Improved assessment of pasture availability in semi-arid grassland of South Africa

Satellite remote sensing technology has been successfully used to monitor grassland productivity, especially for estimating the green component of biomass using popular indices such as Normalized Difference Vegetation Index (NDVI). The non-green component, which includes senescent and dead standing material, has not been widely quantified. Our study aimed at devising a satellite remote sensing-based method that can distinguish between green and non-green herbage, in order to improve the accuracy of total aboveground biomass (TBM) estimations. This in turn can minimise under-estimations of pasture availability in semi-arid grasslands. MODIS satellite data was used to determine the relations of various indices to ground-measured green aboveground biomass (GBM) and non-green aboveground biomass (NBM) in South African semi-arid grasslands. We found a strong correlation of GBM to NDVI. We were then able to detect a correlation of NBM to Normalized Difference Water Index (NDWI), but a robust relationship was between NDWI and the ratio of NBM to TBM. NDVI and NDWI were used to estimate long-term TBM, which varies inter- and intra-seasonally. During the non-rainy season, NBM is important to maintain livestock grazing and in this regard monitoring of pasture availability in terms of green and non-green herbage is critical for sustainable grassland management.