The long and winding road of salmon trucking in California

Spawning Chinook salmon at Nimbus Hatchery on the American River (image by Barry Lewis)

By Dr Anna Sturrock

Trucking juvenile hatchery salmon downstream is often used in the California Central Valley to reduce mortality during their perilous swim to the ocean. But is it all good? Researchers at UC Berkeley, UC Davis, UC San Francisco and NOAA Fisheries published an article in Fisheries this month exploring the history and implications of salmon trucking in a changing climate.

When I moved from England to California in 2012 to start a postdoctoral position at UC Santa Cruz, I distinctly remember the feeling of awe. Everything was on a bigger scale, and anything seemed possible. The patches of green English fields were replaced by never-ending rows of crops and fruit trees, gold-scorched hills, and craggy mountains. Our calmer, smaller waterways were replaced by fast-flowing rivers fed by dams the size of skyscrapers. Even the ocean waves were bigger here. When I started hearing about behemoth pumps that sucked so much water they reversed the river direction, and baby salmon hitching rides in trucks, I wasn’t sure if people were being serious. It turns out they were.

California is a land of plenty, boasting diverse, beautiful, vast landscapes and an abundance of natural resources. It also hosts a diverse and ambitious population of visionary thinkers with an unparalleled can-do attitude. Undeterred by its wildly variable Mediterranean climate of long, hot, dry summers and multi-year droughts, or wet, stormy winters that often flooded major towns, Californians have created a network of imaginative engineering solutions to store and move water around the state. One of the most transformative solutions to manage its boom-bust weather patterns was to build huge dams along the foothills of the Sierra Nevada to store winter rains and snow, providing water ‘banks’ to sustain domestic and agricultural users through the long, dry summers and droughts. These dams provide significant benefits to humans, supplying water to extensive agri-business and urban populations. However, the diversion of water, particularly during droughts, and the fragmentation of river habitats has drastically affected many aquatic species. For example, migratory fishes like Chinook (or King) salmon are often prevented from ascending the rivers to high elevation habitats where they would literally chill over the summer, eliminating the historically dominant spring run salmon from most Central Valley streams.

To mitigate for lost salmon production above the dams, five production hatcheries were built between 1944 and 1970 – Coleman National Fish Hatchery, and four state-operated hatcheries – Nimbus, Mokelumne, Feather, and Merced River Hatcheries. When I first moved here I wasn’t totally sure what hatcheries were: European media tends to focus on fish farms (where fish are kept in pens until they end up on your plate), generating horror stories about over-crowding, sea lice and escapees. So I was pleasantly surprised when I first visited a hatchery and talked to the people working there. I liken hatcheries to salmon IVF clinics (Fig. 1). Only these IVF clinics are not just addressing parent fertility issues, but providing luxury birthing suites, daycares and calorific dinners to the babies that would have otherwise been supported by the upstream habitat. Hatcheries then release the babies into the wild, typically once they have grown into smolts (‘fat toddlers’) to complete an otherwise natural life cycle.

The objectives of this study were to explore (a) when, where, and how many hatchery juveniles were released each year, (b) whether release patterns had changed through time, and (c) where these fish ended up if they were lucky enough to survive to adulthood.

Spawning Chinook salmon at Nimbus Hatchery on the American River (image by Barry Lewis)

In the early years (1940s to 1960s) everything was fairly low key compared with today – hatcheries tended to allow their youngsters to swim directly into the adjacent river where they would mingle with the wild fish, and they would swim to the ocean together (Fig. 2; also see our accompanying web app). In intermediate years (1960s to 1990s), the hatcheries intensified their role. For example, it was not a popular option to cull excess production (more babies than the hatchery could support), so – particularly in wet years – juveniles were often trucked to small, remote creeks that do not typically support their own salmon populations (e.g., Secret Ravine, Doty Ravine). Most of these fish were tiny and unmarked so we don’t really know what became of them – their offspring are probably an interesting study in themselves. However, over time managers and stakeholders got wise to the poor survival of juveniles during their perilous migration through the Sacramento-San Joaquin River Delta – historically, a diverse, mosaic of wetland habitats, now a channelized water conveyance system packed with introduced predators and contaminants. Thus, from 1980 onwards, trucking hatchery salmon directly to the ocean really took off, particularly during droughts. The argument is that chauffeuring fish past mortality hotspots (typically worse during droughts) gives them a survival advantage that boosts commercial and recreational fisheries. The idea of putting fish into a truck was a real culture shock for me. Faced with the same issues in the UK, I am pretty sure Boris Johnson would shrug his weary shoulders, mumble something incoherent, and we would all just have to tolerate low returns (of both hatchery and wild salmon) after droughts. Here, in historic drought year 2015, almost all the hatchery salmon produced (about 26.5 million individuals) were loaded into trucks and driven to the Delta, bays, or ocean (Fig. 2). That takes a lot of people, gas, and money. It also creates an even wider survival gap between hatchery and wild fish, contributing to the dominance of hatchery fish on nearly all Central Valley rivers.

