Analyzing Fish Population Dynamics in the Bay-Delta

I have been analyzing the declines in Bay-Delta and Central Valley fish populations for over 40 years. Fish population dynamics were the focus of my college education and my 50-year career in environmental impact assessment. I have participated in all the major efforts to understand the Bay-Delta fish population declines. From all of these efforts, it is clear to me what has caused the major fish population crashes.


First and foremost are the well known historic factors, the original sins pre-1970s of diverting water, building levees and dams, urban development, gold mining, cutting forests, polluting rivers, over-fishing, and introducing non-native species. These explain many of the major native fish population declines and extinctions such as the Sacramento perch and San Joaquin spring-run Chinook salmon, and the near extinctions of Delta smelt, green sturgeon, winter-run and spring-run salmon, and steelhead.


Since 1970, there have been dramatic declines in salmon, steelhead, smelt, sturgeon, splittail, and striped bass, often described as “recruitment failure” or failure to reproduce. While some of the blame most certainly is on continuing effects of the aforementioned original sins, the major post-1970 shifts were the consequence of a new array of stresses that hit the whole fish community, especially native fish populations. Most certainly the droughts of 76-77, 87-92, 01-02, 07-09, and 12-15 were a major underlying factor; however, it was the man-made responses to the droughts that caused most of the damage. Asian clam and other non-native aquatic invertebrate invasions to the Bay-Delta in the 80s were another stress, in part brought on by the aforementioned factors. Poor water management response to these new threats has caused further damage. The big culprits of change were the water management stresses described below.

1. State Water Project

The addition of the State Water Project (SWP) in the mid-1970s nearly tripled Delta export capacity (4400 to 11,400 cfs pumping rate1) and annual exports (2 million acre-feet to 6 million acre-feet annual exports). The additional Delta exports had huge fish population effects in the mid-70s from salvage mortality and entrainment of young fishes, as well as on fish habitat conditions in the rivers, Delta, and Bay. These stresses resulted in major population declines, which in turn resulted in the imposition of export restrictions in new water quality standards in 1978 (D-1485), and eventually to species listings under the Endangered Species Act in the 1990s.

2. Reservoir Operations

The increase in exports changed reservoir operations, including within-year reservoir release strategies and long-term multiyear reservoir storage patterns. Reservoir storage was depleted faster in droughts because of higher water supply demands. These effects continue today.

3. Water Supply Demands

Ever-increasing water supply demands from agricultural and municipal users have reduced river flows, Delta outflow, and reservoir storage. It’s not only the Delta’s 6 million acre-feet of exports, but the more than 20 million acre-feet from other Central Valley water diversions.

4. Invasive Species

Invasions of non-native clams, shrimp, fish, and zooplankton species since the 1970s have occurred in-part due to changes in Bay-Delta hydrology and water quality, as well as physical and biological habitat conditions. Delta pelagic (open water) habitat is now dominated by low-productivity reservoir water. The low salinity or mixing zone of the estuary became far less productive because of species invasions and reservoir water moving through to the south Delta export facilities, taking productive low-salinity habitat with it. The Delta is warmer from higher warm river inflows from spring through fall to feed water project exports, further favoring non-native warm-water fishes. Turbidity is lower, favoring non-natives. Invasive aquatic vegetation benefits from low turbidity, and the vegetation further favors non-native fishes over native fishes.


Since 1990, there have been steps backward that have undermined effective strategies and actions that had been undertaken beginning in the late 1970s to help depressed fish populations. Below are five examples in a long list of actions/changes.

1. Changes to D-1485

Beginning In 1978, Delta water quality standards in Decision 1485 placed restrictions on Delta exports, improved Delta outflows, and set salinity standards that had benefits for native fishes. Beginning in the 1990s, these post-1970 constraints on water diversions were changed, ignored, or eliminated. For example, new standards in D-1641 (1995 Accord) dropped the D-1485 June-July export restrictions.

2. Eliminating VAMP Export Restrictions and Higher Outflow Requirements in April and May

The Vernalis Adaptive Management Plan (VAMP) from 2000-2009, and its operational precursors under the CVPIA (1991) and the 1995 Accord, sought to protect Central Valley salmon and Delta native fishes by reducing April-May Delta exports and increasing spring Delta inflows and outflows. During the VAMP years, exports were restricted to less than 2000 cfs in April-May to protect fish (Figure 1). In the post-VAMP decade, restrictions were lifted and exports increased, especially in post-drought recovery wet years 2011 and 2017 (Figure 2).

