Richard S. Sojda, Wildlife Biologist, USDI - Geological Survey, Biological Resources Division and Biology Department, Montana State University

Thousands of trumpeter swans once migrated from wintering grounds in the Arctic to breeding grounds in the southern United States. However, during the early 1900's, overhunting, habitat loss, and human disturbance reduced swan populations in the Lower 48 to less than one hundred birds. Through aggressive management actions (such as the establishment of national wildlife refuges, management of wetland habitats, limiting disturbance from recreationists to nesting birds, and trapping and transporting) swan populations began to increase.

Today, trumpeter swans in the United States are divided into three groups: the Pacific Coast, Interior, and Rocky Mountain populations. The Pacific Coast population of Alaska and western Canada is healthy and numbers over 15,000 birds. In the Lower 48, trumpeter swans originally bred from Michigan and Illinois west to Washington and Oregon, and as far south as Kansas and Missouri. Transplants in the Midwest have been relatively effective, with several scattered, small breeding populations becoming established and helping rebuild the Interior population. Progress in establishing trumpeter swan breeding populations in the Rocky Mountain West has been less successful. The research detailed below is aimed at better understanding this and exploring management options related to trumpeter swan migratory pathways.

Current Management Issues
Overall, the Rocky Mountain population of trumpeter swans is doing quite well, numbering about 2,700 birds primarily in Montana, Wyoming, and Idaho. However, the number of Rocky Mountain birds breeding in the United States is still low. In 1932, breeding Rocky Mountain trumpeter swans were found exclusively in the Greater Yellowstone Area (GYA) and numbered a mere 57 birds. By 1967, the population had increased to 564. However, since then, there has been more than a 30 percent decline.

In addition, much of the population breeding in the GYA no longer migrates to what are thought to be historical wintering grounds in Utah, Nevada, and California. The reasons for few migratory traits in this group are not completely understood, but likely include past overhunting, habitat loss, artificial feeding during winter, and attractiveness of warm spring habitats in the GYA. Ironically, some of these same factors have encouraged swans that nest in Canada to winter in the GYA. This has added additional competition for existing habitat with local birds and as the Canadian portion of the population continues to increase, more wintering grounds will become necessary.

The most secure management situation would likely have some birds continuing to winter in the GYA, and others migrating to less harsh environments farther to the South and West. Thus, should catastrophe from extreme winter weather, disease, pollution, or other factors affect one wintering area, other groups of birds would remain secure.

The successful recovery and migration of the Rocky Mountain Population of trumpeter swans is dependent upon the collaboration of all those involved in swan management. The Pacific Flyway Council is comprised of representatives from all the states and provinces with land west of the Continental Divide (Rocky Mountains) plus biologists from the U.S. Fish and Wildlife Service and Canadian Wildlife Service. They promote efforts among agencies to set and meet common management goals and objectives for migratory birds in the Pacific Flyway. The Council has a technical subcommittee charged with examining issues related to the Rocky Mountain Population of trumpeter swans. Currently, the Subcommittee is working to increase the Rocky Mountain trumpeter swan population so that it is both a secure and migratory population.

Making Decisions and the Value of New Technology
Today's managers often face a multitude of potential management goals and actions from which they must base their management decisions. In the case of trumpeter swan management, the level of knowledge and understanding of swan ecology and migration often varies among managers, with little information published in the scientific literature. Because of this, the Pacific Flyway Council's Rocky Mountain Trumpeter Swan Subcommittee (Subcommittee) has endorsed the development of a Trumpeter Swan Decision Support System. Such a system will allow managers the ability to use the best available science and quantitative decision analysis when examining the effects of various management actions on trumpeter swan populations.

A decision support system, or DSS, is an interactive, computer-based system designed to help decision-makers deal with complex problems. Scientists with the U.S. Geological Survey's Biological Resources Division propose to develop such a system which will allow for realistic and ecologically-based trumpeter swan management at multiple geographic and temporal scales. It will do this through the use of artificial intelligence methods such as expert systems, blackboards, and cooperative distributed problem solving. These methods address problems where complexity often prevents an individual decision-maker from conceptualizing, or otherwise dealing with an entire problem.

