THE STATUS OF STEELHEAD POPULATIONS IN CALIFORNIA IN REGARDS TO THE ENDANGERED SPECIES ACT

SPECIAL REPORT

February 1995

Submitted to
National Marine Fisheries Service
on behalf of
Association of California Water Agencies

S.P. Cramer & Associates, Inc.
300 S.E. Arrow Creek Lane
Gresham, OR 97080

AUTHORSHIP

This report was prepared as a collaborative effort between biologists that were funded by a variety of different sources. Responsibility for assembling information reported here was divided into geographic areas, and questions regarding specific streams should be directed to the appropriate Regional Coordinator. The biologists, their employers, and their regional responsibility are listed below. The first person listed under each region served as the Regional Coordinator.

Technical Lead and Writer
Steven P. Cramer, S.P. Cramer & Associates, Gresham, OR

North Coast
Aldaron Laird, Trinity Restoration Associates, Arcata, CA
Keith Barnard, Trinity Restoration Associates, Arcata, CA

North Central Coast
Douglas B. Demko, S.P. CRAMER & Associates, Chico, CA
Erick Van Dyke, S.P. CRAMER & Associates, Gresham, OR
Keith Barnard, Trinity Restoration Associates, Arcata, CA

Central Valley
William T. Mitchell, Jones & Stokes Associates, Sacramento, CA
Jeff Hagar, Consultant to East Bay Municipal Utility District, Alameda, CA
Robert C. Nuzum, East Bay Municipal Utility District, Alameda, CA
Philip A. Unger, Jones & Stokes Associates, Sacramento, CA

South Central Coast
Thomas L. Taylor, Trihey and Associates, Concord, CA
David H. Dettman, Monterey Peninsula Water Management District, Monterey, CA
Donald W. Alley, D.W. Alley & Associates, Brookdale, CA
Jerry J. Smith, Habitat Restoration Group, Felton, CA

South Coast
Jean E. Baldrige, Trihey and Associates, Concord, CA
Randall Orton, Las Virgenes Municipal Water District, Calabasas, CA
Thomas P. Keegan, Entrix, CA
Donald W. Alley, D.W. Alley & Associates, Brookdale, CA

Hatchery and Environmental Data
Douglas B. Demko, S.P. CRAMER & Associates, Chico, CA
Erick Van Dyke, S.P. CRAMER & Associates, Gresham, OR
Brian Farrell, S.P. Cramer & Associates, Woodland, CA

ACKNOWLEDGMENTS

Assembly and review of the information contained in this report required the efforts of many people not listed as authors. We thank Jennifer Persike-Becker of ACWA who invested substantial energy as administrative coordinator of this project. We thank Chip Wullbrandt and Dr. Jim Hughes, who represented ACWA members, for their input and support at coordination and technical meetings, and for their editorial review of this report. We thank Tim Leathers whose efforts at drawing the attention of ACWA members to the steelhead issue played a key role in initiating this project. Glenn Yoshioka of Jones & Stokes Associates assembled substantial data on steelhead in the Sacramento River basin.

EXECUTIVE SUMMARY

This report was prepared to assemble and synthesize the best available information on status of steelhead runs in California. Information presented in this report was prepared by a team of fisheries consultants with extensive experience throughout California. The report is intended to provide information to NMFS that will be useful in their determination as to whether to list steelhead as threatened or endangered under the federal Endangered Species Act (ESA). We often refer to steelhead in this report as "steelhead/rainbow" to reflect the fact that resident and anadromous forms can be produced by the same parents. Little attention is given in this report to streams north of the Eel River, because NMFS has already completed an internal review of those populations (although their findings have yet to be published). Information presented in this report is organized under three objectives:

1. Determine which steelhead population groupings in California qualify as Evolutionary Significant Units (ESU's)
2. Determine the extent and causes of run depression
3. Assess the risks that threaten persistence of each ESU

In general terms, NMFS has defined an ESU to be a population that meets two criteria: (1) It must be substantially reproductively isolated from other conspecific population units, and (2) it must represent an important component in the evolutionary legacy of the biological species.