3. Temporary Urgency Change Petitions (TUCPs) and Orders

Temporary urgency change orders during the recent drought allowed April-May Delta outflow to fall to around 5000 cfs in 2014 and 2015, from the normal near-10,000 cfs lower limit (Figure 3). Such low outflows in combination with Delta exports are devastating to Delta native fishes and Central Valley salmon and steelhead.

4. Delta Channel Barriers

The operation of the Delta Cross Channel, Head of Old River, South Delta, and False River barriers helps to keep export salinity down by funneling the fresher Sacramento River water to the south Delta export pumps. This increases the efficiency of exports in taking reservoir water in drier years and seasons. With the exception of the Head of Old River, barrier operation also funnels Delta native fish production (pelagic eggs and juveniles) and migrating young salmon (and their low salinity habitat and food sources) directly to the export pumps instead of to the Bay.

5. Suisun Marsh Salinity Control Gates

Since the installation of the Suisun Marsh Salinity Control Gates (SMSCG) in Montezuma Slough in 1989, the Slough and Marsh no longer function as critical low salinity habitat in drier years and seasons. Without high freshwater inflow, the Slough and Marsh no longer maintain the high biological production the once contributed to the Bay. The following excerpt from a DWR 2019 blog post inadvertently describes how limited the benefits of Suisun Marsh have become in the absence of flow:

DWR launched a pilot project last year that directed more fresh water flow into Suisun Marsh. The action involved opening salinity control gates in the summer months instead of during fall and winter, as is customarily done to reduce salinity in the marsh for migrating ducks and other waterfowl. The Delta smelt relies on low-salinity water – opening the salinity control gates allowed the smelt to enter the marsh from the Sacramento River, where it can access greater amounts of food and shelter.

Extinction looms so closely over the Delta smelt population that the project could have been considered a success even if it didn’t lure any countable Delta smelt to the marsh, said DWR Lead Scientist Ted Sommer. Just creating the conditions that allow smelt to thrive – that is, low salinity levels, lots of food, and high turbidity or muddy water that magnetizes smelt – would have been a cause for celebration.


There are many, many other examples of adverse changes that have put fish population dynamics in the Delta in a perpetual downward spiral. Since 1970, almost of all them involve reduction of Delta inflow and outflow, elimination of measures to mitigate the effects of reduced Delta inflow and outflow, and/or the biological response to reduced Delta inflow and outflow.

Figure 1. State south Delta exports (Harvey Banks pumping plant) in spring 1997-2010.

Figure 2. State south Delta exports (Harvey Banks pumping plant) in spring 2011-2019.

Figure 3. Delta outflow April-May 2007-2009 and 2013-2015 droughts.




  1. Initially exports were even higher with the new 11,000 cfs export capacity of the State Water Project. Total exports reached 12,000-14,000 cfs

Klamath River Salmon – the Wrong Advice!

In a June 2019 article in the LA Times , also posted in Maven’s Notebook, JACQUES LESLIE suggests that “hatcheries don’t belong in this picture” once the planned removal of four dams on the Klamath River is complete.  Based on my decades of work in the Klamath watershed, this post suggests a different approach.  A conservation hatchery could accelerate and improve the outcome of the recovery of Klamath River salmon.  I respond below to a few statements in the article.

“Allowing hatchery salmon to mix with struggling native salmon after removing the dams is like rescuing a dying man only to slowly poison him.”

Native salmon are nearly extinct or already extinct over much of the Klamath River watershed.  A small population of spring-run Chinook remains only in the Salmon River, and is about to be listed as endangered.  Small declining runs of listed Coho salmon remain in several tributaries.  Modest runs of wild fall-run Chinook continue in the Scott and Shasta Rivers, but they are not native to the upper watershed above the mainstem dams slated for removal.  Remaining salmon in the Klamath River are the progeny of hatchery salmon or of interbred hatchery and wild salmon.   Remaining wild Klamath River steelhead are also not native to the upper watershed, and many of them spawn in tributaries downstream of Iron Gate Dam, the lowest Klamath River dam.  Wherever they come from, salmon and steelhead that re-populate the upper watershed will not be native to the upper watershed, at least not initially.

