Biological information related to risk assessments is presented below. This section is organized by broad geographic regions, with general information for each region summarized before the specific analysis for each ESU within the regions.
Historically, chinook salmon were abundant in the Central Valley. Early estimates did not differentiate run timing, so the following estimates are assumed to be totals for all runs. Eggs were collected from at least 30,000 adults in the upper Sacramento River in 1905; the total run in the Sacramento River could have been 10 times higher (ca. 300,000) (Reynolds et al. 1993). Gill-net catches suggest peak Central Valley chinook salmon in-river runs may have been 800,000 to 1,000,000 fish, with average run size about 600,000 fish prior to 1915 (Reynolds et al. 1993). Total Central Valley chinook salmon spawning escapement was estimated in 1965 to be about 421,000 fish (332,000 fall- & late-fall-run, 61,000 winter-run, and 28,000 spring-run) (CDFG 1995).
Chinook salmon in this region have been strongly affected both by losses and alterations of freshwater and estuarine habitats and by a long history of hatchery production. Reynolds et al. (1993) discussed habitat problems extensively. They reported a 95% loss of Central Valley freshwater salmon habitat due to damming, migration blockages, or severe degradation. The most severe losses began in 1849 with the discovery of gold, and culminated in the 1970s with the completion of major water diversion and conveyance facilities. Hydraulic mining caused sedimentation of spawning grounds, water diversions blocked migrations and depleted flows, and explosive human population growth led to major settlement and disturbance (including logging and agricultural activities) along Central Valley streams and rivers (CSLC 1993). Construction of levees for flood protection reduced off-channel habitat availability. By the 1930s, only 25% of the valley floor was subject to periodic inundation. Dam and water project construction further reduced habitat substantially between the 1930s and 1960s.
Direct relationships exist between water temperature, water flow, and survival of juvenile salmonids. Elevated water temperature in the Sacramento River has limited the survival of young salmon (Mitchell 1987, DWR 1988). Survival of juvenile salmon in the Sacramento River is also positively correlated with June streamflow and June and July delta outflow (Dettman et al. 1987).
Since 1872, chinook salmon have been continuously produced at a number of hatchery facilities. Millions of eggs were exported from the region during the 1800s. The majority of fish released prior to 1913 were unfed fry, whose contribution to the run was probably minimal (Clark 1929). By 1919, some 1.3 billion chinook salmon fry had been released into the Sacramento River Basin (Cobb 1930). Artificial propagation resources have been devoted primarily to fall-run chinook salmon. In the last 50 years, 1.6 billion fall-run fish have been released into the Central Valley; this is approximately 40 times more than the number of spring-run fish and 600 times more than the number of winter-run fish released (Table 6, Appendix D). The production of spring- and winter-run chinook salmon has been limited by the lack of suitable facilities for holding returning adults during the summer months.
Three hatcheries--Coleman NFH (1946), Feather River Hatchery (1969), and Nimbus Hatchery (1955)--have been responsible for most of the chinook salmon produced in the latter half of this century. Fish from these hatcheries have been released throughout the Sacramento and San Joaquin River Basins and in San Francisco Bay.
This ESU has been extensively reviewed by NMFS (1987, 1989, 1990a,b, 1994b), and that information is briefly summarized and updated here.
Historically, the winter run was abundant and comprised populations in the McCloud, Pit, Little Sacramento, and Calaveras Rivers. Construction of Shasta Dam in the 1940s eliminated access to all of the historic spawning habitat for winter-run chinook salmon in the Sacramento River Basin. Since then, the ESU has been reduced to a single spawning population confined to the mainstem Sacramento River below Keswick Dam (Reynolds et al. 1993). The last documented sighting of adult winter-run chinook salmon in the Calaveras River was made in 1984 (CDFG 1984).
Historic abundance has been estimated from anecdotal accounts, habitat capacity, and river gillnet fishery landings, but quantitative estimates of run-size are not available for the period prior to the completion of Red Bluff Diversion Dam in 1966. CDFG (1965) estimated spawning escapement of Sacramento River winter-run chinook salmon at 61,300 (60,000 mainstem, 1,000 in Battle Creek, and 300 in Mill Creek) in the early 1960s, but this estimate was based on "comparisons with better-studied streams" rather than actual surveys. Fish ladders at Red Bluff Diversion Dam permitted counting of the spawning runs after 1966. During the first 3 years of operation of the counting facility (1967-69), the spawning run of winter-run chinook salmon averaged 86,500 fish. The most recent 3-year (1994-96) average run-size wa s 830 fish. Since counting began in 1967, the population has been declining at an average rate of 18% per year, or roughly 50% per generation (Fig. 29). The trend in the most recent 10 years has been the same as the trend over the entire 27 years of data (Fig. 30, Appendix E).
