Content area
Abstract
To manage dryland salinity, one needs to know how changed land use affects groundwater recharge. Few techniques are available for comparing 'deep drainage' under different land uses. Soil-tracer methods, although good for replication and remote field sites, are subject to spatial variability. Lysimeters are good for comparisons but are difficult for drier areas and sloping land. Agronomic water-balance studies, where appropriate soil-water measurements exist, may be used with a soil-vegetation model to estimate long-term deep drainage. Complex models are required to analyze specific land-use differences, such as perenniality and root and leaf area dynamics, but models require intensive and extensive data for calibration. This approach is time-consuming, labour-intensive, and difficult in remote locations. Because of the one-dimensionality of most soil-vegetation models and the small fraction of the total water balance that is deep drainage, little success has occurred in extrapolating beyond the research plot, or to spatially heterogeneous systems such as alley farming. Some 'top-down' modelling and landscape disaggregation approaches have been partially successful in making catchment or regional-scale predictions. The direction for further work depends on the level of recharge reduction that is required for most groundwater systems and difficulties that it imposes.[PUBLICATION ABSTRACT]