Lessons in American River Hatchery Management Based on Returns from 2016 and 2017 Salmon Smolt Releases

The American River provides approximately 20-40% of the Central Valley fall Chinook run each year. Most of the American River run is derived from hatchery smolt releases from the Nimbus American River Hatchery. In this post, I suggest hatchery and water management actions to increase the ratio of adult returns per hatchery smolt releases to enhance fishery benefits from the American River hatchery.

This post builds on earlier posts. In a July 2020 post, I described how the American River was “shortchanged” on hatchery production, resulting in declining runs and less contributions to ocean and river fisheries. In a May 2019 post, I supported a hatchery strategy that included smolt releases to the river in wet years because river releases survived as well as smolts trucked to the Bay in wet year 2011. I also promoted trucking to the Bay in drier years, because of poor adult returns from river smolt releases in dry years.

Recent data indicate that trucking provides benefits even in wet years, with limited drawbacks. Water year 2017 was the first wet year since 2011. Returns from the 2017 releases indicate a distinct advantage of trucking smolts to the Bay over river releases (Figure 1). Trucking to the Bay enhanced returns of adult salmon by 250%, although river smolt releases had reasonably good survival (>1%). Returns from below-normal water year 2016 smolt releases (Figure 2) were poor for river releases as expected, but were also lower for Bay releases compared to 2017 releases.

Detailed results from the 2016 and 2017 release groups support trucking smolts to the Bay even in some wet years. Water years 2016 and 2017 provide a good contrast in hatchery smolt return patterns for dry and wet year smolt release strategies, focusing on the differences observed between river and Bay releases. In both years, all smolt releases occurred in late May and early June. Past strategies of late April or early May releases have proved less successful, despite the potential concerns of poor late spring river conditions. It appears that larger smolts have a greater chance of survival regardless of the release strategy.

2016 Smolt Releases

  • River smolt releases in 2016 were in mid-May. River conditions were good, with flows of 4,000-5,000 cfs and water temperatures below 60oF (Figure 3). These smolts then had to contend with marginal Delta and Bay conditions (Figures 4-6) on their journey to the ocean. Water temperatures in the Delta in the two weeks following the releases rose from 66 oF to 73 oF (stressful to near-lethal levels), as Delta outflow declined from 14,000 cfs to 7000 cfs.
  • Bay smolt releases in 2016 were in early June, with Delta outflow around 7000 cfs (Figure 4). Water temperatures were stressfully warm around the Bay release site at 66-70 oF (Figure 6).
    2017 Smolt Releases

2017 Smolt Releases

  • River smolt releases in 2017 were in late May. American River conditions were good, with flows of >6000 cfs and water temperatures below 60oF (Figure 3). These smolts also had high flows (>50,000 cfs) and near optimal Delta and Bay conditions on their journey to the ocean.
  • Bay smolt releases in 2017 were in early June under near optimal conditions, with Delta outflow around 20,000-30,000 cfs (Figure 4).

River vs Bay Releases

  • Returns from river smolt releases continued to be very poor in dry year 2016 (0.1-0.3%, Figure 2), while being measurably improved in a wet year 2017 (1.03-1.23%, Figure 1).
  • Returns from Bay smolt releases in dry year 2016 (0.9-1.67%, Figure 2) were lower than in wet year 2017 (2.64-2.71%, Figure 1).
  • Overall, Bay releases in dry year 2016 had return rates 3 to 10 times higher than river releases. In wet year 2017, Bay release returns were 2.5 times higher than returns of river releases.

Bay releases in a wet year and a dry year

  • Bay releases provide higher returns with higher Delta outflow (Figure 7), continuing the pattern previously noted. Survival in wet year 2017, when outflow exceeded 10,000 cfs-20,000 cfs, was nearly double that in dry years 2016. A combination of lower net transport flows and higher water temperatures (Figure 6) in 2016 likely explains why returns from Bay smolt releases in 2016 was 50% of returns from Bay smolt releases in 2017.

Straying of adult returns from Bay and river releases

Straying of adult returns originating from the American River hatchery to other rivers remains an issue with off-site hatchery smolt releases (trucking to the Bay). Though the proportion of strays is low (10-30%, Figures 8 and 9), it is a concern for two reasons:

  1. The number of adult spawners reaching the American River is lower.
  2. Hatchery strays can reduce the genetic integrity of the salmon population of the American River as well as other river populations, with unknown but potentially serious long-term consequences.