“Salmon hatcheries don’t belong in this picture. They are relics of an outdated worldview that maintains that technology can conquer and control nature. They curtail salmon runs on the river, and instead of diverse stocks of fish that possess varied abilities enabling them to return to spawn — and die — at spots all along the river where they were born, hatchery fish’s birthplace is a single place: the hatchery. The identical life histories of these fish make them more susceptible to disease and predators than their native relatives.”

The modern view of hatcheries, and of conservation hatcheries in particular, is that they (and “technology”) can work with nature rather than controlling it.  One problem is that the life histories of salmon that have survived the dams are not lined up with the likely best life histories for the 400 miles of migration, spawning and rearing habitat of the upper Klamath watershed that will soon become accessible.   Existing life histories of Klamath salmon are lined up with the habitat that was left to them, largely in the few remaining large Klamath tributaries that enter the mainstem downstream of Iron Gate Dam.  Managers of a conservation hatchery can select from the few remaining fish that have the most desirable life histories.  Outplanting these hatchery-bred juveniles in the upper watershed and similar strategies can provide source stock for wild populations that can then better adapt to the habitats of the upper Klamath watershed.

“In fact, maintaining the salmon hatcheries amount to a federal subsidy for commercial and recreational fishing, a subsidy that is supposed to be justified by the fishery’s economic benefits.”

Hatcheries are mitigation for a loss to society and culture, not a “subsidy.”  Those who benefit from the loss commit to paying for the loss.  It is absolutely true that the mitigation has created its own set of problems.  That does not absolve the beneficiaries of responsibility, and it should not disallow the opportunity to improve or accelerate the transition to the robust self-sustaining wild fisheries that every responsible stakeholder seeks.

“The salmon hatcheries on the Klamath should be phased out as quickly as possible. Even if the post-dam comeback of wild salmon is slow, river managers should resist pressure to continue or even expand hatchery operations.”

The hatcheries as they now exist should be phased out if the need to mitigate ends.  Sad thing is that the hydro dams will leave a legacy of degraded habitat and species diversity loss.  It remains to be seen how far habitat restoration can go.  Conversion of the hatcheries to species conservation would help the recovery effort.

In conclusion, a conservation hatchery program could help to restore populations of coho, spring-run Chinook, fall-run Chinook, and steelhead to the areas of the watershed to which dam removal will restore access.  Recovery efforts for native green and white sturgeon, bull trout, redband trout, and suckers could also benefit from modern conservation hatchery programs.  Conservation hatcheries can also preserve the genetic diversity of these native fishes for the future when and if habitat is restored or altered by climate change.



Feather River Fall-Run Status through 2018

In my last post on the status of Feather River fall run salmon in May 2017, I analyzed recruitment through the fall-run in 2016 that included survival of brood years through 2013. In a recent May 2019 post, I discussed the survival of hatchery brood year releases through 2013. After near record low escapement/recruitment in 2008 and 2009, there was a strong recovery from 2010-2014, followed by lower runs in 2015 and 2016 (brood years 2012 and 2013), the product of the 2012-2016 drought. Brood years 2012 and 2013 suffered from poor juvenile river survival of hatchery and wild salmon in critical drought years 2013 and 2014. Overall production was sustained by Bay and coastal hatchery smolt releases (trucking and pen releases).

In this post, I update the status of the run through 2018 with the addition of escapement estimates for the 2017 and 2018 runs. I also provide information on returns of hatchery brood year 2014. In addition, I provide a prognosis for the fall 2019 run.

2017 and 2018 Escapement

Poor river escapement in 2017 (Figure 1) likely reflects poor survival of naturally spawned salmon from fall through spring of critical drought water year 2015. The strong hatchery escapement in 2017 (Figure 2) reflects higher survival of brood-year 2014 hatchery releases. Escapement improved in 2018 with higher river and hatchery contributions from brood year 2015.

