Northwest Fisheries Science Center

Ecosystem Analysis Projects


Dr. Beth Sanderson
Program Manager
Staff Directory

Conserving thermal habitat for Pacific salmon

Photo of researcher installing temperature sensors in a Snoqualmie watershed stream. Researcher Colin Sowder installs sensors, which record stream temperatures at 30–minute intervals for evaluations of temporal variation in thermal habitat.  Sensors were installed in over 30 locations throughout the Snoqualmie watershed. 

In aquatic animals, temperature can affect metabolism, growth, and reproduction, as well as response to pathogens and sensitivity to pollutants.  Thus temperature is a key driver of the ecological processes that control population and community structure in aquatic ecosystems. 

Across large watersheds, temperature partly defines species distributions.  At the reach or stream level, temperature can define the extent and connectivity of suitable habitats, including those used during critical life stages such as foraging and breeding. 

Human activities that have altered the thermal regime of streams and rivers include operation of hydropower dams, development of land adjacent to streams, and water withdrawal for irrigation. 

Climate change is predicted to further alter thermal and hydrologic conditions in streams and rivers across the Pacific Northwest.  For coldwater species such as Pacific salmon, real challenges to conservation can result from increases in temperature, altered variability in thermal regimes, and changes to the configuration of thermal refuges

Two charts showing temperature in the Middle Fork Snoqualmie River.  Top panel shows temporal variation and bottom shows spatial variation at one point in time for the same location. Two charts showing temperature in the Middle Fork Snoqualmie River.   Top panel shows temporal variation at one location and bottom panel shows spatial variation at one point in time for the whole river. 

To ensure success, many restoration plans seek to incorporate climate adaption strategies, which consider threatened stocks in terms of their vulnerability to expected changes.  

Recent advances in sampling technology enable us to compare patterns in temperature across hundreds of rivers throughout the Pacific Northwest.  With these data, we can characterize patterns in stream temperature that are important to salmon within and across reaches, rivers, or watersheds

Below we briefly describe completed, ongoing, and planned analyses that will enable us to conserve thermal habitat for Pacific salmon.  Our collaborative partners include the University of Washington, U.S. Geological Survey, and U.S. Forest Service, as well as specialists from the private sector.

Spatial Heterogeneity in Water Temperature


Our challenge is to characterize present and future spatial patterns of thermal habitat in order to map locations where Pacific salmon may be vulnerable in responding to expected changes in temperature.  By learning about features of thermal habitat that are important to fish, managers can better understand how to conserve and restore key habitats.  Research questions that guide our analysis include: 

  • How frequently do observed stream temperatures in rivers reflect our theoretical expectation of gradual downstream warming?
  • Are cold–water patches sufficiently large and closely spaced to meet the needs of salmon?
  • What are the environmental features that control spatial variability in water temperature?
  • Can we predict the influence of climate change on spatial patterns in thermal habitats?
Map showing with remotely-sensed stream temperatures on the Molalla River, Oregon.  Colors correspond to temperatures shown in key, with reds indicating warmer temperatures and blue, cooler temperatures). Map showing stream temperatures collected during August 2004 in the Molalla River, Oregon.  Red colors correspond to warmer stream temperatures, blue to cooler (flown by Watershed Sciences Inc.).

To address these questions, we are using remotely–sensed summertime stream temperature from hundreds of watersheds across the Pacific Northwest and California. 

We compared whole–river thermal profiles (stream temperature vs. distance) of over 60 rivers.  At right is an example of such a profile for the Molalla River, Oregon.  Both the map and graph at right illustrate thermal heterogeneity at multiple spatial scales.

Chart illustrates thermal heterogeniety across spatial scales within the Molalla River, Oregon.  Colors correspond with those shown the map above, with reds indicating warmer temperatures and blues, cooler temperatures. Graph of longitudinal thermal profile illustrates thermal diversity across spatial scales in the Molalla River, Oregon.

We also investigated spatial patterns of water temperature within rivers, looking at metrics such as the length and spacing of thermal "patches " of river below a biological threshold.  For example, we quantified what proportion of the river fell below 15°C, the length of each cold patch, and how far apart cold patches were spaced.  We looked at how these metrics differed among salmon species and with landscape characteristics such as stream size and elevation.

