Malheur Lake is the largest freshwater marsh in the western contiguous United States and is one of the main management units of the Malheur National Wildlife Refuge in southeastern Oregon. The marsh provides excellent waterfowl production habitat as well as vital migration habitats for birds in the Pacific flyway. Water shortages have typically been a problem in this semiarid area; however, record snowfalls and cool summers have recently caused Malheur Lake to rise to its highest level in recorded history. This has resulted in the loss of approximately 57,000 acres of important wildlife habitat as well as extensive flooding of local ranches, roads, and railroad lines. Because of the importance of the Refuge, any water management plan for the Malheur-Harney Lakes Basin needs to consider the impact of management alternatives on the hydrology of Malheur Lake.

The facilitated modeling workshop described in this report was conducted January 14-18, 1985, under the joint sponsorship of the Portland Ecological Services Field Office and the Malheur National Wildlife Refuge, Region 1, U.S. Fish and Wildlife Service (FWS). The Portland Field Office is responsible for FWS reporting requirements on Federal water resource projects while the Refuge staff has management responsibility for much of the land affected by high water levels in the Malheur-Harney Lakes Basin. The primary objective of the workshop was to begin gathering and analyzing information concerning potential fish and wildlife impacts, needs, and opportunities associated with proposed U.S. Army Corps of Engineers (COE) flood control alternatives for Malheur Lake. The workshop was structured around the formulation of a computer model that would simulate the hydrologic effects of the various alternatives and any concommitant changes in vegetation communities and wildlife use patterns.

The simulation model is composed of three connected submodels. The Hydrology submodel calculates changes in lake volume, elevation, and surface area, as well as changes in water quality, that result from the proposed water management projects (upstream storage, upstream diversions, drainage canals) and the no action alternative. The Vegetation submodel determines associated changes in the areal extent of wetland and upland vegetation communities. Finally, the Wildlife submodel calculates indices of abundance or habitat suitability for colonial nesting birds (great egret, double-crested cormorant, white-faced ibis), greater sandhill crane, diving ducks, tundra swan, dabbling ducks, and Canada goose based on hydrologic and vegetation conditions. The model represents the Malheur-Harney Lakes Basin, but provides water quantity and quality indicators associated with additional flows that might occur in the Malheur River Basin. Several management scenarios, representing various flood control alternatives and assumptions concerning future runoff, were run to analyze model behavior. Scenario results are not intended as an analysis of all potential management actions or assumptions concerning future runoff. Rather, they demonstrate the type of analysis that could be conducted if the model was sufficiently refined and tested.

Early in a model development project, the process of building the model is usually of greater benefit than the model itself. The model building process stimulates interaction among agencies, assists in integrating existing information, and helps identify research needs. These benefits usually accrue even in the absence of real predictive power in the resulting model. This workshop initiated interaction among the primary State and Federal resource and development agencies in a nonadversarial forum. The exchange of information and expertise among agencies provided the FWS with the best information currently available for use in the Planning Aid Letter it will develop at the Reconnaissance state of the COE study. If the COE subsequently initiates a Feasability Study, this information will be refined further and will aid the FWS in preparing its Coordination Act Report on any flood control alternative proposed by the COE.

The model building and testing process also helped identify model limitations and more general information needs that should be evaluated for further study prior to preparation of an FWS Coordination Act Report. Major needs associated with the Hydrology submodel include a more detailed representation of hydrologic units (separately consider Harney Lake, Mud Lake, and Malheur Lake or the three hydrological units within Malheur Lake, rather than a combined lake system) and explicitly representation of groundwater storage and discharge in water budget calculations. A better representation of the hydrological units will require more detailed topographic data for the basin, capacity-elevation and elevation-surface area curves for each unit, and better water flow data between the units. Additional water quality parameters and constraints on proposed canal operation due to conditions in the Malheur River might also be added. Key Vegetation submodel needs include fine-tuning existing vegetation relationships in the model and adding relationships to address the influence of historical conditions on vegetation development, effects of very rapid changes in lake level, effects of wildlife populations (e.g., carp, muskrat), responses of vegetation to habitat management actions (e.g, haying, grazing, burning), and better representation of sago pondweed dynamics. A complementary geographic information system might also be developed for spatial analyses. Major needs that should be evaluated for the Wildlife submodel include addition of other wildlife species that have important effects on habitat on the Refuge (e.g., carp, muskrat) and consideration of additional life-cycle requisites and controlling variable for species presently in the model. Some of these limitations could perhaps be overcome if historical data on habitat conditions were developed to use with historical data on wildlife populations.