Fig. 2 Densities (numbers) of hatchery salmon released during years characterized by (1) On-site releases (represented by each hatchery’s first release year); (2) non‐natal releases (spreading them out – e.g., in wet year 1998); or (3) trucking them to the bay (e.g., drought year 2015). ‘Bay’ indicated by a dashed line. Hatchery codes: COL = Coleman, FEA = Feather, NIM = Nimbus, MOK = Mokelumne, MER = Merced. Figure from Sturrock et al. (2019). If you want to download the release data or explore it in more depth, check out the accompanying web app.

Another consequence of trucking is that trucked salmon are much more likely to stray into other hatcheries or rivers when they return to spawn. Normally, young salmon create olfactory maps when they swim to the ocean, sequentially recording the smells they encounter along the way. If they are among the lucky few that survive to adulthood, they use these olfactory memories like street signs to find their way home. Trucked salmon have large gaps in their maps, making it harder for them to navigate home. We analyzed tagging data for salmon released by the five hatcheries since the start of the Constant Fractional Marking Program (an excellent program involving system-wide tagging of 25% of hatchery releases and increased tag recovery efforts to see who came back and where). We found that the further the salmon were trucked from the hatchery, the more likely they were to stray into a different river when they came back (Fig. 3). The effect was most extreme for hatcheries on smaller, more distant rivers, and when natal stream flows were lower during the period of return. We also found that fish returning at older ages were more likely to stray. These older individuals may simply be more forgetful, but it is more likely that their longer ocean residence time correlated with larger changes in the freshwater environment (i.e., their map became ‘outdated’), particularly in the Central Valley’s highly engineered waterways.

 Why do we care about straying? Some level of straying is natural and can help maintain genetic diversity, expand range boundaries, and recolonize impacted habitats. However, in a natural system, most individuals return home, allowing populations to adapt to local stream conditions (e.g., thinner individuals in shallower spawning grounds) and increase fitness (i.e., more babies per adult). The main concern is that excessive straying (often >80% of trucked juveniles strayed as adults) eliminates existing local adaptation, makes it harder for local adaptation to re-evolve, and can introduce maladapted genes into recipient populations. Excessive straying can also have more immediate impacts. For example, after Coleman (the biggest producer in the system) trucked all their fish to the Delta in 2015, so many of these fish strayed elsewhere when they returned in 2017 that Coleman was unable to meet its production goals (i.e., make enough babies for the next generation). Some argue that Central Valley salmon – being at the edge of the species distribution and more prone to drought and disturbances – might have higher baseline straying rates than their northerly counterparts. However, we found that the hatchery fish released on-site had “normal” straying rates, averaging 0.3% to 9.1% (i.e., if reducing straying rates were the only objective, then releasing on-site does seem to work).

Fig. 3 Observed (circles) and predicted (lines) straying indices of California Central Valley hatchery fish as a function of transport distance and return age (other model covariates averaged). Indices based on coded wire tag recoveries from brood years 2006–2012 (this study; circles) and 1980–1991 (Niemela 1996; crosses). Circles sized by the logged number of tag recoveries (used to weight model). Predicted straying rates for 3‐year‐old fish at minimum and maximum natal stream flows during the return period indicated by dashed lines (specifically, the range of mean October–November flows in return years 2008–2015, displayed above each plot). River distance from each hatchery to the bay indicated by an arrow (note, MER did not perform bay releases during the years examined). Hatchery codes defined in Fig. 2. Figure from Sturrock et al. (2019).

Hatcheries are often controversial and not everyone likes them. But without the Central Valley production hatcheries there would not be much of a salmon fishery in California, and – managed well – they could hold the key to salmon persistence in a rapidly changing climate. Trucking may help supplement the fishery in a given year, but it comes at a cost (impeding local adaptation and increased competition for food, mates, habitat – both from straying and the survival advantage provided by the trucking itself). Long-term, the combination of such high straying rates and such a large survival imbalance could reduce the stability of these populations and the fishery.

Fig. 4 Ranse Reynolds (retired CDFW Nimbus Hatchery Manager) at his home after interviewing him about some of the more vaguely described release locations.

Today, we are at an ecological tipping point, and California’s climate is predicted to become increasingly volatile and prone to hotter, longer droughts. To counter this uncertainty perhaps we should consider managing our fish and environment using a cautionary, risk-spreading approach that promotes long-term resilience, even if this occasionally leads to short-term losses in returns. Ideally, we would manage salmon stocks for both resilience and abundance, by (1) reducing straying rates of hatchery fish (e.g., using flow-through barges, segregation weirs, terminal fisheries, and attraction flows) and (2) enhancing the abundance and survival of natural-origin salmon (e.g., by increasing habitat carrying capacity via restoration and flow management). Many fish and water agencies, NGOs, stakeholder groups, and hatcheries are already exploring ways to achieve both objectives. Coleman and Mokelumne hatcheries have led the way in experimenting with alternative release strategies, and we are encouraged to see other hatcheries also trying broader release periods, and releasing fish closer to the hatchery in recent years. Such experiments may be expensive and challenging to perform, but – if carefully coordinated and repeated across water years – they will be crucial to informing management decisions in the future.

