The Science of Fish Eating Fish

Are introduced predatory fish a major obstacle to restoration of salmon runs? Or are the levels of predation seen just part of the normal attrition that occurs between egg and returning adult? A workshop to answer questions about predation on juvenile salmonids in the Central Valley was conducted as a joint effort by the California Department of Fish and Wildlife (DFW), the National Marine Fisheries Service (NMFS), and the Delta Science Program (DSP) July 22-23 on the University of California, Davis campus.

Striped Bass - Reclamation photograph by Rene Reyes.
Striped Bass

A panel of independent experts answered a series of questions about the state of the science on fish predation on salmonids in the Bay-Delta and its watershed. The panel’s report was released on Oct. 1, 2013. Key findings included:

  • General Observations

    Although much information has been generated about losses of juvenile salmon in the Delta, that information does not provide “unambiguous and comprehensive” estimates of predation rates. Likewise, the information does not clearly show the overall effect of predation on Central Valley salmon and steelhead populations. Much of the information about predation rates comes from studies that use acoustic tags surgically implanted into hatchery-raised salmon and steelhead (see Science News June 2010 article), a technique that has some inherent biases. Other factors such as water temperature, loss of habitat, disease, and water pollution may also be contributing to predation losses by making fish less fit and more vulnerable to predation.

    One key recommendation was to develop a more comprehensive and standardized way of collecting, organizing, and sharing data about salmonids, predators, and environmental characteristics such as water temperature and turbidity.

  • Technical Challenges

    Predation can be an important factor affecting fish populations but is difficult to assess. Scientists can observe that predatory fish are present and that they eat the fish of concern—salmon and steelhead in this case. But it is very difficult to determine exactly how many are eaten and even more difficult to determine if this has an appreciable effect on the number that survives to adulthood. There are a number of scientific approaches for estimating predation rates and the impacts of predation on fish populations. These include:

    • Developing estimates of predator populations and the amounts of prey they consume

    • Bioenergetics approaches that are based on the amount of food (energy) needed for predatory fish to thrive and grow

    • Salmonid population modeling that use estimates of predation rates at different life stages

    All of these methods require a considerable amount of data that are rarely sufficient.

    Experiments to measure predation rates can help but also have their limitations. For example, it is possible to assess the effects of predator removal on salmonid survival within a stream segment, but it shouldn’t necessarily be assumed that these results can be extended to the entire length of a stream because local conditions affecting predation can vary widely.

  • Regional Differences and Hotspots

    The panel’s report summarizes the available information about salmonid predation rates in each of six Delta regions. The highest appears in the eastern, central, and southern Delta regions and is less of a factor in the northern Delta, the Cache Slough complex and western Delta. Hotspots for predation could have an impact on salmonid survival. Predation hotspots include Clifton Court Forebay and the Stockton Ship Channel. More study, however, is needed to better evaluate predation risk at both hotspots and other locations and identify ways to reduce it where high.

  • Proximate Causes vs. Ultimate Causes

    Although being eaten by another fish may be the immediate (proximate) cause of death, other factors such as harsh environmental conditions, may make a fish weaker or slower and may be the ultimate cause of death. Conditions that may make juvenile salmonids more susceptible to predation include the presence of toxic substances, elevated water temperature, altered physical habitat, and low dissolved oxygen. High mortality rates may have many causes and there is often disagreement about the relative importance of them.


“Conditions that may make juvenile salmonids more susceptible to predation include the presence of toxic substances, elevated water temperature, altered physical habitat, and low dissolved oxygen.”

--Predation Panel


  • A Caution about Tagging Studies

    Scientists frequently assume that most, if not all, losses of tagged salmon and steelhead are due to fish predation. However, the effects of handling and the physiological stress of the implanted tag must also be considered when attempting to determine predation rates from tagging studies. Another limitation is the use of hatchery fish that are often larger than the wild fish they are meant to emulate. Hatchery fish perform differently than wild fish, and larger fish generally have higher survival rates than smaller fish. Tagging studies are useful for determining relative differences in predation at different locations but are less useful for determining absolute predation rates.

  • Effects on Populations

    Predation at a given life stage of any fish species may or may not have a significant effect on the adult population. Heavy predation at any life stage can reduce the adult population but this is not always the case. Lower survival at an intermediate life stage can be offset by higher survival at later life stages. Fish population models can help scientists understand the effects of juvenile salmonid predation on the number of surviving adult fish. Different modeling approaches can be used; each with its own pros and cons. The panel points out that the available data and data management tools are not up to the job of providing scientifically reliable answers about the impacts of predation or the likely results of predation management.

  • Lessons from Other Systems

    Control of predator populations is a common approach for protecting species of concern across the United States. The panel cited examples from lamprey control in the Great Lakes to the pikeminnow control program on the Columbia River. Lamprey control, using chemicals that are selectively toxic to larvae, has been successful. Other efforts have had mixed results. The program to control the predatory pikeminnow in the Columbia River has successfully reduced the pikeminnow population. It has not, however, resulted in any detectable increase in salmonid populations. Because other predators might rush to fill the gap, the panel cautioned against assuming that reducing the population of any predatory species will increase the adult population of salmonids.

  • Monitoring and Research Recommendations

    The panel noted some important gaps in the current understanding of the effects of predation on salmonids in the Delta system and provided advice about how to improve the science related to predation. Scientists need a better understanding of both salmonid populations and the populations of their major predators. While it’s known that striped bass are likely important predators of juvenile salmon in the Delta system, and there is more information about them than other predatory fish species, some of the basics about their life cycle are still not fully understood. For example, the numbers of striped bass less than a year old have dropped dramatically, yet the adult population remains relatively high. Likewise it is not fully understood what controls populations of Chinook salmon.

    Proposals to reduce fishing regulation protections for predatory fish introduced into the Delta have been controversial because many of these are prized game fish.

    The panel recommends a coordinated modeling and monitoring effort that includes consistent sampling methods. Coordinated modeling supported by consistently collected monitoring data will provide managers with tools to predict likely outcomes resulting from management actions. A system-wide geographic information system (GIS) to organize physical (e.g. salinity and temperature) and biological data that is used consistently across studies and programs would be particularly useful. Thirteen specific research topic recommendations were also provided that include better estimates of predator and prey populations and larger scale predator removal experiments.