VII. Evaluating the Effects to Sacramento Winter-Run Chinook

A. Biological Requirements

For this consultation, the biological requirements of winter-run chinook can be best described in terms of the population sizes needed to ensure the continued survival and recovery of the population.

Survival Requirements. Over the short-term, the survival of winter-run chinook will largely depend on its ability to retain sufficient abundances that enable the population to persist in the face of random events that could drive it to extinction. Chance events operate at several levels that affect the likelihood of extinction, including demographic, environmental and genetic stochasticities (Shaffer 1981).

When populations become small, there is concern that qualitative changes in population dynamics can take place which make the populations more susceptible to extinction and less able to recover. One example is a decline in the per capita reproductive success at very small population sizes, which is variously known as depensation, an Allee effect, and inverse density dependence. Average productivity may decline due to a skewed sex ratio, and from decreasing spatial and temporal overlap between male and female spawners. Myers et al. (1995) found in a survey of 128 fish stocks of various species that significant depensation occurred in Pacific salmon stocks which had been driven to extremely low levels by fishing and habitat loss. From this survey, Myers et al. (1995) suggested that depensation may occur in salmonids at population abundances of less than 100 females. Such depensatory dynamics in a population, like winter-run chinook, where abundance has been severely reduced may preclude the population from recovering even when mortality is reduced (Neave 1954; Myers et al. 1995).

Environmental stochasticity usually refers to unpredictable events, such as changes in the weather, food supply, and the populations of competitors, predators and parasites. Some of the largest, past examples of recurrent environmental variability include El Niño events and prolonged drought conditions. El Niño events have likely significantly contributed to an order-of-magnitude decline in the winter-run chinook population (such as with the 1981 broodyear). Prolonged drought conditions, combined with inadequate water management, have also resulted in extreme losses of eggs and larvae, resulting in order of magnitude declines in the winter-run chinook population (1976, 1977, 1986, 1987, and 1988 broodyears). The current very small population probably does not have the buffering capacity to withstand another large mortality event, and would be highly susceptible to extinction in the face of the next El Niño or protracted drought conditions.

Winter-run chinook may be particularly vulnerable to environmental stochasticities because of its narrow age composition. Winter-run chinook return to spawn primarily as three-year olds with few four-year old spawners (Hallock and Fisher 1985), such that the population is very dependent upon this one dominant age-group (Botsford and Brittnacher 1995). This concentration of spawning at age three likely leads to higher interannual fluctuations in abundance, and greater risk to the loss of a brood-cycle. In the event that a environmental stochasticity eliminated or severely reduced a single year-class, the winter-run population would have only a few four-year old spawners to contribute towards rebuilding a severely depleted year-class. In contrast, a chinook population with a broader composition of age-classes would have a larger portion of four-and five-year olds, or three-and five-year olds, to contribute towards rebuilding a depleted year-class and restoration would likely occur more quickly.

Winter-run chinook is also more susceptible to extinction because the run is limited to a single, isolated, undivided spawning population, and does not have multiple subpopulations. Species composed of many subpopulations are likely more resilient to extinction because reservoir subpopulations can repopulate other depleted subpopulations. The winter-run chinook population, however, is isolated receiving no immigrants from other subpopulations, and therefore, it has a very limited ability to withstand random environmental events.

Genetic risks include the loss of genetic variation in a population, which results in decreased fitness through random genetic drift (Ewens et al. 1987). A population remains viable when it maintains sufficient genetic variation for evolutionary adaptation to a changing environment. The genetically effectivepopulation size conveys information about expected rates of inbreeding and genetic drift, which can affect fitness and adaptive potential (Hedrick and Miller 1992).

