Table 2. Major anthropogenic factors limiting, or potentially limiting, viability of populations of Central Valley fall Chinook salmon in California, where a factor rated high is a major limiting factor, a factor rated intermediate is a factor that has the potential to be a major limiting factor but has had only a moderate effect so far on population viability, and a factor rated low has a low or unknown effect on population viability. Certainty of these judgments was 4 on 1-4 scale.
Dams. Large dams on the Sacramento River and its tributaries have denied fall run Chinook salmon access to historic spawning grounds and altered habitats below the dams in ways that are often not suitable for spawning and rearing, by reducing flows, increasing temperatures, embedding the substrate (so it cannot be moved by spawners), reducing cover for juveniles, and generally increasing the impact of other factors affecting salmon abundance (Moyle 2002). Most large dams now have special flow releases for salmon but their effectiveness varies. Dams also reduce or eliminate recruitment of spawning gravels into the river beds below dams. Loss of spawning gravels can limit salmon spawning success and large quantities of gravel are now trucked to spawning areas below dams and dumped in to provide spawning habitat. Techniques for adding spawning gravels to rivers for successful Chinook spawning are especially well developed for Central Valley streams but their effectiveness at the population level is not well documented (Mesick 2001, Wheaton et al. 2004).
Agriculture. There are literally hundreds of agricultural diversions along the Sacramento and San Joaquin Rivers and their tributaries, as well as in the Delta, which have the potential to impact salmon populations through the entrainment of juvenile salmon. Because diversions are an obvious source of loss of fish, most of the larger ones are screened to prevent entrainment. Moyle and Israel (2005) point out, however, that fish screens on rivers are subject to failure and may create holding areas for salmon predators. They also note that despite their numbers, small diversions, even cumulatively, probably do not kill many salmon, unless they are on small tributaries. In general, the higher percentage of flow taken by a diversion, the more likely the diversion is to have a negative impact on local salmon populations through entrainment of juveniles.
The largest diversions in the Central Valley are those of the State Water Project (SWP) and the federal Central Valley Project in the south Delta, which take water for both agricultural and urban use. They entrain large numbers of fall Chinook salmon (as well as salmon of other runs) especially from the San Joaquin River tributaries. These diversions are screened and salmon are salvaged from the projects by capturing, trucking, and then releasing them downstream in the Delta. However, mortality is likely high, both directly and indirectly. Kimmerer (2008) calculated about a maximum 10% loss of juvenile Chinook to direct entrainment, recognizing the high degree of uncertainty associated with any such estimate. Direct mortality is caused by high predation rates in Clifton Court Forebay from which the SWP pumps water (prior to running it through the salvage facility), by the stress of salvage, and by predation after they are released, disoriented, into predator-rich areas. Indirect mortality is likely considerably higher than direct mortality and is caused by changes in Delta hydrology due to project operations, created by both the pumping itself and by the dam releases (or lack thereof) to provide water for the water project pumps. The salmon essentially can be diverted into unfavorable parts of the Delta in which they are much more likely to die of environmental stress or predation. In general, when flows are higher and salmon avoid the pumps, survival of outmigrants tends to be higher, although there is no simple relationship between the amount of water being diverted per se and salmon survival (Brandes and McClain 2001). However, San Joaquin fall Chinook salmon are most likely affected by South Delta Pumping, especially when their populations are at low ebb, because they are most vulnerable to the pumps by sheer proximity.
Agriculture also contributes to loss of juvenile habitat in the rivers by denying access to the shallow riverine habitats needed for feeding and protection from predators during migration. Construction of levees to contain rivers has had multiple effects, including simplifying bank structure through use of rip-rap and removal of trees, reduction in shade, and reduced access to floodplains. This whole process of bank hardening has been made much easier by the reduction of peak flows by dams. A particularly severe problem has been the loss of access of fish and water to suitable floodplains. Recent studies have demonstrated their importance for increased juvenile salmon growth and survival (Sommer et al. 2001, Jeffres et al. 2008).
Urbanization. Urbanization simplifies and pollutes Chinook salmon habitats. By diverting water and denying access to floodplain areas, the simplification process is similar to that discussed above for agriculture only impacts are more intense locally. Here we discuss pollution which also is a joint urban and agricultural problem.
