2009 Progress Report: Innovative Management Options to Prevent Loss of Ecosystem Services Provided by Chinook Salmon in California: Overcoming the Effects of Climate ChangeEPA Grant Number: R833017
Title: Innovative Management Options to Prevent Loss of Ecosystem Services Provided by Chinook Salmon in California: Overcoming the Effects of Climate Change
Investigators: Moyle, Peter , Thompson, Lisa C , Truan, Melanie L. , Mosser, Christopher M , Engilis Jr., Andrew , Purkey, David , Yates, David , Escobar, Marisa
Current Investigators: Moyle, Peter , Thompson, Lisa C , Truan, Melanie L. , Mosser, Christopher M , Engilis Jr., Andrew , Purkey, David , Escobar, Marisa
Institution: University of California - Davis , Stockholm Environmental Institute , National Center for Atmospheric Research
Current Institution: University of California - Davis , Stockholm Environmental Institute
EPA Project Officer: Packard, Benjamin H
Project Period: October 1, 2007 through September 30, 2010 (Extended to September 30, 2011)
Project Period Covered by this Report: October 1, 2008 through September 30,2009
Project Amount: $722,963
RFA: Nonlinear Responses to Global Change in Linked Aquatic and Terrestrial Ecosystems and Effects of Multiple Factors on Terrestrial Ecosystems: A Joint Research Solicitation- EPA, DOE (2005) RFA Text | Recipients Lists
Research Category: Global Climate Change , Ecosystems , Climate Change
This is an interdisciplinary study to investigate the implications of climate change for spring-run Chinook salmon and terrestrial ecosystems adjacent to spawning grounds, based on modeled changes in water temperature and flow conditions. Chinook salmon are a keystone species that link aquatic and terrestrial ecosystems and support directly or indirectly many ecosystem services in California watersheds. Annual runs of Pacific salmon historically transferred large quantities of marine-derived nutrients into the California interior through the decay of salmon carcasses, supporting forests, terrestrial animals, and other aquatic species. The geographic focus of the study is Butte Creek, which was selected because of its critical role in plans for the protection of threatened spring-run Chinook populations and the apparent vulnerability of the system to climate change. This vulnerability stems from the fact that the low elevation headwater regions of Butte Creek, and the West Branch Feather River from which water is transferred into the Butte Creek watershed, could potentially limit snow accumulation and melt which are important for the maintenance of suitable water temperature conditions for salmon. The complex set of non-linear interactions between habitat and ecosystems is being represented by coupling models of physical and biological processes. The integrated watershed hydrology, water management, and water quality model, WEAP (Water Evaluation and Planning System), is being applied to address: a) rainfall-runoff watershed processes, b) water supply, use, withdrawal, return flows, outflows, and reservoir operations; and c) water quality, most notably water temperature in streams and reservoirs which will be a critical determinant of the presence of the cold water conditions required by salmonids. WEAP outputs of water temperature and streamflow are passed to the SALMOD model, which simulates the population dynamics of anadromous and resident freshwater salmonid populations under variable ecological conditions. SALMOD results related to salmonid survival in Butte Creek under a range of future climate scenarios will provide information to an expert panel that will assess the plausible implications of these results for terrestrial ecosystem services in the communities adjacent to spring-run spawning grounds. Using model runs which assume alternative water management arrangements, the expert panel will also comment on the potential utility of strategies designed to better manage cold water resources and to increase the resilience of the ecosystem in the face of climate change.
Our research objectives are formulated as a set of questions, first for the aquatic and then for the terrestrial systems:
For the aquatic system:
· Could gradual climate change lead to abrupt change in the Butte Creek ecosystem through severe reduction or even extinction of its spring-run Chinook salmon population?
· If the spawning habitat of the spring-run were restricted to the upstream portion of the watershed, would salmon populations remain viable? How often would this restriction occur under climate change and what would this mean for the spawning spring run?
· In light of climate change and other stressors, how would suitable habitat for the salmon shift under the assumption that there were to be no management interventions (e.g. the “natural” watershed)?
· What management options are available to ameliorate climate change impacts on the spring-run Chinook, and at what cost to other ecosystem services (municipal and industrial water supplies, irrigated agriculture, hydropower, and recreation)?
For the terrestrial system:
· What would be the implications for the terrestrial ecosystem, particularly wildlife communities in the riparian corridor, if the spring-run Chinook salmon population were abruptly lost as the result of climate change?