Several minimum effective population sizes have been proposed as general recommendations for species to maintain population numbers and genetic variation (Franklin 1980; Land and Barrowclough 1987). An effective population size (including males and females) of 50 has been prescribed to prevent inbreeding depression (Franklin 1980), which is viewed as the main threat to short-term survival (Shafer 1987). An effective population size of 500 has been recommended to avoid long-term genetic losses (Franklin 1980; Land and Barrowclough 1987), which is considered the primary threat to the loss of genetic variation essential to continuing adaptation (Shaffer 1987). While these are merely 'rules of thumb' and the necessary sizes will vary from species to species, it has been strongly recommended that effective population sizes of at least hundreds of individuals be maintained to preserve evolutionarily important amounts of genetic variation (Lande and Barrowclough 1987).

NMFS has identified a threshold escapement level of 300 natural spawners annually for larger subpopulations of Snake River spring/summer chinook and for fall chinook (NMFS 1995b). This threshold level represents the escapement at which uncertainties about processes or population enumerations are likely to become significant, and at which qualitative changes in processes are likely to occur (BRWG 1994). That is, if the population drops below the threshold, significantly more uncertain, uncontrollable, irreversible or deleterious effects or processes take hold.

For winter-run chinook, NMFS believes that a somewhat higher threshold escapement level than that of Snake River chinook is appropriate. A higher number is considered necessary due to the unique demographic, biological and genetic characteristics of the winter-run chinook population. As discussed, winter-run chinook have a narrow age structure with most spawning adults returning at age three, few as four-year olds and none as five-year olds. In contrast, Snake River chinook populations return as three, four and five year olds and thus, have a greater overlap between year-classes to help buffer the impact on a single brood-year from a potential large mortality event. Also, winter-run chinook is a single, isolated, undivided population which spawns only in the mainstem Sacramento River, and lacks immigration from anyreservoir subpopulations. In contrast, Snake River spring/summer chinook salmon have 39 subpopulations. Snake River fall chinook is somewhat comparable to winter-run chinook having one main spawning population below Hells Canyon Dam, but Snake River fall chinook also spawn in the lower reaches of the Imnaha, Grande Ronde, Clearwater, Salmon and Tucannon rivers. Without the potential reservoir of spawning adults from other tributary streams, winter-run chinook may be less resilient in their ability to withstand a substantial environmental event. Hence, the demographic features of winter-run chinook are considered to make the population particularly vulnerable to the potential environmental stochasticities that could drive the population to extinction.

Also, winter-run chinook have a lower fecundity (average of 3,353 eggs per female) than most other chinook populations, including Sacramento fall-run chinook (average of 5,498 eggs per female) (Fisher 1994), Columbia River chinook (average of 5,032 - 5,453 eggs per female), and other chinook populations from the Klamath River north through Alaska (ranging from 3,634-10,622 average number of eggs per female but most with an average of 5,000 eggs per female and higher) (Healey and Heard 1984). Winter-run chinook salmon therefore have a lower reproductive potential than most other chinook salmon, and probably are less capable of sustaining a similar mortality level, whether from in-river or ocean conditions.

In addition, consideration of genetic variation is important towards determining an appropriate threshold escapement level for winter-run chinook. An genetically effective population size that is adequate for maintaining evolutionary potential has not been quantitatively determined for winter-run chinook, but the relatively conservative value of 500 is recommended to guard against further loss of genetic variation. To calculate the census population size needed to achieve this effective population size, the effective population size is divided by the ratio of the effective population size to census size (Nb:N). The Nb:N ratio for winter-run chinook has been estimated to range between 0.1 and 0.333 (Hedrick et al. 1995). Also, the effective population size of 500 is the value for the total spawning population, which includes the sum of all three of the main year classes. Accordingly, the census size of the total winter-run chinook spawning population to achieve an effective population size of 500 is estimated to range between 1,515 and 5,000 spawning adults. This equates to an annual escapement between about 500 - 1,660 spawning adults.

