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

APPENDIX B
STOCK ABUNDANCE AND TRENDS

Stock Abundance and Trends
Methods
Stock Summaries
Elk River, OR
Euchre Creek, OR
Rogue River, OR
Applegate River, OR
Illinois River, OR
Hunter Creek, OR
Pistol River, OR
Chetco River, OR
Winchuck River, OR
Smith River, CA
Klamath River, CA
Trinity River, CA

STOCK ABUNDANCE AND TRENDS

This appendix contains stock abundance and trend summaries for coastal steelhead trout (anadromous Oncorhynchus mykiss) spawning in coastal streams from Cape Blanco, Oregon, to, and including, the Klamath River Basin, California. Stocks are aggregated within major coastal drainage basins, listed from north to south. Minor drainages for which we have little information are not considered.

The goal of this summary is to provide sufficient information to assess the viability of natural steelhead populations in this region. While there are several quantitative techniques used for population viability (or vulnerability) analysis (PVA), such analyses for Pacific salmon are not sufficiently well developed to serve as a basis for ESA listing decisions. Instead, we consider a variety of information in evaluating the level of risk faced by a species. Important factors include 1) absolute numbers of fish and their spatial and temporal distribution; 2) current abundance in relation to historical abundance and current carrying capacity of the habitat; 3) trends in abundance, based on indices such as dam or redd counts or on estimates of spawner-recruit ratios; 4) natural and human-influenced factors that cause variability in survival and abundance; 5) possible threats to genetic integrity (e.g., from strays or outplants from hatchery programs); and 6) recent events (e.g., a drought or changes in management) that have predictable short-term consequences for abundance of the species. All of these factors need to be considered in terms of populations in the natural environment.

Information presented here includes estimates of historical and recent levels and trends in adult spawner populations and adult abundance indices derived from adult counts at dams or weirs, sport catch records, and spawner surveys. No consistent data for other life-history stages are available.

Historical abundance information for this geographic area is largely anecdotal, as is information relating to habitat capacity. This information is summarized in the main report and is not repeated here. Time-series data for adult abundance are available for most populations only since 1970. We compiled and analyzed this recent information to provide several summary statistics of natural spawning abundance, including recent total spawning run size, percent annual change in total run size, recent natural run size, and average natural return ratio.

Because the ESA (and NMFS policy) mandates that we focus on viability of natural populations, we attempted to distinguish natural 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; "natural" fish are progeny of naturally spawning fish.

Although the quantitative evaluations presented here used the best data available, it should be recognized that there are a number of limitations to these data and not all summary statistics were available for all populations. For example, spawner abundance was generally not measured directly; rather, it often had to be estimated based on catch (which itself may not always have been measured accurately) or on limited survey data. In many cases, there were also limited data to separate hatchery production from natural production.

Methods

Data Sources

Information was compiled from a variety of state and federal agency records. We believe it to be complete in terms of long-term adult abundance records for steelhead in the region covered. Principal data sources were angler catch estimates, dam or weir counts, and stream surveys. None of these provides a complete measure of adult spawner abundance for any of the streams. Specific data sources and problems are discussed below for each data type.

Angler catch--Availability of sport harvest information differs between Oregon and California. In 1952, Oregon instituted a punchcard system to record all salmon and steelhead caught by species. However, methods of estimating and reporting catch changed in 1970, so earlier data are not directly comparable to those since 1970. Our analyses for Oregon river basins focusses on data for the 1970 to 1992 run years (ODFW 1980, 1992c, 1993b). California began using questionnaires to estimate steelhead catch in 1953 but discontinued regular reporting after 1956, so no time-series of catch data are available for California river basins.

