|Bird Conservation Planning in
the Interior Low Plateaus
Robert P. Ford1 and Michael D. Roedel2
Bird Conservation Areas (BCAs) can be identified over the landscape of a physiographic area using geographic information system (GIS) technology, breeding bird atlas data, Partners in Flight (PIF) species prioritization scores, and land management partnerships. The physiographic area is a useful geographic scale at which to set conservation and management objectives, because physiographic areas often have similar land forms and land uses, have similar bird distribution patterns and bird conservation needs, and transcend traditional political boundaries, such as state lines. The objectives of this paper are to describe the method we used for the first phase of bird conservation planning in a specific physiographic areathe Interior Low Plateaus (ILP)and to present the results, including the identification of potential BCAs for bird species assemblages that nest in the ILPs mature hardwood forests, early successional forests or old fields, and open lands.
The ILP encompasses more than 12 million ha in southern Ohio, Indiana, and Illinois; central Kentucky; central Tennessee; and northern Alabama. The rolling topography of this area has been dominated historically by oak-hickory forests, with areas of rock outcrops and glade habitats, prairies, and barrens (Martin et al. 1993). All habitats now are highly fragmented. Currently, this project includes only Kentucky, Tennessee, and Alabama; the remainder of the ILP will be incorporated later. Distinct subdivisions within the current scope of the project include the Bluegrass region and Shawnee Hills in Kentucky; the Western Highland Rim, Eastern Highland Rim, and Central Basin of Tennessee; and the Tennessee River Valley of Alabama (Figure 1). These subdivisions serve as distinct conservation planning units. About 95% of the land base consists of non-industrial forest lands, open lands for agriculture (pasture), and urban areas. Public lands and lands managed by the forest products industry make up less than 5% of the total area (Vissage and Duncan 1990).
Figure 1. Interior Low Plateaus and its subdivisions in Kentucky, Tennessee, and Alabama.
The Tennessee Conservation League, state affiliate of the National Wildlife Federation, has served to facilitate PIF objectives in the ILP with leadership from state wildlife agencies and the forest products industry. A steering committee developed to provide oversight, direction, and implementation included representatives of the Alabama Department of Conservation and Natural Resources, Game and Fish Division; Champion International Corporation; Kentucky Department of Fish and Wildlife Resources; Tennessee Department of Environment and Conservation; Tennessee Conservation League; Tennessee Ornithological Society; Tennessee Valley Authority; Tennessee Wildlife Resources Agency; U. S. Fish and Wildlife Service; U. S. Forest Service; Westvaco Corporation; and Willamette Industries Incorporated. In addition, a technical working group was formed that included land managers and biologists from the above groups.
The first phase conservation planning objectives were (1) to identify centers of abundance for species and species assemblages by subdivision within the ILP, (2) to identify specific lands managed by the project's current cooperators where integration of nesting songbird management is a high priority, (3) to identify areas with the highest potential for restoration of degraded habitats, (4) to identify and prioritize areas for potential acquisition and/or public-private partnerships for conservation, and (5) to identify areas that require more extensive breeding bird inventories.
Tools and information
The conservation planning objectives were accomplished by using the best available information from vegetation maps generated by interpretation of satellite imagery from the Landsat Thematic Mapper; data from each state's breeding bird atlas; PIF species prioritization values (Carter et al. this volume); and known boundaries from public and private cooperators' lands. This process was modified from the Gap Analysis protocol (Scott et al. 1993). A standard vegetation classification and habitat patch size distribution, including areal coverage and spatial location information, were fundamental layers of information. For this planning phase, the satellite imagery was processed using GIS (ARC INFO) to provide broad landcover categories that included deciduous forest, coniferous forest, mixed forest, wetlands, agriculture, open water, and urban areas. These analyses were completed in Tennessee by the Tennessee Gap Analysis (Jones et al. 1995) and Biodiversity (Reid 1993) projects, and in Alabama by the Tennessee Valley Authority. This process is not complete in Kentucky, although Gap Analysis has been initiated by the Kentucky Department of Fish and Wildlife Resources and Murray State University.
Breeding bird distribution and relative abundance were documented in the ILP by breeding bird atlas projects in each state. Tennessee (Nicholson in press) and Kentucky (Palmer-Ball 1996) have completed breeding bird atlas projects; observers have completed about 70% of the coverage for a breeding bird atlas in northern Alabama. Generally, these atlas projects followed national protocols. One breeding bird atlas block was designated for a sixth of each of 820 USGS 7.5 minute topographic maps in the ILP. During the atlas periods, over 95% of the atlas blocks were completed. For this analysis, we used breeding bird atlas information from 810 of the topographic maps. Vegetation maps based on GIS technology were available for all of middle Tennessee and northern Alabama, but not available for most of Kentucky at the time of this analysis. Although the coverage for breeding bird atlas and vegetation mapping were incomplete at the time of this writing, the analysis has provided preliminary results, clarification for future planning needs, and a test of the process. Although observer effort differed among states, results provided sufficiently consistent data for this process in the ILP. Breeding bird species, bird species assemblages, and habitats were prioritized using the PIF concern scores for the ILP (Hunter et al. 1993a, Carter et al. this volume). Finally, the boundary maps of current cooperator lands were included in the GIS.
