Decision-making under uncertainty: Prioritizing freshwater habitat restoration for salmonid recovery

Jody Brauner
Ph.D., Quantitative Ecology and Resource Management

Introduction
The pursuit of clean water, clean air, and endangered species protection has raised nationwide interest during the late 20th century. As impacts to our natural resources continue to grow, the need to understand ecological systems and manage them effectively becomes increasingly important. Threats to declining salmon populations in the Pacific Northwest have resulted in the listing of 15 salmon species as threatened or endangered under the Endangered Species Act (ESA). With more listings anticipated in the near future, every county in the region will be directly or indirectly affected. In response, federal and state agencies in collaboration with Native American tribes and local watershed groups are rapidly developing recovery strategies. However, current salmonid recovery efforts are plagued with scientific uncertainty. Research has demonstrated that species survival is linked to the freshwater habitat health and function, but which specific components? To what extent? And under what conditions? These are the key questions that decision makers are asking. Without such knowledge, they are forced to allocate funds arbitrarily, without knowing the likelihood of achieving ecological success.

In the pursuit of watershed habitat protection and restoration, it is imperative to understand the causes of habitat degradation and quantify the effects on species survival. Responding to this resource management imperative, this project will synthesize current studies of riparian habitat and stream channel dynamics, and develop analytical methods for prioritizing restoration alternatives.

 

Project Rationale and Objectives
There have been many notable studies of habitat function and species suitability conducted throughout the Pacific Northwest (Abbe and Montgomery 1996, Montgomery et al 1995, Bilby and Ward 1991, Bisson et al. 1987, Grette 1995, Murphy and Koski, 1989). However, these studies typically lack quantitative measures of essential habitat conditions, species-specific requirements, and land use impacts. In addition, current estimates of smolt production and fishery escapement are commonly based on a limited number of studies, most of which were conducted in a wide range of watershed conditions, and very few of which have been published in peer-reviewed literature (Baranski 1989). While scientists agree the inherent uncertainties are problematic, most studies continue to apply the conventional and deterministic methods of regression analysis. Using such methods assumes no uncertainty in scientific knowledge. This is clearly unrealistic. Thus a scientific question evolved: Can quantitative tools be used to prioritize restoration alternatives given the uncertainties inherent in endangered species recovery efforts?

Here in the Pacific Northwest, timber harvest, agriculture and urban development have drastically altered freshwater systems and the species they support. To demonstrate the application of probabilistic tools coupled with a decision-making framework, this study is focused on steelhead trout habitat degradation. The research objectives of the study are threefold: 1) estimate the historic and current carrying capacity of juvenile steelhead trout, 2) simulate the effects of various restoration alternatives on channel morphology (pool formation), and 3) prioritize various restoration scenarios using Bayesian decision analysis.

 

Site Selection
The South Santiam watershed, located in eastern Linn County, Oregon, encompasses approximately 3200 km2 in the western Cascade headwaters of the Upper Willamette Basin. Dating back to the 1850's the watershed has supported diverse land use activities including urban, agricultural, domestic, hydroelectric, recreational and industrial development, predominantly in the form of timber harvest. Extensive land use and the construction of Foster Dam have significantly impacted the watershed health and function. However, despite these modifications, the basin remains comparatively less disturbed than other watersheds in the region.

Steelhead in the Upper Willamette Basin were listed as threatened under the ESA in March 25, 1999. This regulatory mandate, along with the availability of steelhead counts and associated habitat data made the South Santiam Watershed a candidate study site. Ultimately it was selected above other sites in the Willamette Basin due to the active public interest of local watershed councils, Oregon State University and the Environmental Protection Agency. For the past several years, this triad has been developing a broad scale decision-making strategy for the Basin (refer to http://biosys.bre.orst.edu/restore/default.cfm); however, the scope of their efforts did not include measures of fish abundance or capacity. Given the 1999 ESA listing, the triad welcomed the addition of my study and has generously provided data and scientific support. Such widespread and community-based support increases the likelihood applying my research findings and restoring critical freshwater habitat.

 

South Santiam Background
 Fires, floods and long-term forest harvest have virtually eliminated wood from the channels in the South Santiam Basin. Without instream structure to trap substrates and organic matter, sediment is rapidly flushed through the high-energy system, scouring the once complex systems down to a uniform bedrock base. In addition to facilitating sediment transport, the lack of wood simplifies the hydrological character of the stream, replacing pools and ponds with riffles and rapids. The consequence is a system-wide reduction in channel complexity and overall steelhead carrying capacity. This change in habitat conditions has most significantly impacted the rearing lifestage of steelhead trout [pers. comm. Wayne Somes, Forest Service (FS) district biologist and Gary Gallovich, Oregon Department of Fish & Wildlife (ODFW) district biologist]. Spawning habitat has also been reduced, but to a lesser extent in terms of production potential.

