Project Description and Objectives of Research:
The goal of this project was to develop an integrated, replicable, and theoretically sound methodology for identifying/suggesting "superior" configurations of stopover habitat locales to augment or enhance flyways for migratory birds. We develop alternative mathematical decision models for identifying strategic stopover locales. The methodology is demonstrated in an application to the Atlantic Flyway. The methodology is general and can be applied to other flyways as well.
Summary of Findings:
The Atlantic flyway is a complex system of migration routes of many migratory bird species. This flyway extends from northeastern Canada to the West Indies and Central and South America. The flyway supports migratory journeys that may take months and involve distances of several thousand miles. "Stopovers" are a critical part of this migration process. It is during stopovers that birds feed, rest, and rebuild energy reserves for the remainder of the trip. The locations of stopover sites can be crucial to the success of migratory journeys.
Human activities—urban and suburban development—have resulted in the fragmentation and degradation of stopover habitat within the Atlantic flyway. For many species, habitat deficiencies now exist or may appear in the future. Ensuring that adequate amounts (area) and distributions (locations) of habitat remain in perpetuity within the Atlantic flyway is critical for the survival of the species that use the flyway. Some lands within the flyway already have protected status and are managed for wildlife conservation among other uses. These "managed areas" include public lands at the federal, state, and local levels (e.g., parks and wildlife refuges), as well as private lands held in preservation trusts. Many managed areas are heavily used as stopover sites by birds during their migrations, but birds use other areas for stopovers as well—areas that may not have protected status and whose habitat values may be lost to future development (e.g., agricultural lands). The existing network of managed areas, by itself, is unlikely to sustain the current or desired population levels of migratory species as more and more unprotected habitat is lost. The goal of this research project is to develop a decision support methodology for identifying/suggesting new habitat locales for stopover sites. These new areas would augment and enhance the existing system of managed areas within the United States portion of the Atlantic flyway.
This report is the final report of this 2-year research project. In this final report, we briefly review the activities of the project's first year (see also the first Annual Report), and provide a full account of activities and results of the second year. The intended audience of this report includes natural resource and conservation scientists, planners, and decision makers.
During the first year, the researchers investigated the flyway problem as a multi-objective decision problem in land conservation. The researchers formulated alternative problem statements, assessed data needs, and developed and tested preliminary decision models. An advisory committee of experts in avian ecology and wetlands policy provided advice and guidance throughout this process. During the second and final year of the project, emphasis was placed on developing a general, replicable methodology that could be applied to the Atlantic flyway as well as to other major world flyways. A final Flyway Model was formulated, and data on the Atlantic flyway and representative species were gathered from a variety of sources. The Flyway Model then was used to generate many alternative solutions/ suggestions for enhancing the U.S. portion of the Atlantic flyway.
The intent of the research was to formulate a methodology for identifying and evaluating alternative configurations of habitat locales—primarily wetland and associated upland areas—from which specific sites might be set aside or restored as major protected stopovers. These new stopovers would augment and enhance existing managed areas within the flyway network by filling in the gaps between managed areas. Through the strategic selection of a sufficient number of new stopovers, a connected, braided network of "stepping stones" would be created, extending from Maine to Florida, and from the Atlantic coast several hundred miles inland.
The methodology consists of a decision model—the Flyway Model—which is a mathematical statement of the problem, as well as a method for solving the problem by executing the model on a computer. Functionally, the Flyway Model identifies broad areas (counties) that may represent suitable stopover locales from ecological, economic, and location perspectives. (The selection of specific sites or parcels to set aside within counties would be left to local planners and decision makers.) In creating a network of stopover locales, the Flyway Model seeks to achieve two objectives. The first objective is to maximize the geographic "coverage" of the stopover network. Under this objective, the model seeks to maximize the land area that is within a specified distance (e.g., 75 miles) of either a new stopover or an existing managed area. In achieving this objective, the likelihood is maximized that, wherever a bird lands, it will never be more than the specified distance away from a stopover.
This first objective prioritizes the location aspects of candidate stopover sites, but other attributes also are important, such as habitat suitability and economic considerations. These other attributes are addressed in the second objective of the model, which seeks to maximize the "quality" of the newly selected stopovers. Under this objective, counties that have a relatively high area in wetlands, as well as relatively low land costs, are sought as new stopovers. Both the location aspects for geographic coverage (Objective 1) and the quality aspects (Objective 2) of stopovers are important for constructing a network of stopovers. However, tradeoffs are likely to exist between the two objectives, as it is unlikely that a solution exists that optimizes both objectives simultaneously. These tradeoffs were evaluated as part of the study. Key elements of the Flyway Model are discussed next: geographic units, managed areas, bird species, flight distance standards, wetlands, land costs, and a "quality" index.
Geographic Units. Because the geographic scope of this research is very large it was necessary to use a relatively "coarse" spatial resolution. Counties were ultimately chosen as the geographic unit for reasons of scale suitability and data availability. In total, 715 counties were included in this study from the 17 Atlantic flyway states (Maine, New Hampshire, Vermont, Massachusetts, Rhode Island, Connecticut, New York, New Jersey, Pennsylvania, Delaware, Maryland, West Virginia, Virginia, North Carolina, South Carolina, Georgia, and Florida).
Managed Areas. Publicly managed lands, such as wildlife refuges, provide an existing backbone or baseline system of protected habitat for the Atlantic flyway. Of the many types of public lands that exist, only certain types of federal managed areas were included in this study. Specifically, large national parks, national wildlife refuges, and national wilderness areas comprised the baseline system. We also considered a second, expanded baseline system that included national forests, federal military bases, and wild and scenic rivers in addition to national parks, refuges, and wilderness areas.
