Brown trout in the Lees Ferry reach of the Colorado River—Evaluation of causal hypotheses and potential interventions
Over the period 2014–2016, the number of nonnative brown trout (Salmo trutta) captured during routine monitoring in the Lees Ferry reach of the Colorado River, downstream of Glen Canyon Dam, began increasing. Management agencies and stakeholders have questioned whether the increase in brown trout in the Lees Ferry reach represents a threat to the endangered humpback chub (Gila cypha), to the rainbow trout (Oncorhynchus mykiss) sport fishery, or to other resources of concern. In this report, we evaluate the evidence for the expansion of brown trout in the Lees Ferry reach, consider a range of causal hypotheses for this expansion, examine the likely efficacy of several potential management interventions to reduce brown trout, and analyze the effects of those interventions on other resources of concern. The brown trout population at Lees Ferry historically consisted of a small number of large fish supported by low levels of immigration from downstream reaches. This population is now showing signs of sustained successful reproduction and is on the cusp of recruiting locally hatched fish into the spawning class, based on analysis with a new integrated population model. The proximate causes of this change in status are a large pulse of immigration in the fall of 2014 and higher reproductive rates in 2015–2017. The ultimate causes of this change are not clear. The pulse of immigrants from downstream reaches in fall 2014 may have been induced by three sequential high-flow releases from the dam in November of 2012–2014, but may also have been the result of a unique set of circumstances unrelated to dam operations. The increase in reproduction may have been the result of any number of changes, including an Allee effect, warmer water temperatures, a decrease in competition from rainbow trout, or fall high-flow releases. Correlations over space and time among predictor variables do not allow us to make a clear inference about the cause of the changes. Under a null causal model, and without any changes to management, we predict there is a 36-percent chance the brown trout population at Lees Ferry will not show sustained growth, and will remain around a mean size of 5,800 adults, near its current size; in contrast, we predict there is a 64-percent chance that the population has a positive intrinsic growth rate and will increase 3–10 fold over the next 20 years. A humpback chub population model linked to the brown trout model suggests an increase of brown trout of this magnitude could lead to declines in the minimum adult humpback chub population over the same time period. Forecasts of rainbow trout abundance, however, suggest that increased abundance of brown trout in the Lees Ferry reach does not pose a threat to the rainbow trout fishery there. There are interventions that may be effective in moderating the growth of the brown trout population in the Lees Ferry reach of the Colorado River. Across causal hypotheses, we predict that removal strategies (for example, a concerted electrofishing effort or an incentivized take program targeted at large brown trout) could reduce brown trout abundance by approximately 50 percent relative to status quo management. Reductions in the frequency or a change in the seasonal timing of high-flow releases from Glen Canyon Dam could be even more effective, but only under the causal hypotheses that involve effects of such releases on immigration or reproduction. Brown trout management flows— dam releases designed to strand young fish at a vulnerable stage—may be able to reduce brown trout abundance to some degree, but are not forecast to be the most effective strategy under any causal hypothesis. We predict that the alternative management interventions would have effects on other resource goals as well, and the pattern of these effects differs across causal hypotheses. The removal strategies would incur direct costs (on the order of $7 million over 20 years) and the mechanical removal strategy is unethical from the perspective of several tribes. The strategies that involve reducing the frequency of high-flow releases from Glen Canyon Dam would decrease the ability to transport and store sediment in the ecosystem, potentially undermining goals associated with sandbar building, recreation, and riparian vegetation, but would increase hydropower revenue. Trout management flows would reduce hydropower revenue. From the standpoint of humpback chub, the alternative strategies largely follow the effect on brown trout; when brown trout abundance is reduced, predation pressure decreases, and humpback chub viability is predicted to increase, but the variation in predicted chub viability is not large across strategies or causal hypotheses. To design a response to brown trout, management agencies will need to navigate both the tradeoffs among resources goals and the uncertainty in the causes of the brown trout expansion. Continued monitoring, possibly coupled with new research or experimental management actions that better inform demographic and ecological dynamics, can help to reduce the causal uncertainty.