Straying occurs from Bay releases especially in dry years, because the hatchery smolts partially imprint on Mokelumne River water occurring at the release site near the mouth of the Napa River in eastern San Pablo Bay. Returning adults get mixed signals in late summer and early fall in the west Delta at the confluence of the Sacramento and San Joaquin rivers. Flows from both rivers have their signature at the confluence, because the Delta Cross Channel (DCC) is open all summer and early fall. Half the American River water can pass through the DCC and come out the San Joaquin channel. As a result, many American fish head up the San Joaquin channel and then up the Mokelumne forks.

Another theory is that because about a third of Mokelumne Hatchery smolt releases to the Delta and Bay wind up returning to the American River to spawn, their offspring are genetically coded to the Mokelumne.

What if?

What if more of the American hatchery smolts were released in the Bay or along the coast? That would increase returns about 50% in wet years (assuming all the returns were 2.5% in Figure 1) and more than double returns in dry years like 2016 (assuming all the returns were 1.25% in Figure 2.) Assuming conservatively that the hatchery contribution to the run is about 80-90 percent, the American fall-run salmon escapement of adults (Figure 10) would increase by over a third. A similar increase in fishery contributions from the American hatchery would also be likely. Since the American contributes about 20% to the overall Central Valley fall run escapement (Figures 10 and 11), Valley-wide contributions to fisheries and escapement would increase 5-10% with more Bay releases.

What if Delta outflow were increased in dry years in late spring during Bay hatchery smolt releases? That could increase dry year returns 50% or more (2015-2018 in Figure 10).

What if hatcheries only spawned their own river genetic coded adults and strays were removed on their way to spawning reaches? And what if Bay smolt releases were barged to Bay release sites instead of trucked? Such actions would reduce straying and improve the genetic integrity of each river’s populations.

What if hatchery smolts were not released in rivers? There might be less competition with wild smolts and reduced effects of predation and competition on wild steelhead fry that predominately emerge from redds in late spring.

Summary and Conclusion

The contribution of American River hatchery salmon smolts to adult spawner returns and fisheries would be significantly higher if these hatchery smolts were all trucked to the Bay. Straying to other Central Valley rivers would increase. However, competition with and predation on wild salmon and steelhead in the American River would decrease.

Figure 1. Percent survival (adult returns) of six tag groups of American River Hatchery releases of fall run salmon smolts in spring 2017. Data source: https://www.rmpc.org

Figure 2. Percent survival (adult returns) of six tag groups of American River Hatchery releases of fall run salmon smolts in spring 2016. Data source: https://www.rmpc.org

Figure 3. Comparison of 2016 and 2017 flows and water temperatures in spring.

Figure 4. Delta outflow in spring 2016 and 2017. Dots represent fall-run hatchery smolt releases. Blue outlined dots are river releases. Red outlined dots are Bay releases.

Figure 5. Sacramento River water temperature in the west Delta near Sherman Island in spring 2016.

Figure 6. North Bay water temperature near Bay hatchery salmon smolt release site in 2016.

Figure 7. Percent return of selected tagged hatchery smolt groups released to Bay by Delta outflow at the time of release with 2016 and 2017 noted. Data sources: https://www.rmpc.org and CDEC.

Figure 8.  Returns of American River origin adult hatchery fall-run to four state hatcheries from 2016 drier year releases to river and Bay.  Data source: https://www.rmpc.org

Figure 8. Returns of American River origin adult hatchery fall-run to four state hatcheries from 2016 drier year releases to river and Bay. Data source: https://www.rmpc.org

Figure 9. Returns of American River origin adult hatchery fall-run to four state hatcheries from 2017 wet year releases to river and Bay. Data source: https://www.rmpc.org

Figure 10. American River salmon escapement 1975-2019. Data source: https://nrm.dfg.ca.gov/FileHandler.ashx?DocumentID=179102&inline=1

Figure 11. Central Valley escapement of four salmon runs 1975-2019. Note the fall run Chinook dominate. Data source: https://nrm.dfg.ca.gov/FileHandler.ashx?DocumentID=179102&inline=1

Scott River Chinook Salmon Update

The Scott River Chinook salmon, a key contributor to the overall Klamath River salmon run, are in major trouble.  In a November post, I had a “mixed” prognosis for this year’s fall run.  Well the numbers are now in – a record low, bleak run of 117 spawners observed (Figure 1) at the weir downstream of Fort Jones.