IDENTIFICATION OF ESU'S

We found provisional evidence of genetic structuring of California steelhead into at least six groupings (ESU's) over broad geographic areas: (1) summer steelhead in the Rogue and Klamath rivers, (2) winter steelhead in coastal rivers from the northern border to the Gualala River, (3) spring/summer steelhead in the Eel and Mad rivers, (4) winter steelhead in coastal rivers from the Russian River to Aptos Creek near Santa Cruz, (5) winter steelhead and resident rainbow in coastal streams from the Pajaro River and south, and (6) the fall-run steelhead in the Sacramento River basin (Figure S-1). The location of boundaries dividing these groups are not definitive, and the uncertainty of boundaries is greatest on the southern portion of the coast. It appears probable that south coast populations should be further split into two groups by a dividing line somewhere in the vicinity of Point Conception, perhaps as far north as Point Buchon (near San Luis Obispo). The best evidence we could find regarding the San Joaquin Basin indicates that rainbow populations there were not and are not anadromous.

Steelhead from the inland portions of the Rogue and Klamath basins are distinguished by their unique "half-pounder" life history. Although immature half-pounder steelhead are regularly observed in neighboring coastal streams, particularly the Eel River, these fish are apparently nonspawning wanderers, because we found no evidence of adult spawning populations with previous half-pounder migrations in any basin other than the Rogue and Klamath. There are several unique aspects of the half-pounder steelhead populations of the Klamath and Rogue Basins that distinguish them as an Evolutionary Significant Unit (ESU). Their location at the northern edge of Region 2 of the California Current (Cape Blanco is the dividing point) places them in an ocean area with high productivity near shore in the spring and summer, due to upwelling, but low productivity near shore during winter, due to downwelling. The Rogue and Klamath Basins provide dependable flows during summer and fall, because they drain mountain ranges with snowpack through most of the year. Thus, near-shore rearing in the ocean during spring and early summer, and freshwater refuge during fall and winter, is uniquely available to steelhead in these streams, when considered on an evolutionary time scale. Most other coastal streams in the area frequently have sandbars that block the stream mouth for extended periods during the summer and fall. In larger streams, such as the Russian and Eel, water temperatures in the lower portions of the River reach highly stressful levels in years of drought. Juvenile life history and spawning time of Rogue and Klamath steelhead is similar to other populations up and down the coast. Gene flow into the Rogue and Klamath basins is restricted by the presence of the disease Ceratomyxa.

Winter-run steelhead in coastal rivers north of and excluding the Pajaro River (near Santa Cruz) are characterized by their ocean-migrant life history, their winter return timing, and by freshwater habitat with more consistent access to the ocean. Winter-run steelhead are predominantly 2-salt ocean migrants (2 years at sea) in north-coast streams, and the proportion of 1-salt migrants (1 year at sea) increases southward. Nearly all north and central coast streams are blocked from the ocean by sand bars during summer and fall, and juvenile rearing in the estuarine lagoon can be an important life-history strategy. Time of river entry and spawning extends over a similar time period for all of these populations, with the exception of spring/summer-run steelhead in the Middle Fork Eel River and Mad River (we consider spring/summer-run steelhead to be a separate ESU). Mixing of winter-run stocks through hatchery transfers and outplants has been extensive on the north and central coast, so there has likely been some homogenization of genetic diversity.

Data on variation in mitochondrial DNA types indicate these north and central coast populations should be further into a northern group beginning with the Gualala River at the southern edge, and central group beginning on the north with the Russian River. We found no obvious life-history differences between steelhead north and south of this line, but there is a unique oceanographic feature at this point that may limit straying of steelhead north and south of the line. A unique and pronounced "coastal jet" current carries water offshore for up to 200-300 km near Point). Because the California Current flows southward, an eddy develops just south of Point Arena, and this eddy may draw steelhead smolts north from the Gualala River (but not from the Russian River further south) into the offshore jet. The coastal eddy would account for the genetic similarity of steelhead from the Gualala River with those of north coast streams. Additionally, the Russian River Basin is lower elevation and drier than large coastal streams to the north.

The spring/summer-run steelhead of the Middle Fork Eel and upper Mad River represent a completely different strategy, with extended ocean migration and return in the spring one year prior to spawning. The return timing and state of immaturity appears to be an adaptation to ensure access to the spawning and rearing areas above falls which are only passible at high flows by fish with great leaping ability. Most coastal streams do not have the combination of passage barriers that only allow passage within a narrow range of flows, and excellent pool habitat for holding adults over summer. Because of this unique habitat and life history adaptation, we suspect that spring/summer-run steelhead from the Middle Fork Eel and upper Mad River may be appropriate for designation as a distinct ESU, in combination with other spring/summer-runs that exist in the Klamath Mountain geologic province. The passage barriers above which these fish spawn isolate them from winter-run steelhead.