The focus of artificial propagation efforts for winter-run chinook salmon has been a supplementation and captive broodstock program initiated in 1989. Recently, hatchery efforts may have resulted in the hybridization of spring- and winter-run chinook salmon (Hedgecock 1995). Furthermore, the fish reared at Coleman NFH (Battle Creek) were released into the mainstem Sacramento River where the winter run naturally spawns (USFWS 1996b), but rather than returning to their point of release they returned to Battle Creek where no suitable spawning habitat exists.
Freshwater harvest is negligible, but there is moderately high ocean harvest on this stock. In 1994, the ratio of ocean harvest to ocean harvest plus escapement (catch /(catch + escapement)) was estimated from CWT recoveries to be 0.54. This estimate was similar to one developed in the early 1970s from a fin-clip study. The recent reductions in ocean harvest are intended to insure that winter-run chinook salmon have a positive population growth rate, on average.
Historically, contribution of hatchery fish to this population has been negligible. Recently a captive-broodstock and smolt supplementation program has been initiated as part of recovery efforts.
The fact that this ESU is comprised of a single population with very limited spawning and rearing habitat increases its risk of extinction due to local catastrophe or poor environmental conditions. There are no other natural populations in the ESU to buffer it from natural fluctuations.
This ESU is currently listed as endangered under the California Endangered Species Act and was listed as threatened in 1989 and reclassified as endangered in 1994 under the US Endangered Species Act (NMFS 1990a, NMFS 1994b). The only other assessment of risk to stocks in this ESU was that made by Nehlsen et al. (1991), who identified one stock (Calaveras River) as extinct. Due to lack of information on chinook salmon stocks that are presumed to be extinct, the relationship of this stock to the existing Sacramento River winter-run is uncertain. It is listed here based on geography and to give a complete presentation of the stocks identified by Nehlsen et al. (1991) (Appendix E).
Historically, spring-run chinook salmon were abundant in the Sacramento River system and constituted the dominant run in the San Joaquin River Basin (Reynolds et al. 1993). Clark (1929) estimated that there were historically 6,000 stream miles of salmonid habitat in the Sacramento-San Joaquin River Basin, but only 510 miles remained by 1928. Subsequently, elimination of access to spawning and rearing habitat resulting from construction of impassable dams has extirpated spring-run chinook salmon from the San Joaquin River Basin and the American River. Construction of impassible dams has also curtailed access to habitat in the upper Sacramento and Feather Rivers.
In 1939, an estimated 5,786 spring-run chinook salmon passed the Cottonwood-Anderson Dam (Redding) on the upper Sacramento River (Hanson et al. 1940). Calkins et al. (1940) estimated a spawning escapement of 38,792 fish for the Sacramento River based on fishery landings. In the mid-1960s, CDFG (1965) estimated total spawning escapement of spring-run chinook salmon to be 28,500, with the majority (15,000) spawning in the mainstem Sacramento River and the remainder scattered among Battle, Cottonwood, Antelope, Mill, Deer, Big Chico, and Butte Creeks and the Feather River. CDFG (1965) reported spring-run chinook salmon to be extinct in the Yuba, American, Mokelumne, Stanislaus, Tuolumne, Merced, and San Joaquin Rivers. Today, spawner survey data are available for the mainstem Sacramento River, Feather River, Butte Creek, Deer Creek and Mill Creek (Big Eagle & Assoc. and LGL Ltd 1995). Small populations are also reported in Antelope, Battle, Cottonwood, and Big Chico Creeks (Campbell and Moyle 1990, Reynolds et al. 1993, Yoshiyama et al. 1996).
