Ecosystem Health and Its Measurement at Landscape Scale: Toward the Next Generation of Quantitative Assessments

The purpose of this paper is twofold: (A) to describe the challenges of reporting on changes in ecosystem health at landscape scales, and (B) to review the statistical and mathematical techniques that allow the derivation of landscape health assessments from a variety of data consisting of remote sensing imagery, demographic and socioeconomic censuses, natural resource surveys, longterm ecological research, and other geospatial information that is site specific. We draw upon seven innovative and integrative concepts and tools that together will provide the next generation of ecosystem health assessments at regional scales. The first is the concept of ecosystem health, which integrates across the social, natural, physical, and health sciences to provide the basis for comprehensive assessments of regional environments. The second consists of innovative stochastic techniques for representing human disturbance and ecosystem response in landscapes, and the corresponding statistical tools for analyzing them. The third constitutes representation of spatial biocomplexity in landscapes through application of echelon analysis to assessment. The fourth concerns innovative combination techniques of upper-echelon-based spatial scan statistic to detect, delineate, and prioritize critical study areas for evaluating and prioritizing causal factors and effects. The fifth involves the capability of comparing and prioritizing a collection of entities in light of multiple criteria, using poset mathematics of partial order with rank frequency statistics, to provide multicriterion decision support. The sixth lies in extending data mining and visualization techniques to determine associations between geospatial patterns and ecosystem degradation at landscape scales. The seventh encompasses comprehensive studies conducted on different types of regional ecosystems. Our focus is to show how the integration of recent advances in quantitative techniques and tools will facilitate the evaluation of ecosystem health and its measurement at a variety of landscape scales. The challenge is to characterize, evaluate, and validate linkages between socioeconomic drivers, biogeochemical indicators, multiscale landscape pattern metrics, and quality of human life indicators. Initial applications of these quantitative techniques and tools have been with respect to regions in the eastern United States, including the U.S. Atlantic Slope and mid-Atlantic region. Address correspondence to: G. P. Patil, 421 Thomas Building, University Park, PA 16802 USA; E-mail [email protected].