 

Methodology
The study is composed of seven tasks:

  • Assess historical South Santiam habitat conditions in an effort to characterize "predevelopment" watershed function. This will bound the channel complexity simulated in Task 5
  • Characterize current South Santiam habitat conditions.
    • Identify habitat types according to the ODFW survey protocol
    • Compile relevant survey data including channel width, slope, and gradient, large woody debris (LWD) frequency and size, and substrate composition.
    • Characterize riparian vegetation and management strategies
  • Identify rearing habitat requirements for winter-run steelhead trout based on existing literature
  • Estimate current carrying capacity of juvenile steelhead trout based on Tasks 2 & 3
    • Employ the Marshall et al. (1980) steelhead production model for age 1+ steelhead
  • Habitat Response Models - simulate changes in channel complexity under alternative restoration scenarios
    • LWD abundance model- estimate the effects of modified forest management scenarios using a tree growth model (Organon) coupled with the Riparian-in-a-Box II (RIAB). Together these models will calculate riparian stand development, LWD recruitment, and LWD depletion (decay and transport)
    • Substrate composition model - estimate changes in substrate composition and abundance under modified forest management scenarios
    • Channel morphology model - estimate changes in pool frequency, area and spacing under modified forest management scenarios
  • Bayesian Decision Analysis (Probabilistic Approach)
    • Calculate the posterior probabilities for each model parameter using the Markov-Chain Monte-Carlo numerical integration method.
    • Use posterior probabilities to calculate the distribution of consequences of different restoration activities. Carrying capacity for a particular alternative will be projected 100 years into the future and conditioned on the values and uncertainty of model parameters
    • Measure biological success of carrying capacity for each restoration alternative (median and confidence bounds) calculated over a 100-year time horizon
  • Costs-Benefit Analysis for each restoration alternative
    • Calculate which restoration alternative or combination of alternatives will produce the most salmon for any level of expenditure. This will be of particular interest to decision groups needing to prioritize a suite of restoration projects given a fixed budget.

Restoration scenarios will range from the mechanical placement of LWD to long-term changes in riparian management zones (i.e., buffer widths, thinning cycles, harvest strategies). The analysis will enable land managers and decision-makers to evaluate the long-term effect of physical changes on biological organisms. Whenever possible, scenarios will reflect actual projects that have been proposed/constructed.

 

Project Value
 The methodology and results of this study will advance the application of Bayesian decision analysis to complex environmental management issues. It is imperative to not only understand the physical and biological requirements of salmonids, but to connect an emerging scientific understanding to current management goals and practices. Environmental protection issues, such as habitat degradation or endangered species, are filled with data gaps and scientific uncertainty. To successfully steward the preservation and restoration of our ecosystems, solutions must cross disciplines and incorporate new methods.

The proposed research is a case study that will illustrate the potential of data synthesis and analysis in the context of habitat restoration and endangered species recovery efforts. The techniques and lessons learned in the South Santiam will be applicable to other watersheds facing similar management challenges. By employing a probabilistic approach rather than the traditional methods of regression analysis, the Habitat Response Models will portray a more realistic assessment of our current knowledge. Relying on analytical methods that ignore uncertainty falsely reflects our scientific knowledge and influences management decisions. Using a more realistic representation of functional relationships, decision-makers can more accurately evaluate tradeoffs and make a fully informed judgement

 

Reference
Abbe, T. B. and D. R. Montgomery. 1996. Large woody debris jams, channel hydraulics and habitat formation in large rivers. Regulated Rivers: Research and Management 12:201-221.

Baranski, C. 1989. Coho smolt production in ten Puget Sound streams. Washington Department of Fisheries Technical Report #99. 29p.

Bilby, R. E. and J. W. Ward. 1991. Characteristics and function of large woody debris in streams draining old-growth, clear-cut, and second-growth forests in southwestern Washington. Canadian Journal of Fisheries and Aquatic Sciences 48:24-2508.

Bisson, P. A., R. E. Bilby, M. D. Bryant, C. A. Dolloff, G. B. Grette, R. A. House, M. L. Murphy, K. V. Koski, and J. R. Sedell. 1987. Large Woody Debris in Forested Streams in the Pacific Northwest: Past, Present, and Future. In Streamside Management: Forestry and Fishery Interactions. Edited by E. O. Salo and T. W. Cundy. College of Forest Resources, University of Washington, Seattle, Washington. pp. 143-190.

Grette, G. B. 1985. The role of large organic debris in juvenile salmonid rearing habitat in small streams. Master's thesis. University of Washington, Seattle.

Marshall, D., H. Mundie, P. Slaney, and G. Taylor. 1980. Preliminary review of the predictability of smolt yield for wild stocks of chinook salmon, steelhead trout, and coho salmon. SEP Management Committee. Report based on a workshop held in Vancouver, B.C. June 17, 1980.

Montgomery, D.R., Buffington J.M, and R.D. Smith. 1995. Pool Spacing in forest channels. Water Resources Research, Vol. 31, No.4, pp. 1097-1105.

Murphy, M.L and K.V. Koski. 1989. Input and depletion of woody debris in Alaska streams and implications for streamside management. North American Journal of Fisheries Management. 9:427-436.

 

Scientific Collaborators: Environmental Protection Agency, National Research Center for Statistics in the Environment National Oceanic and Atmospheric Administration, Northwest Fisheries Science Center