declines in steelhead returns, including a combination of low ocean productivity in the Gulf of Alaska, competition for food due to increased salmonid hatchery smolt releases and increased pink and sockeye salmon stocks, and catch of steelhead in high-seas driftnet fisheries.

Table 11. Estimates of average annual steelhead runs in the mid-1980s (Light 1987).
Adults (thousands) Percent
Region Hatchery Wild Total wild
Alaska 2 73 75 97
British Columbia 34 190 224 85
Washington Coast/Puget Sound 151 64 215 30
Columbia Basin 330 122 452 27
Oregon Coast 222 108 330 33
California 60 215 275 78
Regions combined 799 772 1,571 49

McEwan and Jackson (in prep.) provide an overview of steelhead abundance and trends in California. Despite the lack of any reliable abundance estimates, they note that angler catch rates, fishway counts, and survey estimates show substantial recent declines throughout the state. They also note widespread habitat loss and extirpation of several runs (especially in southern California) over the last two decades.

Historical Abundance in Southern Oregon and Northern California

Information on steelhead abundance in southern Oregon before the 1950s is sketchy, coming primarily from Rivers' (1957, 1963) studies of Rogue River Basin steelhead. Regarding late 19th-century fisheries in the Rogue River Basin, Rivers (1963, p. 56) reported that

cutthroat and downstream migrant steelhead were abundant and easily caught by the hundreds from streams all through the settled portions of the basin.... The headwaters of the Applegate River, the Illinois River, Jumpoff Joe Creek, and Grave Creek were sections of the basin preferred for trout fishing because of the easy access afforded by mining roads.

Historical information for northern California is even more scarce, although Snyder (1925) noted that trout (including steelhead) were declining in the Klamath River Basin at that time.

The Threshold Question
In considering whether the ESU containing Illinois River winter steelhead is threatened or endangered according to the ESA, we evaluated both qualitative and quantitative information. Recent information regarding steelhead stock abundance and trends are summarized at a river-basin level in
Appendix B. In compiling that summary, we sought to include all available assessments, both qualitative and quantitative, of steelhead populations in the region.

Qualitative assessments--Qualitative evaluations considered recent published assessments by agencies or conservation groups of the status of steelhead stocks from Cape Blanco to the Klamath River Basin (Nehlsen et al. 1991; Nickelson et al. 1992; USFS 1993a,b; McEwan and Jackson in prep.). Results of these assessments are summarized in Table 12; more detail can be found in Appendix B. Most winter steelhead stocks in the region are considered to be depressed and/or declining. Of the exceptions (those from the Rogue, Winchuck, Smith, and some subbasins of the Klamath and Trinity Rivers), most are heavily influenced by hatchery production. Only the Smith River appears to have healthy and largely natural production of winter-run steelhead in this region. For summer steelhead, the best assessment for any stock in this region is "depressed," and most are considered to be at moderate or high risk of extinction by the above authors.

Quantitative assessments--Historical abundance information for the geographic area of the ESU is largely anecdotal. Within this area, time series data are available for most populations only since 1970. We compiled and analyzed this information to provide several summary statistics of natural spawning abundance, including recent total spawning run size, percent annual change in total run size, recent naturally-produced spawning run size, and average natural return ratio (described below). Complete methods and results are given in Appendix B.

Because the ESA (and NMFS policy) mandates that we focus on viability of natural populations, we attempted to distinguish naturally produced fish from hatchery produced fish in compiling these summary statistics. All statistics are based on data for adults that spawn in natural habitat ("naturally spawning fish"). The total of all naturally spawning fish ("total run size") is divided into two components (Fig. 5): "Hatchery produced" fish are reared as juveniles in a hatchery but return as adults to spawn naturally; "naturally produced" fish are progeny of naturally spawning fish.