Spawning escapement has been estimated by a combination of methods, including snorkel surveys, aerial surveys, boat surveys, foot surveys, and fishway counts at Red Bluff Diversion Dam (Reavis 1985). The California Department of Fish and Game has estimated spawning escapement since the late 1940s or 1950s for the remaining populations except those in the mainstem Sacramento River, which has been counted at Red Bluff Diversion Dam since 1967. The sum of the 5-year geometric mean escapements for this ESU is 6,700 spawners, of which 4,300 (64%) have returned to the Feather River (Fig. 29, Appendix E). The Feather River Hatchery releases several million spring-run chinook salmon annually, with the bulk of their production released off-site into the Sacramento River Delta. Therefore, the origin of the fish returning to the Feather River is uncertain, and fish from these releases may stray to other parts of the valley. Of the remaining 2,400 spawners, 435 are in the mainstem Sacramento River where their spawning overlaps in both time and space with the more abundant fall run. Sacramento River mainstem spawners have declined sharply since the mid-1980s, from 5,000-15,000 to a few hundred fish. The Feather River population is believed to be hybridized with the fall run in the Sacramento River (Reynolds et al. 1993), and probably includes many hatchery strays from the Feather River Hatchery program. The remaining three natural populations (Butte, Deer, and Mill Creeks) are small, and all have long-term declining trends in abundance (Fig. 30, Appendix E).
Efforts to enhance runs of Sacramento River spring-run chinook salmon through artificial propagation date back over a century, although programs were not continuously in operation during that period. We found no recent records of introduction of spring-run fish from outside the Sacramento-San Joaquin River Basin. In the 1940s, trapping of adult chinook salmon that originated from areas above Keswick and Shasta Dams may have resulted in stock mixing, and further mixing with fall-run fish apparently occurred with fish transferred to Coleman Hatchery. Deer Creek, one of the locations generally believed most likely to retain essentially native spring-run fish, was a target of adult outplants from the 1940s trapping operation, but the success of those transplants is uncertain. Since 1967, artificial production has focused on the program at the Feather River Hatchery (discussed above). Cramer (1996) reported that half of the hatchery-reared spring-run fish returning to the Feather River did not return to the hatchery, but spawned naturally in the river. Given the large number of juveniles released off station, the potential contribution of straying adults to rivers throughout the Central Valley is considerable. The termination of CWT marking programs for hatchery-derived spring-run fish and the absence of spring-run carcass surveys for most river systems prevented the accurate estimation of the contribution of naturally spawning hatchery strays. Cramer (1996) reported that up to 20% of the Feather River spring-run chinook salmon are recovered in the American River sport fishery. Furthermore, the use of a fixed date to distinguish returning spring- and fall-run fish at the Feather River Hatchery may have resulted in considerable hybridization between the two runs (Campbell and Moyle 1990).
Harvest rates appear to be moderate. Ocean fishery management focuses on the fall run, with no defined management objectives for spring-run fish. Because of the similarity in ocean distribution with fall-run fish and smaller average size, spring-run harvest rates are probably lower than those for the fall run.
Reynolds et al. (1993) reported that spring-run fish were likely to have interbred with fall-run fish in the mainstem Sacramento and Feather Rivers, but the extent of hybridization was unknown. They also reported that pure strain spring-run fish may still exist in Deer and Mill Creeks.
The only previous assessment of risk to stocks in this ESU is that of Nehlsen et al. (1991), who identified several stocks as being at risk or of special concern (Appendix E). Four stocks were identified as extinct (spring/summer-run chinook salmon in the American, McCloud, Pit, and San Joaquin [including tributaries] Rivers) and two stocks (spring-run chinook salmon in the Sacramento and Yuba Rivers) were identified as being at a moderate risk of extinction. Due to lack of information on chinook salmon stocks that are presumed to be extinct, the relationship of these stocks to existing ESUs is uncertain. They are listed here based on geography and to give a complete presentation of the stocks identified by Nehlsen et al. (1991).
The historical abundance of Central Valley fall- and late-fall run chinook salmon is poorly documented. For the San Joaquin River, Reynolds et al. (1993) reported recent abundance to be only a remnant of the historical abundance. They estimated that production (ocean-run size) of San Joaquin River fall- and late-fall-run chinook salmon historically approached 300,000 adults and probably averaged approximately 150,000 adults. In the mid-1960s, escapement to the San Joaquin River Basin totaled only about 2,400 fish, spawning in the Stanislaus, Tuolumne, and Merced Rivers.