Interpreting population abundance from angler catch data presents several problems. First, numbers of fish caught do not directly represent abundance, which must be estimated from catch by applying assumptions about fishing effort and effectiveness. Fishing effort is largely determined by socioeconomic factors, including fishery regulations. Fishing effectiveness is a function of both the skill of the anglers and environmental conditions which affect behavior of both fish and anglers. Both effort and effectiveness may exhibit long-term trends and interannual fluctuations that can obscure the relationship between catch and abundance. Second, estimates of catch may not be accurate. In Oregon, catch is estimated from returns of punchcards corrected for nonreporting bias. While catch estimates are generated separately for each stream basin, the bias correction is calculated statewide and may not be accurate for any particular stream due to local variations in the tendency to return punchcards. Third, when fishing effort varies across a river basin, catch may reflect only local abundance rather than the total basin population. However, statewide salmon and steelhead fishing effort (as indexed by number of punchcards issued) has been relatively constant since the late 1970s (Fig. B-1), and winter steelhead catch rates (calculated by comparing catch estimates with dam passage counts) for the upper Rogue and upper North Umpqua Rivers have shown only small variation over the last several years (Fig. B-2).

The following analysis assumes that catch trends reflect trends in overall population abundance. We recognize that variations in effort and effectiveness introduce a certain amount of error, and that the index may not precisely represent trends in the total population in a river basin, but we believe that changes in catch still provide a useful indication of trends in population abundance.

Dam and weir counts--Dam and weir counts are available in two river basins: at Gold Ray Dam on the upper Rogue River, and at Bogus Creek and Shasta River weirs in the Klamath River Basin. These counts are probably the most accurate estimates available of total spawning run abundance, but they only represent small portions of the total population in each river basin. As with angler catch, these counts represent a combination of hatchery produced and natural fish, and these types are counted separately only at Gold Ray Dam.

Stream surveys--The California Department of Fish and Game and the U.S. Forest Service have conducted multiyear summer surveys for steelhead in several streams in northern California. For most of these streams, these are the only observations of steelhead abundance available. Unfortunately, these surveys count only fish that are "holding" in the streams during the summer, and so reflect only the early summer run, not the late summer (fall) run or the winter run. In addition, methods were not standardized in early surveys, and many streams were not surveyed each year, so analysis of these data is limited. The Oregon Department of Fish and Wildlife has conducted seine surveys for summer steelhead since 1976 at Huntley Park near the mouth of the Rogue River, from which total run size for the entire Rogue River Basin (including the Applegate River) has been estimated (ODFW 1994). The accuracy of these estimates is unknown.

Population Abundance Estimates

Historical abundance information is not available for individual river basins. Recent natural run-size estimates were compiled from various sources, including dam or weir counts and expansions of angler catch estimates, as described below.

Kenaston (1989) estimated average run sizes for Oregon winter steelhead returning to coastal streams from 1980 to 1985. These estimates were calculated by dividing estimated angler catch in each stream by an assumed exploitation rate based on classifying the local fishery as high, moderate, or low intensity. Kenaston also divided total run size into hatchery and "wild" components based on historical scale-analysis for individual streams or aggregate averages (when individual stream information was unavailable). We calculated similar estimates for the 1987 to 1991 run years including summer as well as winter steelhead. For winter steelhead, we used the exploitation rates used by Kenaston (1989, appendix A); for summer steelhead, we assumed the rates reported for moderate intensity fisheries (Kenaston 1989, Table 2). To estimate natural and hatchery components of the total run-size estimates, we applied average estimates of hatchery composition in the fishery from Chilcote et al. (1992), supplemented with estimates made by ODFW at Gold Ray Dam (summer and winter runs) and Huntley Park (summer run only). Resulting estimates of total and natural run sizes are of course only approximate, and should be interpreted only as approximate indicators of true population abundance.

Population Trend Estimates

As an indication of overall trend in steelhead populations in individual streams, we calculated average (over the available data series) percent annual change in adult spawner indices within each river basin. Trend estimates were calculated using exponential regression of spawner abundance indices against time with a generalized linear model (GLIM) (McCullagh and Nelder 1983) technique assuming Poisson observation errors. The GLIM technique was used rather than simple log-linear least squares regression because it is robust to zero counts in the population index and reflects the tendency of variance in population observations to be related to abundance. The regressions provided direct estimates of mean instantaneous rates of population change (r); these values were subsequently converted to percent annual change, calculated as 100(er-1). No attempt was made to account for the influence of hatchery produced fish on these estimates, so the estimated trends include any supplementation effect of hatchery fish.