Table 12. Summary of recent qualitative assessments of steelhead abundance for all river basins reviewed. Blanks indicate that a particular run was not evaluated.
River basin Run-type Nehlsen et al. risk
USFS assessmentc
Elk River Winter Healthy
Euchre Creek Winter
Rogue River Winter Healthy Healthy
Summer Moderate Depressed Depressed
Applegate River Winter
Illinois River Winter Moderate Depressed Depressed
Hunter Creek Winter
Pistol River Winter Depressed
Chetco River Winter Depressed Depressed
Winchuck River Winter Healthy Healthy
Smith River Winter Healthy Low abundance
Summer High Depressed
Klamath River Winter Low abundance,
Summer Moderate Depressed,
moderate to high
Trinity River Winter Stable, depressed
Summer Stable, high risk

a - Risk of local extinction, as defined in Nehlsen et al. (1991).
b - Assessments in state agency documents: Oregon, Nickelson et al. (1992); California, McEwan and Jackson (in prep.).
c - General assessments of condition of portions of runs on U.S. Forest Service lands (USFS 1993a,b).

Figure 5
Figure 5. Schematic diagram of mixing of naturally and hatchery produced fish in natural habitat. Ovals represent the total spawning in natural habitat each generation. This total is composed of naturally produced (N) and hatchery produced (H) individuals in the previous generation.

The natural return ratio is used here as an indicator of the production of natural fish in the natural environment. This ratio is an estimate of the ratio of naturally produced spawners in one generation to total natural spawners (both naturally and hatchery produced) in the previous generation. This provides a rough index of natural production with the current-year contribution of hatchery spawners removed. As an example, the upper Rogue River summer steelhead run, counted at Gold Ray Dam, has been increasing at an average rate of 2% per year since 1971. However, it is estimated that between 11 and 73% of these have been hatchery fish, and the estimated average natural return ratio (based on a 4-year life cycle) for this period is only 0.68, indicating that natural production is not maintaining the population.

Results of these quantitative evaluations are summarized in Table 13. Most of the stocks in the region are in significant decline, even with hatchery production included. Natural production appears to be below replacement for all stocks for which we have this information; given the qualitative assessments, there is little reason to believe that other stocks are in better condition (with the possible exception of the Smith River winter run mentioned above). We are unable to demonstrate that any steelhead stocks in this region are naturally self-sustaining.

Total abundance varies widely among populations within the proposed ESU, with several populations having run sizes of 10,000 or more fish. The heavily hatchery influenced summer/fall run from the Klamath River may total 100,000 or more fish. At the other extreme, there are a number of populations with less than 1,000 spawners per year.

Table 13. Summary of recent steelhead population status for all river basins reviewed. Ranges in estimates reflect either multiple data sources or variation in estimates of percent hatchery composition. Question marks (?) indicate insufficient information. See individual stock summaries for details and data sources.
River basin Run-type Approximate
total run size
Data years Annual
change (%)
mate natural
run size
Elk River
Winter 850 1970-91 -8 540 0.44 36
Euchre Creek
Winter 140 1970-91 -5 90 ? ?
Rogue River, upper
Winter 5,300-11,000 1943-91 -5 to 0 8,500 0.16-0.79 47-81
Summer 8,900-14,000 1942-91 +2 to +3 7,000 0.68 18-49
Applegate River
Winter 5,300 1970-91 -2 1,900 0.18-0.49 47-81
Summer 1,600 1970-91 0 1,300 ? ?
Illinois River
Winter 5,900 1970-91 -10 5,500 0.60 7
Hunter Creek
Winter 380 1970-91 -6 130 0.17 67
Pistol River
Winter 1,500 1970-91 -3 910 0.53 38
Chetco River
Winter 5,100 1970-91 0 2,600 0.47 49
Winchuck River
Winter 540 1970-91 -4 350 0.44-0.60 25-45
Smith River
Winter ? ? ? ? ?
Summer 50 1981-91 +9 to +38 ? ? ?
Klamath River
Winter 20,000 ? ? ? ?
Summerb 110,000 1977-91 -15 to +4 ? ? ?
Trinity River
Winter ? ? ? ? ?
Summerb 15,000 1977-91 +5 to +16 ? ? ?
a - NRR: Natural Return Ratio (see Glossary,
Appendix A).
b - "Summer-run" estimates for Klamath and Trinity River Basins include "fall-run" steelhead.

Estimates of percent annual change indicate that most of the populations in the region are in significant decline, even with hatchery production included. We considered that this assessment may be influenced by the recent coastwide decreases in survival noted above. However, excluding these recent years (1987-present) from the trend analysis did not substantially change overall conclusions for the stocks considered here. Of those populations that are not declining, most have a large (ca. 20-80% of the run) hatchery produced component, so the apparent stability of these populations cannot be directly attributed to natural production.

