This is the accessible text file for GAO report number GAO-08-797 entitled 'Wildlife Refuges: Changes in Funding, Staffing, and Other Factors Create Concerns about Future Sustainability' which was released on September 24, 2008. This text file was formatted by the U.S. Government Accountability Office (GAO) to be accessible to users with visual impairments, as part of a longer term project to improve GAO products' accessibility. Every attempt has been made to maintain the structural and data integrity of the original printed product. Accessibility features, such as text descriptions of tables, consecutively numbered footnotes placed at the end of the file, and the text of agency comment letters, are provided but may not exactly duplicate the presentation or format of the printed version. The portable document format (PDF) file is an exact electronic replica of the printed version. We welcome your feedback. Please E-mail your comments regarding the contents or accessibility features of this document to Webmaster@gao.gov. This is a work of the U.S. government and is not subject to copyright protection in the United States. It may be reproduced and distributed in its entirety without further permission from GAO. Because this work may contain copyrighted images or other material, permission from the copyright holder may be necessary if you wish to reproduce this material separately. Report to Congressional Requesters: United States Government Accountability Office: GAO: September 2008: Wildlife Refuges: Changes in Funding, Staffing, and Other Factors Create Concerns about Future Sustainability: GAO-08-797: GAO Highlights: Highlights of GAO-08-797, a report to congressional requesters. Why GAO Did This Study: The National Wildlife Refuge System, which is administered by the Fish and Wildlife Service in the Department of the Interior, comprises 585 refuges on more than 96 million acres of land and water that preserve habitat for waterfowl and other migratory birds, threatened and endangered species, and other wildlife. Refuges also provide wildlife- related activities such as hunting and fishing to nearly 40 million visitors every year. GAO was asked to (1) describe changing factors that the refuge system experienced from fiscal years 2002 through 2007, including funding and staffing changes, and (2) examine how habitat management and visitor services changed during this period. We surveyed all refuges; visited 19 refuges in 4 regions; and interviewed refuge, regional, and national officials. In commenting on a draft of this report, the Department of the Interior made technical comments that we have incorporated as appropriate. GAO is not making recommendations in this report. What GAO Found: Between fiscal years 2002 and 2007, the refuge system experienced funding and staffing level fluctuations, the introduction of several new policy initiatives, and the increased influence of external factors such as extreme weather that threaten wildlife habitat and visitor infrastructure. Although core funding—measured as obligations for refuge operations, maintenance, and fire management—increased each year, inflation-adjusted core funding peaked in fiscal year 2003 at about $391 million—6.8 percent above fiscal year 2002 funding. Inflation-adjusted core funding ended the period 2.3 percent below peak levels, but 4.3 percent above fiscal year 2002 levels by fiscal year 2007. Core refuge staffing levels peaked in fiscal year 2004 at 3,610 full-time equivalents—10.0 percent above the fiscal year 2002 level—and then declined more slowly than funding levels. By fiscal year 2007, staffing levels fell to 4.0 percent below peak levels, but 5.5 percent above fiscal year 2002 levels. Through fiscal year 2007, the number of permanent employees utilized by the refuge system declined to 7.5 percent below peak levels. During this period, refuge system officials initiated new policies that: (1) reduced staff positions and reallocated funds and staff among refuges to better align staff levels with funding; (2) required refuge staff to focus on a legislative mandate to complete refuge conservation plans by 2012; (3) shifted to constructing a larger number of smaller visitor structures, such as informational kiosks, and fewer large visitor centers to spread visitor service funds across more refuges; (4) increased the number of full- time law enforcement officers and their associated training and experience requirements; and (5) resulted in additional administrative work. During this period, external factors that complicate refuge staffs’ ability to protect and restore habitat quality also increased, including severe storms and development around refuges. Our survey showed that the quality of habitat management and visitor service programs varied across refuges during our study period. Habitat conditions for key types of species improved about two times more often than they worsened, but between 7 percent and 20 percent of habitats were of poor quality in 2007. Certain habitat problems increased at more than half of refuges during this period, and managers reported that they increased the time spent on certain habitat management activities, such as addressing invasive plants, despite declining staffing levels. However, several managers we interviewed told us that staff were working longer hours without extra pay to get work done, and managers expressed concern about their ability to sustain habitat conditions. While the quality of four key visitor service programs was reported to be stable or improving between fiscal years 2002 and 2007 at the vast majority of refuges, the other two key programs—environmental education and interpretation—were considered poor quality at one-third of refuges in 2007. Changes in the time spent on visitor services varied considerably across refuges, and managers noted that visitor services generally are cut before habitat management activities when resources are limited. Managers are concerned about their ability to provide high-quality visitor services in the future given staffing and funding constraints. To view the full product, including the scope and methodology, click on [hyperlink, http://www.gao.gov/cgi-bin/getrpt?GAO-08-797]. For more information, contact Robin Nazzaro at (202) 512-3841 or nazzaror@gao.gov. [End of section] Contents: Letter: Results in Brief: Background: Refuge Funding and Staffing Levels Fluctuated, New Policies Were Introduced, and the Influence of Various External Factors Affecting Refuges Increased for Fiscal Years 2002 through 2007: Several Changes in Habitat Management and Visitor Services Occurred at Refuges from Fiscal Years 2002 through 2007, Raising Managers’ Concerns About Future Sustainability: Concluding Observations: Agency Comments and Our Response: Appendix I: Scope and Methodology: Appendix II: Statistical Analysis of Habitat Change: Appendix III: Deferred Maintenance: Appendix IV: Refuge Operating Needs System: Appendix V: Total Core Obligations, Nominal and Inflation-Adjusted (in 2002 Dollars), and Core FTEs, Fiscal Years 2002 through 2007: Appendix VI: Comments from the Department of the Interior: Appendix VII: GAO Contacts and Staff Acknowledgments: Tables: Table 1: Core Refuge System FTEs, Fiscal Years 2002 through 2007: Table 2: Total FTEs That Supported the Refuge System, Fiscal Years 2002 through 2007: Table 3: Permanent and Total Refuge System Employees, Fiscal Years 2002 through 2007: Table 4: Change in Habitat Quality by Species Type, Fiscal Years 2002 through 2007: Table 5: Habitat Quality by Species Type, Fiscal Year 2007: Table 6: Extent to Which Habitat Is Meeting the Needs of Species, Fiscal Year 2007: Table 7: Common Refuge Problems Affecting Habitat and Trends in These Problems, Fiscal Years 2002 through 2007: Table 8: Habitat Management Activities that Increased the Most at Refuges, Fiscal Years 2002 through 2007: Table 9: Change in Time Spent by Type of Worker on Habitat Management Activities, Fiscal Years 2002 through 2007: Table 10: Percent Increase or Decrease in Time Spent on Habitat Management by Nonpermanent Workers as a Function of Permanent Staff Time Spent on Habitat Management, Fiscal Years 2002 through 2007: Table 11: Change in Quality of Visitor Services Programs, Fiscal Years 2002 through 2007: Table 12: Quality of Visitor Services Programs, Fiscal Year 2007: Table 13: Infrastructure Quantity and Condition Changes, Fiscal Years 2002 through 2007: Table 14: Change in Time Spent on Visitor Services, Fiscal Years 2002 through 2007: Table 15: Change in Time Spent on Visitor Services by Type of Worker, Fiscal Years 2002 through 2007: Table 16: Percent Increase or Decrease in Nonpermanent Worker Time Spent on Visitor Services as a Function of Permanent Staff Time Spent on Visitor Services, Fiscal Years 2002 through 2007: Table 17: Summary Statistics for Habitat Change and Refuge Characteristics, Fiscal Years 2002 through 2007: Table 18: Change in Quality of Waterfowl Habitat by Various Characteristics of Refuges, Fiscal Years 2002 through 2007: Table 19: Change in Quality of Other Migratory Bird Habitat by Various Refuge Characteristics, Fiscal Years 2002 through 2007: Table 20: Regression Results for Change in Quality of Waterfowl Habitat, Fiscal Years 2002 through 2007: Table 21: Regression Results for Change in Quality of Other Migratory Bird Habitat, Fiscal Years 2002 through 2007: Table 22: Deferred Maintenance Backlog by Region, Fiscal Years 2002 through 2007: Table 23: RONS Project Backlog, Fiscal Years 2002 through 200: Table 24: RONS Projects Selected for Funding, Fiscal Years 2002 through 2007: Table 25: Ratio of RONS Project Backlog to Funded Projects, Fiscal Years 2002 through 2007: Figures: Figure 1: Map of the National Wildlife Refuge System: Figure 2: Principal Funding Sources for Wildlife Refuges: Figure 3: Refuge System Core Obligations, Nominal and Inflation- Adjusted Dollars (in 2002 Dollars), Fiscal Years 2002 through 2007: Figure 4: Total Obligations for the Refuge System in Nominal Dollars, Fiscal Year 2007: Figure 5: Total Obligations for the Refuge System, Nominal and Inflation-Adjusted Dollars (in 2002 Dollars), Fiscal Years 2002 through 2007: Figure 6: Total Obligations for the Department of the Interior, Fish and Wildlife Service, and National Wildlife Refuge System, Nominal Dollars, Fiscal Years 2002 through 2007: Figure 7: Total Obligations for the Department of the Interior, Fish and Wildlife Service, and National Wildlife Refuge System, Inflation- Adjusted Dollars (in 2002 Dollars), Fiscal Years 2002 through 2007: Figure 8: Sources of Total FTEs that Supported the Refuge System, Fiscal Year 2007: Figure 9: Comparison of Cumulative Percentage Change in the Refuge System’s Core Funding, Core FTEs, and Permanent Employees, Fiscal Years 2002 through 2007: Figure 10: Visitor Facilities at National Wildlife Refuges: Figure 11: Results of Activities to Remove Knotgrass at Tualatin River National Wildlife Refuge: Figure 12: Change in Quality of Waterfowl Habitat by Refuge Tier, Fiscal Years 2002 through 2007: Figure 13: Change in Quality of Other Migratory Bird Habitat by Refuge Tier, Fiscal Years 2002 through 2007: Figure 14: Change in Quality of Waterfowl Habitat by Change in Time Spent on Habitat Management Activities, Fiscal Years 2002 through 2007: Figure 15: Change in Quality of Other Migratory Bird Habitat by Change in Time Spent on Habitat Management Activities, Fiscal Years 2002 through 2007: Figure 16: Purple Loosestrife: Figure 17: Nutria and Streambank Damage: Figure 18: Nisqually National Wildlife Refuge Approved Acquisition Boundary: Figure 19: Change in the Quality of Waterfowl Habitat by Change in the Contribution of External Factors, Fiscal Years 2002 through 2007: Figure 20: Change in the Quality of Other Migratory Bird Habitat by Change in the Contribution of External Factors, Fiscal Years 2002 through 2007: Figure 21: Examples of Visitor Services Infrastructure for Wildlife Observation and Photography: Figure 22: Examples of Visitor Services Infrastructure for Environmental Education and Interpretation: Figure 23: Fee Box at Occoquan National Wildlife Refuge: Abbreviations: FHWA: Federal Highway Administration: FTE: full-time equivalent: FWS: U.S. Fish and Wildlife Service: GDP: gross domestic product: Improvement Act: National Wildlife Refuge System Improvement Act of 1997: Interior: Department of the Interior: MMS: Maintenance Management System: NWR: National Wildlife Refuge: NWRS: National Wildlife Refuge System: RONS: Refuge Operating Needs System: SAMMS: Service Asset and Maintenance Management System: WMA: Wildlife Management Area: WMD: Wetland Management District: [End of section] United States Government Accountability Office: Washington, DC 20548: September 22, 2008: The Honorable Nick J. Rahall, II: Chairman: Committee on Natural Resources: House of Representatives: The Honorable Madeleine Z. Bordallo: Chairwoman: Subcommittee on Fisheries, Wildlife, and Oceans: Committee on Natural Resources: House of Representatives: The Honorable Ron Kind: House of Representatives: The National Wildlife Refuge System, administered by the Department of the Interior’s U.S. Fish and Wildlife Service (FWS), comprises about 585 refuges on more than 96 million acres of land and water that provide habitat for millions of waterfowl and other migratory birds, threatened and endangered species, and other plants and wildlife. Each year, nearly 40 million visitors take part in one or more of the refuge system’s six wildlife-dependent visitor activities—hunting, fishing, wildlife observation, wildlife photography, environmental education, and environmental interpretation—and other recreational activities. The refuge system spans all 50 states, American Samoa, Guam, Puerto Rico, the U.S. Virgin Islands, and other U.S. territories. FWS manages its dispersed refuges through its headquarters office in Washington D.C., eight regional offices, and hundreds of field offices located on or near refuge lands. Individual refuge offices may report directly to a regional office (these refuges are referred to as “stand-alone” refuges in this report), or may be grouped with other offices into a “complex” under the common management of a project leader. The 585 refuges have been organized into 126 complexes and 96 stand-alone refuges. Staff at refuges may include refuge managers, project leaders, wildlife biologists, law enforcement officers, park rangers, and administrative or maintenance staff, among others. What is recognized as the first wildlife refuge was established in 1903 as a federal bird reservation to protect brown pelicans and other waterbirds. Over time, refuges have been added to the system for various specific purposes such as providing habitat for one or more specific endangered species, or for broader purposes such as providing habitat for waterfowl and other migratory birds in general. In 1997, the National Wildlife Refuge System Improvement Act (Improvement Act) provided a unifying mission for all the units in the refuge system—to conserve, manage, and, where appropriate, restore fish, wildlife, and plant resources and their habitats for the benefit of present and future generations.[Footnote 1] The act requires refuges to give priority to wildlife-dependent recreational uses that are deemed compatible with the purposes of the refuge in refuge planning and management. In addition, the act generally requires refuges to complete comprehensive conservation plans—long-range plans for managing, among other things, habitats and providing visitor services—by 2012. You asked us to (1) describe changing factors that the refuge system experienced from fiscal years 2002 through 2007, including funding and staffing changes, and (2) examine how habitat management and visitor services changed during this period. To address these objectives, we surveyed 538 units within the refuge system—including stand-alone refuges and refuges within complexes—that we determined to be within the scope of our review and obtained an 81 percent response rate. Survey respondents were primarily refuge managers or project leaders (for the purposes of this report we refer to both of these groups as refuge managers). We also obtained and analyzed funding data, as measured by obligations, and staffing data, as measured by the number of full-time equivalents (FTE) and the number of permanent employees, from the Federal Financial System and the Federal Personnel Payroll System, and refuge planning and performance data from FWS’s Refuge Annual Performance Planning System.