干旱区生态环境知识资源中心

In African savanna, spatio-temporal variability in moisture availability, fire regime and land transformation related to exploitative land uses are the main drivers of changing vegetation greenness patterns. Deconstructing the role of these drivers at local scale is critical for managing the impact of projected climate and land use changes on savanna ecosystems. Focusing on an area encompassing Africa’s largest terrestrial protected area, this study utilized time-series MODIS NDVI (2000-2014) and employed a robust trend analysis technique to detect significant temporal trend in key greenness metrics including overall greenness, peak and timing of the peak of annual greenness and examined how spatial variability in vegetation morphology and land use impacted the distribution of these greenness parameters. To access causation of change, we linked detected greenness trends to precipitation trends derived from Tropical Rainfall Measurement Mission (TRMM), annual burnt area extent and fire frequency and compared multi-temporal imagery in Google Earth. Results show that distribution of both overall and peak annual greenness was vegetation morphology specific as with decreasing woody cover the proportion of pixels showing negative overall and peak annual greenness trend increased consistently. Majority of the area with significant trend in the timing of the peak of annual greenness showed late greenup, implying changing phenological patterns. Trend analysis of TRMM derived mean annual precipitation showed no significant change in precipitation over the study period. At local scale, increase in overall and peak annual greenness was associated with woody plant encroachment due to increased moisture availability and reduced fire frequency due to construction of fire breaks in areas under farm/ranch land use. In protected areas, negative overall and peak annual greenness was driven by unmanaged high intensity mega fire events and outside due to anthropogenic land clearing for pastoral agriculture. These results further our understanding of the complex interaction of multiple ecological factors and resultant spatial variability in savanna inter-annual vegetation greenness patterns and are useful for implementing local scale management strategies. (C) 2015 Elsevier Inc. All rights reserved.