· Are there longitudinal attributes of the riparian ecosystem of Butte Creek that make particular reaches more or less suitable for enhanced biodiversity with salmon present (e.g. are there local access restrictions due to physical landscape attributes such as extreme slopes)?
Our tasks for the overall project fall into three categories:
- Integrated Water Resource Modeling to determine habitat condition under current and future conditions.
- Population dynamics modeling to explore the ecological role of spring-run Chinook salmon in Butte Creek.
- Convene an Expert panel on possible terrestrial ecosystem impacts; run alternative management scenarios with select stakeholders
In the second year of the project our tasks involved model calibration and validation, scenario generation, and advance awareness:
Task A- WEAP Butte and Feather Watershed Model
A1- Refined water quantity and quality simulations of Butte and Feather systems, including stream and reservoir simulations of water quantity and temperature. Meso-habitat defined in Task B were used as the channel reaches for WEAP model of Butte Creek. Refined water management operations in system to be sure water is being appropriately allocated and reflects hydrologic realities.
A2- Began assembling climate change scenarios which will be downscaled to the sub-catchments and aggregated to a weekly time step. We ran model scenarios for simple 2oC, 4oC, and 6oC temperature increases in WEAP and SALMOD in order to provide a preliminary test of WEAP-SALMOD climate change predictions. We acquired the more detailed California Climate Change Center scenarios (http://meteora.ucsd.edu/cap/cccc_model.html), which are already downscaled to the sub-catchment level, which will be run in Year 3.
Task B-Spring Run Chinook
B3- Based on lessons learned from statistical analysis of fish and habitat data, evaluated the appropriateness of SALMOD relationships for spring-run Chinook in Butte Creek. Where appropriate, modified SALMOD to reflect appropriate bio-dynamics; ran and evaluated model simulations of spring-run Chinook populations.
B4- Began to extract flow and water temperature data from WEAP simulations of current and future conditions, and passed them to the SALMOD model for simulating population dynamics.
Task C- Conceptual Model/Food Web
Analyzed plant and animal tissue samples for marine-derived nutrient (15N) content. Finalized draft conceptual model of impacts of climate change to salmon and salmon nutrient footprint (food web diagram).
Task A Results - WEAP Butte and Feather River Watershed Model
Model Development – Rainfall-runoff Hydrology and Operations
The focus of our work during the second year of this project was on refining the rainfall-runoff hydrology and operations model of the Butte and Feather systems, implementing the stream temperature module of the spawning reach (Figure 1), developing a reservoir temperature module, and assembling climate change scenarios. We extended the rainfall-runoff model, which originally ended at week 39-2003, to week 39-2005, because of the existence of relevant biological data for spring-run Chinook salmon, then re-calibrated the rainfall-runoff routine using PEST, a non-linear numerical estimator that assists in adjusting model parameters. The final set of parameters allowed a more precise representation of baseflows, which are critical for the capacity of adult salmon to survive before spawning (Figure 2).
Figure 1. Location of Butte Creek watershed in California. Subwatersheds were defined upstream from management points. Reaches and subreaches were defined according to the California Department of Fish and Game delimitations.
Figure 2. Observed (black line) versus simulated (gray line) streamflow value at Butte Creek Nr Chico USGS station with new parameter set (top panel). The lower panel shows the baseflows in more detail.
Model Development – Water Temperature
Watershed-river Temperature Model
We implemented the water temperature module in the spawning section from reaches A to E. For this, we assigned specific stage-discharge-width relations to each pool-riffle-run sequence and used a linear relation between air and water to obtain water temperature for all the catchments in the model. We used PEST to calibrate these parameters to obtain the best fit for temperatures in the spawning reach. The statistics of this calibration indicate an overall good fit for the spawning reach, with a slight under prediction as indicated by the bias. This calibration was used to obtain temperature outputs for all habitat units within the spawning reach for SALMOD calibration.
Reservoir Temperature Model
The vertical temperature distribution within a reservoir is subject to seasonal variation, and the water can become highly stratified during the summer months. Of the two reservoirs in the system only Philbrook Creek Reservoir shows stratification (Figure 3), and constitutes an important management opportunity for the cold water pool within the Butte and Feather systems. Consequently, we implemented a one-dimensional reservoir temperature model to estimate the temperature distribution. This feature will be used to observe water management options and the sensitivity of the system to changes in water temperature imposed by climate change.
Figure 3. Modeled temperature (oC) profile versus Philbrook Reservoir depth (m) for weeks 4 and 46.
