Abstract: With RS and GIS, spatial principal component analysis was used to construct the eco-environmental quality index of Gansu, a region with relatively fragile eco-environment, by selecting 11 indicators including natural and anthropogenic factors. The spatiotemporal distribution and influencing factors of eco-environmental quality were analyzed. The results showed that eco-environmental quality in 2000 and 2017 showed a polarization phenomenon in term of spatial distribution. The areas with poor and fair grades of eco-environmental quality were mainly concentrated in northwest Gansu, while the areas with good and excellent grades of eco-environmental quality were mainly concentrated in southeast Gansu. Temporally, there was a trend of improvement for eco-environmental quality in the past 17 years. Generally, climate, vegetation cover, land use type, and soil matrix were dominant factors affecting eco-environmental quality, while the role of population economy, landform and hydrological environment were relatively minor.

Key words: precipitation exclusion, specific root respiration, nonstructural carbohydrate., fine root