Brood Year 2014 Hatchery Survival

Survival estimates based on hatchery coded-wire-tag returns for brood year 2014 (released in spring of critical drought year 2015) were high (3-5%) for coastal releases, good (1-2%) for Bay releases, and poor for river releases (Figure 3). Approximately 6 million smolts were released, of which 4.2 million were released to San Pablo Bay net pens. Of the remainder, 1.6 million were released to the Feather River, 10,356 at Tiburon near the Golden Gate,1 and 331,000 to Half Moon Bay on the coast south of San Francisco.

Prognosis for Brood Year 2016 (Fall 2019 Run)

River flows and Delta outflow conditions were much better in winter-spring 2017, a wet water year, than in 2015, a critical drought year (Figure 4). Hatchery brood year 2016 releases totaled nearly 5 million smolts, with 750,000 released in the Bay, 264,000 to the Golden Gate at Tiburon, and 3.5 million to the Feather River (Figure 5). River natural and hatchery release survival should be good given the wet year conditions. Hatchery Bay release survival should be good given high outflows. All indications look good. “Finally there are larger numbers of big king salmon showing in the upper river on the Sacramento River. … Over on the Feather River flows are great and water temperatures have finally dropped to excellent levels to catch the fresh king salmon migrating towards the Feather River hatchery. More salmon have come through this week than the entire month of August.”2

Figure 1. River spawner estimates 1953-2018.

Figure 1. River spawner estimates 1953-2018.

Figure 2. Hatchery spawner estimates 1964-2018.

Figure 2. Hatchery spawner estimates 1964-2018.

Figure 3. Feather River hatchery smolt release survival to adults from 2008-2014 brood years based on coded-wire-tag returns. Data Source:

Figure 3. Feather River hatchery smolt release survival to adults from 2008-2014 brood years based on coded-wire-tag returns. Data Source:

Figure 4. Delta outflow winter-spring 2015-2017.

Figure 4. Delta outflow winter-spring 2015-2017.

Figure 5. Brood year 2016 Feather River Hatchery smolt releases spring 2017. Source:

Figure 5. Brood year 2016 Feather River Hatchery smolt releases spring 2017. Source:


Reclamation’s Proposed Delta Smelt Fall Habitat Action In 2019

The US Bureau of Reclamation (Reclamation) is proposing a “Delta Smelt Fall Habitat Action” that would eliminate the requirement to increase outflow from the Delta in the late summer of this wet water year.1 The “Fall X2” flow increase that Reclamation has placed on the chopping block is a major provision in the 2008 Delta smelt biological opinion (BiOp). Reclamation describes the substitute proposed Action in a summary its webpage:

The Proposed Fall Habitat Action (Proposed Action) for Delta Smelt habitat in Water Year (WY) 2019 will achieve the Action 4 objective. Action 4 of the 2008 BO requires adaptive management to ensure that the implementation addresses the uncertainties about the efficiency of the action. Action 4 also states that as new information is developed and as circumstances warrant, changes by the Service to the Fall X2 action itself may be necessary. The Proposed Action is a plan to adaptively manage and modify its operation of the CVP/SWP under RPA Action 4.2

In brief, the “circumstances” that “warrant” this “adaptive management” are that getting rid of Fall X2 will allow Reclamation to export more water from the Delta this fall.

In August 2019, Reclamation issued an Environmental Assessment (EA) of the Proposed Fall Habitat Action. Though the Action will unravel a major component of the 2008 Biological Opinion for smelt, the comment period was 15 days. CSPA submitted comments; some of them are restated below, in response to citations from the “Effects Analysis,”3 an appendix to the Environmental Assessment.

The Effects Analysis quotes the BiOp to portray the Proposed Action as a scientific investigation, stating:

“[T]here is a high degree of uncertainty about the quantitative relationship between the size of the Action described above and the expected increment in Delta Smelt recruitment or production.” (p. 1).

Comment: After 2011, it has been nearly impossible to measure population response to changes in management because the smelt population has become so low. Likewise, a negative response cannot be detected at the present population level. The only certainty to be gained from harming a nearly extinct population even more is that Reclamation will increase exports from the south Delta.