Conservation planning process
Of the 15 species for which prioritization scores exceeded 23 in the ILP, six species were typical of mature hardwood forests, six of old fields and early successional forests, and three of open lands. To more fully represent these habitat types, we selected a species assemblage that included species with concern scores of 19 or above. Birds within each assemblage were selected to represent different habitat conditions within the broad habitat category, and loosely followed management assemblages as described by Block et al. (1995).
Three bird species were chosen to represent the species assemblages within each of three broad habitat groups (Table 1). The mature hardwood forest bird species assemblage was represented by Cerulean Warbler, Worm-eating Warbler, and Black-and-white Warbler. Both Cerulean Warbler and Worm-eating Warbler were of high concern score for PIF (Hunter et al. 1993a); although Black-and-white Warbler scored of moderate concern, this species was included because of its area sensitivity (Hamel 1992). The early successional forest and old field bird species assemblage was represented by Prairie Warbler, Blue-winged Warbler, and Yellow-breasted Chat, all of which scored 23 or above. The open lands bird species assemblage was represented by Grasshopper Sparrow, Dickcissel, and Loggerhead Shrike, all of which scored 23 or above. Nine additional species that were too rare or locally distributed to fit into a species assemblage were listed for single species management or monitoring priority.
Table 1.(A) Species assemblages selected for each of three broad habitat types in the Interior Low Plateaus physiographic province and; (B) bird species of single species management or monitoring priority. The number represents the Partners in Flight concern score in the Interior Low Plateaus (Hunter et al. 1993).
Breeding bird atlas blocks were then identified in which at least two of the three species of each assemblage had been recorded. Each of the blocks were placed in one of four conservation planning categories based on proximity to cooperator lands, inventory needed, and landscape habitat characteristics. These categories were designed to set direction for management and maintenance of existing populations on or near current cooperator lands, and to identify additional locations where management or restoration of habitats will be a priority.
The first category includes atlas blocks on which all three species of the assemblage were recorded and cooperator lands are present. The conservation objective here is to work with traditional land management objectives to sustain or increase target populations. The second category includes atlas blocks on which all three species in the assemblage occur and cooperator lands are nearby; that is, no more than three atlas blocks (12 miles) away. The conservation objective here is to identify the best opportunity for public-private partnerships to increase core bird populations. The third category includes atlas blocks on which all three species occur and vegetation maps indicate that habitat cover will sustain populations, yet the block is far-removed-from cooperator lands (greater than three atlas blocks away). The conservation objective here is to conduct further bird inventories and to search for opportunities for public-private partnerships. Conservation and management actions include an increased opportunity for biologists and land managers to cooperate with nearby landowners to investigate populations and recommend management options to sustain or increase the sizes of core populations. The fourth category includes atlas blocks on which only two of the three species within an assemblage were recorded, and which are far removed from cooperator lands. The conservation objective here is to prioritize for further bird inventory and monitoring.
Conservation plans for priority atlas blocks that are far removed from any cooperator's land depend on whether two or three species in an assemblage have been recorded there. Blocks having all three species, and where existing vegetation maps from GIS indicate that habitat patch size or land-use patterns would sustain populations of the species assemblage, will be prioritized for cooperative management agreements or incentive programs. Atlas blocks on which two of three species occurred, and where vegetation maps indicate that habitat size or land use is marginal to sustain populations, will be explored for restoration opportunities. Known locations of rare and locally distributed species (see Table 1) will be prioritized for management or monitoring, and will be placed into one of the above categories as the planning process continues.
The mature hardwood forest bird species assemblage (Cerulean Warbler, Worm-eating Warbler, Black-and-white Warbler) occurred in 46 of 810 (6%) atlas blocks in the ILP; two of these three species occurred in 72 of the 810 blocks (9%) in the region (Figure 2). This species assemblage was most common in the Western Highland Rim, which accounted for 50% of the blocks in which all three bird species of the assemblage occurred and 53% of those with at least two of the three bird species. The Eastern Highland Rim ranked second among subdivisions for the distribution of the mature hardwood forest bird species assemblage; approximately 11% of the atlas blocks with this assemblage occurred in Kentucky's Eastern Highland Rim.