Natural Production Estimates

The important role of artificial propagation (in the form of hatcheries) for Pacific salmon requires careful consideration in ESA evaluations. Waples (1991) and Hard et al. (1992) discuss the role of artificial propagation in ESU determination and emphasize the need to focus on natural production in the threatened or endangered status determination. However, they do not address the specific methods for evaluating natural production in the threatened or endangered status determination. This second problem is addressed here.

Because of the ESA's emphasis on ecosystem conservation, the threshold determination focuses on naturally reproducing salmon. An important question in the threshold determination is thus: Is natural production sufficient to maintain the population without the constant infusion of artificially produced fish? To answer this question, we need a method of estimating natural production with the contribution of hatchery reared fish removed. The natural return ratio (NRR) described below provides a rough measure of this.

Terminology--It is important to carefully distinguish stock components in populations that are derived from a mixture of natural and artificial production (Fig. 5). The natural component consists of fish that complete their entire life cycle in essentially natural habitat; the artificial component consists of fish that spend part of their early life cycle under artificial conditions. Note that these definitions refer only to the conditions under which fish live, not their heritage; natural fish may be the progeny of artificially produced parents, and vice versa. The two components will mix across generations: naturally spawning fish in one generation may be derived from both natural and artificially produced parents, and natural fish may be removed from natural habitat as broodstock for artificial propagation.

Production of a population is defined here in terms of return ratios ( lambdat) per generation and the closely related annual instantaneous rates of change (rt), both of which typically vary across brood years (t). Return ratios are simply the ratio of returning adult spawners to adult spawners the previous generation; an average ratio of 1 indicates a stable population. Annual instantaneous rates of change are calculated as

equation

where alpha is the mean age of spawning in the population. A value of 0 for the mean instantaneous rate of change indicates a stable population. Approach--The general approach to estimating NRR consists of three steps:
1) identifying natural stock abundance through time,
2) estimating the returning offspring from natural spawners, and
3) calculating return ratios and instantaneous rates of change for each brood year (or set of time-averaged brood years).
The average return ratio serves as an index of trend in the natural stock component, and variation about the average indicates the degree of variation in natural stock production.

Estimating return ratios--Because we rarely have age information on returning adults, we have estimated NRR by using returns at a fixed time-lag corresponding to the most common spawning age for the stock (assumed to be age 4 for these steelhead stocks; this assumption has little influence on the estimated average values). Because we have no direct counts of naturally spawning adults, we have used the best available index of natural spawning: dam counts, spawner survey counts, or angler catch estimates.

Estimation of return ratios depends on the type of information available for the population. Here, we consider only two typical scenarios: high information, with separate annual counts of natural fish and artificially produced fish on the spawning grounds, and minimal information, with only an annual index of total run size and an estimate of the average proportion of artificially produced fish in the spawning population. Estimates for the second scenario are of course more approximate. Among stocks considered in this review, the high-information scenario applied only to adult counts at Gold Ray Dam on the Rogue River; the minimal-information methods were used for other Oregon stocks. No California stocks had even the minimal information needed for this analysis.

Under the high-information scenario, the calculation proceeds as follows. Define Tt as the total (hatchery produced plus naturally produced) natural spawners in year t and Nt as the naturally produced natural spawners in year t. Then, assuming a 4-year life cycle, average NRR may be calculated as the geometric mean of Nt+4 /Tt. For the minimal-information scenario, we have data for only Tt and the average proportion of the run that is hatchery produced (h). We note that, on average, Nt = (1-h)Tt , so average NRR may be approximated as the geometric mean of (1-h)Tt+4 /Tt.