Juvenile salmon are continuously exposed to toxic materials discharged in to rivers from both urban and agricultural sources. The latter are particularly likely to affect juvenile San Joaquin fall Chinook salmon and a potentially major source of mortality is the toxic, anoxic water associated with the Stockton Deepwater Ship Channel which results from pollutants from agricultural wastewater, discharges from the Stockton sewage treatment and storm drains, and other sources. A new threat is the use of pyrethroid pesticides which are particularly toxic to fish. The effects of these diverse pollutants on wild juvenile salmon abundance is largely unknown but mortality is periodically recorded. Even if pollutants are not directly lethal, they (or poor water quality in general) can stress both adult and juvenile fish, making them more susceptible to diseases that are always present in the environment.
Mining. Both gold and gravel mining left deep pits in a number of Central Valley rivers (e.g. Tuolumne, Merced, San Joaquin). These pits can trap flows and provide a place for predatory fish to concentrate, increasing mortality of juvenile salmon moving downstream. Likewise, hydraulic and dredge mining in the 19th and early 20th centuries dramatically altered a number of rivers, reducing salmon spawning and rearing habitat, from which the rivers are still recovering. In the Sacramento River, a potential major problem is highly acidic water laden with toxic heavy metals from the Iron Mountain mine site, if Spring Creek retention reservoir spills or bursts. These highly toxic wastes could wipe out either migrating adults or, more likely, juveniles foraging in the river.
Estuarine modification. There is growing appreciation of the importance of biocomplexity for the persistence of salmon in a variable environment (Hilborn et al. 2003). Biocomplexity is the existence of multiple variations on the life history theme that characterizes successful salmon populations. Thus, historically fall Chinook presumably had juveniles that entered the estuary at different months and spent different amounts of time there, with some strategies being more successful in some years than others but all strategies maintained because both estuarine and ocean environments could be very different from year to year. Loss of diverse habitats in the San Francisco Estuary has essentially eliminated aspects of life history diversity and the best strategy for juvenile salmon today seems to be to move through the estuary as fast as possible.
Harvest. The effects of harvest on Central Valley salmon is discussed at length by Williams (2006). Chinook salmon are harvested in both ocean and in-river fisheries. Hatchery fish can sustain higher harvest rates than wild fish, but fisheries do not discriminate between them. The fisheries thus presumably take wild and hatchery fish in proportion to their abundance; a harvest rate that is sustainable for hatchery fish may be unsustainable for wild fish. This can lead to hatchery fish replacing wild fish in the fishery rather than just supplementing them (as they were supposed to do).
Commercial fisheries also likely affected Chinook populations indirectly through continual removal of larger and older individuals. This selectivity results in spawning runs made up mainly of two and three-year-old fish, which are smaller and therefore produce fewer eggs per female. The removal of older fish also removes much of the natural cushion salmon populations have against natural disasters, such as severe drought, which may wipe out a run in one year. Under natural conditions, the four- and five-year-old fish still in the ocean help to keep the runs balanced and make up for the fish lost during an occasional catastrophe. In order to protect declining stocks of Chinook salmon, marine salmon fisheries were greatly restricted in 2006-2009 by the National Marine Fisheries Service and the Pacific Fisheries Management Council (Congressional Record, 50 CFR Part 660); Sacramento salmon fisheries were banned completely in 2008 and 2009. This has resulted in the return of a higher proportion of larger and older fish than in previous years, although numbers of fish are nevertheless exceptionally low.
Hatcheries. After an exhaustive review of the literature on hatchery practices in California, Williams (2006) concluded that hatcheries almost certainly have deleterious effect on wild populations of salmon, which may run contrary to recovery goals for wild fish. The effects can stem from competition by hatchery fish with wild juveniles when hatcheries flood the environment with juveniles that are bigger and more numerous than wild fish. This can displace wild fish, making them more vulnerable to predation and reducing growth rates. Such competition potentially can exist at all phases of the life cycle, including during ocean feeding and on the spawning grounds. As indicted above, the presence of large numbers of hatchery fish also resulted in unsustainable harvest rates of wild Chinook salmon because stocks are mixed in the ocean, further reducing the viability of wild populations.