Considering these various demographic, biological and genetic features, the provisional threshold escapement for winter-run chinook is no fewer than 500 spawning adults annually. At this level, we anticipate that winter-run chinook will not decline further or drop to zero in the short term while conditions arebeing changed to recover the population. This threshold level should also avoid any population condition that would make recovery impossible or significantly more difficult to achieve.

Recovery Requirements. Demographic or extinction modeling has been conducted to determine the winter-run chinook population abundance that is needed to ensure a low probability of extinction (Botsford and Brittnacher 1995). The model results suggest that mean annual spawning abundances of 10,000 females over any 13 consecutive years would reduce the probability of the population going extinct within 50 years to less than 0.1. These provisional recommendations are being considered by the Winter-run Chinook Recovery Team as the recovery goal for the winter-run chinook population.

B. Environmental Baseline

The winter-run chinook salmon population has remained very low since the 1989 emergency listing. Escapement has steadily declined from 533 spawning adults in 1989, to 441 and 191 adults in 1990 and 1991, respectively. In 1992, the population rebounded to 1180 adults, but declined again in 1993 and 1994 to 341 and 189, respectively, before rebounding again in 1995 to 1361 adults. Hence, four of the seven year classes since listing were below the threshold escapement level of 500. Also, assuming a 1:1 sex ratio, the 1991 and 1994 year-classes fell below 100 females, the level where depensatory effects may occur.

To evaluate whether these population abundances represent an increasing or decreasing trend, the cohort survival of several year classes can be examined based on the winter-run chinook population’s age structure. Winter-run chinook return to spawn primarily as two- and three-year olds (25% return as two-year olds, 67% as three-year olds, and 8% as four year olds) (Hallock and Fisher 1985). To evaluate cohort survival, the proportion of spawning females in the population can be tracked, by assuming that there are sufficient spawning males to fertilize the eggs of all the females. To estimate the proportion of females in the winter-run chinook population, two assumptions are made: 1) the sex ratio is 1:1, and 2) returning two-year olds are all males. Accordingly, the proportion of females returning in the population as two, three, and four-year olds is estimated as 0, 0.89 and 0.11, respectively (while the proportion of males as two, three and four-year olds is 0.50, 0.44 and 0.06, respectively). Thus, the cohort survival of winter-run chinook can be examined by tracking the number of spawning females from a parental cohort that return mainly as three-year olds (89%) plus a small fraction returning as four-year old (11%).

As such, the cohort survival can be represented as a cohort replacement rate (CRR), or the ratio between the number of spawning adults in one generation to the number of spawningadults in the next generation. For the years since listing, approximate calculations of CRR for the 1989, 1990, and 1991 year classes are 2.04, 0.74, and 1.66, respectively (Table 1). The geometric mean of the CRRs for 1989-1991 broodyears is 1.36, suggesting an overall increase in the population since listing. It is important to note that this mean estimate is very uncertain because it is based on only three samples. A CRR cannot be calculated for the 1992 year class until the number of four year olds returning in 1996 is estimated. Considering the broodyears over the past eight years, it appears that the steep downward trend observed in the population before listing may be stabilized.

Table 1. Estimates of winter-run chinook escapement and corresponding cohort replacement rates.

This increasing trend, however, may be primarily a function of one relatively strong year class, which begins in 1989 (533 adults), increases in 1992 (1180 adults), and continues to increase in 1995 (1361 adults). This one relatively strong year class is probably influencing the apparent trend of a positive population growth rate.

It is also important to evaluate the accuracy in these run estimates. Prior to 1986, the entire winter-run chinook population was monitored during the course of their upmigration past RBDD. Beginning in 1986, the gates at RBDD have been raised for various time periods during their migration to enable freer passage to spawning grounds. Since 1990, the gates have been raised for up to 85% of the winter-run upmigration period, such that about 15% of the run has been monitored rather than theentire run. This monitoring level equates to a sampling precision with a variance of 1.0 (in logarithms), such that the ratio of estimated to actual values varies between 0.36 and 2.72 (±1 standard deviation)(L. Botsford, pers. comm.). For example, the 1994 year class had an escapement estimate of 189 spawning adults, but the accuracy of this estimate is fairly low, such that the actual run-size may have varied between 68 and 514 adults.