Assumptions--In interpreting average NRR as a quantitative indicator of population status, a number of simplifying assumptions need to be recognized. These include:
1) The population consists of a single unit, closed to all migration and immigration except for interaction with the included artificial stock.
2) Per capita contribution of artificially produced natural spawners to future generations is equal to that of naturally produced natural spawners.
3) Density dependence is not important.
4) Artificially produced fish have no effect on the production of natural fish.
Departures of real populations from these assumptions will of course affect the utility of NRR as an indicator of population status. The effect of the first assumption (closure to migration) could be either positive or negative, depending on whether emigration or immigration is larger for a particular population. The second assumption (equal reproductive success of natural and hatchery fish) is intentionally conservative (i.e., leading to a lower estimate of NRR than would other assumptions). There is some evidence for steelhead that artificially produced fish may have lower per capita contribution to future generations than do natural fish (Reisenbichler and McIntyre 1977), but the extent to which such effects depend on specific stocks or hatchery practices is unknown. The effect of the third assumption (lack of density dependence) is also conservative in that, if a mixed stock is near carrying capacity, the apparent NRR may be substantially lower than would be observed for the same stock at lower abundance levels (or if there were no hatchery). The final assumption (no effect of hatchery fish on natural fish) is also probably in a sense conservative: if the effect of hatchery fish is negative (e.g., through competition, disease transmission, or lowered hybrid fitness), then the observed NRR would be lower than return ratios for the natural stock in the absence of hatchery fish. Considering all the assumptions together, it is likely that average NRR provides a somewhat conservative estimate of natural stock production.

Finally, it must be recognized that these estimates of NRR are only approximate. Especially for estimates derived from angler catch data, there is a potentially high level of error in estimates of both spawner numbers and hatchery proportions. Because of these errors and the various assumptions in interpreting return ratios, these ratios should not be viewed as formal statistical estimates of true population parameters, and we have not tried to provide error estimates or confidence intervals for them.

Trend and Production Estimates

Results of the quantitative analyses are summarized in Table B-1. Other information, including qualitative assessments of population status, are given in the individual stock summaries that follow.