In addition, studies on other salmonids, especially steelhead, have shown that fitness (ability to produce young that survive to reproduce) decreases rapidly when fish are raised in hatcheries. Araki et al. (2007) estimate that fitness of steelhead decreases almost 40% per generation of hatchery culture. In a follow up study, Araki et al. (2009) showed that introduction of hatchery steelhead into wild populations reduced fitness for several generations, i.e. a naturally spawned fish that had only one grandparent of hatchery origin still had significantly reduced fitness. The loss of fitness due to hatchery effects has not been experimentally determined for Chinook. The implications, however, are almost certainly serious, given that the fish spawning in Central Valley rivers are overwhelmingly of hatchery origin and recent studies show that the entire population is genetically homogeneous (Lindley et al. 2009). The introgression of hatchery genetics may mean that the entire CV fall run is suffering from the effects of reduced fitness due to domestication selection. This negatively impacts the possibility of recovering a self-sustaining wild population because it means that the reproductive capacity of all naturally spawning individuals is compromised. Reduced reproductive capacity in the wild would severely impact programs such as the Anadromous Fish Restoration Program (AFRP) which tasked by the Central Valley Project Improvement Act to make "all reasonable efforts to at least double natural production of anadromous fish in California's Central Valley streams on a long-term, sustainable basis".
Hatchery practices may also make Chinook less adapted to adverse conditions in both fresh and salt water (e.g., physiologically less capable of surviving on less food, less sensitive to changing ocean conditions, less able to avoid predation). It is also possible that the fairly uniform nature of Central Valley fall Chinook has reduced variability in response to environmental conditions, making them more vulnerable to mortality under variable ocean conditions (Lindley et al. 2009).
Alien species. For the past 150 years, Chinook salmon have been faced with an onslaught of potentially deleterious alien species yet have managed to persist despite their presence. Probably most significant are fish that are predators, including striped bass, largemouth bass, smallmouth bass, and spotted bass. Striped bass have not been implicated directly in any salmon declines, perhaps because they arrived early enough on the scene so they mainly replaced native predators. They can consume large numbers of juvenile salmon, however, below diversions such as Red Bluff Diversion Dam, or where hatcheries release large numbers of naïve fish. The three centrarchid (black) basses can also be locally important as predators, especially when they inhabit in-channel gravel pits and other obstacles the juvenile salmon have to pass through on their way downstream. Fortunately, their metabolic processes are relatively slow, due to low temperatures, when peak juvenile salmon out-migrations occur, which reduces predation. One of the reasons CDFG made such a huge (and apparently successful) effort to eradicate northern pike (Esox lucius) from Davis Reservoir on a tributary to the Feather River is that pike are cool-water predators, so are likely to be much more effective predators on juvenile salmon than existing alien predators. Their potential invasion of the Sacramento River system could be disastrous for already depleted salmon runs.
Factors Affecting Status- An Integrated View: Ever since the Gold Rush, Central Valley Chinook salmon populations have been in decline. Historic populations probably averaged 1.5-2.0 million (or more) adult fish per year (Yoshiyama et al. 1998). The high numbers resulted from four distinct runs of Chinook salmon (fall, late-fall, winter, and spring runs) which took advantage of the diverse and productive freshwater habitats created by the cold rivers flowing out of the Sierra Nevada. When juveniles moved seaward, they found abundant food and good growing conditions in the wide valley floodplains and complex San Francisco Estuary, including the Delta. The salmon smolts then reached the ocean, where the southward flowing, cold, California Current and coastal upwelling together created one of the richest marine ecosystems in the world, full of the small shrimp and fish that salmon require to grow rapidly to large size. In the past, salmon population no doubt varied as droughts reduced stream habitats and as the ocean varied in its productivity, but it is highly unlikely the numbers ever even approached the low numbers we are seeing now.