Figure 1. Scott River fall run salmon escapement 1978-2020. Source: CDFW unpublished data.

The poor run can be directly attributed to lack of fall river flow, a fact that I had addressed in a 2017 post.  Salmon simply cannot ascend the lower Scott River into Scott Valley spawning grounds from the Klamath River because of lack of streamflow.  Some may spawn in the steep canyon below the Valley (and counting weir), but poor spawning habitat and low flows in the canyon offer little solace for the salmon.

Poor fall flows (Figure 2) can be directly attributed to fall groundwater extraction and surface water diversions for hay-pasture irrigation.  The State Water Board should stop crop irrigation after October 1.  This irrigation practice has been getting worse over the past several decades, aided by improved well extraction and sprinkler technology and greater demand and higher revenues.  Present water use permits allow irrigation into December, which ranchers have been taking advantage of to get an extra crop of hay (with the help of climate change).

Unlike 2020 (Figure 2), water use in past drought years tapered off earlier and flows increased during October (Figures 3-5).  This allowed fall-run salmon access to the Valley.  In contrast, recent wet and normal years see a combination of precipitation and reduced water use, which enhances fall flows (Figures 6 and 7).

In conclusion, the State Board should limit fall water irrigation in Scott Valley to save the salmon.  The Sustainable Groundwater Management Act (SGMA), passed in September 2014, requires local agencies to develop Groundwater Sustainability Plans (GSP) that will assess and project future groundwater conditions, and provide management and monitoring activities.  The Scott River basin is a priority basin.  Siskiyou County is required to develop and submit a GSP for the Scott River basin by January 31, 2022.  A preliminary plan recently developed by the advisory group suggests reducing irrigation acreage (Figure 8) to increase streamflow (Figure 9).  That would help, but what salmon need is a cutoff of irrigation by October 1.  An option for further augmentation is to employ unused groundwater extraction wells in the fall to add water to the river  for short periods.  Stored water in the tailings ponds (red area in Figure 8) could also be gravity-fed or pumped into the river at critical times.

Figure 2. Scott River flow fall 2020. Water year 2000 was a drought year.

Figure 3. Scott River flow fall 2000 and winter-spring 2001. Water year 2001 was a drought year.

Figure 4. Scott River flow fall 2013 and winter-spring 2014. Water year 2014 was a dry year.

Figure 5. Scott River flow fall 2014 and winter-spring 2015. Water year 2015 was a normal year.

Figure 6. Scott River flow fall 2016 and winter-spring 2017. Water year 2017 was a wet year.

Figure 7. Scott River flow fall 2017 and winter-spring 2018. Water year 2018 was a below normal water year.

Figure 8. Baseline (present) and preliminary action alternative for Scott Valley irrigation. Source: preliminary plan.

Figure 9. Analysis of preliminary action alternative. Source: preliminary plan.



American River Nimbus Hatchery – Fish Ladder and Public Health Update

The new fish ladder at the Nimbus Fish Hatchery is coming along.  It will help in hatchery operations and allow salmon and steelhead a bit more spawning habitat because it no longer requires installation of a seasonal weir that keeps fish from migrating to the short stretch of river that is upstream of the hatchery and downstream of Nimbus Dam.

I visited the site on the Saturday after Thanksgiving.  I missed the river in this Covid year.  I thought with the recent limited stay-at-home order the popular viewing area would not be too crowded for a brisk walk with a dog to see the new ladder and a whole lot of salmon.  I was surprised by the crowd, but found a parking spot.  I donned my mask and headed for the viewing area.  Hundreds of viewers with dogs, bikes, and strollers.  There was limited mask wearing and social distancing.  It was actually kind of scary given my many months of minimal human contact.  Having a masters in public health did not give me greater confidence.  I did my best to avoid everyone and follow the recommendations adopted by the Sacramento County Health Department.

I suggest CDFW put up appropriate warning signs and require masks and social distancing at its popular venue.