Table B-1. Summary of estimated abundance statistics for individual data series, listed by state and river basin. Recent run size estimates reflect an average of the most recent 5 years of data, or the most recent published estimate. Blanks (--) indicate lack of information. Where ranges are given, these reflect ranges in estimates of the hatchery produced proportion of spawning stocks. Sources of information are given in the individual stock summaries.
River basin Run-type Data
typea
Data years Recent
total run
Annual
change (%)
(mean(s.e.))
Percent
hatchery
Recent
natural
run
Average
NRRb
Oregon
Elk River winter AC 1970-91 850 -8.4(0.1) 36 540 0.44
Euchre Creek winter AC 1970-91 140 -4.7(0.5) - 90 --
Rogue River, upperc winter AC 1970-91 5,300 -5.3(0.0) 47-81 1,900 0.16-0.45
winter DC 1942-91 11,000 +0.3(0.0) 47-81 8,500 0.79
summer AC 1970-91 8,900 +1.5(0.0) 18-49 5,200 --
summer DC 1942-91 14,000 +3.2(0.0) 18-49 6,900 0.68
Rogue River, lowerc winter AC 1970-91 14,400 -- -- 5,200 --
summer AC 1970-91 13,200 -- -- 10,300 --
summer BS 1976-91 18,000 -2.5(0.0) 22 14,000 0.57
Applegate River winter AC 1970-91 5,300 -1.7(0.1) 47-81 1,900 0.18-0.49
summer AC 1970-91 1,600 -0.1(0.1) -- 1,300 --
Illinois River winter AC 1970-91 5,900 -10.2(0.1) 7 5,500 0.60
Hunter Creek winter AC 1970-91 380 -5.8(0.3) 67 130 0.17
Pistol River winter AC 1970-91 1,500 -3.2(0.3) 38 910 0.53
Chetco River winter AC 1970-91 5,100 -0.2(0.1) 49 2,600 0.47
Winchuck River winter AC 1970-91 540 -3.9(0.2) 25-45 350 0.44-0.60
River basin Run-type Data
typea
Data years Recent
total run
Annual
change (%)
(mean(s.e.))
Percent
hatchery
Recent
natural
run
Average
NRRb
California
Smith River
Middle Fork summer SS 1982-91 -- +38.0(10.6) -- -- --
South Fork summer SS 1981-91 -- +9.4(2.3) -- -- --
Klamath River
winter UK 1980s 20,000 -- -- -- --
summer and
fall
UK 1980s 110,000 -- -- -- --
Salmon River, North Fork summer SS 1980-91 -- -12.8(1.5) -- -- --
Salmon River, South Fork summer SS 1980-91 -- -9.0(0.9) -- -- --
Wooley Creek summer SS 1980-91 -- -2.6(0.6) -- -- --
Bluff Creek summer SS 1980-91 -- +3.7(1.1) -- -- --
Redcap Creek summer SS 1980-91 -- -1.8(2.0) -- -- --
Dillon Creek summer SS 1980-91 -- -8.2(0.6) -- -- --
Clear Creek summer SS 1980-91 -- +2.2(0.4) -- -- --
Elk Creek summer SS 1980-91 -- -3.9(0.9) -- -- --
Combined Klamath River Basin summer SS 1980-91 -- -3.3(0.3) -- -- --
Shasta River fall DC 1977-92 -- -14.9(0.5) -- -- --
Bogus Creek fall DC 1984-92 -- -1.1(4.6) -- -- --
Iron Gate Hatchery fall and
winter
HR 1963-91 -- +1.5(0.1) -- -- --
Trinity River
above Willow Creek RR 1980-91 15,000 -- -- -- --
Trinity River, South Fork summer SS 1982-91 -- +5.3(1.9) -- -- --
Trinity River, upper summer SS 1980-91 -- +16.4(3.6) -- -- --
New River summer SS 1980-91 -- +5.5(0.4) -- -- --
Trinity River, North Fork summer SS 1980-89 -- +11.4(0.6) -- -- --
Canyon Creek summer SS 1980-91 -- +4.7(2.5) -- -- --
Combined Trinity River Basin summer SS 1980-89 -- +7.6(0.4) -- -- --
Trinity River Hatchery summer and
winter
HR 1958-90 -- -1.5(0.0) -- -- --
a - AC--angler catch; BS--beach seine; DC--dam or weir count; HR--hatchery return; RR--run reconstruction; SS--stream survey; UK--unknown method (see stock summary).

b - NRR: Natural Return Ratio (see Glossary, Appendix A).

c - Includes some upper Rogue, Applegate, and Illinois River steelhead.


STOCK SUMMARIES

Elk River, OR

The Elk River has only winter-run steelhead. We have no historical (pre-1900s) steelhead abundance estimates specific to the Elk River. Recent abundance estimates are derived from angler catch estimates (ODFW 1980, 1992c, 1993b). Kenaston (1989) estimated average 1980-85 winter steelhead run size of ca. 1,400 total and 800 natural fish; updated run-size estimates (1987-91 average) are 850 total and 540 natural fish. Angler catch declined at an average rate of ca. 8% per year between 1970 and 1991 (Fig. B-3). Hatchery fish have recently averaged 36% of the angler catch (Chilcote et al. 1992), and average NRR based on angler catch is ca. 0.44. (Chilcote et al. estimate that less than 10% of fish on spawning grounds are of hatchery origin, so actual NRR may be higher than that estimated from angler catch.) Biologists with the U.S. Forest Service (USFS) report that this steelhead population appears healthy (USFS 1993a,b).

Euchre Creek, OR

Euchre Creek has only winter-run steelhead. We have no historical (pre-1900s) steelhead abundance estimates specific to Euchre Creek. Recent abundance estimates are derived from angler catch estimates (ODFW 1980, 1992c, 1993b). Kenaston (1989) estimated average 1980-85 winter steelhead run size of ca. 300 total and 200 natural fish; updated run-size estimates (1987-91 average) are 140 total and 90 natural fish. Angler catch declined at an average rate of ca. 5% per year between 1970 and 1991 (Fig. B-4). No estimate of the proportion of hatchery fish in the run is available (Chilcote et al. 1992), so we cannot estimate NRR.

