Developing a Watershed Database for Risk Assessment and Analysis of Cumulative Impact of Land Use Change to Aquatic Habitat. D. F. Bleil, J. G. Ahl, and F. N. Schenerman, Maryland Department of Natural Resources, Tidewater Administration, C-2 Tawes State Office Building, Annapolis MD 21401
Population growth in Maryland is placing great stress on the state's natural resources, particularly on watersheds discharging into the Chesapeake Bay. State and local governments in Maryland are realizing that population growth and associated land use changes must be managed more effectively than in the past. With the advent of new technology such as Geographical Information Systems (GIS) and statistical packages using digital geographical files, Maryland's DNR is building a statewide watershed database which 6an be used to analyze cumulative impacts of development on existing wildlife habitat especially aquatic habitat. DNR can also use this database to undertake quantitative assessment of risk to aquatic ecosystems at the sub-watershed level. The first phase involves bringing together information in digital form from multiple sources, comprising three major groups: (1) socioeconomic, such as the 1990 Census data; (2) institutional variables, such as Land Use and Land Use Change data and; (3) biological, such as water quality, habitat assessment scores and fish data. Phase two will use the database to analyze cumulative impact to watersheds and to undertake ecological risk assessment. Characterization of risk to the watersheds will be based on a variety of quantitative techniques including multivariate clustering algorithms (to determine the degree of aquatic habitat similarities) and Bayesian techniques to select appropriate probability distributions. Correlations between land use and development patterns and aquatic habitat scores will enable us to build probabilistic relationships between development patterns and aquatic habitat damage. Then when a particular development is proposed, the probable damages to aquatic habitat can be forecast with due allowance being given to weather induced fluctuations in stream flow which are independent of land use impacts. The mechanisms for damage to aquatic habitat from development are mostly already known. What has yet to be learned is the relative magnitude of land use change which produces a qualitative or quantitatively identifiable damage to aquatic habitat. By combining several new technologies, we expect to characterize risk to aquatic habitat through knowledge of changes in land use. The outcome of the analysis of this database will focus on evaluating the significant variables associated with change rather than isolating specific impacts of specific causative agents. The relative sensitivity of an unstudied watershed will be estimated by its degree of multivariate similarity to other watersheds which have been studied. Multivariate clustering techniques can rank watersheds and document their sensitivities to disturbance and damage. Programs to minimize development pressures on aquatic habitats can become more focused. This will permit more rational response to requests for new development while providing protection for Maryland's natural resources.