Hillslope runoff and erosion on duplex soils in grazing lands in semi-arid central Queensland. II. Simple models for suspended and bedload sediment

Abstract

The use of simple models of soil erosion which represent the main effects of management in grazing lands in northern Australia is limited by a lack of measured parameter values. In particular, parameters are needed for erosion models (sediment concentration v. cover equations) used in daily soil-water balance models. For this research, we specifically avoided equations that use rainfall and runoff rates (e.g. peak flow), as current daily models are limited in their ability to estimate these rates. The resulting models will therefore give poor estimates of soil losses for individual events, but should give good estimates of long-term average erosion and management influences. Runoff and erosion data were available for 7 years on 12 hillslope plots with cover of 10–80%, with and without grazing, with and without tree canopy cover, on a variety of soils according to various soil classification systems. Soils were grouped into those derived from sandstone (SS), mudstone (MS), and eroded mudstone (MSe). These data were used to determine two parameters, i.e. (i) efficiency of entrainment for bare soil and (ii) a cover factor, for simple models of bedload and suspended sediment concentrations. Methods used to fit parameters affected the results; optimising to obtain the minimum sum of squares of errors in soil losses gave better results than fitting an exponential equation to sediment concentration–cover data. The use of a linear slope factor in the sediment concentration models was confirmed with data from plots with slopes 4–8%. Parameters for the bedload sediment concentration model were the same for SS, MS, and MSe soils. Parameters for the suspended sediment concentration model were the same for SS and MS soils, but the MSe soil had a greater efficiency of entrainment for bare soil (about double). The sediment concentration–cover relationships and fitted cover factors were different for suspended and bedload sediment. Thus, the resulting modelled proportion of sediment as suspended load changed with cover, from ~0.3 for bare soil to 0.9 at 80% cover, mimicking the measured data. The cover factor was lower than published values for cultivated soils, indicating less reduction in sediment concentration with greater cover. A compilation of parameter values for the sediment concentration model from published and unpublished sources in grazing and cropping lands is provided.