The Effects Analysis selectively calls out results of the first implementation of the Fall X2 requirement in 2011:

“Abiotic habitat did increase in 2011 as predicted from the AMP, but other variables such as zooplankton abundance were too variable to draw a conclusion, and Delta Smelt growth rate comparisons remain incomplete as of 2019.”  (p.2)

Comment:  Following implementation of Fall X2 flows, smelt abundance increased sharply in the fall 2011 index. A quick look at zooplankton 2011 vs 2010 (Figure 1) also indicates an increase in zooplankton (key smelt food source) in Suisun Bay/Marsh in 2011.

Figure 1. Zooplankton (key Delta smelt food source) in September 2011 and 2010.
Note increased abundance downstream in 2011.

The Effects Analysis tries to explain the lack of response of Delta smelt to the 2017 Fall X2 action as a function of water temperature, concluding that the action was just futile:

In 2017, a Fall X2 adaptive management action was implemented. The results of the 2017 monitoring program were evaluated in the IEP’s 2019 draft FLOAT-MAST, which concluded that summer water temperatures were a major factor in the condition of Delta Smelt in 2017, stating at p.102: Given the long periods in July and August >22C we are confident that water temperature had a major negative effect on Delta Smelt in 2017 and is likely a primary factor in the lack of response of the Delta Smelt population to the high flows.  And at p. 104: Dynamic biotic components were somewhat better in 2017; however, the lack of response of the Delta Smelt population suggests that any benefits of changes in the habitat were minimal. (p. 3)

Comment:  It is true that recruitment of Delta smelt in 2017 was exceptionally low, despite the Fall X2 action.  This is because the number of adult spawners in 2017 was at a record low (Figure 2).

Fall 2017 water temperatures were slightly higher in the west Delta at Jersey Point compared to 2011 (Figure 3).  However, fall 2017 water temperatures were not unusually high compared to fall 2011 for Freeport or for Rio Vista in the Sacramento River channel of the Delta (Figure 4).  June and July Sacramento River water temperatures were substantially higher in 2017 than in 2011, because June and July Sacramento River flows were much lower in 2017 compared to 2011 (Figure 5). 

In general, spring habitat conditions were poorer in 2017 than in 2011:  2017 had lower spring Delta outflows (Figure 6) and much higher south Delta exports (Figure 7).

In sum, the 2017 fall index for Delta smelt was unusually poor because of poor conditions for Delta smelt in the spring and summer.  However, the response of longfin smelt to the implementation of Fall X2 in 2017 had a better outcome (Figure 8).  Despite poor number of spawners, the 2017 recruitment of longfin smelt per spawner was high. 

These data undermine Reclamation’s conclusion that the 2017 Fall X2 action had no benefit. 

Figure 2. Log vs Log plot of fall FMWT Index of Delta smelt (recruits) vs previous fall index (spawners). Blue years are wet years. Red years are dry and critical water years. Year types are determined by the California Department of Water Resources for the Sacramento River runoff to the Bay-Delta Estuary (

Figure 3. Comparison of Jersey Pt habitat conditions in 2017 versus 2011. Tidally filtered flow data were not available for 2011.

Figure 3. Comparison of Jersey Pt habitat conditions in 2017 versus 2011. Tidally filtered flow data were not available for 2011.

Figure 4. Freeport and Rio Vista water temperatures 2013-2019.

Figure 5. Freeport flow and water temperature summer 2011 and 2017.

Figure 6. Summer Delta outflows in 2011 and 2017.

Figure 7. South Delta federal exports (TRP) and state exports (HRO) in 2011 and 2017.

Figure 8. Longfin smelt spawner-recruit relationship, with improved recruitment in 2011 and 2017. Wet year blue, dry year red. Source:

So, after stating inaccurate and misleading reasons why the Fall X2 action is not effective, the Effects Analysis offers inaccurate and misleading reasons why not implementing Fall X2 and instead fussing with the Suisun Marsh Salinity Control Gates (SMSCG) will be a positive switch.

“Forecast of salinity conditions in the Delta indicate that operating to an X2 of 80 km along with SMSCG operations in September and October would result in suitable salinity conditions (< 11,000 uS/cm) in the western Delta including Suisun Marsh, Grizzly Bay, and Honker Bay during these two months.”  (p. 17)

Comment:  Operating the SMSCG tide gates, while not implementing Fall X2, will push more Delta outflow into Suisun Marsh, with less outflow reaching eastern Suisun Bay.  Both actions would potentially negatively affect Delta smelt compared to implementing the Fall X2 Action per the existing BiOp.