Genetics and climatic data indicate a change in steelhead/rainbow populations in streams south of Santa Cruz, starting with the Pajaro River, where the climate is arid and annual rainfall is highly variable. The Pajaro River also marks the first stream beyond the southern range of coho salmon. Based on genetics data, a provisional ESU boundary appears to occur at the Pajaro River, with its population and those south having more similarity in mitochondrial DNA types among themselves than similarity with populations from the San Lorenzo River, and north. Streamflow data also indicate a change beginning at the Pajaro and Salinas rivers, where ocean access during drought periods is difficult or impossible. The Pajaro and Salinas rivers drain primarily more arid interior regions. These drainages have less surface flow at present than before turbine pumps and reservoirs, and may have had serious access problems during extended drought of drier centuries. However, the streams draining the wetter Santa Lucia Range to the south (i.e., Carmel River, Little Sur River, Big Sur River, Willow Creek, Arroyo de la Cruz, and Santa Rosa Creek) probably remained accessible for steelhead during most critically dry years. The Carmel River at the north end and Chorro Creek at the south end of the Santa Lucia presently have access problems in dry years as a result of groundwater pumping. A coastal climatic change occurs with streams south of the Santa Lucia from Point Buchon to Point Conception and beyond, where steelhead access becomes a severe problem during drought, and where more arid climate causes highly variable annual rainfall. Streamflow becomes more flashy with longer dry periods in.


The major streams south of Point Buchon consistently flow across wide, flat valleys before entering the ocean, making steelhead access to the ocean more problematic than streams flowing out of the Santa Lucia to the north. Such streams include San Luis Obispo Creek, Santa Maria River, Santa Ynez River, Ventura River, Santa Clara River, Los Angeles River, all of which have remnant anadromous steelhead populations at best. Water development has accentuated the passage difficulty in this century. Combined with these climatic, topographic, hydrologic differences that occur near Point Conception and south, the ocean current also changes. This has likely isolated ocean populations of steelhead north and south in the past. However, analysis of genetic data available to us has been insufficient to substantiate this southern most boundary.

Because of the climatic, topographic, hydrologic and oceanographic current changes that occur in southern San Luis Obispo County, another provisional ESU boundary appears appropriate in the vicinity of Point Conception, perhaps as far north as Point Buchon. Further genetics data should be gathered to substantiate this ESU boundary.

Steelhead/rainbow in many of these streams, particularly in the Pajaro south of Point Buchon, have probably had to adapt to alternating periods of resident and anadromous life histories. Long-term hydrographs and rainfall records indicate runoff is exceptionally low during sequences of multi-year drought. Analyses of tree ring spacings indicates that rainfall during 1850 to 1950 was higher than previous centuries, and that there have been many droughts of greater severity and longer duration during the last 2,000 years than the recent drought (Haston and Michaelsen 1994; Hughes and Brown 1992). During severe drought sequences, the migratory link between freshwater and the ocean may have been broken for extended periods, and populations reproduced in freshwater. This scenario applies with greater certainty to streams south of the Santa Lucia Mountains, and may not have applied to the Pajaro and Salinas basins prior to groundwater pumping by man. Resident rainbow are probably the foundation of rainbow/steelhead populations in streams south of the Santa Lucia Range, because these streams drain areas of low rainfall and flow across broad coastal plains. Steelhead/rainbow with anadromous tendencies are successful and quickly multiply in sequences of wet years. The last streams to have access problems for steelhead, and therefore most likely to receive strays from nearby blocked streams, would be those draining the Santa Lucia Range (e.g. Big Sur River and Santa Rosa Creek). We found little evidence of steelhead straying during years of adequate stream flow, and genetics data show substantial differences between proximal populations within the south coast group. Because most of the alleles analyzed are believed to be selectively neutral, genetic differences reflect isolation of breeding and random genetic drift. Thus, long-term rainfall records and genetics data indicate that drought serves to isolate many small populations.

We found numerous examples demonstrating that California steelhead have the ability to readily convert from anadromy to complete freshwater residency and vice-versa. Titus et al. (1994) found that resident rainbow occur with increasing frequency from north to south in California steelhead streams, with a coastwide average of 36% of steelhead streams supporting resident rainbow. We found several examples of resident populations that developed from steelhead after dams blocked access to the ocean. Following the recent drought, we found that anadromous smolts were produced in the Carmel River from fish that had reproduced after a complete life cycle in freshwater.

