Abstract
Degradation in Australia's arid grazing lands can be a multi-faceted problem involving soil erosion, soil degradation and pasture species change which includes shrub invasion. Each can reduce grazing productivity and all three interact in space, and over time. An added complexity is that some forms of degradation may occur naturally but also are accelerated by poor grazing management. Surveys of degradation frequently rely on "point" survey techniques. Selection of these sample "points" can quite often ignore the linkages between key landscape components, e.g. the areas that shed soil, and those that accumulate soil. A regional or whole-landscape approach which looks at grazing areas in their entirety is therefore essential, if new surveys are not to repeat the errors of flawed predecessors. Grazed landscapes might be assessed in terms of their vegetation cover, the spatial distribution of that cover, and its variability in time as a function of both climate and management. Remote sensing is the only way to conduct such assessments over vast areas if the problems associated with point sampling are to be overcome.