A Simplified Look at the Complex World of Fish Population Dynamics

I have a simplified approach in analyzing fish population dynamics from which I review the status of populations of smelt and salmon. It looks at the dynamics of the relationship between the number of spawning adults and their returning adult recruits one to several years later (Figure 1). In the fish science vernacular, it is sometimes referred to as the “spawner-recruit curve” or “stock-recruitment relationship” or simply “S/R relationship”. The major features of a S/R relationship are shown in Figure 1 (A, B, and C):

A. The blue and red curves show a standard spawner-recruit relationship, with higher spawners bringing more recruits – more eggs, more young, more smolts, more returning spawners, etc. It tails off when too many adults result in competition for food or spawning habitat, or higher rates of communicable disease – density-dependent effects.

B. The variability around the blue and red curves, shown by the vertical lines through the curves, is caused by density-independent effects such as drought, fishing harvest, or pollution that vary from year to year.

C. The difference between the blue and red curves, shown in the example as a yellow arrow, is a shift in the S/R curve that is a result in a fundamental shift in the relationship. Examples of such changes are the amount or quality of habitat from a dam being built, watershed destruction from a fire, loss of streamflow from new water diversions, loss of prey base, etc. The blue curve shows the S/R relationship before a fundamental shift; the red curve shows the S/R relationship after the fundamental shift.

Some environmental factors can affect one or more of the three features. For example, hatcheries can increase recruits (A and B), or they can cause a fundamental shift in the relationship (C) by imposing genetic changes in the population. Hatcheries benefit egg viability and fry survival, producing more smolts to the ocean per spawner in salmon populations, but may alter the wild component’s genetic viability.

The winter-run salmon population’s S/R relationship (Figure 2) exhibits these features, as well as the overall complexity in the relationship. Hatchery smolt introductions have propped up the population over the past two decades and increased its variability (red curve and vertical line), especially during periods of drought.

For longfin smelt, a state-listed species, there is a strong S/R relationship (Figure 3) to the features described in A-C above. There is a strong positive S/R relationship (A). There is a strong effect of the climate (B). And there appears to be a fundamental shift in recent years (C).

For Delta smelt (Figure 4), a state- and federally-listed species, which I consider nearly extinct at least in the wild, there was a strong S/R relationship (A), a climate effect (B), and a fundamental shift (C). The latter proved simply not sustainable, leading to a population crash that is not recoverable without supplementation (hatchery inputs) or drastic changes In environmental conditions.1 Note that 2016 is the last year in this figure, because the population since 2017 has been too close to zero to evaluate.

The largest salmon population, the Sacramento fall-run salmon, long sustained by hatchery inputs, is mainly controlled by feature B (Figure 5). Climate and water management are the dominant control of survival of hatchery and naturally-produced smolts reaching the ocean.

In conclusion, I recognize that S/R relationships represent a simplified view of extremely complex and changing relationships in the real world. Estimates of the number of spawners and recruits are often crude. But the relationships are real and statistically significant. It is up to us to interpret them by relating causal factors and developing hypotheses that can be tested with further scientific study and experiments. Unfortunately, managing fishery resources in the face of complex ecology, difficulty monitoring, natural variability, and statistical measurement errors is inherently difficult, even before political and economic factors get into the mix.

Figure 1. Spawner-Recruit relationships with three main features (A-C). See text for explanation of the features. In figures 2-4 below, the blue curve represents the historical S/R relationship. The red curve represents the new historical S/R relationship following a fundamental shift in the relationship, including long-term drought. The vertical lines through the curves show the range of the annual variability of the S/R relationship attached to each curve, excluding the density-dependent variability that is incorporated into the curve. In this example figure, the yellow curve tracks a fundamental shift in the S/R relationship. Spawners are shown on the x-axis; recruits are shown on the y-axis. The numbers on the axes are log transformed in order to make size of the figures manageable; log transformation does not alter the statistical relationships.

