The effect of vegetation dynamics on erosion processes, sediment dynamics, and landscape evolution in semiarid areas with sparse plant cover

Vegetation not only controls but is also controlled by erosion processes. This tight feedback effect leads to the coevolution of vegetation and erosion patterns that modulate landform shape, and regulate many other landscape processes. These tight interactions are particularly important in semiarid landscapes. We have studied these interactions using a landform evolution model that accounts for the effect (and feedbacks) of spatially and temporally varying hydrologic and vegetation patterns.

We apply the modelling framework to improve our understanding of the coevolution of landforms and vegetation patterns in different semiarid landscapes in Australia. The vegetation of the selected sites is Acacia Aneura (Mulga) which covers vast areas of Australia.  These sites display a sparse vegetation cover and strong patterns of water redistribution, with sources located in the bare areas and sinks in the vegetation patches which characterize the observed hydrologic connectivity. This effect triggers high spatial variability of erosion/deposition rates that affects the evolving topography and induces feedbacks to the dynamic vegetation patterns. We run simulations for 1000 years using local rainfall and erosion and vegetation parameters previously calibrated for similar sites in the Northern territory. Our numerical modelling results are validated by comparing simulated and observed patterns of vegetation and landforms obtained from satellite, airborne remote sensing and field data. We further investigate the effect of alterations in hydrologic connectivity induced by climate change and/or anthropogenic activities, which affect water and sediment redistribution and can be linked to loss of resources leading to degradation.

Our simulations are able to reproduce observed banded vegetation and landform patterns for the Northern territory in Australia. We show that an increase in hydrologic connectivity can trigger changes in vegetation patterns inducing feedbacks with landforms leading to degraded states. These transitions display non-linear behaviour and in some cases can lead to thresholds with an abrupt reduction in productivity. Critical implications for effective long-term restoration efforts are discussed.