Rogue River, OR

The Rogue River has both winter- and summer-run steelhead. We have no historical (pre-1900s) steelhead abundance estimates specific to the Rogue River. Recent abundance estimates are derived from angler catch estimates (ODFW 1980, 1992c, 1993b), adult passage counts at Gold Ray Dam on the upper Rogue (ODFW 1990, 1994), and summer steelhead surveys at Huntley Park near the river mouth (ODFW 1994). From angler catch data, Kenaston (1989) estimated average 1980-85 winter steelhead run sizes of ca. 7,400 total and 3,200 natural fish in the lower Rogue, and 4,000 total/1,500 natural fish in the upper Rogue; corresponding updated run-size estimates (1987-91 average) are 14,400 total/5,200 natural fish in the lower Rogue and 5,300 total/1,900 natural fish in the upper Rogue. For summer steelhead, estimated average 1987-91 run sizes were 13,200 total/10,300 natural fish in the lower Rogue and 8,900 total/5,200 natural fish in the upper Rogue. Recent (1981-91) counts at Gold Ray Dam had the following ranges: 4,300-16,200 total and 2,900-12,700 natural winter-run steelhead; 4,400-26,300 total and 3,200-13,000 natural summer-run steelhead. Between 1970 and 1991, angler catch of winter-run steelhead declined at an average rate of ca. 5% per year while catch of summer-run steelhead increased ca. 2% per year (Fig. B-5). Over a similar period, counts at Gold Ray Dam increased by less than 1% (winter run) and ca. 3% (summer run) per year (Fig. B-6), while estimates of adult summer-run steelhead passing Huntley Park declined by ca. 3% per year (Fig. B-7). Estimated average return ratios (see Table B-1) have shown similar variation among the data sets. Nehlsen et al. (1991) listed summer-run steelhead in the Rogue as at "moderate risk of extinction." The ODFW described Rogue River winter steelhead as "healthy" and summer steelhead as "depressed" (Nickelson et al. 1992); USFS biologists concurred with this assessment (USFS 1993a,b).

Applegate River, OR

The Applegate River has both winter- and summer-run steelhead. We have no historical (pre-1900s) steelhead abundance estimates specific to the Applegate River. Recent abundance estimates were derived from angler catch estimates (ODFW 1980, 1992c, 1993b). Kenaston (1989) estimated average 1980-85 winter steelhead run size of ca. 2,200 total and 800 natural fish; updated run-size estimates (1987-91 average) are 5,300 total and 1,900 natural fish. Recent (1987-91 average) run-size estimates for summer steelhead are 1,600 total and 1,300 natural fish. Summer-run angler catch showed no significant decline between 1970 and 1991, while winter-run catch declined at an average rate of ca. 2% per year (Fig. B-8). Hatchery fish have recently averaged 47-81% of the winter run (Chilcote et al. 1992), and average winter-run NRR is ca. 0.18-0.49. No hatchery composition estimate is available for summer-run steelhead.

Illinois River, OR

The Illinois River presently has only winter-run steelhead. Rivers (1957) noted a small summer run, but whether these summer fish actually spawned in the Illinois River is unknown. We have no historical (pre-1900s) abundance estimates specific to the Illinois River. Recent abundance estimates were derived from angler catch estimates (ODFW 1980, 1992c, 1993b), which reflect primarily the upper basin (above Illinois Falls). Kenaston (1989) estimated average 1980-85 winter steelhead run-size of ca. 10,300 total and 6,300 natural fish; updated run-size estimates (1987-91 average) are 5,900 total and 5,500 natural fish. Angler catch declined at an average rate of ca. 10% per year between 1970 and 1991 (Fig. B-9). Hatchery fish have recently averaged only 7% of the run (Chilcote et al. 1992), and average NRR is ca. 0.60. Nehlsen et al. (1991) listed winter-run steelhead in the Illinois River as at "moderate risk of extinction." ODFW described this population as "depressed" (Nickelson et al. 1992, ODFW 1992a); USFS biologists concurred with this assessment (USFS 1993a,b).