[Footnote 2] We visited headquarters, 4 regional offices, and 19 refuges, and conducted phone interviews with officials at the other 4 regional offices and about 50 additional refuges. We selected refuges in order to see a range in geographic location, visitation level, refuge prioritization, and type of management activities and challenges. We conducted our work from July 2007 to September 2008 in accordance with generally accepted government auditing standards. Those standards require that we plan and perform the audit to obtain sufficient, appropriate evidence to provide a reasonable basis for our findings and conclusions based on our audit objectives. We believe that the evidence obtained provides a reasonable basis for our findings and conclusions based on our audit objectives. More detailed descriptions of our scope and methodology and the statistical analysis of our survey data are presented in appendixes I and II, respectively. Results in Brief: For fiscal years 2002 through 2007, funding and staffing levels for the refuge system fluctuated, several new refuge system policy initiatives were introduced, and the influence of external factors such as extreme weather and human development that affect refuge operations increased: * Funding. Inflation-adjusted funding (in 2002 dollars) for refuge operations, maintenance, and fire management—considered “core” refuge activities by refuge officials—peaked in fiscal year 2003, for the celebration of the refuge system’s centennial, at about $391 million—6.8 percent above fiscal year 2002 levels—and then declined to 2.3 percent below peak levels in fiscal year 2007; it ended 4.3 percent above fiscal year 2002 levels. In nominal dollars, core funding increased each year over the time period from about $366 million in fiscal year 2002 to about $468 million in fiscal year 2007. At the refuge level, inflation-adjusted core funding varied considerably during the time period, with about as many losing funding as gaining since fiscal year 2002. Specifically, core funding decreased at 96 of 222 complexes and stand-alone refuges and increased at 92, with funding remaining about the same at 34. The magnitude of the changes in core funding at the refuge level also were more pronounced than for the trend overall. For example, from fiscal year 2002 through fiscal year 2007, the Kootenai National Wildlife Refuge in Idaho lost 66 percent of its core funding. Overall, core funding for 39 complexes and stand- alone refuges decreased by more than 25 percent during this time period. * Staffing. Staffing levels for core refuge activities (core staffing), measured in FTEs, peaked 1 year later than core inflation-adjusted funding and then declined more slowly. Specifically, core staffing peaked in fiscal year 2004 at 3,610 FTEs—10.0 percent above fiscal year 2002 levels—and then declined to 4.0 percent below peak levels by fiscal year 2007; staffing levels ended 5.5 percent above the fiscal year 2002 level. In addition to FTEs, refuge officials said that permanent positions are an important staffing measure because they represent a predictable workforce for managers to rely upon from one year to the next. Like FTEs, the number of permanent employees peaked in fiscal year 2004, then declined to 7.5 percent below 2004 levels by fiscal year 2007—a loss of 275 employees—and 1.7 percent below fiscal year 2002 levels. Though 38 complexes and stand-alone refuges gained staff since 2004, more than three times as many lost permanent employees. * Policy initiatives. Several new refuge system policy initiatives were implemented during this period. First, recognizing that funding declines after 2003 were exacerbating an already high proportion of staff costs in refuge budgets, regional offices began to (1) reduce staff positions through attrition and by further consolidating some stand-alone refuges into complexes, and (2) categorize refuges into three tiers for the purpose of prioritizing funding and staffing allocations among refuges. These measures are primarily responsible for the decline in FTEs and permanent employees from fiscal year 2004 peak levels and the shifts in staffing among complexes and stand-alone refuges. Also, in fiscal year 2004, refuge system officials at headquarters recognized that the refuge system was not on pace to meet the Improvement Act mandate to complete conservation plans for each refuge by 2012. To help meet this deadline, refuge system officials created a completion schedule and required staff at refuges to turn their attention to completing the plans. While refuge officials believe that they can meet the deadline, current information shows that some plans are behind schedule. In addition, during fiscal years 2002 through 2007, to help spread visitor service funds across as many refuges as possible, refuge officials began placing a greater emphasis on constructing smaller visitor facility structures, such as informational kiosks and restrooms, at a larger number of refuges rather than constructing a smaller number of traditional visitor centers. Furthermore, refuge system management began an initiative to increase the number of full-time law enforcement officers and their associated training and experience requirements to improve safety and address other concerns. Finally, during this period, various refuge system, FWS, and Department of the Interior policies increased administrative work for nonadministrative refuge staff by requiring additional data entry into certain systems and responses to numerous data calls. Refuge system officials are beginning to implement changes to reduce some of these administrative burdens. * External factors. The influence of external factors—those outside the control of the refuge system that complicate refuges’ abilities to protect and restore habitat quality, including extreme weather and development on adjacent lands—increased over this period. For example, our survey found that from fiscal years 2002 through 2007, the influence of development—such as the expansion of urban areas and the conversion of off-refuge land near refuges to agriculture or industrial use—increased around refuges and contributed to refuge habitat problems for almost one-half of the refuges. Such development can pollute refuge lands and waters and make it more difficult to maintain viable, interconnected habitat in and around a refuge’s borders. Survey responses and interviews with refuge managers indicated that the quality of habitat and visitor service programs, as well as the amount of time devoted to these activities, varied across refuges during our study period. Given recent funding and staffing changes, and other factors affecting refuges, managers are concerned about their ability to provide quality habitat and visitor service programs into the future: * Habitat management. Twenty-eight percent to 40 percent of habitats on refuges for several types of key species, such as waterfowl and other migratory birds, improved between fiscal years 2002 and 2007, but conditions of 11 percent to 18 percent of refuge habitats worsened and 7 percent to 20 percent were in poor condition in 2007. Complicating habitat management is growing pressure from increasing habitat problems occurring on refuges and the influence of external factors. Our survey found that invasive plant species and habitat fragmentation were the leading problems, affecting 55 percent and 44 percent of refuges, respectively, and both were increasing on more than half of refuges. At the same time, managers reported increasing the time spent on habitat management at many refuges. Of note, we estimated that refuges where staff time increased were 3.0 times more likely than refuges where staff time decreased to report improved, rather than worsened, habitat for both waterfowl and other migratory birds. However, 93 percent of refuge managers also noted increases in administrative workload on nonadministrative staff from fiscal years 2002 through 2007; managers said that such tasks take away from the time staff can devote to habitat management, and some managers reported that staff are working longer hours without overtime pay to address habitat needs. Many refuge managers expressed concern about their long-term ability to maintain high-quality habitat in light of decreasing permanent staff levels at refuges between fiscal years 2002 and 2007 and increasing pressures on refuges and refuge staff. * Visitor services. Our survey found that the quality of all six wildlife-dependent visitor services was stable or improving between fiscal years 2002 and 2007 at the vast majority of refuges. And while four of the six visitor services were of moderate or better quality in fiscal year 2007 at more than three-quarters of refuges, environmental education and environmental interpretation programs were of poor quality at about one-third of refuges. Some refuges reported that they expanded their visitor services infrastructure, for example, by adding informational kiosks and trails and tour routes, yet more than one-half of refuges reported no change. The change in the time spent by refuges on visitor services varied considerably throughout the system. Refuge managers we interviewed told us that visitor services generally are reduced first—before habitat management activities—when a refuge faces budget constraints, and several told us that they have become more dependent on volunteers to staff their visitor centers or run their programs. Our survey and site visits found that refuge managers are very concerned about their ability to provide high-quality visitor services to the public given recent funding and staffing changes. In commenting on a draft of this report, Interior made technical comments that we have incorporated as appropriate. Background: The National Wildlife Refuge System (NWRS)—the only system of federal lands protected specifically for wildlife conservation—provides more than 96 million acres of habitat for over 700 species of birds, hundreds of threatened or endangered species, and a wide variety of other species. Each year, millions of birds stop to rest at refuges strategically located along their migration routes. In 1903, President Theodore Roosevelt established what is now recognized as the first refuge, the Pelican Island National Wildlife Refuge (NWR) in Florida. During the more than 100 years since, the refuge system has grown to include 548 wildlife refuges and 37 wetland management districts that address a variety of wildlife purposes.[Footnote 3] For example, the Merced NWR in California was established in 1951 with the broad purpose of serving as a sanctuary for migratory birds, while the Antioch Dunes NWR, also in California, was established to protect three specific endangered species—Lange’s metalmark butterfly, Contra Costa wallflower, and the Antioch Dunes evening primrose. The refuge system employs more than 4,000 staff dispersed in its offices across the country. Individual refuges may report directly to a regional office, or may be grouped with other refuge units into a complex under the common management of a project leader.[Footnote 4] Complexes range in size from 2 to 19 refuges, and one of the refuges in each complex usually serves as the complex headquarters.[Footnote 5] Complexing has reorganized the 585 refuges into 126 complexes and 96 stand-alone refuges. Officials in headquarters serve as advisors to regional refuge chiefs and to refuge managers. Figure 1 shows the location of the 585 refuges comprising the NWRS. Figure 1: Map of the National Wildlife Refuge System: [Refer to PDF for image] This figure is a map of the United States depicting the National Wildlife Refuge System. Indicated on the map are the geographic locations of Regions 1-8 as well as national wildlife refuges within each region. Source: Fish and Wildlife Service. [End of figure] Until the passage of the National Wildlife Refuge System Improvement Act of 1997, the refuge system was the only major federal public lands network without a basic statute providing a mission for the system, policy direction, and management standards for all of its units. The Improvement Act gave the refuge system a unifying mission—to administer a national network of lands and waters for the conservation, management, and, where appropriate, restoration of the fish, wildlife, and plant resources and their habitats within the United States for the benefit of present and future generations of Americans. The legislation also called for FWS to plan and direct the continued growth of the system in a manner designed to accomplish this mission. In addition, the Improvement Act required that the biological integrity, diversity, and environmental health of the refuge system be preserved. The act generally requires refuges to complete comprehensive conservation plans—long-range plans for managing, among other things, habitats and providing visitor services—by 2012. An important component to the act was that it recognized six wildlife-dependent recreational uses of the refuge system—hunting, fishing, wildlife observation, wildlife photography, environmental education, and environmental interpretation—as appropriate uses that are consistent with the mission of the refuge system, when they are determined to be compatible with the purposes of individual refuges. While hunting and fishing have always been popular uses on refuges, wildlife observation is the most prevalent activity on refuges today, and attracted over 23 million visitors in 2006. For the most part, refuges generally perform similar activities that are compatible with the mission of protecting wildlife and habitat and providing visitor services: * Habitat management. Refuges manage their lands to provide adequate habitat for the species they were established to conserve and to maintain biological diversity and integrity. Management activities may include (1) performing habitat management work such as maintaining water levels in water impoundments and ponds and treating invasive species; (2) performing fire management activities including conducting prescribed burns; (3) restoring habitat to make it more useful for wildlife purposes; (4) monitoring species and habitat through surveys and other studies; (5) managing volunteers doing habitat- or wildlife- related work; and (6) coordinating habitat management efforts with outside entities, such as private land owners, state agencies, and other groups. * Visitor services. Nearly 40 million people visit refuges each year, and the vast majority of refuges provide visitors with the opportunity to participate in one or more of the six wildlife-dependent recreational uses outlined in the Improvement Act. To support these activities, refuges install roads, trails, docks, and boat ramps, and develop interpretive and educational exhibits, among other things. Refuges also perform work that supports both habitat management and visitor services: * Maintenance. The refuge system maintains more than $18 billion in real property, including more than 41,000 facilities such as buildings, visitor infrastructure, and roads; more than 4,000 vehicles; and almost 4,000 pieces of heavy equipment. Refuge staff perform preventative maintenance on their