The status of the winter-run chinook population can also be evaluated from a genetic perspective. Bounds on the genetically effective population sizes for winter-run chinook have been determined for 1991, 1992, and 1993 year classes with minimum estimates of 21.9, 127.3, and 39.0 and maximum estimates of 61.6, 401.0 and 108.6, respectively (Hedrick et al. 1995). The total effective population size for these three runs is approximated by summing the yearly estimates, such that the minimum estimates of the effective population size is 188.2 and the maximum estimate is 571.2. Both estimates are above the limit of 50 proposed to avoid inbreeding depression, but the minimum estimate is below the limit of 500 recommended to retain long-term adaptive variation. This suggests that the winter-run chinook population may have already been genetically affected by low abundance, which may affect the population’s ability to recover.

Considering these trends and the status of winter-run chinook under the environmental baseline, the likelihood of the survival and recovery of winter-run chinook can be mainly qualitatively assessed. Based on the historical record of CRRs, the probability of winter-run chinook going extinct within 50 years is 1.0, i.e. extinction is imminent (Botsford and Brittnacher 1995). However, since the listing, many important actions have been taken to improve freshwater habitat conditions for winter-run chinook (Table 2). Many of these improvements may have contributed to increased survival and the increasing population trend observed in recent years. Ongoing improvements are expected to further improve survival, but the incremental benefit of these actions is not yet known. Yet, even with these improvements to freshwater habitat, current conditions probably do not represent a healthy environment. Some of the remaining problems include: discharges of heavy metals from the Superfund Iron Mountain Mine site; passage problems at the Anderson-Cottonwood Irrigation District’s dam and the Red Bluff Diversion Dam for portions of the migrant population; entrainment from thousands of unscreened or inadequately screened diversions; poor rearing conditions in the river due to loss of riparian habitat and poor water quality conditions; poor rearing conditions in the Bay-Delta due to water project operations, reduced primary and secondary productivity, and the loss of riparian and wetland habitat. These continuing habitat problems likely limit the population’s ability to recover.

To consider extinction risks under the environmental baseline, we also need to evaluate what the population’s future status could be without harvest. The returning run-sizes of winter-run chinook cannot be projected for the future under the environmental baseline. However, it is possible to generally comment on the potential for increased escapement without harvest impacts. The anticipated level of harvest impacts in the near future, based on the proposed action, is estimated as a harvest fraction (catch/catch + escapement) of about 0.5 (see Section VII.C.). Without this substantial source of mortality, the likelihood of survival would likely improve substantially for winter-run chinook. In the absence of any harvest, it is expected that the very weak year classes (1993 and 1994) would not immediately increase above the threshold level, but would over the course of up to two generations.

Although escapement is expected to improve under the baseline conditions (i.e. without harvest), the likelihood of survival and recovery is still considered low for two main reasons. One, the two weak year classes are expected to remain critically small for several more generations. While these year classes remain below the threshold level, the potential is higher for uncertain, uncontrollable, irreversible or deleterious effects or processes to occur. Second, until the freshwater and estuarine habitats are improved to a healthy condition, in-river survival of winter-run chinook will be limited, and the likelihood of recovery will be tenuous. Table 2. Actions taken to improve freshwater habitat for winter-run chinook salmon; the year that these actions were implemented and should have improved survival during spawning and rearing; and the year that those actions are expected to contribute to increasing escapement.