Figure 2. Spawner-Recruit relationship for winter-run Chinook salmon in the Sacramento River. Numbers shown represent the brood year of recruits (number of returning adults) for year displayed. For example, “11” represents fish produced in wet year 2011. The color of the number shows the conditions when brood was spawned and reared in the upper Sacramento River below Shasta Dam before emigrating to the ocean. A red number shows a dry year during spawning and early rearing. A blue number designates wet year spawning and rearing conditions. A green number designates normal water year conditions. For example, 15 represents brood-year 2015 recruits that returned in 2018, while its red color designates drought conditions in 2015. In this figure, numbers on axes are log-2 transformed.

Figure 3. The longfin smelt S/R relationship. The number and color represents the brood year’s fall index (recruits) and its water year type during its spawning run and first year of rearing. The spawners are the index from two years earlier. For example, the red number 15 represents the fall index for brood-year 2015 under water-year 2015 drought conditions, with spawners being the recruits from 2013. In this figure, numbers on axes are log-log transformed.

Figure 4. The Delta Smelt S/R relationship. I added two curves and a vertical line to an original figure to show the hypothesized S/R relationship; there is too little variability in the red curve for a vertical line to be meaningful.

Figure 5. Spawner-Recruit relationship for upper Sacramento River mainstem fall-run Chinook salmon. Number is recruitment year (escapement). Spawners are recruits from three years prior. Numbers are log minus 3 transformed. A red number shows a dry water year two years prior during rearing and emigration. A blue number shows a wet year two years prior. A green number shows for a normal water year two years prior. For example: red 17 represents 2017 run that reared in drought year 2015, with spawners (parents) being the 2014 green run number. Note that only one curve is shown. in gray, for this run of salmon, which is almost entirely dependent on hatchery production.

Water Projects’ Temporary Urgent Change Petition 3/18/22 Comment on Provision #1 – Spring (April-June) Delta Outflow and Salinity Intrusion

The U.S. Bureau of Reclamation and the California Department of Water Resources (Reclamation and DWR) filed a “Temporary Urgency Change Petition” (TUCP) on March 18, 2020.  If granted, the TUCP reduce Delta outflow requirements.  The proposed averaging requirements in the TUCP pose a problem in addition to the problem of too little overall outflow.

The TUCP states:

Reclamation and DWR are requesting to modify certain terms as the Projects’ storage and inflow may not be sufficient to meet D-1641 requirements and additional operational flexibility of the Projects is needed to support Reclamation and DWR’s priorities, which include: operating the Projects to provide for minimum health and safety supplies (defined as minimum demands of water contractors for domestic supply, fire protection, or sanitation during the year); preserve upstream storage for release later in the summer to control saltwater intrusion into the Sacramento-San Joaquin Delta (Delta); preserve cold water in Shasta Lake and other reservoirs to manage river temperatures for various runs of Chinook salmon and steelhead; maintain protections for State and federally endangered and threatened species and other fish and wildlife resources; and meet other critical water supply needs. (3/18/22 TUCP, p. 1)

The TUCP is requesting reduced Delta outflow requirements for the April 1 through June 30, 2022 period, for the stated primary purpose of preserving storage in Oroville and Folsom reservoirs.   What I term Provision #1 is reduction of outflow requirement from 3-day average of 7,6001 to 14-day average of 4,000 cubic feet per second (cfs).

The requirement in Revised Water Rights Decision 1641 of a 3-day average of 7,600 cfs is meant to keep salt water from encroaching upstream from the Bay into the Delta in drought years like 2021 and 2022.   This helps to protect the beneficial uses in the Delta including fish, fish low-salinity habitat, Delta agriculture, and south Delta water export water quality.  The TUCP’s proposed Delta outflow of 4000 cfs is meant to provide minimum protection in the face of low available water supply (reservoir storage and precipitation).

The overriding problem with the TUCP’s proposed flow reduction is that it does not require enough flow.  Yet, even accepting the need to reduce flow to allow storage of more water in Reclamation and DWR’s reservoirs, the requested change could be modified to better protect beneficial uses.

Under past TUCPs, DWR and Reclamation have used the 14-day averaging window to game operations to skate as closely as possible to the edge of compliance.  This has led to erratic outflows, often below 4000 cfs (Figure 1).  More precise estimates taking into account tides show outflow is lower than intended (Figure 2).  In these circumstances, salinity has increasingly moved up from the Bay into the Delta under such minimum freshwater outflow (Figures 3-6).