Yuba River Salmon 2020

In a 2017 post and a 2018 post, I related the status of Yuba River Chinook salmon runs. The 2017 spawning escapement estimate was a record low. The 2018 and 2019 runs were not much better (Figure 1). While the record low 2017 can be blamed predominately on the 2013-2015 drought, the poor 2018 and 2019 runs cannot. Water year 2016 was a normal water year and 2017 was a wet year.

The likely culprit in the decline in Yuba escapement is the continuing persistent decline of wild spawners and increase in hatchery strays (greater than 90% of the run), leading to the erosion of the locally adapted Yuba River salmon population. This was the diagnosis for the overall Feather River population by Willmes et al., 2018.

Figure 1. Yuba River fall run salmon escapement 1953-2019.

A further look at the composition of the hatchery strays in the Yuba escapement surveys provides added clues about the cause of the recent decline in overall escapement to the Yuba River. In 2016 (Figure 2), nearly half the tag returns were strays from the Battle Creek hatchery, while the other half were a combination of spring-run and fall-run strays from the Feather River hatchery, along with a smattering of strays from the hatcheries on the American and Mokelumne rivers. In 2017 (Figure 3), Yuba tag returns featured an even greater proportion of strays from Battle Creek, the American, and the Mokelumne. In 2018 (Figure 4), about half of the tag returns were from the Mokelumne River hatchery, and a third were of Feather River spring-run hatchery origin.

Feather-tagged spring-run show up consistently in the Yuba escapement surveys. This is unusual, because spring-run make up only about 10% of the Feather hatchery smolt production, with fall-run being about 90%. One reason is that spring-run smolts are 100% tagged, while fall-run hatchery smolts are only 25% tagged. Another reason is that all the spring-run smolts are released near Gridley just upstream from the mouth of the Yuba, whereas most of the fall-run smolts are released in the Bay. A third reason is that in most years springtime flows in the Yuba are higher and colder than those in the Feather, and are thus more likely to attract returning adults. Regardless of the reason, the fact that a significant portion of Yuba “fall run” escapement is derived from spring-run hatchery smolts is cause for concern.

Battle Creek hatchery fall-run made up about half the tag returns in 2016 and 2017. Much of the smolt production from the Battle Creek hatchery was trucked to the upper Bay in 2014 and 2015, a strategy prone to increased straying.1 No Battle Creek hatchery smolts were trucked to the Bay in 2016, and none showed in the Yuba in 2018. With over 10 million fall-run smolts produced at the Battle Creek hatchery (federal Coleman Hatchery near Red Bluff), nearly double the Feather hatchery production, it is easy to see why Battle Creek hatchery salmon could dominate the Yuba escapement. Without the Battle Creek, American River, and Mokelumne River hatchery strays in 2017, the record low escapement in the Yuba River would have been far worse. The fact that most of the 2017 spawners were hatchery strays from rivers other than the Feather should also be cause for concern.

In conclusion, the escapement of fall-run salmon to the Yuba River has declined over the past five years, approaching record-low levels. Spawners are now predominately hatchery strays from smolts released in the Bay and along the coast. Natural “wild” Yuba River smolt production is virtually nonexistent. This is a crying shame for what is widely regarded as one of the best non-hatchery salmon rivers in the Central Valley. I have spent many days on the Yuba River over the past two decades. I observed the big runs at the turn of the century and in 2013 and 2014. The river’s pools turned purple with adult salmon. Dead and dying salmon filled the backwaters, feeding eagles and buzzards. The odor was prevalent. Spawning salmon and redds were everywhere. Such occurrences are now rare. Our Yuba River needs so many fixes, a subject for another post.

Figure 2. Composition of tag returns from Yuba River spawners in 2016. FRS = Feather River spring run. MRF = Mokelumne River fall run. FRF = Feather River fall run. ARF = American River fall run. BCF = Battle Creek fall run. MeRF = Merced River fall run. Data source: rmis.org.

Figure 3. Composition of tag returns from Yuba River spawners in 2017. FRS = Feather River spring run. MRF = Mokelumne River fall run. FRF = Feather River fall run. ARF = American River fall run. BCF = Battle Creek fall run. Data source: rmis.org.

Figure 4. Composition of tag returns from Yuba River spawners in 2018. FRS = Feather River spring run. MRF = Mokelumne River fall run. FRF = Feather River fall run. ARF = American River fall run. Data source: rmis.org.