Land Degradation Assessment in Central Australia Using a Grazing Gradient Method.

Abstract

: how well are they structured? J. Am. Soc. Inform. Sci. 42(7), 528-31. Stmnk, W. and White, E.B. (1972). 'The elements of style'. Macmillan, New York, and Collier Macmillan. London. Manuscript received 1 March 1993, accepted 22 March 1993 Rangel. J. 15(2) 1993,190-216 LAND DEGRADATION ASSESSMENT IN CENTRAL AUSTRALIA USING A GRAZING GRADIENT METHOD G.N. Bastin, G. Pickup, V.H. Chewings and G. Pearce CSIRO Division of Wildlife and Ecology, Centre for Arid Zone Research, PO Box 21 11, Alice Springs, NT 0871 Abstract Rangeland monitoring using ground-based methods has been fraught with difficulty because of the lack of accurate and repeatable techniques capable of separating grazing impact from both seasonal variability and natural landscape heterogeneity. The results presented here show how these problems can be overcome by analysing the entire grazed landscape using remotely-sensed data and 'grazing gradient' methods which separate grazing effects from natural variation. Land degradation is defined as 'a reduction in the capacity of landscapes to produce vegetation cover from rainfall'. After a period of grazing, vegetation cover typically decreases as water is approached producing a spatial pattern known as a grazing gradient. Vegetation increases across the whole landscape following rain and the extent of vegetation recovery in the vicinity of watering points . after significant rainfall is determined for each mapped land system. A 'percentage cover production loss' index is then derived allowing grazing management effects across land systems (and paddocks) to be compared. We report on the analysis of 38,000 krn2 of grazed country on all or parts of 16 pastoral leases. Land systems containing a high proportion of palatable forage have high index values and are most adversely affected by grazing. These land systems are frequently the most intensively stocked due to their original productivity. In some cases, parts of these land systems have also had a relatively long history of grazing. Some land systems of lower pastoral value show no gradient after rainfall but a pattern of decreasing cover with distance from water after a period of grazing. This is associated with an increase in cover of unpalatable shrubs. We advocate the use of grazing gradient methods which utilise remotely-sensed satellite data as an operational means of determining grazing impact in the arid rangelands. Our methods compare favourably with conventional ground-based monitoring in terms of cost and labour requirements. The remote sensing-based method is capable of providing comprehensive, objective and repeatable analysis of grazed country after future significant rainfalls. It also offers the significant additional advantage of immediate information about landscape condition through the analysis of archived remotely-sensed data.