Steelhead in the Sacramento River basin have developed a unique run timing in the fall that is probably an adaptation to migratory conditions through the mainstem Sacramento River. Winter-run steelhead have been introduced, but are not native to the basin. Unlike other coastal streams in that latitude, access to the Sacramento River is maintained throughout the year. However, water temperatures in the lower river during mid summer were likely to be lethal to salmonids in many years. Major flooding was a frequent occurrence during winter. Thus, it appears that fall (October-November) offered the best window for successful upstream migration. The additional 2 months of ocean rearing, compared to Klamath half-pounders returning in mid August, results in 1-salt returning fish being 2-3 inches longer than half-pounders, but 2-3 inches shorter than winter-run 1-salt migrants. One-salt fish are mature and compose about half of the returns. Returning steelhead in the Sacramento River are deeper bodied, at a given length, than winter steelhead. A notable portion of the offspring mature in freshwater without going to sea. Gene flow into the Sacramento Basin is restricted by the presence of the disease Ceratomyxa.

ABUNDANCE AND SURVIVAL TRENDS

We examined several indices of ocean survival of steelhead, including dam counts, hatchery returns, angler catch, diver counts, and juvenile densities. With a few exceptions, there has been no consistent effort to monitor steelhead populations in the state. We found three populations with long term abundance records continuing to present; the upper Sacramento River above Red Bluff Diversion Dam, the upper Eel River above Cape Horn Dam (Van Arsdale Reservoir), and the Middle Fork Eel River. There were additional time series of adult counts at dams, but where fish have not been counted in recent years. There were no systematic data on harvest by anglers. Age of returning adults was not monitored, so steelhead counts were only sufficient to calculate rough indices of combined ocean and freshwater survival. The abundance of steelhead in the upper Sacramento River has been declining since the early 1970's, and in the upper Eel River since the late 1950's. In contrast, diver counts of spring/summer-run steelhead in the Middle Fork Eel River did not decline until after 1987. Much of the Middle Fork Eel River is in wilderness area, consumptive harvest is illegal, and there has been no stocking of hatchery fish.

Warm Springs Hatchery on the Russian River was the only hatchery at which releases of juveniles and returns of adults were managed in a way that smolt-to-adult survival could be indexed. Data there did not begin until the 1981 brood year, and return rate was substantially higher for that brood than any subsequent broods. The pattern of decreased return rate since the 1981 brood is similar to that for most hatchery steelhead in Oregon and Washington.

Data on juvenile steelhead densities were available in recent years for a number of streams and were useful for comparing present rearing densities to those in streams seeded to capacity. We limited our analysis to data that were collected by multiple pass electrofishing and by systematic sampling of the stream area. The length of time series for which we found data varied between regions, but we found no indication of trends in average rearing densities in any region over the last 15-20 years.

We averaged data from 1990 through 1994 for each stream from which we had data to examine regional trends under recent conditions. This comparison showed that juvenile rearing densities were generally less in streams further south. Most of the central coast streams sampled were in the Monterey Bay area (36-37N), where juvenile steelhead/rainbow densities ranged widely. The farthest south stream was one of the highest quality tributaries of the Santa Ynez River, so other streams in that basin probably had lesser densities. Data from Burns (1971) in streams between Crescent City and Fort Bragg suggest that carrying capacity for small coastal streams is between 0.5 and 1.0 fish/m2 for all ages combined, so rearing densities in the north coast streams sampled appear healthy.

Titus et al. (1994) assembled anecdotal and quantitative data from a more comprehensive set of steelhead streams south of San Francisco Bay. Titus et al. (1994) found that 14% of steelhead populations had not changed noticeably from historical levels, 31% were reduced compared to historical levels, and 24% were extinct. Titus et al. found further that the proportion of extirpated populations increased southward, while the proportion of populations that were unchanged from historical levels increased northward. There were almost no remaining anadromous populations at the southern end of the range (Orange and San Diego counties), but all anadromous populations remain north of Monterey Bay. The north-south trend found by Titus et al. (1994) for increasing health of steelhead populations northward towards San Francisco Bay is consistent with the trend we found with juvenile steelhead/rainbow densities.