C. Effects of the Proposed Action

The NMFS proposes to continue implementation of the FMP, as constrained by the 1991 Winter-run Chinook Biological Opinion on Pacific Salmon Ocean Harvest (NMFS 1991). The FMP seeks to achieve specific spawning escapement goals for Central Valley fall-run chinook and Klamath-Trinity river fall-run chinook. The FMP has not established a management goal for winter-run chinook, and therefore, does not directly require that this listed stock be considered in determining maximum fishing rates. The indicator stock for the Central Valley is the Sacramento River fall-run chinook, for which the FMP specifies a spawning escapement goal of between 122,000 and 180,000 combined hatchery and natural adults.

The method used to project annual escapement of Central Valley fall-run chinook has been the Central Valley Index (CVI), which is the annual ocean fishery landings south of Point Arena plus the spawning escapement of adult Central Valley stocks in the same year. Harvest impact on Central Valley stocks is estimated by the Central Valley ocean exploitation index which is the landings south of Point Arena divided by the CVI. The CVI and the ocean exploitation index are recognized as only crude approximations of actual abundance and harvest (NMFS 1996). A considerable level of uncertainty is also associated with pre-season estimates of the CVI and the CV ocean exploitation index (NMFS 1996a).

The 1991 Harvest Biological Opinion concluded that the ocean salmon fishery would not jeopardize winter-run chinook at the impact level observed in the 1990 fisheries, and recommended that management measures be refined to continue to reduce harvest rates. Since different chinook stocks cannot be distinguished in the ocean, it is not possible to directly measure harvest rates specifically on wild winter-run chinook population. The CV ocean exploitation index has been the best and only available tool for estimating relative impacts on winter-run chinook. Accordingly, the primary constraint on the ocean salmon fishery for the protection of winter-run chinook has been to cap the CV ocean exploitation index at the 1990 fisheries level (0.79). In addition, the Incidental Take Statement required that: 1) the ocean recreational fishing season south of Point Arena be reduced by two weeks at the beginning and the end of the normal fishing season, and 2) the commercial fishery not be opened until May 1st.

The recent recoveries of coded-wire tagged (CWT) fish in the ocean and river have provided data to reexamine the impact of ocean harvest on winter-run chinook. The CWT data indicate that the harvest fraction (catch/catch + escapement ratio) on winter-run chinook was 0.54 for the broodyear 1992 (NMFS 1996). The NMFS Biological Assessment indicates that this harvest fractionwas estimated based on relatively limited data due to the small size of the juvenile release group. Nonetheless, the recovery of tagged winter-run chinook both verifies the incidence of harvest and provides a rough approximation of present ocean impacts.

The estimated harvest fraction from the 1992 broodyear compares well to previous harvest estimates from an earlier fin-clip marking study (Hallock and Reisenbichler 1980). This study estimated harvest fractions of 0.47 and 0.56 for the 1969 and 1970 broodyears, respectively (NMFS 1996). Thus, the harvest fraction estimate from the recent CWT data are consistent and within the range of estimates based on the earlier fin clip data. This suggests that the present incidental harvest impact on winter-run chinook has not changed from catch levels 20 years ago, and that contrary to expectations, incidental harvest impacts were not reduced by restrictions imposed by the 1991 Biological Opinion on ocean harvest.

The assessment in the 1991 Biological Opinion determined that harvest levels were acceptable because they were below levels sustained by other chinook stocks, namely California Central Valley chinook salmon south of Point Arena and Washington State chinook fisheries (NMFS 1991). However, since the 1991 opinion was issued, the winter-run chinook population has shown critically low spawning abundances and correspondingly, has changed in its listing status from threatened to endangered. Hence, it is necessary to reassess whether a harvest fraction in the range of 0.5 is acceptable for the winter-run chinook population at its present level.

The abundance data indicate that the present winter-run chinook population is critically low, but due to one relatively strong year class, the population appears to be growing. Can this very small but growing population sustain a 0.5 harvest fraction without being significantly impaired in its ability to recover and without substantial risk of extinction? There are several possible approaches to examining this question: one is assessing whether these harvest levels are at a sustainable level, and another is examining the risks of extinction to the population.