The remaining hardwood species assemblage locations were scattered widely, often along the transition between subdivisions, commonly along a forested elevation gradient, such as between the Eastern Highland Rim and Cumberland Plateau. Other locations were at the edges of the Bluegrass Region and the Shawnee Hills and near the Ohio River or its tributaries.
Each atlas block with the mature hardwood forest species assemblage was assigned to a conservation category (Table 2). Twelve atlas blocks were placed in the existing category, nine of which are in the Western Highland Rim. There were 16 atlas blocks with the bird species assemblage in the close-to cooperator category, 18 atlas blocks in the far-removed-from cooperator lands category, and at least two of three species occurred in 72 atlas blocks and were in the increased inventories category. The designation of blocks in the last two categories remains unclear because of the current lack of a GIS vegetation map in Kentucky.
Table 2. Conservation categories defined with recommended conservation actions.
The early successional or old field bird species assemblage (Blue-winged Warbler, Prairie Warbler, Yellow-breasted Chat) occurred throughout the ILP with nearly uniform frequency; observers recorded all three species in 126 of 810 (16%) atlas blocks. Two of three species occurred in 379 of 810 (47%) blocks (Figure 3). Approximately 43% of the 126 blocks that had all three of these birds are in the Western Highland Rim. Otherwise, the species assemblage was evenly distributed among the Eastern Highland Rim (17%), Bluegrass Region (17%), and Shawnee Hills (15%). The species assemblage occurred least frequently in the Central Basin (7%). Although these birds were widely distributed, notable gaps include much of the agricultural lands in the Tennessee Valley of northern Alabama, parts of the Shawnee Hills, the Western Highland Rim in Kentucky, and parts of the Bluegrass region.
Blocks were distributed evenly among existing, close-to cooperator, and far-removed-from cooperator conservation categories (Table 3). Early successional forest or old field habitat was the only habitat for which the percentage of existing blocks was higher than that of the close-to cooperator category, probably because of the acreage and management of cooperating forest products industries. This species assemblage occurred in all subdivisions except the Tennessee Valley. The Western Highland Rim had the highest number of blocks with the species assemblage (21). The blocks in the increased-inventory category were widespread and occurred uniformly across the region.
Table 3.Preliminary assignment of mature hardwood forest bird species assemblage to conservation categories by subdivision in the Interior Low Plateaus. In the existing category all three species occurred in a Breeding Bird Atlas block which also had cooperator lands. In the close to cooperator category, all three species occurred in a Breeding Bird Atlas block within 12 miles of cooperator lands. In the far removed category, all three species occurred in a Breeding Bird Atlas block greater than 12 miles from cooperator lands. In the category for increased inventory, two of the three species occurred in a Breeding Bird Atlas block. The objective of the process is to move the highest possible number of sites into the existing category through increased partnerships and cooperative land management.
The open lands bird species assemblage (Grasshopper Sparrow, Dickcissel, and Loggerhead Shrike) occurred on 108 of 810 blocks (13%); two of these three species occurred on 177 of 810 (21%) blocks (Figure 4). The 108 blocks where all three species occurred were most widely distributed in the Eastern Highland Rim and the Shawnee Hills (32% and 30% respectively), in Kentucky. The Western Highland Rim in Kentucky had a relatively high percentage of the blocks with all three species (23%). The remainder of blocks for the open lands species assemblage occurred in the Central Basin of Tennessee and the Bluegrass Region of Kentucky. Areas containing historically prairie, barren, or glade habitats (see Martin et al. 1993) were compared to atlas locations for this species assemblage. At least 29 occupied blocks occurred in the historical area of the Kentucky Barrens, four in Tennessee's Central Basin, and eight in Tennessee's barren habitats (a total of 41 or 38% of 108).
Sixteen (6%) of the open lands bird species assemblage blocks occurred in the existing-conservation category, 40 (14%) were in the close-to cooperator category, 52 (18%) in the far-removed category, and 177 (62%) in the increased-inventory category (Table 4). Of the three broad habitat types, birds of the open lands species assemblage occurred the least frequently in atlas blocks with existing cooperators lands. However, the species assemblage occurred in five blocks with existing cooperators lands in the Shawnee Hills and in seven in the Eastern Highland Rim.
Table 4. Preliminary assignment of old field and/or early successional forest bird species assemblage to conservation categories by subdivision in the Interior Low Plateaus. In the existing category all three species occurred in a Breeding Bird Atlas block which also had cooperator lands. In the close to cooperator category, all three species occurred in a Breeding Bird Atlas block within 12 miles of cooperator lands. In the far removed category, all three species occurred in a Breeding Bird Atlas block greater than 12 miles from cooperator lands. In the category for increased inventory, two of the three species occurred in a Breeding Bird Atlas block. The objective of the process is to move the highest possible number of sites into the existing category through increased partnerships and cooperative land management.
