Products This section presents the existing geographic data, publications, news articles, the GAP home page, and meetings, symposia, and workshops that have resulted as GAP products. Not accounted for here is the effect GAP is having on building institutional relationships around information. Although the results from the GAP state-level partnerships are easily of equal value to more tangible products such as data, the products resulting from better integrated institutional relationships are more complex, more difficult to quantify, and require a great deal of time and effort. Analyses Gap Analysis provides a way to assess how well vertebrate species and habitat types are represented in the current network of conservation lands (public and private lands managed for long-term biodiversity maintenance). In addition, it can locate general areas where the most efficient opportunities exist for managing the species and habitat types that are presently under-represented in conservation lands. As the only current synoptic data of natural land cover and species distributions, the GAP data layers are also used for many other purposes, ranging from county-level zoning decisions to basic research. In the original concept, however, the final analytical product from GAP is not a list of "gaps," but a map showing areas of opportunity for managing biodiversity resources in order to avoid extinction crises in the long term (Scott et al. 1987, Scott et al. 1993, Csuti 1994a). Substantial discussion and debate have occurred about specific reserve design, for example, which elements within a particular reserve should be emphasized in determining reserve size and shape (see SoulJ 1987, Simberloff and Abele 1982, Noss and Cooperrider 1994). However, there has been far less effort devoted to strategies for overall system design with the goal of maintaining representative examples of the broad categories of biodiversity, such as vertebrate species and natural community alliances. The analyses of GAP data provide such an opportunity. Because this area of conservation biology is less developed, it is hoped that the GAP activities will stimulate discussion, debate, experimentation, and discovery, leading to continued progress. A brief review of some methods of analyses and actual analyses conducted is provided in the following sections.
Csuti (1994a) reviews past concepts and efforts at reserve selection and then provides detailed recommendations for applying GAP data to reserve selection. He covers:
Wright et al. (1994) used GAP data to examine how the degree to which unprotected habitat types would be conserved under proposals for four new national parks in Idaho. The proposals involved four separate areas of about 220,000 hectares each. They found that the four proposed new park areas would actually add little to improving the conservation status of unprotected habitat types because 67-78 percent of the habitat types found within the proposed new park areas were already protected elsewhere. Between 16-30 percent of the land areas in the proposed new park areas would provide conservation of habitat types which were previously unprotected. The protection provided by each of the new park proposals could be enhanced by altering the configuration of the proposed areas with the addition of relatively few hectares.
Caicco et al. (1995) used the GAP data for Idaho to assess the extent and degree of conservation management afforded to 71 natural land cover types. They identified six types having no representation within the network of conservation lands and five other types that are represented on less than 1,000 hectares. An additional 18 land cover types are represented on less than 5,000 hectares each and 29 other types on less than 10,000 hectares each. Economically valuable forest types, such as Western redcedar, Western hemlock, and grand fir, are poorly represented in conservation lands. Most of the opportunities for meeting the conservation needs of these types occur on land managed by federal agencies. This study showed that Gap Analysis is an efficient and useful method for assessing the adequacy of biodiversity conservation across large areas (i.e., >200,000 km2).
Merrill et al. (1995) used GAP data for Idaho to evaluate four options of proposed wilderness allocation in that state. The options ranged from not establishing any new wilderness areas to the designation of all eligible public lands. Each option was evaluated for its potential to conserve 10 percent of the area occupied by each of 113 native vegetation types and a minimum of 10 percent of the distribution of each of the 368 non-fish native vertebrate species in the state. The authors found that none of the options produced an optimal configuration in terms of conserving the greatest number of species and habitat types with the minimum amount of land area. Because over half of the vegetation types in the state occupy only 1.5 percent of the land area occurring in a small number of distinct areas, their probability of occurring within wilderness areas is low. The option which would conserve the most species and habitat types would cover 56 percent of species and 65 percent of vegetation types, but these same results could possibly be obtained with less land area. This study showed how GAP data can be used to evaluate some ecological parameters of large-scale land use planning.
In their report on the Utah GAP project, Edwards et al. (1995) and Edwards (1995) found that less than 0.1 percent of the state's land area met the Level 1 status of areas having a management plan in operation to maintain a natural state and within which natural disturbance events are allowed to proceed without interference. Almost 4 percent of the state's land area met the Level 2 status of areas generally managed for natural values, but which may receive uses that degrade the quality of existing natural communities. About 70 percent met the Level 3 category of areas for which legal mandates generally prevent permanent land cover conversions from natural or semi-natural habitats to anthropogenic habitats, such as conversions to agriculture, but which are subject to extractive uses such as silviculture or mining. The remaining 25 percent were found to be Level 4-type landsCareas managed for intensive human uses. Table 3 shows the number of land cover types in Utah which have either greater than 10 percent, 6-10 percent, or less than 5 percent of their land area in conservation lands. Of the six types with more than 10 percent of their area within Level 1 or Level 2 management status, four were found at high elevations while the remaining two are wetlands and non-vegetated areas. Five other cover types had 6-10 percent of their total areas in Level 1 or 2 management areas, and the remaining 25 cover types had less than 5 percent in Level 1 or 2 management areas. This analysis reveals that the majority of land cover types have less than 10 percent of their land area within the network of conservation lands.