Although this quantitative evaluation used the best data available, interpreting these results requires consideration of several complicating factors related to data reliability, analytic methods, and natural factors which may affect population abundance and trends. These problems are discussed in Appendix B and are only briefly mentioned here. Much of the quantitative analysis is based on either angler catch or instream adult survey data, which may not accurately reflect trends in population abundance. The methods used to derive natural return ratios from mixed-stock information require several assumptions about population regulation, which may lead to over- or underestimating potential natural production.

Reductions in Available Steelhead Habitat

In this section, we briefly discuss human activities that may have affected anadromous salmonid distribution and abundance within the geographical area of the proposed ESU. This is not meant to be a comprehensive analysis; rather, it is intended to briefly outline the nature and scope of activities that have occurred.

The effects of human activities on salmonids in the Klamath River Basin have been recognized to the extent that in 1986 Congress passed the Klamath River Basin Fishery Resources Restoration Act (16 U.S.C. 460ss-460ss-6, Public Law 99-552) to restore and maintain anadromous fish populations. The Klamath Act (p. 592) states in part:

...floods, the construction and operation of dams, diversions and hydroelectric projects, past mining, timber harvest practices, and roadbuilding have all contributed to sedimentation, reduced flows, and degraded water quality which has significantly reduced the anadromous fish habitat in the Klamath-Trinity River System.


A number of dams and diversions have been constructed within the Klamath Mountains Province during the past century. Dams have been installed for the purposes of flood control, hydropower, recreation, and domestic, industrial, and agricultural water supply. Not all of the dams have survived to the present time. Rivers (1963) and the Klamath River Basin Fisheries Task Force (KRBFTF 1991) chronicled the history of dams and diversions in the Rogue and Klamath River Basins, respectively. This document will discuss only those dams which have had a substantial impact on salmonid distribution and abundance.

Rogue River Basin--Gold Ray Dam (RKm 203) was originally completed in 1905 and rebuilt in 1940 (Rivers 1963). Fish ladders of various types and effectiveness have been installed at Gold Ray since 1906 (Rivers 1963). The Oregon Department of Fish and Wildlife has used the present ladder as a counting station for anadromous fish since 1968 (Evenson et al. 1982). Savage Rapids Dam (RKm 173) was completed in 1922 and has been laddered since 1923 (Rivers 1963). Construction of Lost Creek Dam on the Rogue River (RKm 254) was completed in February 1977; the primary purpose of this dam is flood control (ODFW 1994). Approximately 13% of the Rogue River Basin is located above Lost Creek Dam (ODFW 1994), which has no provision for fish passage (Cramer et al. 1985). Cole Rivers Hatchery operates as a mitigation hatchery for Lost Creek Dam; summer and winter steelhead as well as coho and spring chinook salmon are reared there (Evenson et al. 1982). Applegate Dam began operation in November 1980 at RKm 75 of the Applegate River (a tributary to the Rogue River at RKm 154) (ODFW 1994). Applegate Dam has no fish passage facility. Adult steelhead broodstock are collected below Applegate Dam, and eggs are cultured at Cole Rivers Hatchery.

Lost Creek and Applegate Dams are part of a three-dam flood control project in the Rogue River Basin (Fustish et al. 1989). The third dam, Elk Creek, has not yet been constructed. Elk Creek enters the Rogue River at RKm 244, and the proposed dam would be at RKm 2.7 on Elk Creek (Flesher et al. 1990).

Klamath River Basin--Anadromous fish passage to the upper Klamath River has been blocked at Klamath Falls, Oregon since the construction of the Link River hydroelectric dam in 1895 (KRBFTF 1991). Two hydroelectric dams were built by the California Oregon Power Company (Copco) northeast of Yreka, California in the early 1900s: Copco 1 in 1917 and Copco 2 in 1925 (KRBFTF 1991). No fish passage was provided, but a mitigation hatchery operated downstream on Fall Creek until 1948 (KRBFTF 1991). In 1958, Copco completed another dam below Keno, Oregon, presently called the J. C. Boyle Dam. Iron Gate Dam, completed in 1962, was ostensibly constructed to regulate the adverse flow regimes caused by Copco 1 and 2; however, it is also used for hydropower production (KRBFTF 1991). The dams described above block anadromous fish access to 120 km of mainstem habitat in the Klamath River and tributaries to that part of the river; it is estimated that this could provide spawning habitat to 9,000 chinook salmon and 7,500 steelhead (KRBFTF 1991).

