干旱区生态环境知识资源中心

The principal drivers that effectively determine the vitality of semi-arid grassland ecosystems are the amount and timing of precipitation. In contrast to well-established relationships between precipitation and some vegetation biophysical properties such as vegetation productivity and biomass, the effect of precipitation on percent vegetation cover in semi-arid grasslands are poorly understood. The percent green vegetation cover is an important indicator of ecosystem health, especially in a protected semi-arid grassland where dead material dominates vegetation canopy. Consequently, this study used field data, high spatial resolution SPOT satellite images, and low spatial but high temporal resolution AVHRR images to (1) analyze temporal variations in maximum percent green vegetation cover over three landscape units in a protected grassland from 1988 to 2007, and (2) investigate the performance of three commonly used climatic moisture indexes (accumulated precipitation P; climate moisture index - CMI; and the modified Palmer drought severity index - MPDSI) in estimating percent vegetation cover. The results revealed moderate to strong links between percent vegetation cover and climate conditions in three landscape units. The selected indices were able to explain a relatively high proportion of variation in percent vegetation cover in the upland/sloped grassland (R-2 = 0.41 for P, 0.45 for CMI, and 0.53 for MPDSI), low variation in the riparian shrub (R-2 = 0.25 for P, 0.28 for CMI, and 0.32 for MPDSI), and even lower variation in the valley grassland (R-2 = 0.09 for P, 0.14 for CMI, and 0.21 for MPDI). In comparison with accumulated precipitation, the CMI and MPDSI which considered water availability, were able to explain a much higher variation in percent vegetation cover over the three grassland landscape units. The results from this work provide new insight into the effects of precipitation/climatic factors upon different landscape units and further aid to establish a more complete picture in this field. (C) 2014 Elsevier Ltd. All rights reserved.