FACTORS DRIVING POPULATION TRENDS

We found that survival trends of California steelhead were correlated to physical and biological changes in the ocean. Oceanographic factors that contributed significantly to the multiple regression model for smolt-to-adult survival of North Umpqua summer steelhead were spring upwelling and spring sea level in the year of return, and winter upwelling, spring sea level, winter surface temperature, and wind speed in the year prior to return. Biologically plausible explanations for each correlation were found. Abundance of upper Sacramento steelhead was correlated to catches of each of the pelagic fishes examined (Pacific sardine, northern anchovy, Pacific mackerel, jack mackerel, Pacific herring, and market squid), with correlations ranging from 0.41 to 0.74. Similarly, diver counts of steelhead on the Middle Fork Eel River correlated negatively to catches of sardine and squid (-0.33 to -0.51). Trends in the upper Eel were not correlated to pelagic fish catches, but this run has been supplemented with up to 200,000 hatchery smolts in some years and none in others. We do not imply that these are cause and effect correlations, but they reflect similar trends in fishes exposed to the same ocean environment. We did not include physical oceanographic factors in the regression analysis for California steelhead, because we were unable to obtain data from the appropriate latitudes in time for the analysis. The decline in the returns per smolt released from Warm Springs Hatchery (Russian River), similar to declines in steelhead return rates to Oregon and Washington hatcheries, indicates that ocean survival of steelhead has been low during the last decade.

We found one example of direct evidence that El Ni–o events affect growth of steelhead in south-central California. Shapovalov and Taft (1954) show unusually small size at return in 1941-42 for both coho and steelhead. This corresponds with an El Ni–o event, for which the southern oscillation index indicates was about equal to the strength of El Ni–o in 1986, but less than 1983 (Fisher 1994). We found no means of estimating brood survival for those years.

Migration barriers and availability of water have been obvious freshwater factors affecting fish runs. Migration to a large portion of the spawning and rearing habitat, particularly in the central and southern portions of the state, have been blocked by dams. Availability of water for migration and rearing was a limiting factor in many steelhead producing streams even under pristine conditions, and water diversions for human uses has substantially aggravated the shortage of water in many streams. Habitat damage from logging and floods has also been well documented, particularly in north coast streams.

Drought has played a major role in the present reduced abundance of steelhead/rainbow throughout California, particularly in southern California. The primary mechanisms by which drought impacts steelhead/rainbow are reduced rearing area, reduced productivity of remaining freshwater habitats, reduced productivity of estuarine lagoons, and blocked migration. Examples of each of these impacts are presented.

We found that angler harvest in trout fisheries and in steelhead fisheries may have contributed to steelhead declines in some streams. Because of reduced ocean survival, the harvest rate that steelhead populations can sustain is much reduced. In streams where trout fishing is permitted, numerous studies have demonstrated that anglers capture a high proportion of the juvenile steelhead that are age 1 and older. The impacts of trout angling have likely been greatest in the southern part of the state, where Titus et al. (1994) found that 80% of streams in Ventura and Los Angeles counties were stocked with catchable trout. Evidence indicates that harvest rates on adult steelhead are highly variable and depend on environmental circumstances. Most of the estimated harvest rates we found are sufficiently high to damage steelhead populations during years when ocean and freshwater survival are low.

Hatchery practices have probably contributed to loss of genetic fitness in many streams, particularly on the north coast and in the Sacramento Basin. Steelhead from a select portion of the run have been extensively outplanted in the Sacramento basin and survival rates appear to be quite low. Timing of migration and spawning at Mad River Hatchery have also emphasized a restricted portion of the run, and large surpluses of fry have been dumped in the Mad and Russian rivers.

We found many streams where efforts have begun to restore steelhead and their habitat. The most notable efforts were in the Sacramento River and the Carmel River. Planning and study efforts to preserve steelhead populations are underway on many other streams, particularly on the south and central coast.

EVALUATION OF RISKS THREATENING PERSISTENCE

The flexible life history of steelhead/rainbow has allowed this species to persist in coastal streams further south than other anadromous salmonids. The cyclic oscillations evident in long term records of flow, rainfall, and abundance of ocean pelagic fishes indicate it is likely the southern range of this species has periodically expanded and contracted. The present contraction has been substantially aggravated by human activities, but the record of recurrent droughts indicates this is probably not the first time that anadromous rainbow have disappeared from several streams in southern California. The substantial genetic diversity found between nearby populations of rainbow suggests that they have been periodically isolated in small groups, as would occur during extended years of drought when anadromy would be impossible.

