Multivariate Analysis of In-stream Nutrient Loads and Salinity for a Large Regional Basin in Australia

The Glenelg-Hopkins area is a large regional watershed in south-west Victoria, Australia (Area : ~30,000 km2). The region delivers many socio-economic benefits with extensive national park systems as well as playing a major role in Australian agriculture. Within the region extensive clearing of native vegetation has led to a decline in water quality including increased solute loads and salinisation. The relationships between patterns in land use and total in-stream phosphorus (TP), total nitrogen (TN) loads and salt concentration (indicated by EC) is investigated. Multi-temporal satellite imagery was interpreted and water quality data analysed from 5 available gauge stations within the Glenelg-Hopkins region. Geographical Information Systems (GIS) were used to analyse spatial variations of land use for corresponding gauging stations in the catchment. Multiple regression analysis for a wide range of catchment characteristics was applied with spatial analysis to predict total stream nutrients and salt concentration. The multiple regression analysis demonstrated that the variables, Dryland pasture, Areas subject to inundation, Agricultural land on greater than 3 percent slope and the Ratio of Agriculture to Native vegetation were most strongly related to TP and TN loads. The regression model for salt concentration shows Native Vegetation, Bluegum Plantations, Dryland Cropping and Irrigated Horticulture were the significant explanatory variables. This study shows strong relationships between in-stream water quality parameters and a selected set of watershed attributes easily determined from satellite images.