Climate Change Scenarios
We ran WEAP model scenarios for simple 2oC, 4oC, and 6oC temperature increases, to provide output data for preliminary predictions with SALMOD. We obtained bias corrected and spatial downscaled data for the Butte Creek watershed from the California Climate Change Center (http://meteora.ucsd.edu/cap/scen08_data.html).
Task B Results - SALMOD Modeling
Calibration of SALMOD consisted of two separate evaluation periods of model performance vs. collected data: Adult summer survival and juvenile production and out-migration. We used PEST to estimate SALMOD parameters. The best model fit for SALMOD summer adult temperature based mortality was obtained by performing a logistic regression within PEST using the weekly mean of daily maximum temperatures, with the second best fit using PEST estimation of the weekly mean of the daily mean of temperatures, though the fit was only marginally better. Calibration of SALMOD for juvenile out-migration presented a challenge. Many out-migrants are not accounted for in out-migrant trap sampling, so we used the adult return (capture) of fish tagged as out-migrating juveniles to estimate total out-migrants through a mark-recapture calculation, via a multinomial maximum likelihood estimation procedure.
Climate Change Prediction Results
Four simulations were run with SALMOD using the calibrated parameter set; baseline, and three climate change scenarios, simulating salmon production under 2, 4, and 6°C increases in air temperature. For each climate change scenario, the same number of adults was supplemented into the system as in the baseline scenario. The annual summer mortality of adults is predicted to increase as a result of increases in air temperature (Figure 4).
Figure 4. Annual number of adult spring-run Chinook salmon that died during summer for baseline conditions and under scenarios with 2, 4, and 6°C increases in air temperature, for the years 1997-2004.
Annual out-migration estimates under these scenarios indicates a distinct reduction in recruitment as projected air temperature increases (Figure 5).
Figure 5.Annual number of juvenile spring-run Chinook salmon out-migrating under baseline conditions and for three climate change scenarios, simulating 2, 4, and 6°C increases in air temperature, for the years 1997-2004.
Task C Results – Conceptual Model/Food Web
Expert Panel and Development of a Conceptual Model/Food Web of Chinook Salmon Subsidies to the Terrestrial Ecosystem
We continued development of a site specific, conceptual model/food web diagram. We conducted a literature review to prepare a target list of possible salmon consumers in the watershed, their trophic relationships, and their dependence on salmon-mediated nutrients. In our original proposal we anticipated that there might be gaps in the available secondary data sources. Because several of the floral and faunal studies that were to be performed by Pacific Gas & Electric under the Federal Energy Regulatory Commission relicensing process were never conducted, due to access issues and other difficulties, we took advantage of opportunities to augment this missing data by conducting our own limited field surveys in the watershed. We deployed motion detector scouting cameras during the spawning season to identify salmon consumers present in the watershed. We also collected plant and animal tissue samples, at different taxonomic, temporal and spatial scales, for stable isotope analysis to quantify the relative importance of spring-run Chinook salmon in transmitting marine derived nutrients to the riparian community.
Species presence-absence and attraction to salmon carcasses
Vertebrate species captured on motion detector scouting cameras included: numerous individuals of black bear, mule deer, beaver, gray fox, raccoon, ringtail, striped skunk, Western gray squirrel, common raven, band-tailed pigeon, mourning dove, various waterfowl, and turkey vulture. While it was unclear to what extent salmon served as a food source for these species, many individuals appeared to be at least investigating the carcasses, if not consuming them. Bears were clearly consuming salmon to an appreciable extent.
Conceptual food web model
A generalized, functional group-based conceptual model of nitrogen cycling in a hypothetical stream system following salmon subsidy of marine-derived nutrients has been developed (Figure 6).
Figure. 6. Nitrogen cycling in a hypothetical stream system. Arrows indicate direction of nutrient flow.
In the third year of this project we will conduct comprehensive model simulations, run additional stable isotope analyses, present our results to the Expert Panel, write reports, and give project presentations.
Tasks A, B, C
Finalize integrated modeling results and prepare results for synthesis of Expert Panel. Hold meeting of Expert Panel to review the results of all modeling to improve predictions of impacts of salmon extirpation on the terrestrial ecosystems and formulate recommendations for model testing. Develop Final report on hydrological, fish population, and conceptual modeling; pursue opportunities to present results at public fora and professional meetings through the Sierra Nevada Alliance; and prepare and submit manuscripts for peer review.