A LANDSCAPE MODEL FOR SERDP

A landscape perspective views the spatial aspects of a region in which a site occurs including the ecological, socioeconomic, and political pressure for land use changes and their impacts. Land-use models typically project land use changes using maps and consider spatial implications that may arise such as edge effects or habitat fragmentation implications for species or ecological systems.

A landscape model can be used to assess the impact of land use activities on natural and cultural resources. It can project the loss/alteration of habitat and the resulting impact on biodiversity. Land-use activities can be characterized using a common set of parameters (magnitude, frequency, areal extent, spatial distribution, predictability) that can be applied both to alternative activities and different levels of the same activity. This approach permits the incremental and cumulative effects of diverse activities, such as road building, military maneuvers, grazing, timber harvests, and environmental restoration, to be evaluated. Evaluating the risk posed to habitats and species can be expressed as the probability of a decline or enhancement in the abundance of guilds of species features. Although such an approach is generic, with appropriate databases it can be applied to any site.

1. Quantitatively characterize land-use activities. A matrix of characteristics to describe land-use activities in terms of magnitude, frequency, areal extent, spatial distribution, predictability, and effects on habitat quality can be developed. For example, some types of troop training are low-intensity impacts that are dispersed throughout a site, whereas construction of an industrial facility is a high-intensity activity that occupies a limited area.

2. Develop a land-cover change risk model. A spatially-explicit, land-cover change model can be developed to simulate potential changes in or loss of individual cover types in response to the land-use activities. Inputs to the model include the matrix of parameters describing land-use activities, and gridded (digital raster) maps of site characteristics, such as present land cover, slope, aspect, soils, etc. The model simulates the impact of land use activities on land cover. A probabilistic model allows representation of stochastic aspects of the land-use activity (e.g., the frequency of training maneuvers) or its effect on the habitat (e.g., the degree to which a forest is damaged by artillery fire) can be represented. Land-use activities that are relatively deterministic and depend on the suitability of the land (e.g., location of new runways) also are easily accommodated within a probabilistic model by setting the appropriate probabilities to 1.0 and fixing specified parameters. From the model, tables and maps of potential land-cover change due to land-use activities can be produced for a particular site. The land-cover change projections can be developed for different scenarios of land-use activities and land cover patterns. Stochastic simulations with the model can be replicated many times and the results summarized statistically, thereby providing an estimate of the magnitude and range of potential effects.

3. Develop a natural resource-susceptibility model. The second model can be developed to relate characteristics of species and ecosystems to land-cover patterns resulting from land-use activities, as projected by the land-cover change model. This natural resource model matches land cover characteristics (e.g., frequency of land cover types, abundance of suitable habitat, size of habitat patches, frequency of edges, corridors, etc.) to species and ecosystems characteristics (e.g., home range size, vegetation patterns). For example, activities that cause habitat fragmentation can be detrimental to species that require large blocks of contiguous habitat (e.g. forest-interior species). This model is probabilistic to ensure its compatibility with quantitative risk assessment. Potential effects on species and ecosystems of no management, alternative land-use activities, environmental restoration, or natural events can be examined. The probability of an undesired outcome (such as loss of a population of interest) is estimated by Monte Carlo simulations of the models under particular scenarios and examination of the frequency distribution of outputs. The visualizations that will accompany this spatially-explicit model will also permit managers to "see" the effects of alternative activities on populations of interest.

In addition to its use for management of natural resources, the landscape approach is directly applicable to (1) planning for facility closures and realignment (e.g., identification of facility closures that provide the best conservation opportunities); (2) developing environmental restoration and waste management strategies; (3) supporting compliance with the Endangered Species Act, the National Historic Preservation Act, the National Environmental Policy Act, and the Executive Orders for Floodplains and Wetlands; and (4) developing integrated risk assessments that address cumulative effects.

Return to the Model Categories page

Return to the Web Site Map

Return to Ecological Modeling home page