“However, as explained above, this estimate of abiotic habitat index does not account for the habitat created in Suisun Marsh through the operation of the SMSGC, which would increase the index.” (p. 31)

Comment:  moving outflow through SMSGC reduces habitat in eastern Suisun Bay, and subsequently traps any Delta smelt in Suisun Marsh once the gates are again closed.

Conclusion:  Stock-recruitment models show a strong positive population response for Delta smelt in 2011 (see Figure 2) and longfin smelt in 2017 (see Figure 8).  The strong population responses in 2011 and 2017 offer a strong case for implementing the Fall X2 action in 2019.  Reclamation’s only justification for eliminating the Fall X2 action in 2019 is to increase water available for export.  Biology has nothing to do with it.

It is Fall X2 Time Again

The 2008 Delta Smelt Biological Opinion for Central Valley Project (CVP) and State Water Project (SWP) includes the Fall X2 provision to keep brackish water west of the Delta through October in wet years to protect Delta smelt.  X2 or the low salinity zone is defined as the location where salinity is 2 parts per thousand (about 4000 EC). Keeping X2 at Chipps Island (km 71 from the Golden Gate) benefits the longfin smelt and Delta smelt populations.  In wet years, the smelt are protected by Delta agriculture salinity standards through August 15.  The Fall X2 provision keeps X2 and the smelt west of the Delta through October.

This is the third wet year since 2008 (the others were 2011 and  2017) in which the provision for X2 in the Biological Opinion has applied.  When I last posted about the 2017 implementation, I remarked that the US Bureau of Reclamation had requested not to apply the provision because of the high cost of water and minimal benefit to the smelt populations.  In the end, the provision was applied, but in an unusual way that likely had some new negative consequences on smelt and their Bay-Delta critical habitats.

Reclamation has again requested exemption from the provision in wet year 2019.  Again, Reclamation referenced the high cost of water and minimal benefits to smelt, despite unequivocal evidence that smelt benefit (Figures 1 and 2).

The Fall X2 flows are being applied with extra releases from Shasta, Oroville, Folsom, and New Melones reservoirs (Figure 3).  Without the X2 requirement that is creating inflows to the Delta of about 25,000 cfs, Delta inflow would only need to be 15,000 cfs to meet 65% export-to-inflow requirement.  With lower reservoir releases, river flows and Delta outflows would be lower at the discretion of Reclamation.  In the past, Reclamation has also failed on many occasions to meet water temperature standards in the lower Sacramento River and the American River, as well as south Delta salinity standards.

Application of the Fall X2 provision benefits smelt and the fall upstream migration of salmon in the Sacramento and San Joaquin rivers and their tributaries.  In Suisun Bay and the western Delta, salinities and water temperatures would be higher without Fall X2 (Figures 4-6).  Fall X2 application also helps maintain the water temperature requirements in Reclamation’s water right permits and the water quality standards in the lower Sacramento River between Red Bluff and the Delta.

Figure 1. Delta smelt spawner-recruit relationship. Note strong recovery in 2011. Source:

Figure 3. Bay-Delta hydrodynamics in late August 2019 under Fall X2 operations. Flows are average daily cfs. South Delta exports are in red. Red circle is location of X2.

Figure 2. Longfin smelt spawner-recruit relationship. Source:

Figure 4. Water temperature and salinity (EC) in eastern Suisun Bay, August 1, 2017 through November 1, 2018. Fall X2 was applied in 2017 (red outlined box). Note lower salinity and lower water temperature in 2017 compared to 2018.

Figure 5. Water temperature and salinity (EC) in the lower Sacramento River channel of the west Delta near Decker Island August 1, 2016 through November 1, 2018. Fall X2 was applied in August-September 2017 (red outlined box). Note lower salinity and slightly lower water temperature in 2017 during Fall X2 application compared to 2016 and 2018 without Fall X2.

Figure 6. Water temperature and salinity (EC) in the lower San Joaquin River channels near Jersey Point in the western Delta from August 1, 2016 through November 1, 2018. Fall X2 was applied in August-September 2017 (red outlined box). Note lower salinity and slightly lower water temperature in 2017 during Fall X2 application compared to 2016 and 2018 without Fall X2.