Preserving and Restoring Wild Salmon Populations while Sustaining Commercial and Sport Fisheries with Hatcheries

The Problem

Hatcheries bypass the high mortality life-history phases of wild salmon populations.  As a result, hatcheries contribute far greater salmon smolt production to the ocean per number of eggs than do wild populations.  Without hatcheries, the replacement rate of Central Valley salmon populations would be less than 1-to-1, and the populations would move toward extinction.  Without hatcheries, there would be no commercial or sport salmon fisheries in California today.

But hatcheries bring many real problems for wild salmon.  These problems include in-breeding/domestication, disease transmission, and over-harvest of, competition for, and direct and indirect predation on wild salmon populations.  In-breeding has already had dramatic effects on the salmon populations, leading to the loss or degradation of many important life-history traits and of subpopulations that carry these traits (the “Portfolio Effect”).

Having lost many traits that nature provided over millions of years of natural selection, hatchery salmon today are simply less able to cope with the new world they now face.  They mature younger and smaller.  They are less able to adapt to changes in their food supply.  They often can’t compete and are less able to avoid predators.  Many arrive on spawning grounds too early, and others can’t find their natal streams.  Their offspring are also far less capable of coping with the stress and adversities, including harvest, pollution, and habitat loss and degradation.

Over-harvest, competition, and straying of hatchery fish has led to the dominance of hatchery fish in the Central Valley salmon populations and homogenization among the populations.  Some populations now survive only in hatcheries or in captive breeding programs.

The Solution

Many of elements of the problem have already occurred and are difficult to overcome.  While some elements are irreversible, it is not too late to limit or reduce some of the negative effects.  A comprehensive set of actions and strategies can avoid, minimize, mitigate, or even reverse these effects.  These actions and strategies should include:

1.      Reduce competition between hatchery and wild salmon in spawning, rearing, and migrating habitat.

  • Do not allow hatchery salmon to spawn in prime wild salmon spawning areas. Sorting at weirs can preclude passing hatchery spawners if hatchery fish are all marked.
  • Do not release hatchery juveniles into rearing and migrating habitats heavily used by remaining stocks of wild salmon. Programs throughout the range of Pacific coast salmon, including the Central Valley, now release hatchery smolts into net pens in rearing areas less frequented by wild salmon.  The best fishery returns to the Central Valley have been from smolts released from coastal net pens.

2.      Reduce straying of hatchery origin spawners into other spawning rivers.

  • Barge hatchery smolts to reduce competition and predation on wild juvenile salmon and decrease the straying of adults that results from trucking.  Barging can help imprint smolts on home rivers and hatcheries.
  • Monitor and sort adult salmon returns in rivers and hatcheries to further eliminate straying.
  • Focus more hatchery production on rivers and streams that do not support significant wild salmon.

3.      Increase harvest of hatchery salmon, while reducing harvest of wild salmon.

  • Focus harvest on hatchery stocks to help protect wild stocks. Release hatchery smolts into locations that focus harvest of adults in areas not frequented by wild salmon.  Adult hatchery salmon tend to stay in or return to areas where smolts were released.
  • Increase existing efforts to reduce the mixed-stock harvesting problem by reducing mixed-stock fishery exploitation rates to levels that are sustainable by wild stocks. Promote selective harvest of hatchery fish by permitting sport fishermen to retain only hatchery fish or to retain more hatchery fish than wild fish.  This would require marking most or all hatchery smolts.

4.      Improve disease control.

  • Hatchery fish experience greater susceptibility to infectious diseases due to higher rearing densities, higher levels of stress and poorer water quality. Diseases/infections can be spread to wild population elements, though research is needed to determine the extent of this threat.
  • Improve filtration systems at hatcheries to reduce the disease threat. This will also alleviate concerns about reintroducing salmon and steelhead upstream of hatcheries.

5.      Improve the genetic makeup of hatchery (and wild) salmon

  • Reverse engineer aspects of genetic diversity that has been selected out. Preferentially spawn 4-5 year-old adults at hatcheries.  Diversify timing of adult runs by breeding hatchery fish throughout the spawning run.  The Mokelumne Fish Hatchery is already implementing many such practices.  “Bad alleles can be purged.”
  • Use conservation hatchery actions to enhance the genetic diversity and fitness to help recover depleted wild populations.
  • Use more wild fish for hatchery broodstocks, particularly fish with more favorable traits.
  • Do not allow adult hatchery fish into spawning habitat used by wild fish.
  • Be more selective in choosing spawners for hatcheries.
  • Develop and support pure strains of wild salmon above dams through trap and haul programs.
  • Promote populations and subpopulations that protect or increase diversity (improve the Portfolio).
  • Develop captive stocks with desired natural traits – with less genetic drift, inbreeding and domestication,
  • Increase monitoring, research, experimentation, and adaptive management on the extent and consequences of domestication selection, as well as steps that may be taken to reduce its effects.
  • Evaluate and operate each hatchery program independently to address its program and its contribution to the overall problem.


