Catchment-wide soil loss from pre-agricultural times to the present: transport- and supply-limitation of erosion

Abstract

A high-resolution record of catchment-wide soil loss for the period c. 1806–1990 has been obtained from Little Llangothlin Lagoon on the New England Tablelands of northeast New South Wales, Australia. The mean annual rate of mineral erosion since the time of European contact in the late 1830s was 269 t km −2. The mean rate of mineral denudation immediately prior to this was 25 t km −2a −1. In the 25 years after the arrival of the first sheep in the catchment, erosion rates increased by a factor of over 50 to 1360 t km −2a −1. After c. 1861, however, there was an apparently sharp transition to a new, low and very constant rate of denudation, 52 t km −2a −1. Eighty-five percent of post-contact erosion thus occurred in the first quarter of a century of European land use. The low and constant erosion rates of the last century or more cannot be attributed to stable environmental conditions, to a decrease in land use intensity or to the introduction of soil conservation measures. Instead, it is possible that early colonial erosion almost entirely depleted the catchment of erodible material with the result that erosion moved from a transport-controlled regime to one that was limited by the rate at which catchment material was made available for transport by weathering. Alternatively, the high, early colonial rates of erosion may have been associated with the extension and deepening of the drainage net during the initial phase of European contact. The subsequent establishment of a new drainage net equilibrium may have reduced soil loss to a low and stable level. Much of the evidence available to test these competing hypotheses is equivocal. Nevertheless, the gullying model must be rejected, first because there is no evidence of past or present dissection of the catchment surface, second because gullying would seem incapable of providing the highly constant rate of sedimentation that has prevailed in the basin over the past century or more and third because the gullying model cannot explain the step change from high to low rates of sedimentation in the basin. Further support for the supply-limitation hypothesis comes from the concordance between likely rates of soil formation in the catchment and rates of sedimentation in the lagoon. These conclusions have implications for our cognisance of the role of supply-limitation in geomorphological processes, for soil conservation practice and for our understanding of the long-term impacts of agriculture on soil erosional systems.