Regional-scale models for relating land cover to basin surface-water quality using remotely sensed data in a GIS.

ABSTRACT

Plant-based management systems implementing deep-rooted, perennial vegetation have been identified as important in mitigating the spread of secondary dryland salinity due to its capacity to influence water table depth. The Glenelg Hopkins catchment is a highly modified watershed in the southwest region of Victoria, where dryland salinity management has been identified as a priority. Empirical relationships between the proportion of native vegetation and in-stream salinity were examined in the Glenelg Hopkins catchment using a linear regression approach. Whilst investigations of these relationships are not unique, this is the first comprehensive attempt to establish a link between land use and in-stream salinity in the study area. The results indicate that higher percentage land cover with native vegetation was negatively correlated with elevated in-stream salinity. This inverse correlation was consistent across the 3 years examined (1980, 1995, and 2002). Recognising the potential for erroneously inferring causal relationships, the methodology outlined here was both a time and cost-effective tool to inform management strategies at a regional scale, particularly in areas where processes may be operating at scales not easily addressed with on-site studies.