refuges’ real property to achieve specific performance targets that are tied to the refuge system’s mission, such as restoring wetlands, monitoring wildlife, and providing recreation opportunities. The refuge system currently has a deferred maintenance backlog, which is described in appendix III. Appendix IV discusses the Refuge Operational Needs System, which maintains information on refuge operational requirements such as staff, equipment, and planned projects. * Law enforcement. The refuge system employs law enforcement officers who are tasked with protecting refuges’ natural resources, infrastructure, and the visiting public. Officers also enforce conservation agreements with private landowners. * Conservation planning. Refuges are required to complete comprehensive conservation plans that outline priorities for wildlife and habitat as well as visitor services. * Wildfire suppression. The refuge system supports wildfire suppression needed on refuge lands as well as other federal lands. The refuge system receives most of its funding for core refuge operations and maintenance activities from FWS’s annual resource management appropriation; funds for fire management to restore and improve habitat as well as wildfire suppression come via a separate appropriation. Funding from several other sources supports other types of refuge system activities. For example, the refuge system receives annual allocations from FWS’s construction appropriation to construct, improve, acquire, or remove buildings and other facilities, and from FWS’s land acquisition appropriation to acquire interests in lands, including easements that provide important fish and wildlife habitat. Refuges also may apply for grants from federal, state, and local governments and nonprofit organizations, among others, to supplement their funding. The Department of Transportation’s Federal Highway Administration (FHWA), through its Public Lands Highway-Refuge Roads Program, provides funds to maintain and improve public roads that provide access to or within a refuge. In addition, the refuge system has a permanent appropriation authorizing refuges to use recreation fees they collect and to accept donations, voluntary services, and in- kind contributions from private conservation groups, such as land or equipment donations. The refuge system receives additional funding through other FWS programs, such as Endangered Species or Fisheries and Habitat Conservation. Like the refuge system, these programs also receive allocations from the resource management appropriation and may, in turn, obligate a portion of this funding to support projects occurring on refuge lands. Figure 2 shows the principal sources of funding for the refuge system. Figure 2: Principal Funding Sources for Wildlife Refuges: [Refer to PDF for image] This figure is an illustration of the principal funding sources for Wildlife Refuges. Those sources are: Core Funding: * Allocation from the Resource Management appropriation (funds refuge operations and maintenance costs); * Allocation from Wildland Fire Management (funds fire prevention and suppression costs); Allocation from Construction appropriation; Allocation from Land Acquisition appropriation; Other authorized sources, including funds from FHWA’s Refuge Roads program; Contributed Funds and in-kind donations; Grants; Recreation fees. Source: GAO. [End of figure] While most refuges carry out the same type of activities, key characteristics of refuges such as acreage, visitation levels, and the type of ecosystems they contain—and consequently the challenges they face—can vary. Sixteen refuges in Alaska account for approximately 85 percent—more than 76 million acres—of the refuge system’s total acreage, and these refuges generally operate somewhat differently than others. The Arctic NWR in northeastern Alaska, for example, contains 8 million acres of wilderness that is relatively undisturbed; as such, activities focus primarily on research, monitoring, and education. In contrast, the Tualatin River NWR—located 15 miles from Portland, Oregon—faces the challenge of protecting natural resources amid rapidly increasing visitation levels. Refuges along the southwest border of the United States, meanwhile, face unique law enforcement challenges as they support the Department of Homeland Security’s border control efforts. Refuge Funding and Staffing Levels Fluctuated, New Policies Were Introduced, and the Influence of Various External Factors Affecting Refuges Increased from Fiscal Years 2002 through 2007: From fiscal years 2002 through 2007, the refuge system experienced fluctuations in funding and staffing levels, the introduction of several new refuge system policy initiatives, and increases in the influence of external factors such as extreme weather and development that affect refuge operations. Inflation-adjusted funding for core refuge system activities—measured as obligations for refuge operations, maintenance, and fire management—increased by 6.8 percent from fiscal year 2002 to fiscal year 2003 for the celebration of the refuge system’s centennial, then declined quickly to 4.7 percent below peak levels by fiscal year 2005, before increasing again to 2.3 percent below peak levels in fiscal year 2007, when adjusted for inflation (in 2002 dollars); it ended the period 4.3 percent above fiscal year 2002 levels.[Footnote 6] In nominal dollars, core funding increased each year over the time period from about $366 million in fiscal year 2002 to about $468 million in fiscal year 2007. Core refuge system staffing levels peaked in fiscal year 2004 after increasing 10.0 percent, and then declined more slowly than funding to 4.0 percent below this level by the end of fiscal year 2007; they ended the period 5.5 percent above fiscal year 2002 levels. During the same period, several refuge system policy initiatives were implemented to reduce staff levels and reprioritize funding among refuges, ensure the completion of required conservation plans, shift focus toward constructing a greater number of smaller visitor facilities, and increase the number of full-time law enforcement officers and associated training; other initiatives increased the administrative workload on refuges. Refuges also experienced an increase in the influence of various external factors that may complicate managers’ abilities to protect habitat and provide visitor services, such as extreme weather events and development on adjacent lands. Inflation-Adjusted Core Refuge Funding Peaked in Fiscal Year 2003 Then Decreased, but Ended the Period above Fiscal Year 2002 Levels: Obligations for core refuge activities—operations, maintenance, and fire management—peaked in fiscal year 2003, then decreased and remained below peak levels through fiscal year 2007, when adjusted for inflation (in 2002 dollars), but above fiscal year 2002 levels.[Footnote 7] As shown in figure 3, total nominal obligations for core refuge system activities increased each year from about $366 million in fiscal year 2002 to about $468 million in fiscal year 2007—an average annual increase of 5.1 percent or about $18.5 million. However, when adjusted for inflation, total core obligations peaked in fiscal year 2003 at about $391 million for the wildlife refuge centennial—an increase of 6.8 percent over fiscal year 2002. Core inflation-adjusted obligations then quickly fell back to 4.7 percent below peak levels by fiscal year 2005. By fiscal year 2007, inflation-adjusted core obligations rebounded somewhat to about $382 million—still 2.3 percent below peak levels, but 4.3 percent above fiscal year 2002 levels. While the refuge system did receive an increase in the allocation from the resource management appropriation for fiscal year 2008, we did not include it in our analysis because the fiscal year was not yet complete. Figure 3: Refuge System Core Obligations, Nominal and Inflation- Adjusted Dollars (in 2002 Dollars), Fiscal Years 2002 through 2007, Dollars in millions: [Refer to PDF for image] This figure is a multiple line graph depicting the following data: Fiscal year: 2002; Core obligations, nominal: 365.7; Core obligations, inflation-adjusted: 365.7. Fiscal year: 2003; Core obligations, nominal: 406; Core obligations, inflation-adjusted: 390.7. Fiscal year: 2004; Core obligations, nominal: 413.1; Core obligations, inflation-adjusted: 377. Fiscal year: 2005; Core obligations, nominal: 425.2; Core obligations, inflation-adjusted: 372.2. Fiscal year: 2006; Core obligations, nominal: 442; Core obligations, inflation-adjusted: 373.1. Fiscal year: 2007; Core obligations, nominal: 467.9; Core obligations, inflation-adjusted: 381.6. Source: GAO analysis of Fish and Wildlife Service data. Note: We used obligations as a measure of the funding for the refuge system. [End of figure] At the refuge level, the trends in inflation-adjusted core refuge obligations at the 222 complexes and stand-alone refuges varied considerably during our study period.[Footnote 8] Specifically, from fiscal year 2002 through fiscal year 2007, core inflation-adjusted obligations decreased for 96 complexes and stand-alone refuges, increased for 92, and stayed about the same for 34.[Footnote 9] The magnitude of the changes in core funding at the refuge level also were more pronounced than for the trend overall. For example, the refuge with the largest percentage inflation-adjusted decrease in funding was the Kootenai NWR in Idaho, where obligations fell from $957,506 in fiscal year 2002 to $324,283 in fiscal year 2007, a decrease of 66 percent. The refuge experiencing the largest inflation-adjusted dollar decrease was the Mid-Columbia NWR Complex in Washington state, where obligations fell from about $9.4 million in fiscal year 2002 to about $5.2 million in fiscal year 2007, a decrease of about $4.1 million. [Footnote 10] Moreover, from fiscal year 2002 through fiscal year 2007, core funding for 39 complexes and stand-alone refuges decreased by more than 25 percent. On the other hand, the refuge receiving the largest percentage increase in inflation-adjusted funding was the Caddo Lake NWR in Texas. Its obligations increased by 156 percent, from $95,255 in fiscal year 2002 to $244,094 in fiscal year 2007, largely reflecting an increase in operations at this refuge since it was established in October 2000. The refuge experiencing the largest dollar increase in inflation-adjusted funding was the Okefenokee NWR in Georgia, where obligations for core refuge activities increased from about $7.4 million in fiscal year 2002 to about $15.8 million in fiscal year 2007—an increase of about $8.4 million.[Footnote 11] However, almost 90 percent of this increase consisted of fire management funding provided largely to respond to the wildfires the refuge faced in April 2007. Appendix V presents obligations for core refuge activities for all 222 complexes and stand-alone refuges in both nominal and inflation- adjusted dollars for each of fiscal years 2002 through 2007. Total obligations, in nominal dollars, for the refuge system were about $816 million in fiscal year 2007. As illustrated in figure 4, core obligations comprised about 57 percent of this total, or about $468 million. Obligations of allocations from Interior’s construction and land acquisition appropriations added approximately 19 percent, or about $154 million. Additional obligations of funds received through recreation fees, donations, conservation funds, and all other sources, contributed approximately 18 percent, or about $148 million. Other funds were obligated from grants and allocations from the FHWA and from other FWS programs. Figure 4: Total Obligations for the Refuge System in Nominal Dollars, Fiscal Year 2007 (Dollars in millions): [Refer to PDF for image] This figure is a pie-chart depicting the following data: Core: $468 million; Other refuge, including recreation fees, donations, and conservation fund: $148 million; Construction: $107 million; Land acquisition: $47 million; Other FWS programs: $34 million; FHWA Refuge Roads program: $9 million; Grants: $5 million. Source: GAO analysis of Fish and Wildlife Service data. Notes: We used obligations as a measure of the funding for the refuge system. Obligations sum to $818 million due to rounding. [End of figure] In contrast to the trend in core funding, total inflation-adjusted funding for the refuge system as a whole did not peak in fiscal year 2003, but instead steadily decreased from fiscal year 2002 levels until fiscal year 2005 and rebounded somewhat thereafter (see fig. 5). Even after rebounding somewhat in fiscal year 2005, however, inflation- adjusted total funding decreased to about $666 million in fiscal year 2007—an average annual decrease of 1.6 percent (about $11.2 million) or 7.5 percent below fiscal year 2002 levels. The main driver in the generally decreasing trend in total funding is a sharp drop in funding for land acquisition, which fell from about $101 million in fiscal year 2002 to about $38 million in fiscal year 2007. In nominal dollars, total obligations increased from about $720 million in fiscal year 2002 to about $816 million in fiscal year 2007—an average annual increase of 2.5 percent or about $18.3 million. Figure 5: Total Obligations for the Refuge System, Nominal and Inflation-Adjusted Dollars (in 2002 Dollars), Fiscal Years 2002 through 2007: [Refer to PDF for image] This figure is a multiple line graph depicting the following data: Fiscal year: 2002; Total, nominal dollars: $720.2 million; Total, inflation-adjusted: $720.2 million. Fiscal year: 2003; Total, nominal dollars: $700.4 million; Total, inflation-adjusted: $674 million. Fiscal year: 2004; Total, nominal dollars: $725.4 million; Total, inflation-adjusted: $662 million. Fiscal year: 2005; Total, nominal dollars: $722.3 million; Total, inflation-adjusted: $632 million. Fiscal year: 2006; Total, nominal dollars: $774 million; Total, inflation-adjusted: $652.8 million. Fiscal year: 2007; Total, nominal dollars: $816.4 million; Total, inflation-adjusted: $667.1 million. Source: GAO analysis of Fish and Wildlife Service data. Note: We used obligations as a measure of the funding for the refuge system. [End of figure] Beyond receiving financial resources, refuges also receive in-kind donations from nonprofit groups, for-profit companies, and other organizations. From our survey, we obtained information on donations received by 246 individual refuges—67 percent of refuges responding to our survey. These donations typically consisted of equipment and other supplies that refuges used to help manage habitat or deliver visitor services, though larger donations included land and construction of visitor centers. Donations at most of the refuges totaled $500,000 or less over the entire 6-year time frame; however, several refuges reported that they received more than $1 million over this period. [Footnote 12] For example, a refuge in Washington state estimated that it received in-kind donations totaling about $20 million, consisting primarily of land donations from nonprofit organizations, bridge work, and habitat restoration projects. Another refuge in Michigan estimated that it received about $5 million worth of in-kind donations, including land donations from local industries, as well as shoreline restoration, fence removal, and tree removal projects performed by these industries. The declining trends in refuge system funding were comparable to general declines in obligations for FWS and Interior overall, although Interior fared somewhat better. FWS fared about the same as the refuge system ending in fiscal year 2007 at 7.5 percent below fiscal year 2002 levels, when adjusted for inflation. Interior overall fared somewhat better over the same period, declining 3.9 percent when adjusted for inflation. Figures 6 and 7 show the trends in nominal and inflation- adjusted obligations, respectively, made by the refuge