Calkins et al. (1940) estimated abundance at 55,595 fish in the Sacramento River Basin during the period 1931-39. In the early 1960s, adult escapement was estimated to be 327,000, predominantly in the mainstem Sacramento River (187,000), but with substantial populations in the Feather (50,000), American (36,000), and Yuba (22,000) Rivers and in Battle Creek (21,000); remaining escapement was scattered among numerous tributaries (CDFG 1965). At that time, total Central Valley fall-run chinook salmon escapement (including the Sacramento, Mokelumne, and San Joaquin River Basins) was estimated at 331,700 adults (CDFG 1965).
Much of the historical fall-run spawning area in the Sacramento River was below major dam sites, and therefore the fall run was not as severely affected by early water projects as were spring and winter runs (Reynolds et al. 1993). Extreme stream temperatures are a major limiting factor in juvenile production; gravel depletion, fluctuating flows, flow reversals in the delta, point and non-point source pollution, rearing habitat limitations, and losses at diversions also limit natural production (Dettman et al. 1987, CACSST 1988).
Spawning escapement has been estimated using a variety of survey methods. The larger spawning populations are estimated using modified Schaeffer or Jolly-Seber multiple mark-recapture methods with tagged carcasses (Reavis 1984). The fall and late-fall runs in the mainstem Sacramento River have been monitored since 1967 by counts in the fishways at Red Bluff Diversion Dam. Since 1992, the dam reservoir has been drawn down until May to allow the winter run to pass unimpeded. This has precluded counting the late-fall run since 1992 and has only permitted monitoring the last 15% of the winter run.
The bulk of the spawning escapement has been to the Feather and American Rivers and to Battle Creek (Fig. 29, Appendix E). The long-term trends in escapement are relatively stable, while the recent trends are mixed (Fig. 30, Appendix E). These are all streams with major salmon hatcheries. State hatcheries on the American and Feather Rivers transport their smolts to saltwater for release to avoid mortality in the delta due to flow reversals, unscreened diversion dams, and predators. Transportation of smolts increases the straying rate of adults when they return and makes it more difficult to account for hatchery strays in the spawning escapement (Cramer 1989). In the San Joaquin River Basin, homing fidelity may be more dependent on the presence of sufficient instream flows (CDFG 1997f).
Estimates of the relative contribution of hatchery and natural fish to spawning escapements are difficult to obtain. According to Dettman et al. (1987), for 1978-84 an average of 20% of the ocean catch of Central Valley salmon, originated at Feather River Hatchery and 24% at Nimbus Hatchery. For the same period, total Sacramento River spawning escapement was comprised of 22% Feather River Hatchery origin and 26% Nimbus Hatchery origin; 78% of the total Feather River run and 87% of the American River run were hatchery fish. For this period, natural production averaged only 12,000 fish in the Feather River and 8,000 fish in the American River. An alternative analysis (Cramer 1989) concluded that total hatchery contribution to the Sacramento River run for 1978-87 was only about one-third, and hatchery proportions in escapement were only 26% in the Feather River and 29% in the American River. Methods used in both studies have biases; Dettman and Kelley's estimates were biased toward hatchery fish and Cramer's estimates toward natural fish. Cramer suggested that the true proportions are probably somewhere between the two groups of estimates.
Fall- and late-fall-run chinook salmon in the Central Valley have been propagated for more than a century. In general, a relatively small number of hatcheries have accounted for the tens of millions of fall-run fish planted annually. The overwhelming majority of fish used have come from stocks within this ESU (Table 6, Appendix D). However, the practice of releasing fish off-station, especially into the Sacramento River Delta region, has resulted in widespread straying by hatchery-reared fish (Bartley and Gall 1990, Fisher 1995). Hatchery strays represent a considerable proportion of fish spawning naturally in many rivers, even those without hatcheries. Straying, in conjunction with frequent exchanges of surplus eggs between hatcheries, may be responsible for the low levels of genetic differentiation among fall-run chinook salmon stocks in the Central Valley (Bartley and Gall 1990). The high contribution of hatchery fish to naturally spawning escapement may be due, in part, to the high survival of hatchery fish that are transported to the Sacramento River Delta (Dettman et al. 1987).
