Predicting the effects of land disturbances on stream salinity in southwest Western Australia

Abstract

A model developed by A. J. Peck to predict the effects of land disturbances on stream salinity has been extended to take account of increased stream runoff, and applied to the jarrah forest region with improved parameter estimates. Validation on Wights experimental catchment suggests that the model is capable of reliable predictions in the case of agricultural clearing. However, the extended model did not provide discernibly better predictions than the original model. When applied to rainfall zones of the northern jarrah forest of W.A., the model predicted that agricultural clearing would result in average stream salinity increases of -70 mg 1 for the high rainfall zone (> 1100 mm yr-1), of -270 mg L-1 for the intermediate rainfall zone (900-1100 mm yr-1), and of ~3400 mg L-1 for the low rainfall zone (<900 mm yr-1). In the case of bauxite mining followed by reforestation, the model was limited to consideration of long-term effects, and neglected transient effects. Predicted stream salinity increases were considerably smaller than for agricultural clearing, primarily because mining involves clearing smaller areas, and these areas are reforested. The model is sensitive to variations in parameter values which implies that a wide range of stream salinity increases could occur within any one rainfall zone, due to the variation of local conditions.