Sustainable Harvest Levels

Put simply, a population is harvested at a sustainable level when fish are not harvested at a rate faster than they can be replaced in the population through reproduction. That is, the natural and harvest losses in a population are balanced by recruitment. Since data are so sparse on spawner-recruit relations of winter-run chinook, it is difficult to explicitly evaluate whether this balance is being achieved. However, Riesenbichler (1987) determined that harvest fraction goals (catch/catch + escapement) of 0.60-0.70 result in maximum recruitment for the sustainable harvest of chinook salmon stocks that have an a (the productivity parameter for the Ricker model) of 7-9 and unknown spawner-recruit relations. Riesenbichler's drew these conclusions by evaluating chinook salmon stocks from British Columbia to northern California, including Central Valley fall-run chinook salmon. Thus, the winter-run chinook's harvest fraction of 0.5 is below Reisenbichler's suggested goal for sustainable harvest on chinook salmon.

However, the productivity parameter for winter-run chinook, although not known, is likely below an a of 7 for three reasons. First, the fecundity of winter-run chinook is lower than most other chinook salmon populations, including Sacramento fall chinook and chinook populations from Klamath River north through Alaska (Fisher, pers. comm.; Fisher 1994; Healey and Heard 1984). Second, freshwater survival in the winter-run chinook population although improved is still considered low, and further habitat improvements are needed to attain a healthy, productive spawning and rearing environment. Lastly, several year classes of winter-run chinook have fallen below 100 females, which is the suggested level at which depensation occurs (Myers et al. 1995).

If winter-run chinook are less productive than other chinook salmon stocks, they are less resilient to the suggested harvest-fraction goals of 0.60-0.70. If the a for winter-run chinook is sufficiently low (below 7), then the maximum recruitment level for sustainable harvest of winter-run chinook may be below the 0.5 harvest fraction, suggesting overexploitation at the present harvest level.

Extinction Risks

Another way to examine the impact of the 0.5 harvest-fraction is asking whether the population can sustain this continued level of harvest with an acceptably low risk of extinction. At the presently small population sizes, this seems highly unlikely.

The risks of extinction to the winter-run population become evident upon examining the three main year classes which presently make up the population. Escapements in the 1993 and 1994 year classes (341 and 189, respectively) were both below the threshold escapement level of 500. With a continued harvest fraction of 0.5, these two year classes are expected to remain below the threshold level for up to six more generations (about 18 years), assuming the same population growth rate observed since listing (1989). If by chance, the one relatively strong 1995 year class experiences some large source of mortality, all three year classes would become depressed. Under these circumstances, the entire spawning population could be vulnerableto depensatory and deleterious genetic effects, placing the population at substantial risk of extinction.

Since ocean harvest is removing an estimated 0.5 fraction of the population in the ocean, harvest is limiting population growth and thereby, impeding recovery--but to what extent? The effects of harvest on population growth and recovery can be illustrated by comparing the relative increases in escapement and population growth expected with reductions in the fishery.

Using the California Department of Fish and Game’s (CDFG) Winter-run Chinook Ocean Salmon Harvest Model, we can evaluate the incremental increases to escapement as incidental harvest impacts are reduced (CDFG 1989). The CDFG ocean harvest model was developed as a tool for evaluating the impacts of ocean fishery regulation options on the winter-run chinook salmon and the ocean salmon fisheries in general. This model is best used for comparing the relative impacts of regulation options rather than actual impacts, because of the difficulty in projecting ocean abundance of California chinook salmon stocks (CDFG 1989). The input variables to the model include: ocean natural mortality, which is applied on all age classes and between fishing seasons; shaker mortality; initial stock sizes; minimum size limits; effort response; and fishing contact rates. Assuming a 20% natural mortality rate which is the generally accepted natural mortality rate for salmon (A. Baracco, pers. comm.), the model predicts that eliminating harvest impacts would increase escapement by 84% (Table 3). A 70%, 50% and 30% reduction in incidental harvest of winter-run chinook would increase escapement by 52%, 35%, and 19%, respectively, again assuming a 20% natural mortality rate.