In contrast to the situation with the fall run, the culture of late-fall-run fish has been relatively limited. The majority of production has come from one hatchery (Coleman NFH) and only within the last 20 years. Late-fall-run fish releases constituted less than 2% of the combined fall- and late-fall-run releases for this ESU.
Recent (1990-94) ocean harvest rate indices (Central Valley Index=catch / [catch + escapement]) have been in the range of 71-79% (PFMC 1996b). Freshwater recreational harvest is believed to be increasing and approaching 25% (PFMC 1997). Late fall fish are larger in size and experience higher harvest rates. The Central Valley Index is not a true harvest rate since it does not distinguish between races or cohorts, does not include freshwater catch or ocean catch landed north of Point Arena, California, and does not include shaker mortality (hook and release mortality of undersized fish).
Angler harvest in the Sacramento River Basin was estimated by creel census in 1991, 1992, and 1993 (Wixom see footnote 10, Wixom et al. 1995). The creel census data provide a harvest estimate of approximately 20% in freshwater.
The only previous assessment of risk to stocks in this ESU is that of Nehlsen et al. (1991), who identified two stocks (San Joaquin and Cosumnes Rivers) as of special concern (Appendix E). The Cosumnes River has had no documented spawning escapement of fall-run chinook salmon since 1989, and surveys in 1991 through 1994 have failed to find spawning salmon (Big Eagle & Assoc. and LGL Ltd. 1995).
Historically, chinook salmon were abundant in this region. Early estimates based on peak cannery pack suggest a total run size in excess of 300,000 fish in the 1910s. Total chinook salmon spawning escapement for the California portions of this region was estimated to be about 256,000 (168,000 in the Klamath River Basin and 88,000 elsewhere) in 1965 (CDFG 1995). An escapement of 250,000 fish in 1969 was estimated by expanded angler catch.
Chinook salmon in this region have been strongly affected both by losses and alterations of freshwater habitats and by a long history of hatchery production. PFMC (1995) identified all of the major rivers in this area as having chronic instream flow problems. Bottom et al. (1985) cited low stream flows and high summer temperatures as problems throughout the southern Oregon coastal area. Timber harvesting and associated road building occur throughout the region on federal, state, tribal and private lands. These activities may increase sedimentation and debris flows and reduce cover and shade, resulting in aggradation, embedded spawning gravel, and increased water temperatures (CACSST 1988, NMFS 1996b). The Rogue and Klamath River Basins have been sites of active mining since the mid-1800s and suction dredge mining still occurs.
Hatchery facilities in this area began operations late in the nineteenth century. These early hatcheries were operated by private companies and state and federal agencies with the goal of restoring declining fisheries. With the exception of operations on the Rogue River, which propagated spring-run chinook salmon, these hatcheries primarily reared fall-run chinook salmon. Dam construction and habitat degradation reduced or eliminated several runs and forced the closure of a number of hatcheries. Currently the Cole Rivers Hatchery and Trinity River Hatchery produce the majority of all spring-run chinook salmon in this area. A number of smaller hatcheries release locally derived fall-run chinook salmon, but the major proportion of fall-run releases comes from the Iron Gate Hatchery (197 million since 1966) and Trinity River Hatchery (69 million since 1969) (Appendix D).
The peak historic cannery pack of chinook salmon in the range of this ESU was 31,000 cases in 1917, indicating a run-size of about 225,000 at that time. CDFG (1965) estimated escapement for the California portion of the ESU at about 88,000 fish, predominantly in the Eel River (55,500) with smaller populations in the Smith River (15,000), Redwood Creek, Mad River, Mattole River (5,000 each), Russian River (500), and several smaller streams in Del Norte and Humboldt counties. Based on the 1968 angler catch records for the Oregon portion of the ESU (which estimated escapements of about 90,000 fish), the average escapement for the entire ESU in the 1960s was estimated to be 178,000 fish.
Within this ESU, recent abundance data vary regionally. Dam counts of upstream migrants are available on the South Fork Eel River at Benbow Dam from 1938 to 1975, and at Gold Ray Dam on the Rogue River from 1944 to the present. Counts at Cape Horn Dam on the upper Eel River are available from the 1940s to the present, but they represent a small, highly variable portion of the run.