system, FWS, and Interior from fiscal years 2002 through 2007. Figure 6: Total Obligations for the Department of the Interior, Fish and Wildlife Service, and National Wildlife Refuge System, Nominal Dollars, Fiscal Years 2002 through 2007 (dollars in millions): [Refer to PDF for image] This figure is a multiple line graph depicting the following data: Fiscal year: 2002; Total, Department of the Interior: $13,414; Total, U.S. Fish and Wildlife Service: $1,922.6; Total, National Wildlife Refuge System: $720.2. Fiscal year: 2003; Total, Department of the Interior: $13,881; Total, U.S. Fish and Wildlife Service: $1,904.3; Total, National Wildlife Refuge System: $700.4. Fiscal year: 2004; Total, Department of the Interior: $14,526; Total, U.S. Fish and Wildlife Service: $1,984.5; Total, National Wildlife Refuge System: $725.4. Fiscal year: 2005; Total, Department of the Interior: $15,839; Total, U.S. Fish and Wildlife Service: $1,985.3; Total, National Wildlife Refuge System: $722.3. Fiscal year: 2006; Total, Department of the Interior: $16,122; Total, U.S. Fish and Wildlife Service: $2,049.3; Total, National Wildlife Refuge System: $774. Fiscal year: 2007; Total, Department of the Interior: $15,799; Total, U.S. Fish and Wildlife Service: $2,179.4; Total, National Wildlife Refuge System: $816.4. Source: GAO analysis of Interior data. [End of figure] Figure 7: Total Obligations for the Department of the Interior, Fish and Wildlife Service, and National Wildlife Refuge System, Inflation- Adjusted Dollars (in 2002 Dollars), Fiscal Years 2002 through 2007 (dollars in millions): [Refer to PDF for image] This figure is a multiple line graph depicting the following data: Fiscal year: 2002; Total, Department of the Interior: $13,414; Total, U.S. Fish and Wildlife Service: $1,922.6; Total, National Wildlife Refuge System: $720.2. Fiscal year: 2003; Total, Department of the Interior: $13,357.6; Total, U.S. Fish and Wildlife Service: $1,832.5; Total, National Wildlife Refuge System: $674. Fiscal year: 2004; Total, Department of the Interior: $13,256; Total, U.S. Fish and Wildlife Service: $1,811; Total, National Wildlife Refuge System: $662. Fiscal year: 2005; Total, Department of the Interior: $13,858.9; Total, U.S. Fish and Wildlife Service: $1,737.1; Total, National Wildlife Refuge System: $632. Fiscal year: 2006; Total, Department of the Interior: $13,597.8; Total, U.S. Fish and Wildlife Service: $1,728.4; Total, National Wildlife Refuge System: $652.8. Fiscal year: 2007; Total, Department of the Interior: $12,909.6; Total, U.S. Fish and Wildlife Service: $1,780.8; Total, National Wildlife Refuge System: $667.1. Source: GAO analysis of Interior data. Staffing Levels Increased through Fiscal Year 2004, Then Declined More Slowly Than Funding and Ended the Period Below Peak Levels: Staffing levels, as measured by FTEs the refuge system actually used, peaked later and declined more slowly than funding for both core refuge activities (core staffing) and all refuge activities (total staffing).[Footnote 13] FTEs for core staffing, which includes operations, maintenance, and fire management, increased from 3,283 in fiscal year 2002 to a peak of 3,610 in fiscal year 2004—an increase of 10.0 percent. Core staffing then fell back to 3,464 FTEs by fiscal year 2007—still 5.5 percent higher than the fiscal year 2002 level, but 4.0 percent below peak staffing levels. While operations and maintenance FTEs increased 3.6 percent overall during our study period, they ended the period down 6.9 percent from their 2004 peak. Fire management FTEs, on the other hand, increased 14.3 percent over fiscal year 2002 levels.[Footnote 14] Table 1 shows FTE trends for core refuge system activities from fiscal years 2002 through 2007. Table 1: Core Refuge System FTEs, Fiscal Years 2002 through 2007: Refuge system operations and maintenance FTEs: 2002: 2,702; 2003: 2,882; 2004: 3,005; 2005: 2,960; 2006: 2,885; 2007: 2,798. Refuge system fire management FTEs: 2002: 582; 2003: 610; 2004: 605; 2005: 595; 2006: 660; 2007: 665. Total refuge system core FTEs: 2002: 3,283; 2003: 3,493; 2004: 3,610; 2005: 3,556; 2006: 3,545; 2007: 3,464. Source: GAO analysis of Fish and Wildlife Service data. Note: Some columns do not sum due to rounding. [End of table] In contrast with funding, FTEs for noncore activities account for a relatively small portion of the total FTEs that support the refuge system. In fiscal year 2007, for example, 664 noncore FTEs supported the refuge system—about 16 percent of total FTEs—as illustrated in figure 8. Slightly more than 500 of these FTEs were allocated to the refuge system to manage construction projects, land acquisitions, grants, and donations, and to collect fees, among other refuge activities. FWS employees assigned to other agency programs accounted for about 157 of the 664 noncore FTEs for activities in support of the refuge system. For example, biologists from FWS’s Ecological Services program often monitor various species at refuges, supplementing the refuges’ habitat management activities while also furthering Ecological Services’ mission to conserve and restore threatened and endangered species. Figure 8: Sources of Total FTEs that Supported the Refuge System, Fiscal Year 2007: [Refer to PDF for image] This figure is a pie-chart depicting the following data: Sources of Total FTEs that Supported the Refuge System, Fiscal Year 2007: Refuge system core activities: 3,464; Other refuge system activities: 507; Other FWS programs: 157; Total: 4,128 FTEs. Source: GAO analysis of Fish and Wildlife Service data. [End of figure] Similar to the trend in core FTEs, total FTEs used in support of the refuge system overall also peaked in fiscal year 2004 and then decreased through the remainder of the period. As table 2 illustrates, total FTEs increased 5.8 percent from fiscal year 2002 through fiscal year 2004, then declined through 2007, to close the period 0.9 percent higher than the fiscal year 2002 level. This amounted to a 4.7 percent drop from the peak staffing levels of fiscal year 2004. Table 2: Total FTEs That Supported the Refuge System, Fiscal Years 2002 through 2007: Refuge system FTEs: 2002: 3,933; 2003: 4,101; 2004: 4,178; 2005: 4,139; 2006: 4,108; 2007: 3,971. FTEs from other FWS programs: 2002: 158; 2003: 160; 2004: 152; 2005: 136; 2006: 141; 2007: 157. Total FTEs (including other FWS programs): 2002: 4,091; 2003: 4,261; 2004: 4,330; 2005: 4,275; 2006: 4,249; 2007: 4,128. Source: GAO analysis of Fish and Wildlife Service data. [End of table] In addition to FTEs, the number of employees on board in refuge system positions also declined after peaking in fiscal year 2004.[Footnote 15] Through fiscal year 2007, nearly 375 employees were lost from the refuge system’s peak staffing levels, a reduction of 8.4 percent over this period (see table 3). About three-quarters of this loss came through a reduction in permanent employees. Refuge managers and regional and headquarters officials told us that the number of filled, permanent positions at refuges is a key measure of the effective strength of the workforce available to conduct core refuge activities because they represent employees on board indefinitely.[Footnote 16] Thus, the loss of 275 permanent employees (7.5 percent) since fiscal year 2004—generally through the elimination of vacant positions created by retirements and resignations—has reduced the number of staff available to conduct needed work. For the overall study period, total employees declined 4.0 percent below fiscal year 2002 levels and permanent employees declined 1.7 percent. Table 3: Permanent and Total Refuge System Employees, Fiscal Years 2002 through 2007: Permanent employees; 2002: 3,446; 2003: 3,592; 2004: 3,663; 2005: 3,624; 2006: 3,512; 2007: 3,388. Total employees: 2002: 4,247; 2003: 4,398; 2004: 4,449; 2005: 4,344; 2006: 4,211; 2007: 4,076. Source: GAO analysis of Fish and Wildlife Service data. Note: The counts in the table are the number of employees as of July 1 for each fiscal year, which refuge system officials reported would represent the summer work season—and thus the annual staffing peak—at most refuges. [End of table] The overall fluctuation in staffing levels and the reductions since fiscal year 2004 in particular have affected many refuges. During the first 2 years of our study period, from fiscal year 2002 through 2004, 114 complexes and stand-alone refuges increased their permanent staff by more than 5 percent, while only 49 lost more than 5 percent and 55 stayed about the same.[Footnote 17] However, over the final 3 years, the situation was reversed: from fiscal year 2004 through 2007, the number of complexes and stand-alone refuges that lost more than 5 percent of their permanent staff more than doubled to 122, while only 38 gained at least 5 percent and 58 stayed about the same. Several Policy Initiatives Were Implemented from Fiscal Years 2002 through 2007: The refuge system implemented several policy initiatives from fiscal years 2002 through 2007, including efforts to achieve more sustainable staffing levels, ensure the completion of conservation plans, construct a greater number of small visitor facilities, and modify the refuge system’s law enforcement function. In addition, various refuge system, FWS, and Interior policies increased administrative work for nonadministrative refuge staff during this period. Workforce Management Planning: Because core staffing levels peaked later and declined more gradually than the system’s core inflation-adjusted funding, as shown in figure 9, rising salary and benefit costs for these staff began to account for an increasing share of refuge budgets after fiscal year 2003.[Footnote 18] In many cases, there was an existing imbalance in refuge budgets that meant that personnel costs already were putting pressure on refuges’ ability to operate. Generally, this reduced refuges’ management capability—that is, the percentage of a refuge’s budget available to pay for other operational costs that support its daily work, such as utilities, fuel, supplies, and seasonal labor. Although circumstances varied by refuge, some refuges’ management capability shrank to less than 5 percent of their total budget—a nearly unsustainable operational scenario, according to some refuge managers and regional and headquarters officials we interviewed. Figure 9: Comparison of Cumulative Percentage Change in the Refuge System’s Core Funding, Core FTEs, and Permanent Employees, Fiscal Years 2002 through 2007: [Refer to PDF for image] This figure is a multiple line graph depicting the following data: Fiscal year: 2002; Core funding (inflation-adjusted): 0; Core FTEs: 0; Permanent employees: 0. Fiscal year: 2003; Core funding (inflation-adjusted): 6.8%; Core FTEs: 6.4%; Permanent employees: 4.2%. Fiscal year: 2004; Core funding (inflation-adjusted): 3.1%; Core FTEs: 10%; Permanent employees: 6.3%. Fiscal year: 2005; Core funding (inflation-adjusted): 1.8%; Core FTEs: 8.3%; Permanent employees: 5.2%. Fiscal year: 2006; Core funding (inflation-adjusted): 2%; Core FTEs: 8%; Permanent employees: 1.9%. Fiscal year: 2007; Core funding (inflation-adjusted): 4.3%; Core FTEs: 5.5%; Permanent employees: -1.7%. Source: GAO analysis of Fish and Wildlife Service data. [End of figure] To attain a more sustainable balance between staffing costs and management capability, in fiscal year 2006, each regional office was directed to develop a workforce management plan. According to FWS guidance for these plans and interviews with senior refuge officials, regions were instructed to focus on doing “fewer things better,” that is, to allocate limited resources in such a way as to showcase selected refuges, rather than to allocate them across all refuges such that the level of habitat management and visitor services would be equally degraded. Although workforce plans differed by region, they generally proposed to: * increase management capability to a minimum of 25 percent of refuges’ operating budgets by reducing the share devoted to salaries and benefits to 75 percent or less;[Footnote 19] * reduce staff costs by (1) abolishing staff positions that became vacant through retirements and resignations, and (2) further consolidating refuges into complexes to eliminate redundant positions and reduce administrative costs; * categorize all refuges into one of three tiers—called focus refuges, targeted reduction refuges, or unstaffed satellite refuges—to prioritize them for funding and staffing increases or decreases; and; * realign some vacated positions by moving them from lower- to higher- tiered refuges. Although refuge system management did not intend for regions’ workforce plans to conform to a rigid national standard, program headquarters did provide criteria for regions to use when placing refuges into the following tiers: * Focus refuges, where FWS would strive to maintain or enhance field operations, would be selected because of the significance of their natural resources, important opportunities for wildlife-dependent recreation, or other “highly significant” values. * Targeted reduction refuges, where reductions in operations would occur, also would be selected on the basis of natural resources, recreation, and other values, but would be considered a lower priority than focus refuges. * Unstaffed satellite refuges had no specific criteria, but would include both refuges that have never been staffed and those that were to be destaffed due to budget constraints. Refuge system documents, as well as our interviews and site visits, showed these refuges to be often smaller, more remote, and less complex to manage than those in the upper two tiers. According to refuge system officials, the process for determining staff reductions, realignments, and refuge tiers varied considerably across regions, and refuge managers disagreed over the appropriateness of the methods some regions used. For example, Region 7 (Alaska) designated all of its refuges as focus refuges, while all other regions placed their refuges into each of the three tiers. Of the 275 refuge managers who answered a survey question on this issue, 41 percent responded that the criteria for categorizing refuges into tiers were appropriate to distinguish among the competing priorities in their respective regions, another 37 percent responded that they were not appropriate, and the remainder said they had no basis to judge. While most of the respondents who disagreed with their region’s criteria were from lower- tiered refuges, 25 percent of those who responded this way managed at least one highest-priority, or focus, refuge. Refuge managers acknowledged that additional management capability was necessary for continued operations, and understood that workforce planning decisions that affected funding and staffing levels were inherently difficult. Still, according to a senior regional office official, refuge tiering added to the emotional strain of an already stressed workforce, establishing a “have” versus “have not” mentality that many staff took personally. Implementation of workforce plans shifted funds and staff from lower- priority refuges to higher-priority refuges or from the regional office to the field, and reduced the total number of positions located at refuges. In all, about 375 refuge and regional office positions were either abolished—through elimination of vacant positions—or moved, with managerial, biological, and maintenance positions among those most frequently targeted for reduction or realignment. These changes were responsible for most of the 275 permanent employees who were lost after fiscal year 2004. By design, lower-tiered refuges absorbed a heavier share of these staff cuts and realignments. According to refuge system officials, the $36 million increase in the fiscal year 2008 allocation for the refuge system from the resource management appropriation was being used in part to restore some of these lost positions and funding at targeted reduction refuges. Further, regional officials reported that management capability across each region had reached the desired margin of at least 25 percent of refuges’ operational budgets. However, given that the fiscal year was not complete before the end of our review, we did not obtain additional data on FTE or position changes at refuges. Conservation Planning: From fiscal years 2002 through 2007, refuge system officials implemented steps intended to ensure that comprehensive conservation plans mandated by the Improvement Act are completed on time. In early fiscal year 2004, refuge officials realized that they were not on track to complete the 554 conservation plans required by 2012—the due date mandated by the act.