Given that anadromy of steelhead/rainbow in many streams from the Santa Cruz to the Mexican border has been periodically interrupted by drought over at least the last 2,000 years, we conclude that anadromy is not a necessary trait to enable persistence of steelhead/rainbow in these streams. We deduce that anadromy may only have been an uninterrupted trait in the larger streams with the most reliable perennial flow. These would be streams flowing out of the Santa Lucia Mountains (Carmel River, Little Sur River, Big Sur River, Arroyo de la Cruz, and Santa Rosa Creek) and Corralitos Creek flowing out of the Santa Cruz Mountains into the lower Pajaro River. Anadromous steelhead may also have been reestablished following drought periods from rainbow populations resident in perennial sections of temporarily intermittent streams, such as are found in the upper Santa Ynez River, Ventura River, Santa Clara River, and Malibu Creek. However, because anadromy in these streams was likely to be interrupted by drought, even before human impacts, temporary loss of access to the ocean in such streams should not trigger listing of steelhead as threatened or endangered under the ESA. Loss of access to the ocean should only be regarded as an unusual threat in the south coast ESU when it occurs in the perennial streams listed above.

The many streams in which steelhead/rainbow reproduce and rear is in itself a strong protection against extinction. Therefore, one measure of risk of extinction is the number of streams in which steelhead/rainbow rear and still have suitable habitat. As long as a diversity of such streams is maintained, the species as a whole has little risk of extinction. The more confined the distribution of fish, the greater likelihood that a given environmental sequence or natural disaster might cause extinction. Risk of extinction would be evaluated best by incorporating spatial dispersion into a cause-and-effect model of population trends, but data for such a model are not available.

Given that time-series data sets on steelhead abundance are few, and that a variety of different factors may limit steelhead production in different streams, we believe that risk of extinction should be rated against the probability that good habitats will be maintained in a diversity of areas, and that harvest and hatcheries will be used properly. Habitat modification, harvest, and hatcheries are factors under human control; the ability of the populations to persistent under extreme environmental variation in the natural habitat has already been demonstrated. Therefore, we compiled a partial list of California steelhead streams in which most habitat is still good to excellent, and in which juvenile steelhead populations are healthy. Some of these streams are in nearly pristine condition. Most have been altered only slightly by man. We also listed streams which continue to produce important steelhead populations in spite of reductions in habitat. In addition to streams listed in Table S-1, there are many headwater streams in good condition which support abundant rainbow populations of steelhead ancestry, but which no longer have consistent access to the ocean. It is likely that a NMFS policy decision will be required to determine whether or not steelhead/rainbow populations should be included in a steelhead ESU after their migratory link to the ocean has been permanently severed by man's actions.

Table S-1.

Partial listing of steelhead streams in California which still have good to excellent habitat and support persistent populations of naturally reproducing steelhead. Some streams are also listed which continue to produce important steelhead populations in spite of reductions in habitat. On the north coast, only the streams with at least 10 miles of good to excellent habitat in use are listed. The adult return supportable was assigned, based on professional judgement, to the following categories: <50, 50-200, 200-500, 500-1000, and 1000+, unless specific estimates were available.

The number of streams with good to excellent steelhead habitat and with open access to the ocean is greater on the north coast than on the south coast. On the north coast, we only attempted to list streams with at least 10 miles of good to excellent steelhead habitat and with healthy populations of juvenile steelhead. Many smaller tributaries could also be listed.

Risk of extinction in a given ESU, as determined from the availability of good habitat, should be ranked with the following factors in mind:

Because a thorough status review of coho has just been completed, it is useful to evaluate risks by comparison to findings with coho. In comparison to coho, steelhead remain in many more streams, have a more diverse life history, spawn over a greater span of time, and can survive and reproduce in freshwater. Coho die after spawning, but steelhead can survive to spawn several times. Multiple freshwater ages at smolting offers greater mechanisms for density compensation, and this is demonstrated in the asymptotic stock recruitment curve typical of steelhead. Shapovalov and Taft (1954) intensively studied the two species side-by-side in Waddell Creek, and drew the following contrast, "Unlike silver salmon, steelhead migrate to sea at various ages and over a long period within a season, spend varying amounts of time in the ocean and return over a fairly long period within a season, are capable of spawning more than once, sometimes spawn before their first journey to sea, and may even remain in fresh water for their entire lives." Given these differences between coho and steelhead, it is clear that the present risk of extinction to steelhead is substantially less than that for coho where the two species co-occur.