The Trinity River is the largest tributary to the Klamath River; two diversion dams, Trinity and Lewiston Dams, were built on the upper Trinity River in 1964 to divert water to the Sacramento Basin as part of the Central Valley Project (KRBFTF 1991). The Trinity River Fish Hatchery was constructed at Lewiston Dam to mitigate the loss of fish passage.

Other Activities

It is relatively simple to quantify habitat loss due to dam construction; however, other activities that may effectively render habitat unusable for steelhead (e.g., through sedimentation, gravel mining, or water withdrawal) are more difficult to quantify.

Timber harvesting and associated road building activities occur throughout the Klamath Province on Federal, State, private, and tribal lands. These activities may increase sedimentation and debris flows and reduce cover and shade, resulting in aggradation, embedded spawning gravel, and increased water temperatures. The majority of forest lands in the Klamath Basin are managed by the U.S. Forest Service (KRBFTF 1991). The Klamath Mountains Province includes holdings of the Klamath, Rogue River, Shasta-Trinity, Siskiyou, and Six Rivers National Forests. According to the Forest Ecosystem Management Assessment Team, FEMAT (USFS and BLM 1994), 56% of the land in the Klamath Province is owned by the U.S. Forest Service, and 9% is owned by the Bureau of Land Management. Recognition of the importance of timber management activities on aquatic habitat is demonstrated in the provisions for riparian reserves and key watersheds described in FEMAT (USFS and BLM 1994).

The Rogue and Klamath Basins have been sites of active mining, primarily for gold, since the mid-1800s (Rivers 1963, KRBFTF 1991). Suction dredge mining results in sedimentation, which affects viability of salmonid eggs and juveniles, reduces holding habitat for adult salmonids, and reduces the standing crop of aquatic insects that salmonids prey upon (Rivers 1963, KRBFTF 1991). Suction dredging in the region continues to the present day (KRBFTF 1991). Dry rock (lode) mining introduces cyanide to the water and may cause fish kills (Rivers 1963, KRBFTF 1991). Lode mining for gold, copper, and chromite in the Klamath River Basin continued as recently as 1987 (KRBFTF 1991).

Irrigation in the Rogue Basin began in the late 1880s (Rivers 1963). Loss of salmon and steelhead to unscreened irrigation diversions was recognized as early as 1901 (Rivers 1963); however, the significance of these losses was not generally accepted. Loss of salmonids to unscreened irrigation diversions continues to the present day and is estimated at 1 million juvenile salmonids per year in the Rogue Basin (Palmisano 1992).

Discussion and Conclusions on the Status of the ESU
Threshold Assessment
Our quantitative and qualitative analyses revealed the following:
  1. Although historical trends in overall abundance within the ESU are not clearly understood, there has been a substantial replacement of natural fish with hatchery produced fish.
  2. Since about 1970, trends in abundance have been downward in most steelhead populations within the ESU, and a number of populations are considered by various agencies and groups to be at moderate to high risk of extinction.
  3. Declines in summer steelhead populations are of particular concern.
  4. Most populations of steelhead within the area experience a substantial infusion of naturally-spawning hatchery fish each year. After accounting for the contribution of these hatchery fish, we are unable to identify any steelhead populations that are naturally self-sustaining.
  5. Total abundance of adult steelhead remains fairly large (above 10,000 individuals) in several river basins within the region, but several basins have natural runs below 1,000 adults per year.
The Klamath Mountains Province steelhead ESU is not now at risk of extinction, but if present trends continue, it is likely to become so in the forseeable future. Although steelhead populations within the ESU share many ecological, life-history, and genetic characteristics, they are by no means homogeneous. The ESU contains populations from small streams as well as large rivers, and includes fish with a wide range of run-timing. Conserving existing diversity within this ESU should be a key component of recovery planning, just as it is for the ESU that contains Snake River spring/summer chinook salmon.
Table of Contents