干旱区生态环境知识资源中心

Appropriate rangeland management requires rangeland function analysis at broad scales. This study aimed to examine the potential of remotely sensed function indices extracted from Landsat data to evaluate the function of semi-arid rangelands in central Iran at the sub-basin scale. Three replicate 30-m transects were randomly placed in the dominant slope direction of 14 selected sub-basins. Various structural properties of vegetation (e.g. number and size of vegetation patches and interpatch lengths) and soil surface were scored based on the landscape function analysis (LFA) procedure. The obtained structural and function indices of the LFA, as well as field percent vegetation cover, were compared with the perpendicular distance vegetation index and remotely sensed function indices including proximity, lacunarity, leakiness index, and weighted mean patch size (WMPS). Remotely sensed function indices were found to be capable of discriminating rangeland landscapes with different conditions. Results showed that the structural properties of vegetation considered in the LFA could also be obtained through WMPS and proximity indices (R>0.76; P<0.01). All indices, except for lacunarity, had significant correlations with percent vegetation cover and the strongest correlation was observed between WMPS and proximity. Our findings highlight the usefulness and efficiency of function indices derived from satellite data in the estimation of structural and functional properties of rangeland landscapes at the sub-basin scale.