In the Oregon portion of this ESU, coastal rivers are monitored by surveys of index reaches. Surveys were begun in 1948 with the intent of monitoring trends in escapement rather than estimating total escapement (Cooney and Jacobs 1994). Because the original selection criteria for index reaches included ease of access and availability of spawners, spawner densities in these index reaches are not representative of spawner densities in other areas. Consequently, though the spawner counts in index reaches may be relatively precise, they are not accurate for assessing abundance.
In 1953 Oregon began using catch report cards, called "punch cards," to report angler catch in rivers and estuaries (Nicholas and Hankin 1988). This reporting system provides precise estimates of catch on a river-by-river basis, which can be expanded by the harvest rate for each river to provide estimates of terminal run-size. Unfortunately, freshwater and estuarine harvest rates are poorly known for most rivers, and vary considerably. Harvest rates depend on fishing effort and angler success rates. Fishing effort varies with run-size, weather, river conditions, and angler success rate. Angler success rates, in turn, depend on weather and river conditions, as well as run-size. Nicholas and Hankin (1988) used estimates of average angler harvest rates to convert angler catch to run-size. These estimates, although imprecise, are probably more accurate for estimating average run-size than expansions based on peak index counts.
In assessing abundance and trends we used expansions of angler catch from ODFWs punch card database (ODFW 1993) and Nicholas and Hankin's (1988) average harvest rates to calculate geometric means of terminal run-size and spawning escapement for the most recent 5-year period (1990-94). Trends were calculated from either the peak index counts or from dam counts, where they were available.
Expanded angler catch data produce a 5-year geometric mean spawning escapement of 132,000 (run-size of 148,000) for the Oregon portion of this ESU. The majority of this escapement (126,000) has been the spring and fall runs in the Rogue River (Fig. 31, Appendix E). No total escapement estimates are available for the California portion of this ESU, although partial counts indicate that escapement in the Eel River exceeds 4,000. Data available to assess trends in abundance are limited. Recent trends have been mixed, with predominantly strong negative trends in the Rogue and Eel River basins, and mostly upward trends elsewhere. Longer term trends, where data are available, are flatter (e.g. Rogue River) (Fig. 32, Appendix E).
Habitat loss and/or degradation is widespread throughout the range of the ESU. The California Advisory Committee on Salmon and Steelhead Trout (CACSST 1988) reported habitat blockages and fragmentation, logging and agricultural activities, urbanization, and water withdrawals as the most predominant problems for anadromous salmonids in California's coastal basins. They identified associated habitat problems for each major river system in California. CDFG (1965, Vol. III, Part B) reported that the most critical habitat factor for coastal California streams was "degradation due to improper logging followed by massive siltation, log jams, etc." They cited road building as another cause of siltation in some areas. They identified a variety of specific critical habitat problems in individual basins, including extremes of natural flows (Redwood Creek and Eel River), logging practices (Mad, Eel, Mattole, Ten Mile, Noyo, Big, Navarro, Garcia, and Gualala Rivers), and dams with no passage facilities (Mad, Eel, and Russian Rivers), and water diversions (Eel and Russian Rivers). We expect that such problems also occur in Oregon streams within the ESU. The Rogue River Basin in particular has been affected by mining activities and unscreened irrigation diversions (Rivers 1963) in addition to problems resulting from logging and dam construction. Kostow (1995) estimated that one-third of spring-run chinook salmon spawning habitat in the Rogue River was inaccessible following the construction of Lost Creek Dam (RKm 253) in 1977. Recent major flood events (February 1996 and January 1997) have probably affected habitat quality and survival of juveniles within this ESU. Although we have little information on the effects of these floods in this ESU, the effects are probably similar to those discussed for the Oregon and Washington Coastal Region below.
Artificial propagation programs have been less extensive in the Southern Oregon and Coastal California ESU than in neighboring regions. The Rogue, Chetco and Eel River Basins and Redwood Creek have received numerous releases, derived primarily from local sources. In contrast, releases into the Russian River have been predominately from a variety of sources from outside the ESU (Table 6, Appendix D). In the absence of genetic information, it is not possible to evaluate the long-term impact of these transfers into the Russian River. San Francisco Bay has also received considerable numbers of introduced fish, the majority of which are off-station releases of Central Valley fall-run chinook salmon. Information on the impact of hatchery-derived fish on naturally spawning populations is limited. For the entire ESU, the hatchery contribution to total spawning escapement is probably low. However, the hatchery-to-wild ratio of Rogue River spring-run chinook salmon, as measured at Gold Ray Dam (RKm 201), has exceeded 60% in some years (Kostow 1995). The majority of the hatchery fish counted at Gold Ray Dam probably return to Cole Rivers Hatchery (located above the dam), but rates of straying into natural spawning habitat are unknown.