[Footnote 20] At that time, refuge officials assessed needs and goals with regard to completing the plans, and provided recommendations to encourage timely completion as well as a monitoring and evaluation strategy. At the end of fiscal year 2005, however, refuge officials noted that the completion of these plans was still behind schedule—only 19 percent of the plans were complete even though more than half of the 15-year time frame had elapsed. To help ensure that the plans would be completed on time, refuge system officials required, among other things, refuge managers to identify work that could be set aside to focus on completing conservation plans. They also required regions to develop completion dates and milestones for completing the plans and to use a central database for tracking milestones. To date, about half of the 554 plans have been completed and about one-third are underway. Refuge officials said that they believe that they can meet the deadlines; however, some plans are still behind schedule. Visitor Service Facility Construction: In 2003, in response to discussions with Congress about how best to fulfill the requirement to provide the six wildlife-dependent activities described in the Improvement Act, the refuge system began an initiative to place greater emphasis on constructing small visitor facility structures, such as observation decks, informational kiosks, and restrooms, instead of larger visitor centers. These small structures are less expensive than visitor centers—which in 2007 were estimated to cost an average of $5.7 million each—and can be completed more quickly. Thus, a larger number of refuges can receive visitor facility funds, enabling refuge system investments to benefit a larger number of visitors. For fiscal years 2003 through 2007, the refuge system directed about $28 million toward these projects. Nevertheless, large visitor centers continue to be funded, and the refuge system was appropriated more than $51 million for visitor center construction from fiscal years 2002 through 2007. Figure 10 shows examples of visitor facility infrastructure. Figure 10: Visitor Facilities at National Wildlife Refuges: [Refer to PDF for image] This figure contains the following three photographs: Visitor center: Tualatin NWR; Interpretive kiosk: Great Bay NWR; Outdoor restroom: Great Bay NWR. Source: GAO. [End of figure] Law Enforcement Modifications: In July 2002, in response to safety concerns, the Secretary of the Interior directed the refuge system to begin an initiative to modify its law enforcement program by, among other things, increasing the training requirements for officers, reducing the system’s reliance on dual-function officers—staff with other primary duties who perform law enforcement duties part time—and creating an officer deployment model. Specifically, the refuge system increased the training requirements for law enforcement officers from 18 weeks to about 30 weeks from 2002 through 2007, with the vast majority of that increase coming from a new field training evaluation program. Refuge system officials also required some senior-level staff, and all dual-function officers who were performing law enforcement functions less than 25 percent of their time, to cease performing their law enforcement duties beginning in 2003. As a result, the refuge system reduced the number of dual- function officers from 495 to 164 and hired 76 full-time officers who serve a single refuge. FWS also created a “zone officer” position to serve multiple refuges and has hired 45 of these officers. In 2005, the International Association of Chiefs of Police released a law enforcement deployment model it developed with the refuge system to identify the level of law enforcement personnel needed to provide adequate protection of refuge resources and the public, and where those officers should be deployed. The model recommended a total of 845 law enforcement FTEs for the refuge system—about 600 FTEs more than the refuge system currently has on board.[Footnote 21] Given the refuge system’s current funding situation and the chances of the system attaining such a level, senior refuge officials have identified 450 positions as the minimum number they believe necessary to provide adequate protection. Refuge officials are hoping they can approach that minimum number by hiring an additional 200 officers in the near term. New Administrative Work: During our study period, refuge managers told us that they began feeling the burden of a myriad of new administrative work, especially work that applies to nonadministrative staff, resulting from refuge system, FWS, and Interior policies. Ninety-three percent of refuge managers who responded to our survey said administrative duties for nonadministrative staff have increased since 2002; less than 2 percent of refuge managers reported a decrease in this workload. For example, refuge managers expressed concerns that the number of national reporting requirements and extent of mandatory training classes, among other administrative tasks, were burdensome. Furthermore, the refuge system created a new maintenance database that required much more data entry than the previous system. In addition, managers indicated that they have been receiving an increasing number of data calls over the years. In 2003, refuge managers began an effort to address increasing administrative requirements and more than 300 refuge managers participated in discussions about the problem. Several managers formed an ad hoc committee in 2003 to address the issue and several officials from headquarters and regional offices joined the effort in 2004. Together, they drafted a white paper that provided several recommendations to reduce the burden. A headquarters team took the effort over and, in October 2007, released a report detailing their findings and 17 recommendations for reducing some requirements, such as reviewing national reporting requirements and eliminating those deemed unnecessary, as well as making Web-based training optional. According to refuge officials at headquarters, FWS is beginning to implement these 17 recommendations. The Influence of Various External Factors That Complicate Refuges’ Abilities to Manage Habitats and Provide Services Increased from Fiscal Years 2002 through 2007: A variety of factors that were generally outside the control of refuge system management became more influential between fiscal years 2002 and 2007. Some of these factors were natural occurrences, such as extreme weather, while others were due to the intensification of human activities, such as development. These factors added to refuge workload, complicating managers’ abilities to protect habitat quality and provide visitor services. One commonly cited external factor was extreme weather events such as droughts, floods, and severe winds. Survey results show that the contribution of extreme weather events to habitat problems increased at 52 percent of refuges; only 2 percent reported a decline. Storm damage also increased at many more refuges than it decreased: in particular, hurricanes Ivan, Katrina, and Rita in 2004 and 2005 damaged large parts of refuges in the southeastern United States. Eighteen refuges in three states were temporarily closed to the public as a result of these storms; a 19th refuge—Louisiana’s Sabine NWR—remains closed to public use. According to FWS, storm damage to the refuge system in 2005 alone exceeded $300 million. Development pressures caused by the expansion of urban areas and problems associated with the conversion of off-refuge land to agriculture or industrial use also increased during this period. Refuge managers reported that human settlement infrastructure such as roads, housing, and airports increasingly contributed to refuge habitat problems between 2002 and 2007—around 46 percent of refuges. These development pressures can contribute pollution to refuge lands and waters and make it more difficult to maintain viable, interconnected habitat in and around a refuge’s borders. Moreover, increasing development around refuges can be accompanied by an increase in the demand for recreational uses of nearby refuges, including some uses—such as recreational boating or rock-climbing—that may be incompatible with a refuge’s established purpose. In addition, the influence of off-refuge agricultural and industrial activities increased for many more refuges than it decreased. Because refuges do not exist in isolation, they must be managed in concert with adjacent lands to maintain healthy habitats, a reality that requires managers to allocate time to spend away from their refuges to develop working relationships with adjacent and upstream landowners. Other external factors affecting the refuge system include inadequate water rights and rights-of-way, such as public roads that divide refuge lands; the impacts associated with these factors worsened at more refuges than they improved, though they were stable for almost half of all respondents. Additionally, almost a quarter of the responses to our survey identified impacts associated with climate change as one of the biggest threats to habitat condition throughout the system. Managers reported that they already are seeing effects that they attribute to climate change, including drying of wetlands and wildfires of increased frequency and intensity. In addition to the obvious effects on habitat condition, these disturbances can affect wildlife-dependent visitor services, such as hunting or photography, to the extent that they change waterfowl migration patterns or the ranges of land- and water- based wildlife that historically are native to a given refuge. [Footnote 22] Several Changes in Habitat Management and Visitor Services Occurred at Refuges from Fiscal Years 2002 through 2007, Raising Managers’ Concerns About Future Sustainability: From fiscal years 2002 through 2007, several changes occurred in refuges’ habitat management and visitor services, creating concerns about the refuges’ abilities to maintain high-quality habitat and visitor services in the future. While 28 percent to 40 percent of habitats on refuges for several types of key species improved between fiscal years 2002 and 2007, conditions at some refuges worsened and 7 percent to 20 percent of habitats were in poor condition in 2007. Refuge habitats are facing growing pressure from increasing habitat problems and external factors, and although most refuges increased time spent on habitat management activities, there is increasing concern from managers that staffing and funding constraints will inhibit the ability of refuges to maintain quality habitat in the future in light of increasing habitat problems and resource constraints. The quality of visitor services improved on one-fifth to nearly one-half of refuges between fiscal years 2002 and 2007, but environmental education and interpretation programs were of poor quality at about one-third of refuges in 2007. While some refuges have been able to increase the time spent on visitor services, refuge managers are concerned about their ability to provide high-quality visitor services to the public given recent funding and staffing changes. While Habitats on Refuges for Key Types of Species Improved More Often Than They Worsened between Fiscal Years 2002 and 2007, Others Are in Poor Condition and Many Refuge Managers Are Concerned about Maintaining Habitat Conditions: While 28 percent to 40 percent of habitats on refuges for several types of key species improved between fiscal years 2002 and 2007, conditions of 11 percent to 18 percent of refuge habitats worsened and 7 percent to 20 percent of habitats were in poor condition in 2007. Habitat problems and external factors are increasing at refuges, and most refuges increased the time spent on habitat management activities. However, managers are concerned that staffing and funding constraints will inhibit the refuges’ ability to maintain quality habitat in the future. Between Fiscal Years 2002 and 2007, Habitat Conditions on Refuges for Key Types of Species Improved More Often Than they Worsened, Although Some Refuges Have Poor Quality Habitat: Refuge managers reported that habitats for five key types of species we surveyed refuges about improved between 2002 and 2007 about two times as often as they worsened (see table 4). Table 4: Change in Habitat Quality by Species Type, Fiscal Years 2002 through 2007: Species type: Waterfowl; Percent of refuge habitats where quality improved: 36%; Percent of refuge habitats where quality stayed the same: 47%; Percent of refuge habitats where quality worsened: 18%. Species type: Other migratory birds; Percent of refuge habitats where quality improved: 40%; Percent of refuge habitats where quality stayed the same: 44%; Percent of refuge habitats where quality worsened: 17%. Species type: Threatened and endangered species; Percent of refuge habitats where quality improved: 28%; Percent of refuge habitats where quality stayed the same: 52%; Percent of refuge habitats where quality worsened: 11%. Species type: Candidate threatened and endangered species; Percent of refuge habitats where quality improved: 33%; Percent of refuge habitats where quality stayed the same: 47%; Percent of refuge habitats where quality worsened: 14%. Species type: State species of concern; Percent of refuge habitats where quality improved: 29%; Percent of refuge habitats where quality stayed the same: 54%; Percent of refuge habitats where quality worsened: 13%. Source: GAO. Notes: Refuge managers identified habitat quality for specific threatened and endangered species, candidate species, and state species of concern occurring on their refuges that are aggregated into these general “types of species” categories. Not all species occurred on every refuge. See app. I for a discussion of our methodology. Some rows may not sum to 100 due to rounding and survey responses such as “no basis to judge.” [End of table] Tualatin River NWR outside of Portland, Oregon, for example, saw a marked improvement in wetland habitat, according to the refuge manager, as the refuge has begun to address an invasive weed infestation over the past year. The refuge has been addressing two primary invasive plants—knotgrass and cocklebur—that had infested approximately one- third of the refuge’s 600 wetland acres since 2003, overtaking the native wetland plants that thousands of birds rely on for food during migration. Through herbicide application, mowing and discing, and water level manipulation, the refuge was able to cut infestations in half over the last year, bringing the habitat back up to sufficient quality for use by the migrating birds.