Rain-use efficiency and NDVI-based assessment of karst ecosystem degradation or recovery: a case study in Guangxi, China

Interannual variation in productivity may be caused by ecosystem structure degradation or recovery or by interannual precipitation fluctuation. A new method for ecosystem assessment was proposed based on the combination of rain-use efficiency (RUE) and normalized difference vegetation index (NDVI). This study aimed to alleviate the effect of precipitation fluctuation on ecosystem productivity. The trend of annual precipitation, annual maximum NDVI, and annual RUE of the study area was first analyzed. The relationship between RUE and precipitation at both temporal and spatial scales was then clarified. Finally, the relationship between the trends of RUE and NDVI was recognized. According to this trend, ecosystem degradation or recovery determined, and the mask of precipitation fluctuation were removed. Results showed that annual RUE varied significantly in space from 1999 to 2008. Annual precipitation displayed a decreasing trend, whereas the annual maximum NDVI presented an increasing trend in most karst ecosystems. RUE presented a remarkable declining trend with annual precipitation at both the temporal and spatial scales (P < 0.001). At a significant level of 0.2, most karst ecosystems at 16 weather stations were recovering from 1999 to 2008. Ecosystems of Baise, Hechi, and Mengshan stations were in pseudo recovery, whereas those of Laibin and Duan stations were in pseudo degradation because of precipitation fluctuation. At a significant level of 0.05, most karst ecosystems showed non-significant change except four stations were in pseudo recovery. The method of combining RUE and NDVI for ecosystem assessment may remove the mask of precipitation fluctuations and thus improve single NDVI-based method for vegetation analysis.