I recommend the outflow required be more stable, allowing for only minimal salinity increase over the spring.  A 3-day average of 4,000 cfs measured outflow would provide greater protection of beneficial uses.

Salinity criteria are more easily defined and measured, and more directly related to beneficial uses.  Criteria for Collinsville, Emmaton, Jersey Point, and Old River that have a maximum for a 3-day average or a maximum daily level with a minimum increase over the spring would also be more protective.

Figure 1. Daily Delta outflow estimated from measured Delta hydrology conditions in spring 2014, 2015, and 2021, drought years when previous TUCPs were implemented

Figure 2. Measured Delta outflow into Suisun Bay in spring 2021.

Figure 3. Salinity in eastern Suisun Bay in spring 2021.

igure 4. Salinity in lower end of Sacramento River near Rio Vista in late April 2021.

Figure 5. Salinity and net tidal flow in lower San Joaquin River channel at Jersey Point in spring 2021.

Figure 6. Salinity (EC) and net tidal flow in lower Old River channel in south Delta near Byron in spring 2021.

  1. The normal requirements are stated in Revised Water Rights Decision 1641, Table 3, footnote 10, pp. 185-186.  The April-June requirement in a Critically Dry year is also met if either the daily average or 14-day running average EC (measurement of salinity) at the confluence of the Sacramento and the San Joaquin rivers is less than or equal to 2.64 mmhos/cm (Collinsville station C2).  From May 1 through June 30, if the Sacramento River Index is less than 8.1 million acre-feet, the flows requested in the TUCP would be the same as the required flows under Decision 1641.

Radio-Tracking Study Greatly Advances Central Valley Salmon Science

Over the past five years (2018-2022), federal and state biologists have undertaken a comprehensive study of salmon smolt migrations down the 350 miles of the Sacramento River from below Shasta Dam to the Delta, Bay, and ocean.  The study released thousands of radio-tagged late-fall-, spring-, fall-, and winter-run hatchery and wild smolts in the Sacramento River near Redding (RM 290), then tracked their progress and survival as they proceeded to Butte City (RM 170), the Tower Bridge in Sacramento (RM 60), and Benicia Bridge at the head of San Francisco Bay.  Each release group provides a story that is helping our understanding of salmon science in the Central Valley.  Most important are lessons learned about drought and climate change.  Much has been learned, with more to come.

An Example: Releases of Radio-Tagged Winter-Run Smolts

Radio-tagged winter-run salmon smolts from the federal Livingston Stone Fish Hatchery were released to the Sacramento River at Redding in winter of years 2018-2022.  Their signals were detected from Butte City to the Benicia Bridge at the head of San Francisco Bay.  For example, Figures 1 and 2 show detections from a release at Redding on 2/14/2019 of 650 radio-tagged smolts.  Figure 1 shows detections over the winter and early spring at Butte City, 120 river miles downstream of Redding.  Figure 2 shows detections at the Tower Bridge in Sacramento, 230 river miles below Redding.

Preliminary Findings

A summary of release-group survival in Table 1 indicates the following:

  • Survival was significantly compromised in dry years 2018, 2020, 2021, and 2022 compared to wet year 2019.
  • Survival was poor in the upper river and the Delta.
  • December late-fall-run releases during wet periods in generally dry years had relatively high survival.

Other findings include:

  • Smolts that received thiamine “boosts” in the hatchery had slightly higher survival.
  • Approximately 20-25% of smolts are diverted into the central Delta at Georgianna Slough
  • Survival among release groups within a year was often related to flow and/or water temperature after release (Figures 3 and 4).
  • Fall-run salmon grown in 2020-2021 at the Coleman Fish Hatchery on Battle Creek had significantly higher survival when released at Butte City than when released directly from the hatchery about 80 miles upstream of Butte City.


Table 1.  Percent survival summary for release groups by run and year.