[Footnote 23] We observed the results of some of these activities to remove knotgrass (see fig. 11). Figure 11: Results of Activities to Remove Knotgrass at Tualatin River National Wildlife Refuge: [Refer to PDF for image] Two photographs of Knotgrass at Tualatin River National Wildlife Refuge. Source: GAO. [End of figure] Refuge managers also reported that 11 percent to 18 percent of habitats on refuges for key species have worsened since 2002. Camas NWR in Idaho, for example, has faced a drought for the last several years. According to the refuge manager, the lack of water has negatively impacted a riparian zone of cottonwoods and willows that migrating birds, such as yellow warblers, use during migration. In addition, many of the trees in the riparian area are close to 100 years old and are dying. Currently, the refuge is working on a plan to restore the vegetation, relying in part on wells for irrigation, but expects it will take decades to restore. As might be expected, we found differences in changes in the quality of habitat for waterfowl and other migratory birds between focus and targeted reduction refuges when compared to unstaffed refuges (see figs. 12 and 13). Specifically, we found that managers at focus and targeted reduction refuges were significantly more likely to report that habitat quality for waterfowl improved between 2002 and 2007 than at unstaffed satellite refuges.[Footnote 24] For example, between fiscal years 2002 and 2007, more than twice as many focus refuges experienced improved waterfowl habitat (42 percent) as experienced worsened waterfowl habitat (20 percent). At unstaffed satellite refuges, by contrast, habitat for waterfowl worsened almost as frequently as it improved, with 20 percent of refuges experiencing improved quality and 16 percent experiencing worsened quality. We found a similar situation for other migratory birds. We also found that these relationships generally remain strong in statistical models that simultaneously account for the effects of the change in staff time and the change in external factors, such as extreme weather and agricultural activity, which can contribute to habitat problems. For example, based on these models, we estimate that focus refuges were 3.4 times more likely than unstaffed satellite refuges to experience improved rather than worsened habitat quality for other migratory birds and that targeted reduction refuges were 3.9 times more likely. [Footnote 25] Figure 12: Change in Quality of Waterfowl Habitat by Refuge Tier, Fiscal Years 2002 through 2007: [Refer to PDF for image] This figure is a multiple vertical bar graph depicting the following data: Tier: Unstaffed satellite refuges; Habitat quality improved: 42%; Habitat quality stayed the same: 38%; Habitat quality worsened: 20%. Tier: Targeted reduction refuges; Habitat quality improved: 48%; Habitat quality stayed the same: 35%; Habitat quality worsened: 18%. Tier: Focus refuges; Habitat quality improved: 20%; Habitat quality stayed the same: 64%; Habitat quality worsened: 16%. Source: GAO analysis of Fish and Wildlife Service data. Notes: Results are based on data from 381 refuges that provided usable responses to the corresponding questions on our survey and that reported that providing waterfowl habitat was a priority at their refuge. See app. II for further details on this analysis. [End of figure] Figure 13: Change in Quality of Other Migratory Bird Habitat by Refuge Tier, Fiscal Years 2002 through 2007: [Refer to PDF for image] This figure is a multiple vertical bar graph depicting the following data: Tier: Unstaffed satellite refuges; Habitat quality improved: 42%; Habitat quality stayed the same: 43%; Habitat quality worsened: 15%. Tier: Targeted reduction refuges; Habitat quality improved: 54%; Habitat quality stayed the same: 30%; Habitat quality worsened: 16%. Tier: Focus refuges; Habitat quality improved: 20%; Habitat quality stayed the same: 59%; Habitat quality worsened: 20%. Source: GAO analysis of Fish and Wildlife Service data. Notes: Results are based on data from 407 refuges that provided usable responses to the corresponding questions on our survey and that reported that providing habitat for other migratory birds was a priority at their refuge. See app. II for further details on this analysis. [End of figure] In addition to analyzing the change in quality by tier, we analyzed changes in the quality of habitat as a function of the time spent by permanent staff at a refuge on habitat management activities. We found that refuge managers were more likely to report that habitat quality improved at refuges that increased the time spent on habitat management since 2002 than for those that reduced time, and were less likely to report that habitat quality worsened (see fig. 14). These results were consistent with our analysis of the change in quality of habitat as a function of staffing level changes at complexes and stand-alone refuges.[Footnote 26] Figure 14: Change in Quality of Waterfowl Habitat by Change in Time Spent on Habitat Management Activities, Fiscal Years 2002 through 2007: [Refer to PDF for image] This figure is a multiple vertical bar graph depicting the following data: Tier: Unstaffed satellite refuges; Time spent by permanent staff on habitat management increased: 47%; Time spent by permanent staff on habitat management stayed the same: 39%; Time spent by permanent staff on habitat management decreased: 14%. Tier: Targeted reduction refuges; Time spent by permanent staff on habitat management increased: 32%; Time spent by permanent staff on habitat management stayed the same: 56%; Time spent by permanent staff on habitat management decreased: 12%. Tier: Focus refuges; Time spent by permanent staff on habitat management increased: 30%; Time spent by permanent staff on habitat management stayed the same: 43%; Time spent by permanent staff on habitat management decreased: 27%. Source: GAO analysis of Fish and Wildlife Service data. Notes: Results are based on data from 374 refuges that provided usable responses to the corresponding questions on our survey and that reported that providing waterfowl habitat was a priority at their refuge. See app. II for further details on this analysis. [End of figure] The odds of habitat for waterfowl improving rather than worsening were significantly higher at refuges where staff time on habitat management activities increased rather than decreased between fiscal years 2002 and 2007. For example, among refuges where staff time increased, more than three times as many refuges experienced improved habitat for waterfowl (47 percent) as experienced worsened habitat (14 percent). In contrast, among refuges where staff time decreased, nearly the same number of refuges experienced improved habitat for waterfowl (30 percent) as experienced worsened habitat (27 percent). We found similar results when comparing change in staff time with the change in habitat quality for other migratory birds (see fig. 15). Figure 15: Change in Quality of Other Migratory Bird Habitat by Change in Time Spent on Habitat Management Activities, Fiscal Years 2002 through 2007: [Refer to PDF for image] This figure is a multiple vertical bar graph depicting the following data: Tier: Unstaffed satellite refuges; Time spent by permanent staff on habitat management increased: 51%; Time spent by permanent staff on habitat management stayed the same: 34%; Time spent by permanent staff on habitat management decreased: 15%. Tier: Targeted reduction refuges; Time spent by permanent staff on habitat management increased: 29%; Time spent by permanent staff on habitat management stayed the same: 63%; Time spent by permanent staff on habitat management decreased: 7%. Tier: Focus refuges; Time spent by permanent staff on habitat management increased: 31%; Time spent by permanent staff on habitat management stayed the same: 44%; Time spent by permanent staff on habitat management decreased: 25%. Source: GAO analysis of Fish and Wildlife Service data. Notes: Results are based on data from 400 refuges that provided usable responses to the corresponding questions on our survey and that reported that providing habitat for other migratory birds was a priority at their refuge. See app. II for further details on this analysis. [End of figure] When we developed statistical models of habitat change, refuges where staff time on habitat management activities increased were about 3.0 times more likely than refuges where staff time decreased to report improved, rather than worsened, habitat for both waterfowl and other migratory birds, even after accounting for the effects of tier designation and the change in external factors, such as extreme weather and agricultural activity, that may cause habitat problems. Resource prioritization at refuges obviously influences the ability of refuges to maintain quality habitat. Refuge managers told us that decisions on how many resources to direct to refuges are based on a variety of factors. Some managers weigh the management needs of all the refuges within a complex, taking into account the relative importance of the habitats as well as the amount of time and resources needed for the management activities. Unstaffed satellite refuges generally are smaller and have lower-priority habitats and, in some cases, the refuges are limited in what management can do, according to managers. Because some unstaffed refuges are located some distance from equipment and supplies and from where refuge staff are located, these distances and associated costs are taken into account as well. Other managers told us that they will undertake efforts mainly in response to specific problems identified at these refuges, while otherwise they generally do not expend resources for habitat management. According to refuge managers, it often is difficult to know what needs to be done at unstaffed refuges because staff generally do not visit the refuges very frequently to monitor the habitats. Even though the condition of many habitats is improving, many of these are still not high quality. Specifically, 40 percent of waterfowl habitats that improved since 2002 were still of moderate quality or poorer in 2007, while 65 percent of habitats that stayed in the same condition were of moderate quality or poorer. Similarly, for other migratory birds, 40 percent of habitats that improved in condition since 2002 were of moderate or poorer quality in 2007 and 55 percent of habitats that stayed in the same condition were of moderate quality or poorer. Refuge managers reported that, on average, habitats on about 44 percent of refuges for each of several types of key species—waterfowl, other migratory birds, threatened and endangered species, candidate threatened and endangered species, and state species of concern—were of high quality in 2007.[Footnote 27] A similar percentage of refuges deemed their habitats to be of moderate quality, and 7 percent to 20 percent of refuges reported habitats to be of low quality, depending on the species type (see table 5). Table 5: Habitat Quality by Species Type, Fiscal Year 2007: Species type: Waterfowl; Percent of habitats reported as high quality: 41%; Percent of habitats reported as moderate quality: 39%; Percent of habitats reported as low quality: 20%. Species type: Other migratory birds; Percent of habitats reported as high quality: 47%; Percent of habitats reported as moderate quality: 47%; Percent of habitats reported as low quality: 7%. Species type: Threatened and endangered species; Percent of habitats reported as high quality: 48%; Percent of habitats reported as moderate quality: 40%; Percent of habitats reported as low quality: 12%. Species type: Candidate threatened and endangered species; Percent of habitats reported as high quality: 37%; Percent of habitats reported as moderate quality: 46%; Percent of habitats reported as low quality: 17%. State species of concern 4741 13 Percent of habitats reported as high quality: 47%; Percent of habitats reported as moderate quality: 41%; Percent of habitats reported as low quality: 13%. Source: GAO. Notes: Refuge managers identified habitat quality for specific threatened and endangered species, candidate species, and state species of concern occurring on their refuges that are aggregated into these general “types of species” categories. Not all species occurred on every refuge. Some rows may not sum to 100 due to rounding. [End of table] Habitat quality is determined by the availability of several key components, including fresh water, food sources, and nesting cover, among other things, and the absence of habitat problems, such as invasive species. High-quality habitat generally provides adequate amounts of each of these main habitat components and is not significantly affected by habitat problems, while low-quality habitat generally lacks these components and may have significant problems. Moderate-quality habitat has a mixture of good and bad attributes. For example, a habitat may have an excellent tree canopy that provides good nesting areas and protection, but the underlying vegetation may be inadequate as a food source due to an infestation of an invasive species that has driven out native plants. Some aspects of moderate- quality habitat are acceptable, but the problems must be addressed overall for these habitats to fully support the species that depend on them, according to managers with whom we spoke. Other migratory birds, threatened and endangered species, and state species of concern appear to be faring the best with 47 percent or more of habitats on refuges deemed to be of high quality and 13 percent or less of habitats of low quality. Habitats for waterfowl and species that are candidates for listing under the Endangered Species Act are doing somewhat worse. Refuge managers told us that these findings may in part reflect the difficulty in addressing the very specific habitat needs of these species and a lack of focus on addressing those needs because they are not yet listed under the act. We spoke with some managers who have areas of robust, high-quality habitat on their refuge. For instance, Cache River NWR in Arkansas has approximately 65,000 acres of bottomland hardwood—45,000 acres of which are in pristine condition, with the remaining 20,000 acres in the early stages of regrowth after being restored from prior agricultural use. This refuge serves as an annual wintering area for 250,000 to 500,000 waterfowl including mallards, pintail, widgeon, gadwall, teal, and wood ducks. The refuge eradicated invasive kudzu plants using herbicides and its high-quality habitat provides necessary food, water, and cover for these waterfowl and also supports a variety of other migratory birds including warblers, indigo buntings, bluebirds, shorebirds, and wading birds. However, we also spoke to managers who reported low-quality habitat on their refuges. At Bowdoin NWR in northern Montana, for example, habitat is compromised by water quality and quantity problems as a result of activities on nearby lands, including haying and cropland use. The quantity of water that the refuge receives is insufficient to allow adequate flow-through of the water supply and, as a result, the water available for the refuge contains high levels of residual salt as well as agricultural chemicals, which affect the composition of vegetation and the survival of invertebrates. While a variety of bird species uses the refuge, including waterfowl, shorebirds, bald eagle, peregrine falcons, and piping plover—a federally listed threatened and endangered species—some populations of these species have declined over time. In addition to habitat quality, whether a refuge’s habitat is meeting the needs of key species types is an important indicator as to a refuge’s effectiveness in meeting its conservation mission. Our survey found that refuge managers reported that habitats at a majority of refuges were meeting the needs of key species types to a moderate or large extent in 2007 (see table 6). Table 6: Extent to Which Habitat Is Meeting the Needs of Species, Fiscal Year 2007: Species type: Waterfowl; Percent of habitats meeting the needs of the species to a large extent: 47%; Percent of habitats meeting the needs of the species to a moderate extent: 29%; Percent of habitats meeting the needs of the species to a small or no extent: 25%. Species