Hunter Creek, OR

Hunter Creek has only winter-run steelhead. We have no historical (pre-1900s) steelhead abundance estimates specific to Hunter Creek. Recent abundance estimates were derived from angler catch estimates (ODFW 1980, 1992c, 1993b). Kenaston (1989) estimated average 1980-85 winter steelhead run-size of ca. 800 total and 500 natural fish; updated run-size estimates (1987-91 average) are 380 total and 130 natural fish. Angler catch declined at an average rate of ca. 6% per year between 1970 and 1991 (Fig. B-10). Hatchery fish have recently averaged 67% of the run (Chilcote et al. 1992), and average NRR is ca. 0.17.

Pistol River, OR

The Pistol River has only winter-run steelhead. We have no historical (pre-1900s) steelhead abundance estimates specific to the Pistol River. Recent abundance estimates were derived from angler catch estimates (ODFW 1980, 1992c, 1993b). Kenaston (1989) estimated average 1980-85 winter steelhead run-size of ca. 2,200 total and 1,200 natural fish; updated run-size estimates (1987-91 average) are 1,500 total and 900 natural fish. Angler catch declined at an average rate of ca. 3% per year between 1970 and 1991 (Fig. B-11). Hatchery fish have recently averaged 38% of the run (Chilcote et al. 1992), and average NRR is ca. 0.53.

Chetco River, OR

The Chetco River has only winter-run steelhead. We have no historical (pre-1900s) steelhead abundance estimates specific to the Chetco River. Recent abundance estimates were derived from angler catch estimates (ODFW 1980, 1992c, 1993b). Kenaston (1989) estimated average 1980-85 winter steelhead run-size of ca. 7,200 total and 3,200 natural fish; updated run-size estimates (1987-91 average) are 5,100 total and 2,600 natural fish. Angler catch declined at an average rate of less than 1% per year between 1970 and 1991 (Fig. B-12). Hatchery fish have recently averaged 49% of the run (Chilcote et al. 1992), and average NRR is ca. 0.47. ODFW described this population as "depressed" (Nickelson et al. 1992); USFS biologists concurred with this assessment (USFS 1993a,b).

Winchuck River, OR

The Winchuck River has only winter-run steelhead. We have no historical (pre-1900s) steelhead abundance estimates specific to the Winchuck River. Recent abundance estimates were derived from angler catch estimates (ODFW 1980, 1992c, 1993b). Kenaston (1989) estimated average 1980-85 winter steelhead run-size of ca. 800 total and 400 natural fish; updated run-size estimates (1987-91 average) are 540 total and 350 natural fish. Angler catch declined at an average rate of ca. 4% per year between 1970 and 1991 (Fig. B-13). Hatchery fish have recently averaged 25-45% of the run (Chilcote et al. 1992), and average NRR is ca. 0.44-0.60. ODFW described this population as "healthy" (Nickelson et al. 1992); USFS biologists concurred with this assessment (USFS 1993a,b).

Smith River, CA

The Smith River presently has both winter- and summer-run steelhead, although the historical presence of the summer run is questionable (USFS 1993a,b). We have no historical (pre-1900s) steelhead abundance estimates specific to the Smith River. Spawning escapement was estimated to be ca. 30,000 in the early 1960s (CDFG 1965, Vol. 3(B)), although this estimate is not based on direct observations and should be viewed as approximate. Recent abundance estimates were derived from summer diver surveys (Roelofs 1983; McEwan 1992; Pisano 1992) which index only early summer-run steelhead. Summer-run survey counts increased since 1980 (Fig. B-14), although the data are limited and estimates of the rate of increase have high standard errors (Table B-1). We have insufficient information to calculate a natural return ratio for this stock. Nehlsen et al. (1991) listed summer-run steelhead in the Smith River as at "high risk of extinction." USFS biologists described the Smith River winter- run steelhead population as low, but stable, and the summer-run population as depressed, of questionable viability (USFS 1993a,b). McEwan and Jackson (in prep.) describe this population (no runs differentiated) as healthy, with fully seeded juvenile habitat.

