Land Degradation in the Limpopo Province, South Africa

I declare that this research report is my own unaided work. It is being submitted in partial fulfilment for the degree of Master of Science at the University of the Witwatersrand. It has not been submitted before for any degree or examination at any other university. Figure 1 Map showing the location of the study area. White areas represent privately owned land used for Figure 3: Long-term (1985-2004) average maximum composite NDVI image (NDVI avemax) used in the degradation analysis. (A) Swartruggens Mountain Bushveld, (B) Indigenous and plantation forestry, (C) Commercial cultivation and riparian vegetation along water courses, (D) Areas of low NDVI values, (E) Swaziland Sour Bushveld vegetation type on dark coloured basalt-derived soils with low NDVI values, (F) Mosaic of subsistence cultivation and communal rangeland, (G) Mining area Figure 4: (a) Second order polynomial regression to test the significance of the curvi-linear trend in yearly cumulative NDVI values; 'x' denotes the two anomalous data points excluded from the analysis, and (b) trend in yearly mean NDVI and mean rainfall values for all pixels in Limpopo Province between Figure 5: Non-linear regressions using a fitted Gompertz logistic function between long-term average maximum NDVI and (a) mean annual precipitation, and (b) growth days index. The fitted functions Figure 6: Map of studentized residuals depicting observed deviation from the predicted Gompertz rain-use efficiency model established in the regression of average maximum NDVI and mean annual precipitation. Negative standard deviations (red) depict areas where rain-use efficiency is lower than it should be, while positive standard deviations (green) depict areas that have higher rain-use Figure 7: Maps of degradation based on thresholds of (a) 0.75 standard deviations, (b) one standard deviation, and (c) two standard deviations, less than the predicted model.. Figure 8: Independent validation datasets: (a) the 'degraded' land-cover class of the National Land-cover 2000 project, mapped from Landsat satellite imagery taken in 2001/2002, and (b) the degradation indices from the 1999 National Review of Land Degradation constructed per municpal district through perception-based surveys with field technicians and extension officers.. Table 1: Synthesis of remote sensing research on desertification with summary characteristics of study scale, Table 2: Land-cover characteristics of mapped degraded areas. A threshold of one standard deviation was used to define the degraded area, and the land-cover classes were aggregated from those used in ABSTRACT An estimated 91 % of South Africa's total land area is considered dryland and susceptible …