In the case of trumpeter swan management, decision-makers are often overwhelmed by the myriad of variables and their potential effects on swan populations. Changes in habitat conditions, management actions, and population trends all influence the recovery efforts of trumpeter swan populations. Not only must managers consider the impacts of these variables at specific management sites and times, but they also face the constraints of past scenarios and future options on current swan populations. The DSS will take into account all these components, plus the interactions among them. From inputs of actual data and results from simulations, the DSS will converge on a set of management recommendations which addresses the ecological constraints and opportunities represented in the system.

Benefits of a Decision Support System (DSS)
The foundation of the DSS will be knowledge bases developed from current information on swan ecology and management, including ecological data, expert opinion, and the scientific literature. Databases pertaining to population management objectives and site-specific management options will also be incorporated. The DSS will consider the effects of various management actions (such as hazing, water management, and changes in recreational activities) on migration chronologies and distributions of swans; quantity, quality, location, and timing of availability of wetland habitats; and population levels and trends. DSS approaches recognize that decisions cannot wait for additional information to be collected, but must deal with the current level of knowledge. (Figure 1)

The power of the DSS is in its ability to conduct simulations and examine consequences of future management actions. Management of migratory birds can be so complex that no one person can have a complete view of all activities and conditions; and furthermore, management plans must evolve as time unfolds. The DSS takes into account the interactions between and within sites over time. The impacts of these interactions are incorporated into the model to assist with management decision-making processes. This system will allow local managers to develop individual plans that benefit from a flyway perspective. In addition, the location situations get linked together to form the broader perspective. (Figure 2)

The following are the types of management questions potentially addressed by the DSS. The examples have been kept particularly simple to illustrate how the system will function. In reality, the picture is much more complex.
If a particular management action is implemented at a particular site and time, what are the consequences for that site and other sites in the flyway?

For example: If hazing of swans on the Henry's Fork were curtailed next winter and temperature patterns are warmer than normal for December, how many swans would be expected in that area? Would they be less likely to migrate to Bear River Migratory Bird Refuge in Utah?

For example: If temperature patterns are warmer than normal and objectives from December through February were to have an average of 15 swans on the Snake River between Jackson Lake Dam and Moose, 15 swans on the Madison River upstream from Hebgen Lake, 50 swans on the Henry's Fork in the Island Park Area, etc., what levels of disturbance from winter recreation might be recommended at each site? What outflows from the associated dams would be optimal? Should winter feeding be conducted anywhere? Where and when should hazing programs be conducted?

For example: If, (1) temperature patterns are warmer than normal, and (2) it were an objective from December through February to have an average of 15 swans on the Snake River between Jackson Lake Dam and Moose, 15 swans on the Madison River upstream from Hebgen Lake, 50 swans on the Henry's Fork in the Island Park Area, etc., and (3) outflows from the associated dams are established and unchangeable, and (4) no winter feeding will occur, and (5) levels of disturbance from winter recreation are established for each site, then: When and where should hazing programs be conducted?

Given a satisfactory set of management actions across all sites to achieve an objective for swan distribution, if an alternative management action were to be implemented at a particular site, what are the consequences for that site and other sites in the flyway in terms of reaching their respective objectives?

For example: Let us assume that a set of management actions has been recommended by the DSS to achieve a certain expectation of swan distributions during December through February, similar to the distributions described in the above examples. Such a hypothetical set of actions might be: (1) temperature patterns are warmer than normal and outflows from various dams have been recommended, (2) no winter feeding of swans will occur, (3) levels of disturbance from winter recreation are established for each site, and (4) hazing programs are to be conducted only on the Henry's Fork during December. How might our achievement of those distributions of swans be affected if the outflows from Jackson Lake Dam are reduced by 50 percent?

Support for DSS Development
The U.S. Department of the Interior's Fish and Wildlife Service is the primary federal agency recommending the development of a Trumpeter Swan Decision Support System. Field managers in both the National Park Service and the U.S. Forest Service, as well at state biologists in Oregon, Idaho, Montana, Wyoming, Utah, and Nevada, have expressed an interest and need for this research. The

The Trumpeter Swan Decision Support System is unique in that it uses new computer modeling technology (based on artificial intelligence) from other disciplines to address complex natural resource issues. The knowledge gained from this current research will be applied in the development and use of future natural resource decision-support systems.