type: Other migratory birds; Percent of habitats meeting the needs of the species to a large extent: 54%; Percent of habitats meeting the needs of the species to a moderate extent: 37%; Percent of habitats meeting the needs of the species to a small or no extent: 9%. Species type: Threatened and endangered species; Percent of habitats meeting the needs of the species to a large extent: 43%; Percent of habitats meeting the needs of the species to a moderate extent: 26%; Percent of habitats meeting the needs of the species to a small or no extent: 22%. Species type: Candidate threatened and endangered species; Percent of habitats meeting the needs of the species to a large extent: 37%; Percent of habitats meeting the needs of the species to a moderate extent: 28%; Percent of habitats meeting the needs of the species to a small or no extent: 26%. Species type: State species of concern; Percent of habitats meeting the needs of the species to a large extent: 44%; Percent of habitats meeting the needs of the species to a moderate extent: 30%; Percent of habitats meeting the needs of the species to a small or no extent: 21%. Source: GAO. Notes: Refuge managers identified habitat quality for specific threatened and endangered species, candidate species, and state species of concern occurring on their refuges that are aggregated into these general “types of species” categories. Not all species types occurred on every refuge. Some rows may not sum to 100 due to rounding and survey responses such as “no basis to judge.” [End of table] Some refuge managers indicated that poor-quality habitat could still meet the needs of some species to a large extent, just as high-quality habitat could fail to meet the needs of some species to a large extent depending on the species’ needs. High-quality habitat could fail to meet the needs of a given species if, for instance, the species’ population was too large for the refuge to support or if other species were competing for the same refuge habitat, according to managers. In contrast, a habitat of moderate quality could meet the needs of a species if that species population was small. Species that are candidates for listing under the Endangered Species Act appear to be faring the worst, while other migratory birds appear to be faring the best with regard to how well the habitat is meeting species’ needs. It is important to note that wildlife refuges are not necessarily intended to provide habitat for all types of species—some refuges were established to serve the needs of specific species such as waterfowl or a particular endangered species, for example—and do not necessarily focus on providing habitat for other species. When managers were asked to rate the importance of their habitat for the different types of key species we asked about, some of the refuge managers that reported low-quality habitats also rated the habitat as having low importance or priority for the species in question.[Footnote 28] For instance, about 56 percent of waterfowl habitat that managers reported as low quality, they also considered that habitat on their refuge to be low-priority habitat or not a priority for waterfowl; they considered about 28 percent of low-quality habitat for other migratory birds to be low-priority habitat or not a priority. Habitat Problems and External Threats Are Increasingly Affecting Refuges: Refuge managers reported that many refuges were negatively affected by a number of problems and external factors, including invasive species, habitat fragmentation, water quantity and quality problems, and soil erosion, and that these problems and factors were increasing (see table 7). Table 7: Common Refuge Problems Affecting Habitat and Trends in These Problems, Fiscal Years 2002 through 2007: Problem: Invasive plants; Percent of refuges reporting as a large problem: 55%; Percent of refuges reporting as a moderate problem: 23%; Percent of refuges reporting the problem as increasing: 55%. Problem: Habitat fragmentation; Percent of refuges reporting as a large problem: 44%; Percent of refuges reporting as a moderate problem: 21%; Percent of refuges reporting the problem as increasing: 57%. Problem: Invasive animals; Percent of refuges reporting as a large problem: 21%; Percent of refuges reporting as a moderate problem: 21%; Percent of refuges reporting the problem as increasing: 23%. Problem: Lack of water; Percent of refuges reporting as a large problem: 18%; Percent of refuges reporting as a moderate problem: 13%; Percent of refuges reporting the problem as increasing: 25%. Problem: Soil erosion; Percent of refuges reporting as a large problem: 15%; Percent of refuges reporting as a moderate problem: 13%; Percent of refuges reporting the problem as increasing: 22%. Problem: Water pollution; Percent of refuges reporting as a large problem: 11%; Percent of refuges reporting as a moderate problem: 17%; Percent of refuges reporting the problem as increasing: 12%. Source: GAO. [End of table] Invasive plant infestation was the most frequently reported problem, cited as a large problem on more than half of refuges and a moderate problem on nearly a quarter of refuges, and was reported to be increasing on more than half of refuges. This concern is consistent with information from the most recent refuge system performance report as well, which shows that more than 2.3 million acres of refuge lands are infested by invasive plants and more than 80 percent of refuges have at least some invasive plants present. Refuge managers with whom we met during site visits stressed that invasive plants have become a problem that affects the quality of their refuges’ habitats and threatens the quality of the refuge system as a whole. According to managers, these invading plants overtake native plant species that are used by animals for food and shelter, and have deleterious effects on biological diversity. For instance, the refuge manager at Merritt Island NWR in Florida told us that 30,000 acres of habitat on the refuge are infested with invasive plants including Australian pine, Brazilian pepper, Old World climbing fern, guinea grass, and cogongrass. In fact, the Brazilian pepper has infested every one of the 75 water level control structures on the refuge, with some impoundments more than 50 percent overtaken by the invasive plant. These invasive plants are eliminating native habitat and negatively impacting migrating birds such as rails, bitterns, sparrows, Florida scrub-jays, and other species, according to the refuge manager. We observed a common invasive plant, purple loosestrife, at several refuges we visited. Purple loosestrife crowds out native plants and can dramatically reduce food, shelter, and nesting sites for wetland- dependent species (see fig. 16). Figure 16: Purple Loosestrife: [Refer to PDF for image] Two photographs of Purple Loosestrife. Source: GAO. [End of figure] The refuge manager responsible for Antioch Dunes NWR north of San Francisco, described her refuge as being “in a constant uphill battle” against invasive plants, including vetch, thistles, and various grasses. The refuge is home to two endangered plants that depend on dune habitat on the refuge—the Contra Costa wallflower and the Antioch Dunes evening primrose—as well as the Lange’s Metalmark butterfly, a federal endangered butterfly species that occurs only on this refuge. However, the refuge has been inundated with a variety of invasive plants that, if not constantly addressed, threaten to overtake the native habitat, including the naked stem buckwheat on which the butterfly depends. The refuge manager told us that the butterfly population declined for 4 consecutive years, losing about 50 percent of the population each year, but increased in fiscal year 2007. In addition, refuge officials told us that invasive plants, like many other problems, can worsen if they are not dealt with swiftly. For example, a refuge may be able to completely eradicate an invasive plant if it addresses it early and thoroughly. In some cases this may require actions for several years in a row, as invasive plants frequently require consistent investment and treatment strategies from year to year. If not treated early, the infestation may spread exponentially and become a serious, long-term problem. Gains made in one year can be lost many times over if control efforts are not sustained. Invasive animals also are problematic for refuges and were reported to be a large problem on one-fifth of refuges. For example, nutria, a large rodent species from South America, has infested refuges in east, west, and Gulf coast states.[Footnote 29] Nutria can wreak havoc on water level control at refuges by burrowing into and destabilizing streambanks and damaging water control structures (see fig. 17). Figure 17: Nutria and Streambank Damage: [Refer to PDF for image] Two photographs of Nutria and Streambank Damage. Source: GAO. [End of figure] Habitat fragmentation was the second-most frequently identified problem for refuges, reported as a large problem at 44 percent of refuges, and increasing on 57 percent of refuges. Habitat fragmentation occurs when corridors of continuous habitat are disrupted, often by human development activities, which affects the refuge system’s ability to accomplish its wildlife conservation mission. The seriousness of this issue was highlighted at a recent meeting of the Western Governors’ Association by the release of a report on wildlife corridors.[Footnote 30] Specifically, the report discusses the rapid changes due to development across the United States—but in the West in particular—and how this adversely affects wildlife, and emphasizes the need for habitat connectivity for species survival. Some species, for example, require large areas of homogenous habitat for successful nesting, foraging, or movement. Managers at refuges close to urban centers showed us examples of development adjacent to their refuge that have cut off natural habitat corridors, which can lead to animals trying to cross busy roads or can cut them off from other members of their species leading to genetic homogeneity and inbreeding. For example, the refuge manager at Great Swamp NWR in New Jersey told us that increased development surrounding the refuge has fragmented or eliminated habitat. Valuable woodlots adjacent to refuge lands are decreasing in size or disappearing altogether around the refuge, limiting suitable nesting areas for species such as the red-shouldered hawk—a state threatened species. In addition, the manager said that movement by the bog turtle—a federal threatened species—has been constrained by fragmentation of its habitat. Managers of more rural refuges talked about increasing pressures to convert lands to agricultural uses, citing factors such as the increasing price of corn, or to industrial uses, such as oil and gas development. Habitat fragmentation sometimes occurs within a refuge’s “approved acquisition boundary.”[Footnote 31] A refuge’s approved acquisition boundary delineates an area that has been approved for inclusion in a national wildlife refuge but does not necessarily indicate that the entire area inside this boundary has been—or ever will be—acquired by FWS. An important conservation strategy for the refuge system that was codified in the Improvement Act is the ability to acquire important habitats, when possible. Thus, many refuges have acquisition plans for lands adjacent to or near existing refuge lands to complete or supplement current refuge habitat. For example, the acquisition plan for Nisqually NWR outside of Olympia, Washington, includes 4,470 acres for eventual purchase within its approved acquisition boundary (see fig. 18). The refuge manager at Nisqually reported that increasing urban development is one of the biggest problems facing the refuge. In addition to impacts such as reduced water quality and increased crime, the manager told us that quality habitat around the refuge is being lost to development despite an active refuge acquisition program, because the refuge cannot address all habitat and land protection needs at the pace necessary to offset habitat loss. Overall, the refuge system has purchased only limited amounts of land within the last 5 years, growing the system at approximately 0.25 percent per year. When asked about their ability to manage an even larger refuge system—a logical concern given the funding and staffing concerns currently facing the refuge system—several managers were quick to point out that simply protecting lands from development was a critical first step in conserving wildlife, even if they did not have the resources to actively manage the land. Figure 18: Nisqually National Wildlife Refuge Approved Acquisition Boundary: [Refer to PDF for image] This figure is a map of the Nisqually National Wildlife Refuge with land ownership depicted in the following categories: U.S. Fish & Wildlife service; WA Dept. Fish & Wildlife service; Fort Lewis (U.S. Army); Nisqually tribe; County land; City of Olympia; Private land; No data. Source: Fish and Wildlife Service. [End of figure] As noted previously, refuge managers also reported a number of external factors that contribute to habitat problems on refuges, including extreme weather and development, and that the contribution of these factors increased during our study period. Refuge managers told us that extreme weather has caused water levels to vary, which can result in the drying of wetlands, increased fire, and actual changes in the size and location of species ranges. Development activities also can increase air, soil, and water pollution to refuge lands and waters. For example, the refuge manager at Ridgefield NWR, which lies in a small watershed in southern Washington, told us increased urban development means more impervious surfaces such as roads, driveways, and sidewalks, thus increasing the amount of water polluted with oil, gasoline, yard chemicals, and animal waste, among other things, that runs directly into the refuge during rainfall, and decreases the amount of water that permeates the soil. In addition, development can increase visitation at refuges, which can negatively impact refuge resources. For instance, the refuge manager at the Upper Mississippi NWR told us that recreational boating has led to trash dumping, trampling of habitat, and excessive noise. Likewise, Ridgefield NWR must deal with increased litter, illegal dumping, increased trespassing, and damage to the habitat from increased refuge visitation levels. Agricultural activities near the refuges also can contribute pollutants to refuge lands and waters from runoff from animal waste and fertilizers, for example. Not surprisingly, managers reported that habitat was more likely to worsen at refuges where there was an increase in external factors that contribute to habitat problems, such as extreme weather and off-refuge agricultural activities (see fig. 19).