Unregulated fisheries, hydraulic mining, logging, levees, dams, and other factors discussed above caused precipitous population declines in the 19th century, to the point where the salmon canneries were forced to shut down (all were gone by 1919). Minimal regulation of fisheries and the end of hydraulic mining allowed some recovery to occur in the early 20th century but the numbers of harvested salmon steadily declined through the 1930s. There was a brief resurgence in the 1940s but then the effects of the large rim dams on major tributaries through out the Central Valley began to be severely felt (Yoshiyama et al. 1998). The dams cut off access to 70% or more of historic spawning areas and basically drove the spring and winter runs to near-extinction, making the fall run the principal support of fisheries. In the late 20th century, thanks to hatcheries, special flow releases from dams, and other improvements, salmon numbers (mainly fall Chinook) averaged nearly 500,000 fish per year, with wide fluctuations from year to year, representing around 10-25% of historic abundance (Figure 1). In 2006, numbers of spawners dropped to about 200,000, despite the closure of the fishery to protect Klamath River Chinook runs. In 2007, the number of spawners fell further to about 90,000 fish, among the lowest numbers experienced in the past 60 years, with even lower numbers in fall 2008 (approx. 67,000 fish). The evidence suggests that these runs are largely supported by hatchery production, so numbers of fish from natural spawning are much lower.
What caused this apparently precipitous decline in salmon? Unfortunately, the causes are historic, multiple and interacting. The first thing to recognize is that Chinook salmon are adapted to living in a region where conditions in both fresh water and salt water can alternate between being highly favorable for growth and survival and being comparatively unfavorable. Usually, conditions in both environments are not overwhelmingly bad together, so when survival of juveniles in fresh water is low, those that make it to salt water do exceptionally well, and vice versa. This ability of the two environments to compensate for one anothers failings, combined with the ability of adult salmon to swim long distances to find suitable ocean habitat, historically meant salmon populations fluctuated around some high number. Unfortunately, when conditions are bad in both environments, populations crash, especially when the heavy hand of humans is involved.
The recent precipitous decline has been blamed largely on ocean conditions. Generally what this means is that the upwelling of cold, nutrient-rich water has slowed or ceased, so less food is available, causing the salmon to starve or move away. Upwelling is the result of strong steady alongshore winds which cause surface waters to move off shore, allowing cold, nutrient-rich, deep waters to rise to the surface. The winds rise and fall in response to movements of the Jet Stream and other factors, with both seasonal and longer-term variation. El Niño events can affect local productivity as well, as can other anomalies in weather patterns and Chinook salmon populations fluctuate accordingly.
The 2006 and 2007 year classes of returning salmon mostly entered the ocean in the spring of 2004 and 2005, respectively (most spawn at age 3). Although upwelling should have been steady in this period, conditions unexpectedly changed and ocean upwelling declined in the spring months, so there were fewer shrimp and small fish for salmon to feed on. According to an analysis by Barth et al. (2007), conditions were particularly bad for a few weeks in spring of 2005 in the ocean off Central California, resulting in abnormally warm water and low concentrations of zooplankton, which form the basis for the food webs which include salmon. All this could have caused wide scale starvation of the salmon. While the negative impact of ocean anomalies on salmon is likely, monitoring programs in ocean are too limited to make direct links between salmon and local ocean conditions.
Ocean conditions can also refer to other factors which can be directly affected by human actions, especially fisheries. For example, fisheries for rockfish and anchovies can directly or indirectly affect salmon food supplies (salmon eat small fish). Likewise, fisheries for sharks and large predators may have allowed Humboldt squid (which grow to 1-2m long) to become extremely abundant and move north into cool water, where they could conceivably prey on salmon. These kinds of effects, however, are largely unstudied.
Meanwhile, what has been going on in the Sacramento and San Joaquin Rivers? On the plus side, dozens of stream and flow improvement projects have increased habitat for spawning and rearing salmon. Removal of small dams on Butte Creek and Clear Creek, for example, increased upstream run sizes dramatically. Salmon hatcheries also continue to produce millions of fry and smolts to go to the ocean. On the contrary side:
The pumps in the South Delta have diverted increasingly large amounts of water in the past decades, altering hydraulic and temperature patterns in the Delta as well as capturing fish directly.
