The causes and spatial pattern of land degradation risk in southern Mauritania using multitemporal AVHRR-NDVI imagery and field data

Multitemporal 1 km NOAA/AVHRR Normalized Difference Vegetation Index (NDVI) maximum composite imagery was utilized in combination with rainfall, soil types, and field survey data on dominant rural activities to assess the risk of land degradation in southern Mauritania. Mauritania is one of eight continental West African Sahel countries that stretch from Chad to the northwestern Atlantic coast, and from the southern fringe of the Sahara Desert to the northern limit of the Sudanian climatic zone. The major environmental problem these countries are currently facing is the temporally and spatially erratic character of rainfall, frequently leading to general or local droughts since the late-1960s with subsequent increase of human pressure on the natural resource base and degradation. Image deviation was applied to maximum NDVI composites of the growing season (June to October) for the period 1990–99 to detect temporal and spatial change patterns over the study area. Only pixels whose values were lower than their temporal mean minus 0ċ5 standard deviation were retained as areas under threat. These were combined with soil types and the spatial pattern of deforestation (fuelwood and building material collection areas) grazing, and agricultural land to determine the causes of lowering primary biological productivity. The results show that the below-normal NDVI values generally coincide with the patterns of below-normal rainfall: deforestation, and overgrazing areas, agricultural land, and low primary biological productivity soil types. Thus the observed substantial decrease in biomass production results from the combined impacts of frequent rainfall deficits and uncontrolled resource-base exploitation by the local population. Large chunks of land that show consistent low biomass production are also observed on unproductive soils. Finally, the below-normal rainfall, below-average NDVI values, and human impacts images were cross-classified to produce a three-class (low, moderate, and high) land degradation risk map along with a table containing the area covered by each class. A Boolean mask of the degradation risk map applied to the soils map showed that all soil types in southern Mauritania are at risk of degradation. These products could serve as a strong basis for decision making in regard to planning resource allocation for environmental rehabilitation.