Bird Species. Our advisory committee suggested two species to guide the formulation of the Flyway Model: the Black Duck and the Sora Rail. These species exhibit the migratory characteristics desired (use of wetlands) and have relatively well-known habitat needs.
Flight Distance Standards. The distances birds typically fly between stopovers vary widely across species and also depend on the availability of food and cover, as well as on local weather conditions. In the Flyway Model we employ the concept of a distance standard, which determines whether or not one county is reachable from another county in a single flight. Distance standards are used to determine the extent to which the stopover network is connected and the extent to which it provides coverage for flyway region. Three distance standards were used in this project: 60 miles, 75 miles, and 90 miles.
Wetlands. Wetlands provide a major source of habitat for many migratory birds during their migration cycles. The National Wetlands Inventory (NWI), compiled by the U.S. Fish and Wildlife Service using the Cowardian classification system, is a comprehensive survey of wetlands in the United States. NWI data were gathered for the 715 counties used in this study. We sought to characterize counties by the amount (acres) of wetlands present, summed over all (Cowardian) wetland classes. The prevalence of wetlands in a county was used as a surrogate for the habitat suitability of the county, with more wetlands implying greater suitability.
Land Costs. In addition to characterizing counties by the presence of wetlands, we also sought to characterize counties by land costs. From an economic perspective, counties with lower land costs would be more desirable as stopover locales than counties with higher land costs. We used the per-acre value of agricultural land as a surrogate for land costs—the cost or opportunity cost of purchasing or otherwise acquiring new stopover habitat.
"Quality" Index. Aside from location considerations, the desirability of counties as stopover locales might be based on both the prevalence of wetlands (habitat suitability) and land costs. We combined these two factors into a single, unit-free index of "quality" to have a simple, quantitative indicator or coefficient of the approximate and relative merits of each county with respect to wetland prevalence and land cost. Such indices often are used when incommensurable attributes need to be aggregated into a single criterion (e.g., the Fish and Wildlife Service's Habitat Suitability Index or HSI).
The Flyway Model is a decision model formulated as a "binary integer" or "zero-one" program. The zero-one program format is used because the decisions involved in solving the flyway augmentation problem are yes/no decisions, which can be conveniently modeled by zero-one decision variables. Principally, these decisions are whether or not to select counties as locales for new stopovers, and whether or not counties are "covered" by new stopovers or existing managed areas. Due to the very large size of the flyway problem, we used a "greedy adding" heuristic approach for solving it, instead of an "exact" approach such as linear/integer programming. This heuristic approach, which typically gives approximate or near-optimal solutions, was able to generate alternative solutions in reasonable amounts of time—one to several minutes—on a personal computer.
In total, 45 alternative solutions were created for augmenting and enhancing the Atlantic Flyway. Two types of tradeoffs were considered in generating these solutions. In the first tradeoff, the number of new stopovers was traded off against the total area of covered counties (Objective 1). In developing this tradeoff, we were able to determine the (minimum) number of new stopovers that would be needed to achieve any particular level of geographic coverage (e.g., 90 percent coverage of the flyway region). By increasing the number of new stopovers, higher levels of coverage could be achieved. The number of new stopovers selected ranged from 1 to 38, and the geographic coverage provided by the new stopovers plus existing managed areas ranged from 60 percent to 96 percent of the flyway region. In the second tradeoff, we examined the extent to which an increase in the "quality" of stopovers (Objective 2) would entail a reduction in the covered area (Objective 1), while holding the number of new stopovers constant. The aggregate "quality" of new stopover locales ranged from slightly below average to the highest "quality" possible. In this tradeoff, geographic coverage declined as "quality" improved for a given number of new stopovers, as expected.
The intent of this project was to develop a systems-based decision support methodology that considers, within a single analytic framework, a broad spatial, ecological, and economic context. This methodology addresses the interrelated nature of habitat conservation decisions—decisions that might otherwise be made on a case-by-case basis to the detriment of achieving system-wide objectives. In applying the methodology to the Atlantic flyway, alternatives for strategically augmenting and enhancing the flyway have been developed. These alternatives are intended to serve as approximate plans or policy options, but are not meant to be complete management plans. Their purpose is to inform and assist in guiding the planning and decision processes that affect the Atlantic flyway. It is expected that the methodology will add value to planning and decision making in several ways: by turning raw data into useful information; by suggesting efficient, cost-effective alternatives; and by identifying superior tradeoffs between conflicting management objectives.
We expect that the results of this research may be useful in the development and evaluation of wetland policies and activities administered by natural resource agencies and organizations. These organizations include, among others, the U.S. Environmental Protection Agency, the U.S. Fish and Wildlife Service, the U.S. Army Corps of Engineers, and The Nature Conservancy. Program and policy areas of relevance include: fulfillment of the wetlands habitat management and acquisition recommendations of the North American Waterfowl Management Plan; fulfillment of recommendations of Endangered Species Act recovery plans; wetlands mitigation and ecosystem restoration planning; and new uses for Department of Defense lands.
Malcolm S, ReVelle C. An anchored covering location model. Presented at the INFORMS Conference, Philadelphia, PA, November 1999.
Malcolm S, ReVelle C. Rebuilding migratory flyways using directed conditional covering. Environmental Modeling and Assessment (submitted May 2000).
Williams JC, ReVelle CS. Enhancing the Atlantic flyway: locating new stopovers for migratory birds. Presented at the 47th North American Meetings for the Regional Science Association International, Chicago, IL, November 2000.
Supplemental Keywords: restoration, habitat, decision making, conservation, modeling, analytical, northeast, southeast, Atlantic coast, midatlantic, mathematical programming, birds, migration.