[Footnote 32] For example, among refuges that reported no net increase in external factors, about 9.5 times more refuges reported improved waterfowl habitat (38 percent) than reported worsened waterfowl habitat (4 percent). By contrast, among refuges that experienced a net increase in external factors, the number of refuges that experienced improved waterfowl habitat (38 percent) was much closer to the number that experienced worsened habitat (22 percent). Figure 19: Change in the Quality of Waterfowl Habitat by Change in the Contribution of External Factors, Fiscal Years 2002 through 2007: [Refer to PDF for image] This figure is a multiple vertical bar graph depicting the following data: Contribution of external factors to habitat problems: No net increase; Habitat quality improved: 38%; Habitat quality stayed the same: 58%; Habitat quality worsened: 4%. Contribution of external factors to habitat problems: Net increase; Habitat quality improved: 38%; Habitat quality stayed the same: 40%; Habitat quality worsened: 22%. Source: GAO analysis of Fish and Wildlife Service data. Notes: Results are based on data from 381 refuges that provided usable responses to the corresponding questions on our survey and that reported that providing waterfowl habitat was a priority at their refuge. See app. II for further details on this analysis. [End of figure] We found similar results for the change in other migratory bird habitat (see fig. 20). Figure 20: Change in the Quality of Other Migratory Bird Habitat by Change in the Contribution of External Factors, Fiscal Years 2002 through 2007: [Refer to PDF for image] This figure is a multiple vertical bar graph depicting the following data: Contribution of external factors to habitat problems: No net increase; Habitat quality improved: 39%; Habitat quality stayed the same: 56%; Habitat quality worsened: 5%. Contribution of external factors to habitat problems: Net increase; Habitat quality improved: 41%; Habitat quality stayed the same: 39%; Habitat quality worsened: 21%. Source: GAO analysis of Fish and Wildlife Service data. Notes: Results are based on data from 407 refuges that provided usable responses to the corresponding questions on our survey and that reported that providing habitat for other migratory birds was a priority at their refuge. See app. II for further details on this analysis. [End of figure] Based on our statistical models, which assess the effects of a change in external factors while adjusting for the effects of tier designation and the change in staff time, we estimate that refuges that experienced no net increase in the number of external factors were about 7.0 times more likely to experience improved, rather than worsened, waterfowl habitat quality and 5.1 times more likely to experience improved, rather than worsened, habitat quality for other migratory birds. Time Spent on Certain Habitat Management Activities Increased at Many Refuges: Refuge managers reported increasing the time spent on a number of key habitat management activities on many refuges between fiscal years 2002 and 2007 (see table 8). Table 8: Habitat Management Activities That Increased the Most at Refuges, Fiscal Years 2002 through 2007: Activity: Addressing invasive plants; Percent of refuges that somewhat or greatly increased time spent on activity: 61%; Percent of refuges that somewhat or greatly decreased time spent on activity: 9%. Activity: Conducting comprehensive conservation planning; Percent of refuges that somewhat or greatly increased time spent on activity: 59%; Percent of refuges that somewhat or greatly decreased time spent on activity: 6%. Activity: Coordinating with nearby landowners; Percent of refuges that somewhat or greatly increased time spent on activity: 49%; Percent of refuges that somewhat or greatly decreased time spent on activity: 7%. Activity: Conducting habitat restoration projects; Percent of refuges that somewhat or greatly increased time spent on activity: 48%; Percent of refuges that somewhat or greatly decreased time spent on activity: 14%. Activity: Conducting routine habitat management activities; Percent of refuges that somewhat or greatly increased time spent on activity: 43%; Percent of refuges that somewhat or greatly decreased time spent on activity: 18%. Activity: Conducting inventory and monitoring surveys of habitat conditions; Percent of refuges that somewhat or greatly increased time spent on activity: 41%; Percent of refuges that somewhat or greatly decreased time spent on activity: 19%. Activity: Conducting inventory and monitoring surveys of wildlife populations; Percent of refuges that somewhat or greatly increased time spent on activity: 39%; Percent of refuges that somewhat or greatly decreased time spent on activity: 21%. Source: GAO. [End of table] Not surprisingly, given the number of refuges with invasive plant problems, refuge managers reported somewhat or greatly increasing time spent addressing invasive plant infestations on 61 percent of refuges, while somewhat or greatly decreasing time spent on this activity at only 9 percent of refuges. The refuge system’s national strategy for managing invasive species states that “invasive species are, collectively, the single greatest threat to native plants, fish, and wildlife with the potential to degrade entire ecosystems;” however, eradicating or effectively controlling invasive species through actions such as controlled burning, mowing, manual removal, and herbicide application is often resource intensive. For example, Antioch Dunes NWR is using a combination of managed grazing, herbicide application, manual removal, and tractors and other equipment to prevent several invasive species from harming two endangered plant species and an endangered butterfly species. In fiscal year 2007, about $9.8 million was budgeted for specific invasive species activities. For example, about $2.3 million was budgeted for invasive species strike teams in five specific areas of the country. These teams were designed to eradicate newly identified infestations before they become widespread with the goal of saving substantial funds in the long run. In addition, more than $500,000 was budgeted for mapping and tracking invasive plants on refuges and coordinating volunteer work for invasive species control activities. Eradication programs for specific species also were budgeted funds, such as $700,000 for nutria eradication at the Blackwater NWR and Eastern Neck NWR in eastern Maryland and the Southeast Louisiana Refuges Complex; over $1.3 million for spartina grass eradication at the Willapa NWR and Grays Harbor NWR in western Washington; and about $200,000 for an exotic rodent species at the Pacific Remote Islands NWR Complex. In addition to these specific programs, refuge managers also may spend resources on other invasive species-related activities on their refuges. Despite these investments, however, performance data from fiscal year 2007 show that only about 12 percent of the acreage identified as infested with invasive plants was treated in fiscal year 2007. The estimated cost of unfunded invasive species control projects found in the refuge system’s operational needs database was more than $150 million dollars at the end of 2007. The refuge system’s national strategy for invasive species control states that nearly half of all refuges report that invasive species infestations interfere significantly with their wildlife management objectives. Given that refuge managers reported that invasive plants were increasing at many refuges and that new invasive species gain a foothold in the United States every year, refuges will likely be constantly battling this problem. As might be expected given reported improvements in habitat quality during our study period, refuge managers reported increasing the time spent on basic habitat management activities such as haying, mowing, prescribed burning, or manipulating water levels on 43 percent of refuges and increasing time spent on habitat restoration activities, such as planting native grasses or trees, and creating water control structures, such as levees, at about 48 percent of refuges since fiscal year 2002. Basic management activities represent the day-to-day work that refuge staff perform to protect, conserve, and improve habitat conditions. Refuge managers, biologists, maintenance workers, and others routinely monitor water impoundments to ensure water levels are optimal for migrating birds, for instance. Restoration projects often have longer-term timelines and are meant to re-establish native habitats on the refuge. Such restoration projects—which may be as extensive as a restoration of a 100-acre grove of bottomland hardwood forest that could take nearly 50 years to complete, or as small as constructing a water impoundment to flood a small wetland area that could be constructed in a couple of months—are key to attracting and sustaining wildlife populations. Managers reported decreasing the time spent on basic management and restoration activities at 18 percent and 14 percent of refuges, respectively. Refuge managers’ responses on changes in the time spent on inventory and monitoring surveys of habitat conditions and wildlife populations—which are key activities that allow refuges to identify, report on, and manage wildlife populations and specific problems—were more mixed. Refuge managers reported increasing the time spent on these activities at 40 percent of refuges, while decreasing time spent on 20 percent of refuges since 2002. These surveys are an important way to understand how well wildlife populations and habitats are doing and whether a refuge is accomplishing its habitat management goals. Managers told us that having accurate data on habitat conditions and wildlife populations, among other things, is critical as they develop and deploy their comprehensive conservation plans. In addition, a number of refuge managers told us that their comprehensive conservation planning efforts led them to increase the amount of survey work they conducted, as the planning efforts require baseline data on habitat and wildlife conditions. For example, Rappahannock River Valley NWR in Virginia increased its inventory and monitoring surveys, partly to support development of its comprehensive conservation plan. In addition, accurate data from these surveys are important for correctly reporting data for FWS’s annual performance report and for early identification of problems affecting habitat. In the case of invasive species, for instance, a small infestation can spread exponentially over a very short time period. Some managers indicated that they are conducting fewer surveys, some of which are of lower quality. The managers said that they are depending more on volunteers or temporary workers to do the surveys, which can limit the survey quality because volunteers may not have the requisite background, experience, or training in biological survey methods. Some refuge managers told us that they have had to cut back on needed survey work due to staffing and funding shortfalls, while a few others told us that increasingly available and easy-to-use technologies have helped them increase the amount of survey work being done. Relatedly, refuge managers reported increasing time spent developing comprehensive conservation plans at nearly 60 percent of refuges since 2002—about two-thirds of refuges reported engaging in planning activities in fiscal year 2007. This is not surprising, given the requirement that all refuges must complete their plans by 2012, and the fact that less than 50 percent have been completed to date. The conservation planning process can be time consuming, given the need to hold public meetings, conduct environmental reviews, and coordinate with state and local entities. During our interviews, several refuge managers told us that they generally did not obtain additional staff to develop these plans; instead, they have had to shift responsibilities from existing staff or curtail other refuge management activities to devote time to the plans. For example, one refuge manager reported that the refuge set aside an invasive species eradication project after several years of implementation in order to work on the conservation plan for the refuge. To minimize the time such planning has taken away from other refuge activities, some refuge managers we interviewed told us that they or their staff worked on the plans on their own time. One refuge manager reported that he worked on the comprehensive conservation plan on weekends and converted the refuge’s biologist to a full-time planner on a temporary basis. Another manager stated that refuge staff attended fewer public meetings that were not related to the comprehensive conservation plan so they would not have to cut back on other refuge work. Consistent with the importance that FWS places on working with owners of lands adjacent to and near refuges, refuge managers reported coordinating with landowners at more than 85 percent of refuges in 2007 and increasing the time spent on this activity at 50 percent of refuges since 2002. This coordination with adjacent landowners is increasing as concerns about habitat fragmentation and off-refuge pollution grow, and FWS increasingly is considering the need to deal with the broader ecosystems of which the refuges are a part. In addition, two refuge managers noted that coordination with adjacent landowners has the added benefit of being a good outreach tool and of giving the refuge, and the refuge system, a better image and a more positive status with the broader public. Some refuges have employees specifically designated to undertake these efforts, while others depend on the work of the refuge manager for these efforts. Permanent staff and volunteers increased the time spent on habitat management activities at 48 percent and 45 percent of refuges, respectively (see table 9). However, consistent with shifts in resources as a result of workforce planning, permanent and temporary staff time spent on habitat management decreased at 29 percent and 19 percent of refuges, respectively. Table 9: Change in Time Spent by Type of Worker on Habitat Management Activities, Fiscal Years 2002 through 2007: Type of staff: Permanent staff; Percent of refuges reporting an increase in time spent: 48%; Percent of refuges reporting that time spent stayed the same: 21%; Percent of refuges reporting a decrease in time spent: 29%; Percent of refuges reporting not using this type of staff: 2%. Type of staff: Volunteers; Percent of refuges reporting an increase in time spent: 45%; Percent of refuges reporting that time spent stayed the same: 30%; Percent of refuges reporting a decrease in time spent: 10%; Percent of refuges reporting not using this type of staff: 15%. Type of staff: Temporary staff; Percent of refuges reporting an increase in time spent: 32%; Percent of refuges reporting that time spent stayed the same: 23%; Percent of refuges reporting a decrease in time spent: 19%; Percent of refuges reporting not using this type of staff: 27%. Type of staff: Cooperators; Percent of refuges reporting an increase in time spent: 32%; Percent of refuges reporting that time spent stayed the same: 35%; Percent of refuges reporting a decrease in time spent: 7%; Percent of refuges reporting not using this type of staff: 27%. Type of staff: Contract workers; Percent of refuges reporting an increase in time spent: 25%; Percent of refuges reporting that time spent stayed the same: 23%; Percent of refuges reporting a decrease in time spent: 7%; Percent of refuges reporting not using this type of staff: 46%. Source: GAO. [End of table] Refuge managers have discretion over the activities on which staff spend their time. For instance, individual refuges face different challenges—due to such things as natural weather cycles and increases or decreases in habitat problems or external factors affecting habitat—that managers need to address. Other influences over how to prioritize staff time may result from management decisions, such as the decision to focus more on working with adjacent landowners, or statutory requirements, such as the comprehensive conservation plans that must be completed. In discussing habitat management activities with refuge managers, managers indicated that unstaffed satellite refuges will generally only see habitat management work on an infrequent or “as-needed” basis. For instance, the manager for a refuge complex in southern Washington state told us that the complex’s unstaffed satellite refuge, which is not open to the public, gets management attention only during a few weeks per year when equipment and staff are available. Similarly, an unstaffed refuge in a complex in New Hampshire has been visited by refuge staff only five times in the past 7 years, according to the complex manager. Based on our site visits, we l