Investigating Ecohydrology and Connectivity Thresholds along a Precipitation Gradient in Australia

Monday, 14 December 2015: 17:00
3020 (Moscone West)
Patricia M Saco1, Mariano Moreno-de las Heras2, Samira Azadi1 and Saskia Keesstra3, (1)University of Newcastle, Callaghan, Australia, (2)Institute of Environmental Assessment & Water Research (IDAEA-CSIC), Barcelona, Spain, (3)Wageningen University, Wageningen, Netherlands
Drylands are very sensitive to climatic or anthropogenic pressures and prone to critical degradation thresholds which make rehabilitation efforts considerably difficult. In these systems, the spatial structure of vegetation and the spatial redistribution of overland flow are tightly linked through feedbacks mechanisms. Disturbances, induced by human activities or climate change, can disrupt the spatial distribution of vegetation triggering erosion and substantial water losses by increasing landscape hydrological connectivity and damaging ecosystem function. Characterizing the integrity of these coevolving patterns is especially relevant for the detection of landscape degradation processes and possible threshold behaviour. Here we investigate the impact of degradation processes, induced by grazing pressure, across the Mulga Lands Bioregion in Australia. We combine remote sensing observations and a modelling approach to analyse changes in ecosystem connectivity and the existence of threshold behaviour along a precipitation gradient (250mm to 450mm annual average rainfall). Vegetation patterns are derived from high resolution remote sensing images, and Rainfall Use Efficiency (RUE) estimated from precipitation data and MODIS vegetation indices for a total of 40 plots along this gradient. A critical degradation threshold, associated to loss of vegetation cover appears for sites with lower mean annual precipitation. In these sites, hillslopes with high vegetation cover display low hydrologic connectivity and high RUE, and can be therefore classified as “functional”. However, below a threshold vegetation cover we found highly “dysfunctional hillslopes” characterized by high connectivity and very low RUE. We found that this threshold behaviour tends to disappear for the north-western sites with higher mean annual rainfall. We further analytically investigate the dynamic response of these hillslopes using a coupled landform evolution-vegetation model to simulate coevolving vegetation patterns and geomorphology. We use the model to analyse changes in hydrologic and sediment connectivity responsible for triggering this threshold behaviour for selected plots along this precipitation gradient. The implications for ecosystem resilience and land management strategies are discussed.