Abstract
One of the many contentious issues facing the appropriate and accurate assessment of land degradation is the varying emphasis placed on vegetation degradation and soil degradation processes. This has led to the compartmentalization of land degradation assessment methods, depending on the particular perspective adopted. The land degradation assessment method presented here attempts to take into account both vegetation and soil degradation. This methodology is applied to the southern part of the Monduli District of northeast Tanzania, an area typifying the soโcalled โaffected drylandsโ of Africa. Three sets of land cover maps synchronized against longโterm rainfall data (1960s, 1991 and 1999) were used to assess land degradation in the area. Utilizing these three sets of land cover maps as a basis for change detection, it is possible to distinguish areas that experienced changes in vegetation due to rainfall variability from those areas that were subject to changes consequent upon anthropological factors. All areas that displayed overall depletion of natural and semiโnatural vegetation due to human factors were deemed to have undergone land degradation, whereas areas that experienced interโannual land cover changes due to rainfall variability were classified as experiencing cover change due to ecosystem dynamics. This method provides a complete and appropriate assessment of land degradation in the study area and can be used to improve degradation assessment in other semiarid areas. Copyright ยฉ 2006 John Wiley & Sons, Ltd.