The Delta continues to be an unfavorable habitat for salmon, especially on the San Joaquin side where the inflowing river water is warm and polluted with salt and toxic materials.
Hatchery fry and smolts are released in large numbers but their survivorship is poor, compared to wild fish, although they contribute significantly to the fishery. Nevertheless, they may be competitors with wild produced fish under conditions of low supply in the ocean. Most of the hatchery fish are planted below the Delta, to avoid the heavy mortality there. Unfortunately, the fitness of naturally produced salmon versus hatchery produced salmon is not understood; it is possible that the influence of hatchery-reared fish is so strong today that the progeny of natural and hatchery spawners have similar survival rates in the wild.
Numbers of salmon produced by tributaries to the San Joaquin River (Merced, Tuolumne, Stanislaus) continue to be exceptionally low, in the hundreds, and the promised restoration of the San Joaquin River will take a long time to be effective.
Thus, reduced survival of naturally spawned fish in fresh water, especially in the Delta, combined with the naturally low survival rates of hatchery fish, could make for plummeting numbers of adult spawners. This is especially likely to happen if young salmon also hit adverse conditions in the ocean, as they enter the Gulf of the Farallons. The growing salmon can also hit other periods when food is scarce in the ocean, along with abundant predators and stressful temperatures, at any time in the ocean phase of their life cycle. Once again, our ignorance of how the salmon survive in the ocean is profound. For example, much could be learned about how ocean food supplies are affecting salmon growth and survival by tracking the growth and condition of juveniles once they have moved out to sea.
The overall message here is that indeed ocean conditions have had a lot to do with the recent steep decline of salmon populations in the Central Valley in recent years. However, they are superimposed on a population that has been declining in the long run (with some apparent stabilization in recent decades, presumably due to hatchery production) and that has been genetically altered by hatchery practices. The salmon still face severe problems before they reach the ocean, especially in the Delta. Overall, blaming ocean conditions for salmon declines is a lot like blaming Hurricane Katrina for flooding New Orleans, while ignoring the many human errors that made the disaster inevitable, such as poor construction of levees or destruction of protective salt marshes. Managers have optimistically thought that salmon populations were well managed, needing only occasional policy modifications such as hatcheries or removal of small dams, to continue to go upward. The listings of the winter and spring runs of Central Valley Chinook as endangered species were warnings of likely declines on an even larger scale.
Little understanding of the genetic and competitive impact of hatchery fish on wild populations may be one of the largest reasons for the decline of Central Valley Chinook salmon. While intended to boost salmon numbers, in reality, hatchery operations have likely contributed to the current all time low numbers of returning fall Chinook. The decline of the states wild chinook went largely unnoticed, obscured by massive hatchery production. Even as greater and greater numbers of salmon were being produced by hatcheries, the long-term persistence of CV fall Chinook has become more and more precarious because of the loss of important demographic and genetic diversity. Historically, large variation in wild salmon behavior and life history strategy guaranteed persistence in Californias extremely variable climate. Hatchery selection not only narrowed this behavioral variation in hatchery stocks leaving them extremely vulnerable to climatic anomalies (ocean conditions, drought etc.), but it also cultivated domestic genes, essentially evolving a new strain of salmon, one adapted to hatchery conditions but which is particularly unfit when exposed to natural conditions. One way that domestication selection operates is by shielding individuals that express deleterious genes from natural selection during early development. These shielded domestic genes then escape when a hatchery fish does not return to the hatchery but instead spawns with a wild fish under natural conditions. In the next generation, domestic genes that were shielded by the hatchery environment are now exposed to natural selection resulting in an increased mortality rate for the brood. This becomes a serious threat to wild populations when hatchery fish spawn with wild fish and introduce these domestic genes into the wild gene pool. Fitness of hatchery/wild hybrid broods has been shown to be severely reduced in steelhead (Araki et al. 2007) and these fitness reductions have been documented to be multigenerational (Araki et al. 2009), the effects likely compounding over time. A significant reduction in the reproductive capacity of the entire wild population then, is the likely consequence of the past 60 years of hatchery operation. The implications of population level fitness reduction for recovery of self-sustaining stocks are serious and are in need of intense study.