Key Findings

  • From our comparison across rivers, we learned that many longitudinal river temperature patterns did not fit our theoretical expectations (Fullerton et al. 2015).  Instead, we found that correlations with commonly used temperature predictors varied across surveys, showing complex spatial patterns in stream temperature that had no clear geographic trends.
  • In our analysis of within–river patterns, we found that warm patches were longer than cold patches, but there were fewer of them.  For many rivers, the length of cold patches and spacing between them was generally within distances that juvenile salmon can swim. 
  • When we included expected increases in temperature due to climate change, we predicted that the amount of habitat exceeding optimal thermal zones for salmon would increase in the future.  However, the size and spacing of remaining cold patches was predicted to be similar to those present today.  Perhaps most interesting, we found that the spatial resolution of water temperature measurements can have an even larger effect on predictions about spatial patterns. 
  • Evaluations of landscape controls on spatial thermal heterogeneity are in progress.

Temporal Variability in Water Temperature


To identify the potential implications for salmon of alterations to natural thermal regimes, we addressed research questions that would assess the temporal variability of water temperature at hourly, daily, seasonal, and annual scales. 

Photo of temperature logger and cable used to secure it within a stream. Temperature sensors, also called loggers, were cabled to trees or roots to ensure they would stay in place during winter floods.
  • Can daily, seasonal, or annual variations in temperature provide thermal refuges for fish at reach scales?
  • How will Pacific salmon respond to altered thermal regimes resulting from climate change?

To address the first of these questions, we used year–round temperature data collected from over 30 monitoring stations throughout a small watershed (pictured here), along with continuous spatial data from one summer. 

Focusing on cold–water patches identified with the spatial data, we used this dataset to consider whether fluctuations in temperature differ among sites, and to determine the duration of each cold–water patch. 

These evaluations may tell us how thermal refuges change over time, and whether such changes are consistent throughout the watershed. 

Photo of incubation chambers in the laboratory. Incubation chambers used in a laboratory experiment to evaluate the influence of temperature on survival and emergence timing of Chinook salmon eggs. 

To explore how salmon respond to altered thermal regimes, we used two separate approaches.  First, we conducted a laboratory experiment to determine whether different levels of thermal variation would influence the size, condition, development, or emergence timing of Chinook salmon. 

For this experiment, we exposed Chinook salmon to eight different thermal regimes during egg and alevin incubation.  Each temperature regime differed in frequency, magnitude, and duration, but all had the same mean temperature.  

Top photo shows eggs; bottom shows alevins. Chinook salmon eggs (upper panel) and alevins from the incubation experiment pictured above. 

Our second approach has been to begin development of a spatially explicit, individual–based model to help us evaluate the potential effects of altered thermal regimes on survival and smolt migration timing. 

Key Findings

  • We are continuing efforts to determine the temporal consistency of cold–water refuges.
  • Our experiments to determine the biological consequences of altered thermal regimes have shown that variation in thermal regimes during incubation does impact both hatch timing and fry emergence and development in Chinook salmon (Steel et al. 2012).
  • These impacts were observed independently of changes in mean water temperature.
  • Modeling of potential thermal effects on survival and smolt migration timing continues. 


Funding for this work was provided by the Northwest Fisheries Science Center Internal Grants Program (2009), the NOAA Fisheries Advanced Studies Program (2010–2012), a Eugene Maughan Scholarship from the Western Division of the American Fisheries Society (2013), and the North Pacific Landscape Conservation Cooperative (2014). 


Fullerton, A.H., C.E. Torgersen, J.J. Lawler, R.N. Faux, E.A. Steel, T.J. Beechie, J.L. Ebersole, and S.G. Leibowitz.  2015.  Rethinking the longitudinal stream temperature paradigm:  region–wide comparison of thermal infrared imagery reveals unexpected complexity of river temperatures.  Hydrological Processes 10.1002/hyp.10506.

Steel, E.A., A. Tillotson, D.A. Larsen, A.H. Fullerton, K.P. Denton, and B.R. Beckman.  2012.  Beyond the mean:  The role of variability in predicting ecological impacts of stream temperature.  Ecosphere 3(11). Article 104.

  Reference to trade names does not imply endorsement by the National Marine Fisheries Service, NOAA.