Central Valley Chinook salmon are at the edge of the species range and are clearly a tough breed – having already persisted in the face of multiple human impacts and extreme droughts. By trying new tools and working as a team – coordinating across watersheds, managers and stakeholder groups – we may be able to alleviate some of the issues we have created, and help these tough fish persist in a rapidly changing world.


This study built on the painstaking work carried out by Eric Huber transcribing decades worth of data from hatchery release reports into an electronic database. My first task was to add GPS coordinates to often incredibly vague site descriptions (e.g. “Misc.” or “Dry Creek” – do you know how many Dry Creeks there are in California?!). This involved a lot of interviewing (hassling) current and retired hatchery employees. I remember Anna Kastner digging out hand-drawn maps from the 70s from the Feather River Hatchery basement, Marc Provencher going through piles of physical planting receipts at Coleman, and Ranse Reynolds (retired Nimbus Hatchery manager – Fig. 4) and his wife Joyce, zooming around google maps from their home in Woodland while their grandkids played with my son. The historical insights were fascinating, and I was encouraged by how forthcoming and helpful everyone was. I cannot thank you enough. Thanks also to my dad, Barry Lewis, for the wonderful pictures of Nimbus Hatchery. Thanks also to Arnold Ammann, Walt Beer, Mark Clifford, Laurie Earley, Fred Feyrer, Brett Galyean, Ted Grantham, Scott Hamelberg, Tim Heyne, Paula Hoover, Rachel Johnson, Brett Kormos, Dave Krueger, William Lemley, Joe Merz, Carl Mesick, Cyril Michel, Kevin Niemela, David Noakes, Gary Novak, Bob Null, Mike O’Farrell, Kevin Offill, Jim Peterson, Corey Phillis, Rhonda Reed, Edward Rible, Paco Satterthwaite, Ole Shelton, Jim Smith, Ted Sommer, Bruce Sturrock, Lynn Takata, Mike Urkov, Judy Urrutia, Dan Webb, Peter Westley, Michelle Workman, and Steve Zeug for their comments, advice, support, and/or provision of data. Funding was provided by the CDFW Ecosystem Restoration Grant (E1283002), the Delta Science Fellowship Program (Award no. 2053) and CDFW Water Quality, Supply and Infrastructure Improvement Act of 2014 (CWC §79707[g]) (P1596028). I am also indebted to the massive efforts of all my coauthors – Stephanie Carlson, Will Satterthwaite, Kristina Cervantes‐Yoshida, Eric Huber, Hugh Sturrock, Sébastien Nusslé – and to my other mentor, Rachel Johnson, for giving me my first proper job, taking me to my first hatchery, and inspiring my obsession with salmon!

Anna Sturrock (@otolithgirl) is an Assistant Project Scientist at the Center for Watershed Sciences using natural and applied tags to reconstruct fish growth and habitat use. She is passionate about science communication and data visualization, and providing empirical data to support and inform natural resource management. Senior co-authors Drs Stephanie Carlson and Will Satterthwaite can also be found in the twittersphere as @fishteph and @satterwill.

Further reading

Sturrock, A. M., Satterthwaite, W. H., Cervantes-Yoshida, K. M., Huber, E. R., Sturrock, H. J. W., Nusslé, S., & Carlson, S. M. (2019). Eight Decades of Hatchery Salmon Releases in the California Central Valley: Factors Influencing Straying and Resilience. Fisheries, doi:10.1002/fsh.10267 (linked here)

Web app to visualize the release data across time and space:

Huber, E. R., & Carlson, S. M. (2015). Temporal trends in hatchery releases of fall-run Chinook salmon in California’s Central Valley. San Francisco Estuary and Watershed Science, 13(2) (linked here)

Niemela (1996) – Effects of release location on contribution to the ocean fishery, contribution to hatchery, and straying for brood years 1987-1991 fall Chinook salmon propagated at Coleman National Fish Hatchery. USFWS. Northern Central Valley Fish and Wildlife Office, Red Bluff (linked here)

California Hatchery Scientific Review Group (2012). California Hatchery Review Report. Prepared for the USFWS and PSMFC (linked here)

This entry was posted in Uncategorized. Bookmark the permalink.

2 Responses to The long and winding road of salmon trucking in California

  1. Phil Isenberg says:

    Dr. Sturrock: An impressive and well-written article. My congratulations. Keep writing! Phil Isenberg

  2. Susan Jean Sturrock says:

    A fascinating insight. Thank you!

Leave a Reply