Wild salmon populations in California’s Central Valley are already compromised to various degrees by hatchery salmon, over-harvest, and habitat degradation.  More can be done to protect wild salmon production and minimize the threat from hatcheries, while continuing to provide valuable commercial and sport fisheries supported by hatcheries.  We can save our salmon and eat them too.

For a more comprehensive scientific review of these subjects see Sturrock et al. 2019 and Nash et al. 2007.

Predators versus River Flow

I keep emphasizing the need for fall flows to get Central Valley salmon fry, fingerling, sub-yearling smolts, and yearling smolts to and through the Delta to the Bay. This especially applies to wild spring-run and to wild and hatchery winter-run and late-fall run, the Chinook salmon runs most in danger of extinction. Extinction comes from population decline and loss of genetic diversity from lower river flows and fragmented habitat. 1

The reason river flow is important is that flow affects habitat, growth, migration, and predation of emigrating salmon.

The long, slow reservoirs behind the mainstem dams on the Columbia River studied by Conner and Tiffan (2012)2 have habitat similar to the long, slow reaches of the lower Sacramento and San Joaquin rivers in the Central Valley. Furthermore, the Delta with its tides acts as a “main-stem” dam, slowing the outward movement of water through the Delta and salmon exiting to San Francisco Bay. The Delta has also been described as the place “where predators meet prey” – where the effectiveness of predation and the role played by “Anthropogenic Contact Points” is accentuated by modified freshwater flows.

The Sacramento River channel at Walnut Grove is one of the key “anthropogenic” contact points in the Delta. The major outlets from the Sacramento River channel to the central Delta, the Delta Cross Channel and Georgiana Slough, are located here (Figure 1). Lehman et al. (2019)3 describe the predator contact points at this location in Figure 1, including submerged aquatic vegetation, rip-rapped levees, docks, and diversions. The role of these particular contact points in predation on juvenile salmon is no doubt significant.

Lehman et al. point out the difficulty in removing the predators and the problematic contact infrastructure. However, they don’t address the role river flow and associated hydrodynamics play in modifying the effects of predators or specific contact points.

In the fall during the peak of winter-run emigration, Walnut Grove is the place where the Sacramento River channel in the north Delta slows and is “diverted” into the abyss of the central Delta. Few salmon escape the central Delta’s many predators and its “anthropogenic contact points”, including the south Delta export pumping facilities. Under low Sacramento River fall inflows (around 12,000 daily average flow at Freeport), high tides cause most of the water and salmon coming down the Sacramento River to divert into the central Delta via the Delta Cross Channel (DCC) and Georgiana Slough (Figure 2). Those young salmon remaining in the Sacramento channel are then vulnerable to the contact points and predators under lower water velocities. If river inflows are higher and the DCC is closed, the risks to young salmon is greatly reduced (Figure 3).

In conclusion, the Lehman study funded by the Metropolitan Water District describes the role of predators and contact point infrastructure including submerged aquatic vegetation, docks, riprap, and diversions. However, the Lehman study does not address the key factors in the fall loss of juvenile fish in the Delta: lower flows and the diversion of water into the central Delta for export. Closing the Delta Cross Channel and increasing river flows are the prescriptions needed to cut losses of emigrating endangered Central Valley salmon. Cutting south Delta exports in the fall would also be beneficial.

Figure 1. Predation contact points near Walnut Grove in the north Delta. Source: From Lehman et al. 2019.

Figure 2. Measured streamflows at USGS gages near Walnut Grove on 12/1/2019 at 8:00 am high tide. The DCC was open and the Sacramento River at Freeport inflow to the Delta was 12,500 cfs.

Figure 3. Measured streamflows at USGS gages near Walnut Grove on 12/5/2019 at the noon high tide. The DCC was closed and the Freeport inflow to the Delta was 21,000 cfs.

  1. Sturrock et al. 2019.
  2. Connor, W. P., and K. F. Tiffan. 2012. Evidence for parr growth as a factor affecting parr-smolt-survival. Transactions of the American Fisheries Society 141:1207–1218, 2012.
  3. Lehman, B.M., et al. 2019.

Dutch Slough Tidal Marsh Restoration

The Dutch Slough Tidal Restoration Project,1 newly redesigned (Figure 1), has some improved design elements, but remains flawed and potentially detrimental to Delta native fishes. Unless the flaws are overcome, the project will be a huge waste of limited Delta restoration funds.