Klamath River, CA

The Klamath River has both winter- and summer-run steelhead. We have no historical (pre-1900s) steelhead abundance estimates specific to the Klamath River Basin. Spawning escapement (excluding the Trinity River) was estimated to be ca. 171,000 (150,000 mainstem, 21,000 tributaries) in the early 1960s (CDFG 1965, Vol. 3(B)), although this estimate is not based on direct observations and should be viewed as approximate. McEwan and Jackson (in prep.) cite total run-size estimates for the 1977-78 to 1982-83 run-years ranging from 87,000 to 181,000, with an average of 129,000. For the early 1980s, Hopelain (1987) estimated that winter-run steelhead abundance was between 10,000 and 30,000. Combining these estimates suggests that early 1980s summer-run (including fall-run) abundance was ca. 99,000-119,000. Recent abundance estimates were derived from weir counts at Shasta River and Bogus Creek (Pisano 1992), returns to the Iron Gate Hatchery (Pisano 1992), and summer diver surveys (Roelofs 1983, McEwan 1992, Pisano 1992) which index only early summer-run steelhead. Summer-run survey counts have been fluctuating with an average decline of 3% per year since 1980 (Fig. B-15). Weir counts (Fig. B-16) index natural fall-run steelhead. Shasta River weir counts showed a strong decline (average 15% per year) since 1977; Bogus Creek weir counts were low, possibly with a slight decline (ca. 1% per year, but not significantly different from zero). Returns to Iron Gate Hatchery had been increasing at ca. 2% per year since 1963, but exhibited a strong decline since 1987 (Fig. B-17). Barnhart (1994) noted that recent steelhead catch rates (fish per angler-hour) showed significant downward trends. We have insufficient information to calculate a natural return ratio for these stocks. Nehlsen et al. (1991) listed summer-run steelhead in the Klamath as at "moderate risk of extinction." USFS biologists described Klamath River winter-run steelhead stocks as low and possibly declining (but with insufficient information for a clear assessment), and the summer-run stocks as depressed, with possibly reduced range, and with moderate to high risk of extinction (USFS 1993a,b). Barnhart (1994) noted that "[w]ild stocks of Klamath River steelhead may be at all time low levels ...," and he cited declining total run sizes and increasing hatchery component of the run as evidence of the problem.

Trinity River, CA

The Trinity River has both winter- and summer-run steelhead. We have no historical (pre-1900s) steelhead abundance estimates specific to the Trinity River Basin. Spawning escapement was estimated to be ca. 50,000 in the early 1960s (CDFG 1965, Vol. 3(B)), although this estimate is not based on direct observations and should be viewed as approximate. Recent abundance estimates were for total fall-run steelhead run size and angler catch above Willow Creek in the lower Trinity River (Heubach 1992), returns to the Trinity River Hatchery (Pisano 1992), and summer diver surveys (Roelofs 1983, McEwan 1992, Pisano 1992) which index only early summer-run steelhead. Summer-run survey counts have been increasing at an average rate of 8% per year since 1980 (Fig. B-18), largely due to increases in the North Fork Trinity River and the New River. For fall-run steelhead, run-size estimates above Willow Creek (Fig. B-19) showed fluctuations since 1980 between about 5,000 and 37,000 adults, averaging ca. 15,000, but data are insufficient to estimate a trend. Returns of hatchery fish were quite low (less than 1,000 in all but 2 years) from 1965 to 1985, after which they recovered for a short time before declining again after a peak of 4,800 fish in 1989; average decline was ca. 2% per year between 1958 and 1990. USFS biologists described various Trinity River winter-run steelhead stocks as stable to depressed with heavy hatchery influence in the mainstem and North Fork, and the summer-run stocks as either low but stable or unknown, except for a drastic reduction and "high risk of extinction" in the South Fork (USFS 1993a,b).


Return to Appendix B Table of Contents