We can also examine how these predicted escapement levels may vary with different ocean conditions, by assuming low and high natural mortality levels in the model (Table 3). Assuming relatively low natural mortality in the ocean, the model predicts that elimination harvest impacts would increase escapement by 103%, while a 50% reduction in harvest would increase escapement by 41%. Conversely, by assuming a relatively high natural mortality, the model predicts that eliminating harvest impacts would increase spawner escapement by 48%, and a 50% reduction in harvest would increase escapement by 22%.

Table 3. Projected increases in escapement of winter-run chinook at various levels of harvest reduction (horizontal axis), and at three levels of natural mortality (low, moderate, and high, on vertical axis), as modeled by the Winter-run Chinook Salmon Ocean Harvest Model (CDFG 1989, A. Baracco, pers. comm.). Reduction in model cells applied equally throughout the recreational and commercial fishing season.

The effects of harvest on population growth can be also examined by evaluating the changes in CRR that could have occurred if the incidental impacts of the fishery had been reduced in the 1989 -1991 broodyears. Estimates of these changes are presented in Table 4, which were calculated: 1) by using the projected escapement increases modeled by the CDFG Winter-run Chinook Ocean Salmon Harvest Model (Table 3), and 2) by using the observed escapement estimates from 1989 to 1995.

Based on this evaluation, the estimate of CRR suggests that the population would have grown substantially faster by eliminating harvest impacts, specifically from the observed geometric mean CRR of 1.36 to a predicted CRR of 2.50 (Table 4). With a 50% reduction in harvest impacts, the population would have grown somewhat slower but still reasonably fast, at an estimated geometric mean CRR of 1.83. A 30% reduction in harvest also suggests a relatively high population growth rate, with an estimated geometric mean CRR of 1.63. Considering the variability in the data, however, estimates of population growth with a 30% harvest reduction do not indicate a growing population throughout the distribution of the means (CRR at one standard deviation below the mean is less than one).

These predictions are based on limited data, and should only be considered as relative measures of potential population growth as harvest is reduced. Nonetheless, they suggest that incidental harvest impacts are substantially limiting population growth rate, and thus impeding winter-run chinook from rapidly growing to an abundance level where it would be buffered against harmful chance events that could drive it to extinction.

Table 4. The geometric mean and standard deviations for the cohort replacement rate (CRR) for the broodyears 1989 to 1991 (n=3), including: observed estimates, predicted estimates with thirty, fifty and seventy percent reductions in incidental harvest, and predicted estimates in the absence of harvest impacts. A 20% annual natural mortality rate in the ocean is assumed.

D. Cumulative Effects

Cumulative effects are defined as the “effects of future State or private activities, not involving Federal activities, which are reasonably certain to occur within the action area of the Federal action subject to consultation” (50 CFR 402.02). For the purposes of this analysis, the action area includes ocean fishing areas mainly off the coast of California. The production of salmon by California State hatcheries will likely continue and has the potential to add cumulative impacts on winter-run chinook in the ocean, through competition and predation. State hatchery salmon production may also influence harvest rates in the ocean through increasing salmon abundance above natural salmon abundance. At this time, the extent of cumulative impacts from State hatcheries’ salmon production is not known, but further evaluation is warranted.

E. Critical Habitat Impacts

The designated critical habitat of Sacramento River winter-run chinook does not include the open ocean where PFMC fisheries occur. As a result, PFMC fisheries are not likely to adversely affect the critical habitat of winter-run chinook. Consideration of critical habitat impacts for PFMC fisheries will therefore not affect the conclusions regarding the jeopardy analysis.

	Back		Forward
	Table of Contents		Sacramento River Winter Run Chinook Page