Winter Run Hatchery Chinook – released at Redding

Year To Butte City To Sacramento To Bay
2018 18.4
2019 64.5 23.3 25.6
2020 25.8 13.2 3.5
2021 25.8 10.1 3.6

Late Fall Run Hatchery Chinook – released at Battle Creek Hatchery

Year To Butte City To Sacramento To Bay
2018 0.17
2019 23.5 4.8
2020 76.1 60.4 16.9
2021 36.6 14.3 4.7
2022 75.5 17.1

Battle Creek Hatchery fall run – released at Battle Creek Hatchery

Year To Butte City To Sacramento To Bay
2019 46.0 22.0
2020 36.4 9.1 0.1
2021 0.3 0.0

Battle Creek Hatchery fall run – paired study

Year To Sacramento To Bay
2021 at Battle Ck 1.7 0.0
2021 at Butte City 26.3 4.7

Feather Hatchery spring run – released in lower Feather River

Year To Sacramento To Bay
2019 49.4 26.2
2020 26.8 2.6
2021 28.6 2.2

Wild Chinook Red Bluff Release

Year To Butte City To Sacramento To Bay
2018 3.2
2021 2.2 0.0 0.0

American River Hatchery Chinook – lower American River release

Year To Sacramento To Bay
2018 68.1 2.0

Butte Creek Wild Spring Run – lower Butte Ck

Year To Sacramento To Bay
2018 27.2
2019 16.3 1.5
2021 0.0 0.0

Battle Creek Start-up Hatchery Winter Run Chinook to Battle Creek

Year To Butte City To Sacramento To Bay
2019 23.3 14.0
2020 17.5 9.4 0.0
2021 11.5 3.3 0.2

Figure 1. Detections at Butte City of hatchery winter run smolts released at Redding with flow at Butte City gage 2/14/2019.

Figure 2. Detections at Tower Bridge in Sacramento of hatchery winter run smolts released at Redding with lower Sacramento River flow at Wilkins Slough 2/14/2019.

Figure 3. Survival rate to Sacramento of wild Butte Creek spring run radio-tagged release groups with Butte Slough flow in 2018.

Figure 4. Survival rate to Sacramento of wild Sacramento River fall run radio-tagged smolts, Red Bluff release groups with lower Sacramento River flow and water temperature at Wilkins Slough in 2018.

Wishful Thinking on The Upcoming 2022 Salmon Season

We should be careful about wishful thinking in considering the forecasts for the 2022 California salmon season along the coast and in major rivers.1  The 2022 ocean abundance projection for Sacramento River fall Chinook, a main salmon stock harvested in California waters, is estimated at 396,500 adult salmon, a higher number than the 2021 forecasts. However, we should expect further drought-related fishery and escapement downturns in 2022 and 2023, as occurred after the 2007-2009 and 2013-2015 droughts (Figure 1). We can also expect low fishery catches, especially in rivers from low summer flows and associated high water temperatures that keep river spawners in the Bay until rivers cool in the fall.

The issues relate primarily to remaining stocks of Chinook salmon, not to the nearly extinct endangered Coho. The 2022 season also relates mostly to the availability of brood year 2019 salmon from Central Valley rivers, not coastal streams. Finally, it is important to remember that most of the adult salmon in 2022 will come from hatcheries on the Klamath, Trinity, Sacramento, Feather, American, Mokelumne, and Merced rivers, especially those hatcheries that truck and barge salmon to the Bay or coast.

Hatchery returns, the backbone of California salmon fisheries, are likely to be lower in 2022 than expected. Near 40 million Chinook salmon smolts from brood year 2019 were released from California federal and state hatcheries in 2020. Of the total, about 26 million were raised at Sacramento River watershed hatcheries, 9 million at San Joaquin River hatcheries, and 4 million at Klamath-Trinity hatcheries. Most of the releases were fall-run (34 million), with about 4 million spring-run, 1 million late-fall-run, and 0.4 million winter-run. Their returns of jacks and adults are likely to be lower in 2022 than expected because of low Delta inflows and outflows (Figure 2) and low Klamath-Trinity flows (Figures 3 and 4) in winter-spring 2020 and 2021, which are likely to result overall in relatively poor returns from river and Bay releases of hatchery smolts.2

Wild salmon stocks, already severely depressed, are going to further decline and contribute even less to fisheries. Natural production of brood years 2019 and 2020 was likely poor because wild smolts faced dry-year conditions in winter-spring 2020 and 2021, respectively.