Bird conservation planning for the PIF initiative has emphasized a multiple species as opposed to a single species approach (Finch and Stangel 1993), a planning process to be developed by physiographic area (Hunter et al. 1993b, Carter et al., this volume) with implementation by states (Williams and Pashley, this volume), and an approach to stop population declines before species become critically endangered. Within the ILP physiographic area planning unit, BCAs are continuous blocks of habitat of sufficient size and quality to support healthy populations of all of the birds within one of the three species assemblages. The conservation strategy for existing sites should be management and maintenance to sustain existing populations. Habitat restoration and consolidation will be required to turn marginal areas into functioning BCAs.
Bird species assemblages of three broad habitat types in the ILP were placed into conservation and management categories based primarily on geographic relationship to current cooperator lands. A conservation and management action was associated with each category. This process was designed to be hierarchical, with a goal to move the maximum number of areas into the "existing" category. Species assemblages that occur close to existing cooperators lands are higher in the hierarchy because of opportunities for targeted technical assistance programs. Areas harboring full species assemblages that are far removed from existing cooperator lands and areas in need of increased inventory, where only two of the three species of an assemblage occurred, will be examined for opportunities to maintain or create viable BCAs.
The mature hardwood forest bird species assemblage was the most limited in the ILP, which represented a dramatic reduction in recent years (see Mengel 1965, Nicholson in press, Palmer-Ball 1996). This species assemblage occurred most commonly in the Western Highland Rim of Tennessee, where a management and maintenance strategy has been undertaken to sustain existing populations and secure 50 sites across the landscape within 25 years. These areas are designed to be approximately 4,000 ha of forested lands, open to a variety of ownerships and forest land management, and are based on core bird populations. Management recommendations for cooperator lands integrate bird management into current management objectives of the cooperator (Ford 1995). Many of the areas within the Shawnee Hills, Bluegrass region, Eastern Highland Rim, and Tennessee Valley, where forest birds are fewer, will require habitat restoration and consolidation in order to achieve long term security. This species assemblage was not a historical component of the inner Central Basin avifauna (Ford and Hamel 1988).
The old field and early successional forest bird species assemblage was the most widespread in the ILP, most commonly on the Western Highland Rim. The relatively high number of atlas blocks where this and the mature hardwood forest bird species assemblage occurred reflects the existing managed forest landscape in this subdivision. However, increasing development from residential and other pressures will require a strategy to maintain forest cover in the Western Highland Rim. This strategy is being implemented by forest products industry cooperators who maintain patches of early successional forests in the context of large blocks of mature forest. This species assemblage occurred in certain successional stages of cedar glades and barrens as well, particularly in the Central Basin, and parts of the Shawnee Hills, where habitat maintenance and restoration will be recommended.
The open lands bird species assemblage was scattered primarily in the Shawnee Hills and Western Highland Rim of Kentucky and the southeast section of the Eastern Highland Rim in Tennessee and Alabama. Although these species may have occurred infrequently in native prairies and glades of Kentucky (see Palmer-Ball 1996), these locations reflect the historical distribution of barrens, prairies, and glade habitats, as well as an increase in pasturelands. The open lands bird species assemblage occurs infrequently on cooperator lands because the cooperators primarily manage forest lands. This species assemblage offers an opportunity to better manage areas that were historically prairies, barrens, or other open lands of the region. In addition, some habitat consolidation may be pursued to increase partnerships among landowners that manage large acreage of pasture lands.
The next steps in this process include:
* Increase field inventories to verify occurrences of species assemblages on current cooperator lands
* Develop habitat management recommendations for cooperating land managers
* Decide on criteria to prioritize areas for restoration and/or increased inventories
* Test the predictive value of landscape habitat models for the occurrence of species assemblages
* List the assumptions used in this planning process and develop them as research projects
* Include non-breeding birds in this process
* Further coordinate project recommendations with existing state priorities for bird conservation, management, and monitoring.
We would like to thank Brainard Palmer-Ball Jr., Charles Nicholson, Keith Hudson, and Lisa Spencer for sharing unpublished breeding bird atlas data from their respective states and commenting on the manuscript. We thank Sunni Lawless and Kirk Miles for assistance with technical work, as well as image processors and others with the Gap Analysis teams from Tennessee Tech University, Tennessee Wildlife Resources Agency, Murray State University, Kentucky Department of Fish and Wildlife Resources, and Tennessee Valley Authority. The manuscript was much improved through discussions with Chuck Hunter, Bob Cooper, Paul Hamel, Alan Mueller, and comments from reviewers.