A Model-Based Assessment of the Potential Response of Snake River Spring–Summer Chinook Salmon to Habitat Improvements

—The current recovery strategy for threatened Snake River Chinook salmon Oncorhynchus tshawytscha relies heavily on improvements to the quality of freshwater spawning and rearing habitat; however, the potential survival benefit from these actions is unknown. To address this issue, we created a model for predicting the early freshwater survival rates (egg to smolt) of this species as a function of five easily measured physical habitat variables and used this model to evaluate survival rates under five alternative future habitat states. Model validation showed that the predictions were reasonably accurate for individual stocks as well as for the trend in predictions across stocks. The results for the future habitat scenarios suggest that the potential for improving survival rates through habitat restoration is high for a few populations and low to nonexistent for most others while the potential for reduction in survival rates due to reduced habitat quality is great for all populations. The effects of modeled egg-to-smolt survival rate changes should be evaluated across the entire Chinook salmon life cycle to assess recovery potential. Snake River spring–summer Chinook salmon Oncorhynchus tshawytscha were listed as a threatened evolutionarily significant unit of Pacific salmon under the Endangered Species Act in 1992 due to the decline in spawner numbers observed during the late 20th century (NMFS 1992). Habitat degradation, hydropower development, hatchery practices, and overharvest have all been implicated as causal factors in this decline (Fleming 1994; Lichatowich 1999), and efforts aimed at mitigating the effects of these factors have been numerous and costly (GAO 2002). In recent years, several model-based evaluations of restoration alternatives have been made by various combinations of state, tribal, and federal agency personnel (e.g., the Cumulative Risk Initiative [CRI; Kareiva et al. 2000] and the Plan for Analyzing and Testing Hypotheses [PATH; Peters and Marmorek 2001; Wilson 2003]), with the primary focus on the configuration of the Federal Columbia River Power System dams in the main-stem Snake and Columbia rivers. While some modeling analyses predict that breaching four dams on the lower Snake River has * Corresponding author: [email protected] Received May 27, 2003; accepted October 26, 2003 a high probability of furthering the recovery of Chinook salmon, the potential for improving survival rates in other life stages in the absence of dam breaching remains less certain. The results of the PATH modeling analysis suggested that of the actions under consideration recovery options including dam breach have the highest probability of achieving recovery standards (Peters and Marmorek 2001). On the other hand, the CRI modeling forum concluded that recovery could be achieved without breaching Snake River dams but instead directing efforts at increasing survival rates through the early life stages via restoration of tributary habitats, where the spring–summer Chinook salmon (hereafter referred to as salmon unless indicated otherwise) spend more than a year of their life before going to sea (Kareiva et al. 2000; but see Wilson 2003). While the relative benefits to overall life cycle survival from main-stem and tributary actions remain unknown, the current recovery strategy relies heavily on improvements to tributary and estuary habitats in lieu of dam breach (Federal Caucus on Salmon Recovery 2000; NMFS 2000). The future of salmon populations in the Snake River basin rests heavily on the efficacy of this recovery strategy, as all existing broodlines of spring–summer Chinook salmon have been fore623 SALMON RESPONSE TO HABITAT RESTORATION FIGURE 1.—Map of the Snake River basin showing the general locations of the index stock spawning and rearing streams used in this analysis. Abbreviations are as follows: ELK 5 Elk Creek, SUL 5 Sulphur Creek, MIN 5 Minam River, UGR 5 upper Grande Ronde River, BSC/LIC 5 Big Sheep and Lick creeks, and IMN 5 Imnaha River. casted to be extinct as early as 2012 (Mundy 1999). This, coupled with the considerable amount of time required to realize the benefits of some habitat restoration efforts (e.g., channel changes resulting from the removal of livestock were not observed for 24 years in one stream; Kondolf 1993), demonstrate the importance of completing a fieldbased, quantitative assessment of the potential for improving early life stage survival rates based on the proposed habitat improvements. Such an evaluation could address the feasibility of achieving recovery under the current strategy and may provide a template for use in the prioritization of habitat restoration efforts. Our primary objective was to create a model that could be used in such an evaluation. To do this, we developed a simulation model that predicts egg-to-smolt survival rates as a function of simple physical habitat variables that (1) have strong survival rate linkages (i.e., are mechanistically explainable) and (2) are targeted for improvement under the Final Basinwide Salmon Recovery Strategy (Federal Caucus on Salmon Recovery 2000). Here we report on model development and validation as well as provide a relevant example of the use of our model to inform decisions on recovery plans for Snake River salmon. Methods Study site.—The Snake River Basin covers a large area characterized by a great diversity of geologic, climatic, habitat, and land management conditions. To best capture this diversity, we selected a subset of six indicator stocks for both field sampling and modeling efforts. These index stocks are associated with long-term data sets (nearly 50 years in most cases) on population trends and for this reason have been used in previous modeling assessments (e.g., Kareiva et al. 2000; Peters and Marmorek 2001). For our purposes, we selected a subset of index stocks based on general habitat quality (summarized in Beamesderfer et al. 1997) and the availability of survival rate data; thus, we included streams that were greatly impacted by past and present land management activities as well as those that are nearly ‘‘pristine.’’ Field habitat and survival rate data were collected for six index stocks, namely, those from the upper Grande Ronde River (UGR), Big Sheep Creek and Lick Creek combined (BSC–LIC), the Imnaha River (IMN), Elk Creek (ELK), Sulphur Creek (SUL), and the Minam River (MIN) (Figure 1). Reported habitat quality varies considerably between and within the six streams selected for study.