Adult salmon that return to spawn in 2022 will face warm rivers, as they did in 2021 (Figures 5 and 6). They will be delayed, and many will die before spawning.

In summary, we must be careful about wishful thinking about the future based on the recently released ocean abundance estimate for 2022. Many other factors point in a downward direction. Perhaps the most immediate question is whethersalmon will simmer again in what looks to be a dry, hot summer.

Figure 1. Sacramento River Basin fall-run salmon escapement 1975-2020.

Figure 2. Sacramento-San Joaquin Delta outflow 2014-2022.

Figure 3. Klamath River flow at Orleans 2014-2022.

Figure 4. Trinity River flow at Hoopa 2014-2022.

Figure 5. Flow and water temperature in the lower Sacramento River at Wilkins Slough in 2021. Red line is 20ºC (68F) water quality standard and safe upper level for salmon.

Figure 6. Water temperature at Wilkins Slough in 2021 and recent historical average.

Salmon Released from Central Valley Hatcheries

Hundreds of thousands of smolts of endangered salmon are being released to the lower Sacramento River from federal and state hatcheries this winter, the San Francisco Chronicle reported on March 2.1 It is a drought year, and most of the young salmon will likely perish in the jaws of predators on their 300-mile journey to the ocean through the river, Delta, and Bay. Without the protection of adequate river flow, cold water, and turbidity for the journey, there is little hope. Reservoir storage is depleted from the past two years of overzealous water deliveries. What little rain and snowmelt there has been in the Valley is being stored in reservoirs. Warm weather has come early. The 400 thousand smolts released on March 2 could use some help, but they have not received it and are not likely to get it.

River Flow

Flow from Shasta Reservoir to the upper river via Keswick Dam has been minimal (Figure 1, KWK). 40 miles downstream at the Bend (BND) gage near Red Bluff, the river received some snowmelt from tributaries in January, but even that tapered off by the end of January. Flow into the Delta from the Sacramento Valley at Freeport (FPT) reached minimal levels by the end of January. Delta outflow (Figure 2) settled at its minimal prescribed level near 12,000 cfs after some January export taking.

Water Temperature

Warm weather and low flows have allowed water temperatures in the lower Sacramento River and north Delta migration corridor to reach the mid-fifties (Figure 3), temperatures that activate lower river predators like striped bass, resident trout, and pikeminnow.


Turbidities (Figure 4) are now at seasonal lows from the upper river near Red Bluff (RDB) to the Delta near the Rio Vista Bridge (RVB), reflecting clear waters needed by sight-feeding predators like striped bass.

Delta Salmon Salvage

Hatchery salmon smolts released in December at least reached the Delta during the December storms (Figure 5). It remains to be seen if the February and March smolt releases even reach the Delta.

What’s Needed

The article in the Chronicle mentioned that only one of the 140 radio-tagged hatchery smolts released in February has reached the Delta. The March 2 release of 400,000 smolts could use a pulse of cold turbid water from Shasta, Oroville and Folsom reservoirs to help them reach the Bay. So far, only Oroville releases have increased, from 3500 to 5500 cfs in recent days, with the added 2000 cfs showing at Freeport (Figure 1). While this helps near the Delta, it does little for the upper 200 miles of river the smolts have yet to transit.

Figure 1. River flows near Keswick (River Mile-RM 300), Bend (RM 250), and Freeport (RM 50) in winter 2022.

Figure 2. Delta outflow in tidally averaged daily cfs in winter 2022.

Figure 3. Water temperature in Sacramento River near Bend (RM 250), Wilkins Slough (RM 140) and Freeport (RM 50) in winter 2022.

Figure 4. Water turbidity in Sacramento River near Bend (RM 250), Red Bluff (RM 240), Freeport (RM 50), and Rio Vista (RM 30) in winter 2022.

Figure 5. Salmon salvage observed at Delta Fish Facilities in water year 2022.