Using sediment tracing to assess processes and spatial patterns of erosion in grazed rangelands, Burdekin River basin, Australia

Abstract

Identifying how agricultural practices can be changed to reduce sediment loss requires knowledge of the erosion processes and spatial areas contributing to end of catchment sediment loads. The Burdekin River basin in northeast Australia is a priority for such knowledge because of its large size (130,000km2), ongoing public investment in changing agricultural practices, and because sediment exports are known to affect the health of a significant aquatic ecosystem, the Great Barrier Reef (GBR). This study applied sediment tracing techniques within the Burdekin River basin to identify the contributions of surface versus subsurface soil, and spatial areas to fine sediment export. Tracer properties included fallout radionuclides and geochemistry. The contributions of each sediment source to river sediment were identified with 95% confidence intervals using a Monte-Carlo numerical mixing model. Between 77% and 89% of fine sediment loss in the study area was derived from subsurface soil sources. High-resolution monitoring of river suspended sediment concentrations indicated that sediment sources were in close proximity to the drainage network, since concentrations were higher on the rising limb than the falling limb of large hydrographs. Gully erosion is likely to be the dominant subsurface soil erosion process, although channel bank erosion and hillslope rilling cannot be discounted. The results contrast with previous sediment budget spatial modelling, which predicted that hillslope erosion was the dominant sediment source in the area, thus demonstrating the need to independently verify modelling predictions where input datasets are poor. The contribution of surface soil to river sediment was generally similar between catchments which were currently grazed and two catchments where livestock grazing ceased 7 years ago. Concurrent increases in vegetation cover in the non-grazed catchments indicate that surface erosion rates had declined, suggesting that subsurface soil erosion rates had also declined by a similar amount. The estimated contributions of spatial source areas within the large study catchments had narrower confidence intervals when source areas were defined using sediment from geologically distinct river tributaries, rather than using soil sampled from geological units in the catchment, since tributary sediment had less-variable geochemistry than catchment soil. Programs to reduce fine sediment losses from the Burdekin River basin should primarily focus on reducing sub-surface soil erosion proximal to the basin's drainage network. Understanding the biophysical processes of pollutant generation is important to help guide on-ground activities to improve water quality.