Simulating the effects of native vegetation configurations on water quality using a multi-agent system coupled to a distributed hydrological process model

Within Australia, excessive clearing of native vegetation has resulted in many landscapes becoming increasingly dysfunctional with respect to the retention of water and nutrients, and the maintenance of biodiversity. This problem is being addressed by the National Action Plan for Salinity and Water Quality, which promotes the management of vegetation cover at the farm and catchment scale to: stabilize soils; regulate groundwater to control dryland salinity; minimize chemical residues, nutrients, and sediment run-off to streams and waterways; as well as the maintenance of environmental flows for healthy waterways. However, the performance of potential landscape designs for the retention of water, nutrients and sediments, such as the reestablishment of tree belts and riparian vegetation, must be socially and economically feasible as well as improving landscape function. The implementation of alternative design strategies occurs at the farm-scale, but also must be applicable at the hill-slope and catchment scales, as well as incorporating temporal variability, and be practical to implement where data is limiting. A method is presented based upon a coupled multi-agent system (MAS) simulation and distributed parameter hydrological process model to optimize the integration of native vegetation within agroecosystems, in order to maintain desired outcomes for sustainable landscape function.