First, the proposed project’s location within the Delta (Figure 2) is extremely detrimental.

  1. The location is an eastward extension of Big Break, an open water of the west Delta that is infested with non-native invasive aquatic plants and that breeds non-native fishes.
  2. The project is located on Dutch Slough, detrimentally warm in summer (Figure 3), with net flows that are negative and eastward toward the south Delta export pumps (Figure 4).

Second, and equally important, the project as designed would further contribute to the existing detrimental non-native vegetation and warm water problems.

  1. The extensive new dead-end slough complexes will become infested with invasive plants and will contribute to lowering turbidity and warming.
  2. The new subtidal habitat will further add to that in Big Break with more invasive plants and breeding and rearing habitat for non-native fish.

Third, the new habitat will attract breeding smelt and rearing juvenile salmon into an area where their eventual survival is highly questionable.

Can design changes overcome these flaws? Yes, but only in combination with other regional fixes.

  1. Big Break must first be restored along the lines being considered and studied in the Franks Tract Restoration Feasibility Study.
  2. A tide gate must be installed on east Dutch Slough, similar to that being considered for False River in the Franks Tract restoration. (This would fix the negative net flows toward south Delta exports and reduce salinity intrusion.)
  3. Open-water subtidal habitat should be eliminated. (Make the subtidal element diked-off non-tidal marsh.)
  4. Dead-end sloughs should allow flow-through to increase tidal circulation.
  5. Finally, more freshwater outflow should be allocated by reducing south Delta exports in low outflow conditions, in order to reduce salinity intrusion.

Figure 1. Conceptual design of Dutch Slough restoration project.

Figure 2. Location of Dutch Slough Project in the Delta.

Figure 3. Water temperature in Dutch Slough in 2014 and 2015.

Figure 4. Daily net flows in Dutch Slough 2007-2018.



How do we increase salmon runs in 2018 and beyond?

Over the past few months, I wrote posts on the status of specific runs of salmon in rivers throughout the Central Valley. In this post, I describe the overall status of salmon runs and the general actions to take to increase both escapement and fish available for commercial and sport harvest.

It was just over a decade ago that there were nearly one million adult salmon ascending the rivers of the Central Valley (Figure 1). At the same time, there were a millions more Central Valley salmon being harvested each year in sport and commercial fisheries along the coast and rivers of the Central Valley. Improvements in salmon management in the decade of the 1990s by the Central Valley Project Improvement Act, CALFED, and other programs had paid off handsomely with strong runs from 1999 to 2005. New and upgraded hatcheries, along with trucking hatchery smolts to the Bay, significantly increased harvest and escapement to spawning rivers.

Figure 1. Central Valley salmon runs from 1975 to 2016 including fall, late fall, winter, and spring runs. Source of data: CDFW GrandTab.

By 2008-2009, escapement had fallen by over 90% to a mere 70,000 spawners of the four races of salmon.  Fishery harvests were greatly restricted by 2008.  The winter run, the most threatened of the four runs fell from 17,296 to 827 spawners in just five years.  Drier years from 2001-2005, poor ocean conditions in 2004-2005, record-high Delta water diversions, and the 2007-2009 drought were contributing factors in the declines.  Impacts to coastal communities and the fishing industries were severe.

Extraordinary recovery measures included closing fisheries and trucking most of the hatchery smolt production to the Bay or Delta.  Federal salmon biological opinions (2009, 2011) limited winter-spring water-project exports from the Delta.  Hundreds of millions of new dollars were spent on habitat and fish passage improvements in the Valley to increase salmon survival and turn around the declines in runs.  A look at Figure 1 indicates that these efforts proved effective in limiting run declines from the 2012-2015 drought compared to the 1987-1992 and 2007-2009 droughts.

However, the prognosis for the future is again bleak, especially for wild, naturally produced salmon.  The consequences of the 2012-2015 drought  have not fully played out.  Once again, projected runs are low, and harvests are likely to be restricted.  Actions are needed to minimize long-term effects and to help bring about recovery of wild salmon productivity and fisheries in general.