Ocean harvest rates for this ESU have not been estimated, but should be comparable to ocean harvest rates on Klamath fall-run chinook salmon (21% in 1991 [PFMC 1996a]). Freshwater and estuarine harvest rates are on the order of 25-30% (calculated from data in PFMC 1996b - Table B4).
Previous assessments of stocks within this ESU have identified several stocks as being at risk or of concern (Appendix E). Nehlsen et al. (1991) identified seven stocks as at high extinction risk and seven stocks as at moderate extinction risk. Higgins et al. (1992) provided a more detailed analysis of some of these stocks, and identified nine chinook salmon stocks as at risk or of concern. Four of these stocks agreed with the Nehlsen et al. (1991) designations, while five fall-run chinook salmon stocks were either reassessed from a moderate risk of extinction to stocks of concern (Redwood Creek, Mad River, and Eel River) or were additions to the Nehlsen et al. (1991) list as stocks of special concern (Little and Bear Rivers). In addition, two fall-run stocks (Smith and Russian Rivers) that Nehlsen et al. (1991) listed as at moderate extinction risk were deleted from the list of stocks at risk by Higgins et al. (1992), although the USFWS (1997a) reported that the deletion for the Russian River was due to a finding that the stock was extinct. Nickelson et al. (1992) considered 11 chinook salmon stocks within the ESU, of which 4 (Applegate River fall run, Middle and Upper Rogue River fall runs, and Upper Rogue River spring run) were identified as healthy, 6 as depressed, and 1 (Chetco River fall run) as of special concern due to hatchery strays. Huntington et al. (1996) identified three healthy Level II fall-run stocks in their survey (Applegate and Middle and Upper Rogue Rivers).
Peak run-size in this ESU was estimated to be about 130,000 chinook salmon in 1912 (from peak cannery pack of 18,000 cases). CDFG (1965) estimated spawning escapement of chinook salmon within the range of this ESU to be about 168,000 adults, split about evenly between the Klamath (88,000) and Trinity (80,000) Rivers.
Recent spawning escapements and run-sizes to the Klamath and Trinity Rivers are monitored by a combination of state, federal and tribal agencies. Hatchery returns to Iron Gate and Trinity Hatcheries are enumerated by the state. CDFG has also estimated escapement to the Trinity River, Scott River, Salmon River, and Shasta River using Petersen estimates from marks applied to upstream migrants at weirs, or tags applied to carcasses in stream surveys (Pisano 1993, Aguilar et al. 1996). Escapement to smaller tributaries is generally estimated from redd counts. The fall run on the Klamath River was counted at Klamathon Racks beginning in 1929, but these counts were discontinued when Iron Gate Dam was constructed and the mitigation hatchery began operation in the early 1960s. Escapement of fall-run chinook salmon to the Shasta River has been counted at a weir, or estimated on the basis of recovery of marks applied at the weir, since 1930 by CDFG. Escapement of spring-run chinook salmon to the Salmon River has been estimated by the U.S. Forest Service by snorkel surveys of holding habitat in the summer since 1980. Tribal commercial, subsistence, and ceremonial harvest has been monitored by the U.S. Fish and Wildlife Service, the Hoopa Valley Tribe, and the Yurok Tribe.
The 5-year (1992-96) geometric mean of recent spawning escapements to natural spawning areas was about 48,000 fish (Fig. 31, Appendix E). Fish returning to the two hatcheries in the basin accounted for 38% of the total (natural + hatchery) spawning escapement. Trends in escapement are relatively stable (Fig. 32, Appendix E). The long-term trend statistics mask the fact that minimal abundances were observed in all areas in 1989-91, and populations have increased sharply since then.