Table 3. Land cover types and management status. For native terrestrial vertebrate species, the degree of habitat conservation varied by major taxa (reptiles, amphibians, birds, mammals), as shown in Table 4. No clear pattern emerged from an examination of those species with less than 19 percent of their habitats in Level 1 or Level 2. Almost all of the bird species having greater than 10 percent of their habitat in conservation lands are those associated with wetlands, which are well represented in Level 2 management areas. Most mammal species that have greater than 10 percent of their habitat in conservation lands are those typically found at high elevations and which occupy habitat in wilderness areas. The authors conclude, though, that how the distributions of these habitat types interact with the surrounding Level 3 lands must be studied further before any definitive evaluation about their value for maintaining viable populations can be made.
Table 4. Percent of species in major taxonomic groups with habitats represented in Level 1 and Level 2 lands in Utah, show by icrements of 0-5%, 6-10%, and >10%.
Davis et al. (1995) and Stoms and Davis (1995) conducted an analysis of conservation gaps in the Southwestern California region using the California Gap Analysis data. For land management generally, they found that the majority of land above 2,500 meters is in Level 1 management, yet it comprises less than 1 percent of the total land area of the region. The largest single portion of Level 1 management land is National Forest Wilderness Areas. The conservation status of 76 dominant woody species and 62 natural communities was evaluated. Nineteen of the natural communities were found to be under-represented in conservation lands. Natural communities at low elevations seemed to be at the greatest risk since the vast majority of these lands have already been converted to agricultural and urban uses, and most of the remaining land area is threatened with future conversion. Davis et al. (1994) evaluated the distribution and conservation of the coastal sage scrub natural community in Southwestern California. In this study, they found that 71 percent of the mapped remaining coastal sage scrub vegetation type occurs on Level 3 private lands, 22 percent on Level 2 public lands, and 7 percent on Level 1 lands. Many of the private lands that are zoned as open space are either idle or grazed. Private lands not zoned for open space are under extreme development pressure.
Using the Idaho GAP data set, Stoms (1994b) examined the effects of sampling unit size on patterns of vertebrate species richness. He used species distributions mapped from two one-degree latitude/longitude blocks occurring in separate physiographic provinces of Idaho. Species richness was resampled into five sequentially-sized grids to provide five different sample sizes. To interpret the spatial pattern in terms of ecological factors, the relative contribution of alpha and beta diversity to the overall gamma richness in each sampling unit was determined and compared across scales and between sites (see Whittaker 1960, 1977, Stoms and Estes 1993 for more on alpha, beta, and gamma diversity). Maps of richness patterns at five sampling unit sizes for the two study areas were produced. Variability in richness across sampling unit areas and for both sites were plotted, then the relationship of alpha and beta indices to the overall richness was plotted as a function of sample unit size. Stoms (1994b) concluded that species richness or species density is an ecological factor, and its measurement is dependent on the resolution at which it is sampled. Predicted patterns of gamma (landscape) species richness clearly are scale-dependent. As sampling unit sizes increase, the mapped distribution of richness changes in inconsistent ways. Recommending a single scale for mapping species richness would be inappropriate. Optimal sample size is a function of regional environmental factors.
Kiester et al. (in press) experimented with the GAP prototype data from Idaho to explore methods of identifying general areas for conservation priorities. The occurrences of: (a) individual species, (b) habitat types, and (c) species considered endangered, threatened, or candidates as such (ETC), were each smoothed to the EPA's 635 km2 hexagonal grid. They predetermined that 83 species are under-represented within conservation lands in Idaho and labeled these as "needy species." Their goal was to determine the best method for identifying general areas that, when selected sequentially, would have the greatest positive cumulative impact on attaining adequate representation of these needy species in conservation lands by developing more localized research and management options in the future. This is a basic optimization problem commonly known as the "set coverage problem." By applying a variety of different mathematical algorithms to these data, Kiester et al. showed an exact set coverage of four hexagons for all vertebrate species, habitat types, ETC species, and for the 83 needy species. They showed the locations of the four hexagons that contain the largest number of needy species. In this process, they discovered that there are 16 combinations of 4 hexagons that capture the maximum number of under-represented species. The authors also conducted a "sweep analysis" to compare the number of total vertebrate species that would be carried along by each of the three other priority-setting schemes (i.e., those four hexagons in the state that capture the largest number of habitat types, or of ETC species, or of needy species). This analysis was done in order to measure the effectiveness of each of the priority-setting schemes at covering all other vertebrate species. The needy species exact set coverage swept along the most total vertebrate species (320 out of 357), the ETC sequence swept along the next most (277 out of 357), and the habitat type sequence swept along the least (266 out of 357). The use of GAP data by Kiester et al. (in press) is spatially coarse and exploratory. It must be stressed that this trial was only performed on data constrained by state boundaries which are arbitrary in biological or ecological terms. Direct conservation actions, such as specific changes in land management via any number of options are not possible from this application. However, such work is critical and path-finding. It shows that it is possible to synoptically select general regions for conservation priority-setting while conforming to the three principles of complementarity, irreplaceability, and flexibility described by Pressey et al. (1993) and thus, identify important areas that can be evaluated for the next steps of conservation action (see also Csuti et al. in press). It represents yet another way in which the GAP data can be used, in this case, for basic research. |
||||||||||||||||||||||||||||||||