Actions for 2018:

  1. Reduce harvest: Sadly but necessarily, the Pacific Fisheries Management Council and states are likely to take this first step of– restricting the 2018 harvest in the ocean and rivers to protect wild runs.
  2. Improve spawning, rearing, and migrating conditions: Sadly, this past year’s rearing and migrating conditions in the Sacramento River were unnecessarily compromised.   Water temperature at Red Bluff reached above the 56oF prescribed in the biological opinion and Basin Plan.  The higher temperatures resulted from low Shasta Reservoir releases (less than 5000 cfs – Figure 2) despite a virtually full Shasta Reservoir.  The low flow and higher water temperatures likely affected salmon egg incubation, rearing, and emigration-immigration success.  Reservoir releases will be necessary to meet flow and temperature targets in all Central Valley rivers and the Delta.
  3. Limit Delta exports: Delta exports this past spring reached unprecedented highs not seen in recent decades, resulting in high salmon salvage rates at the Delta fish facilities (Figure 3).1 With high water supplies from this past wet water year 2017, there will be high exports again unless there are some constraints.  If anything, winter-spring exports should be reduced to allow salmon to recover.  April-May exports should be reduced, like they were in the 1990’s and 2000’s, to 1500 cfs.

Near term actions over the coming year:

  1. Transport hatchery smolts to Bay: The transport of millions of fall-run smolts from state hatcheries on the Feather, American, and Mokelumne rivers to the Bay provides higher rates of escapement and contributions to the fishery and low rates of straying.  Barge transport to the Bay offers potentially lower rates of predation and straying for federal hatcheries near Redding.
  2. Raise hatchery fry in natural habitats: Recent research indicates that rearing hatchery fry in more natural habitat conditions increases growth rates, survival, and contributions to escapement and fisheries.  Raising hatchery fry in rice fields is one potential approach.
  3. Restore habitats damaged by recent record high flows in salmon spawning and rearing reaches of the Central Valley rivers and floodplains: In nearly every river, habitats were damaged by the winter 2017 floods, requiring extraordinary repairs and maintenance to ready them again to produce salmon.
  4. Take further actions to enhance flows and water temperatures to enhance salmon survival throughout the Central Valley: Actions may include higher base flows, flow pulses, or simply meeting existing target flow and temperature goals.

In conclusion, managers should take immediate actions to minimize the damage to salmon runs from the recent drought and floods, using this past year’s abundant water supply.  They should avoid efforts to exploit the abundant water in storage for small benefits to water supply at the expense of salmon recovery, and should make every effort to use the water in storage for salmon recovery.

Figure 2. Upper Sacramento River flows and water temperatures in May 2017. The target water temperature for Red Bluff is 56oF. Source of data: USBR.

Figure 3. Export rate and young salmon salvage at South Delta federal and state export facilities in May 2017. The target export rate limit for May should be 1500 cfs. Source of data: USBR.

Sometimes it doesn’t take a lot of water.

In a May 29 post, I discussed how a small diversion of cold water from the West Branch of the Feather River sustains the Butte Creek spring-run Chinook salmon, the largest spring-run salmon population in the Central Valley. In a May 8 post, I described how the Shasta River, despite its relatively small size, produces up to half the wild fall-run Chinook salmon of the Klamath River. In both examples, it is not the amount of water, but the quality of the water and the river habitat that matters. In the former case, man brought water to the fish. In the latter, man returned water and habitat to the fish.

While both examples are remarkable given the relatively small amount of water involved, the relatively small restoration effort required on the Shasta River and the minimal effect on agricultural water supply make it almost unique.

Just take a look at the present late May 2017 hydrology of the Klamath River (Figure 1). There was only 140 cfs flowing in the lower Shasta River. At the same time, there was 25,000 cfs flowing in the lower Klamath, 2000 cfs in the upper Klamath below Irongate Dam, and 2000 cfs in the Scott River. What is different is that most of the Shasta flow is spring fed, some of which is sustained through the summer. Of the roughly 300 cfs base flow in the river in late May 2017, about 200 was from springs (Figure 2). By mid-summer, flow out of the Shasta River into the Klamath will drop to about 50 cfs, with agricultural diversions from the Shasta at about 150 cfs. October through April streamflow is generally sufficient to sustain the fall-run salmon population. Summer flows are no longer sufficient to sustain the once abundant Coho and spring-run Chinook salmon.

Figure 1. Lower Klamath River with late May 2017 streamflows in red. Note Shasta River streamflow was only 140 cfs near Yreka, California. Data source: CDEC.

Figure 2. Selected Shasta River hydrology in late May 2017. Roughly 150 cfs of the 300 cfs total basin inflow is being diverted for agriculture, with remainder reaching the Klamath River. Red numbers are larger diversions. The “X’s” denote major springs. Big Springs alone provides near 100 cfs. Of the roughly 100 cfs entering Lake Shastina (Dwinnell Reservoir) from Parks Creek and the upper Shasta River and its tributaries, only 16 cfs is released to the lower river below the dam. Red numbers and arrows indicate larger agricultural diversions. Up to 15 cfs is diverted to the upper Shasta River from the north fork of the Sacramento River, west of Mount Shasta.