For over a hundred years, hatcheries have operated in the Upper Klamath and Trinity River Basins. Several million chinook salmon eggs were introduced into the region from the Central Valley, but the success of these introductions is doubtful, especially given the practice of releasing fry during the early part of this century. Dam construction on the Klamath and Trinity Rivers led to the construction of two major hatchery complexes (Iron Gate Hatchery and Trinity River Hatchery) to mitigate the loss of spawning and rearing habitat. Within the last 30 years, these 2 mitigation hatcheries have accounted for the overwhelming majority of artificially propagated fish in this region. Between 1964 and 1994, 50 million spring and 236 million fall-run chinook salmon (almost all from local sources) have been released (Table 6, Appendix D). It has been estimated that 11.2% of the spring-run fish and 31.2% of the fall-run fish naturally spawning in the mainstem Trinity River were of hatchery origin in 1994 (Aguilar 1995). Similarly, Barnhart (1995) reported that considerable numbers of coded-wire-tagged fish from the Iron Gate Hatchery are recovered among naturally spawning populations in Bogus Creek, and to a lesser extent in the Shasta River. Information on the contribution of hatchery fish to naturally spawning populations in other tributaries is lacking. Since systematic monitoring of spawning escapement began, the percentage of hatchery returns to total escapement has increased from 18% in 1978-82 to 26% in 1991-95 (PFMC 1996b).
The current management goal for fall-run chinook salmon in the Klamath River Basin is an escapement of 33-34% of potential spawners in each brood while providing a minimum of 35,000 adult spawners to natural spawning areas (PFMC 1994). Because of low abundance, recent management has been for a minimum escapement goal rather than the brood escapement rate. As a result, ocean fishery impact rates have decreased from 44-65% during the period 1986 to 1990 to 21% in 1991. Ocean fishery impact rates have remained below 20% since 1991 (PFMC 1996a).
Habitat loss and/or degradation is widespread throughout the range of the ESU. Upper basin habitat has been blocked by dam construction in both the Klamath and Trinity River Basins (KRBFTF 1991). NMFS (1996b) cited several factors affecting the habitat in this region, including water diversion/extraction, habitat blockages, hydropower development, and logging, mining, and agricultural activities. CDFG (1965, Vol. III, Part B) identified several critical habitat factors: water diversions and resulting low flows and high temperatures (Shasta, Scott, and Trinity Rivers), logging resulting in log jams and siltation (Klamath River), and small dams for present water diversion and at abandoned gold mines (Klamath River). They also cited siltation resulting from past mining activities as a problem in the Scott River, and noted that habitat in the Salmon River Basin was in very good condition. Timber harvesting and associated road building are widespread in the basin and result in increased sedimentation and debris flow and reduced cover and shade (KRBFTF 1991). Fifty percent of the spawning habitat in the Trinity River Basin was lost following the construction of Lewiston Dam at RKm 249 (Moffett and Smith 1950). Gold mining has occurred in this area since the mid-1800s. Lode mining for gold, copper, and chromite, which may introduce cyanide into the water and result in fish kills, continued in the Klamath River Basin until 1987. Suction dredge mining, which directly results in gravel disturbance and sedimentation, still continues in the basin (KRBFTF 1991).
Previous assessments of stocks within this ESU have identified several stocks as being at risk or of concern (Appendix E). Nehlsen et al. (1991) identified seven stocks as extinct, two stocks (Klamath River spring-run chinook salmon and Shasta River fall-run chinook salmon) as at high extinction risk, and Scott River fall-run chinook salmon as of special concern. Due to lack of information on chinook salmon stocks that are presumed to be extinct, the relationship of these stocks to existing ESUs is uncertain. They are listed here based on geography and to give a complete presentation of the stocks identified by Nehlsen et al. (1991). Higgins et al. (1992) provided a more detailed analysis of some of the stocks identified by Nehlsen et al. (1991), classifying three chinook salmon stocks as at risk or of concern. Of the three stocks Higgins et al. (1992) listed as at high risk of extinction, two matched with the Nehlsen et al. (1991) findings (Klamath River spring run and Shasta River fall run), while one stock was added to the list (South Fork Trinity River spring run). Additionally, three chinook salmon stocks were identified as of special concern. Of these, Higgins et al. (1992) classified one (Scott River fall run) in agreement with that of Nehlsen et al. (1991), while two others (Trinity River spring run and South Fork Trinity River fall run)were additions to the earlier list.