Modeling grassland net primary productivity and water-use efficiency along an elevational gradient of the Northern Tianshan Mountains

Mountainous ecosystems are considered highly sensitive and vulnerable to natural disasters and climatic changes. Therefore, quantifying the effects of elevation on grassland productivity to understand ecosystem-climate interactions is vital for mountainous ecosystems. Water-use efficiency (WUE) provides a useful index for understanding the metabolism of terrestrial ecosystems as well as for evaluating the degradation of grasslands. This paper explored net primary productivity (NPP) and WUE in grasslands along an elevational gradient ranging from 400 to 3,400 m asl in the northern Tianshan Mountains-southern Junggar Basin (TMJB), Xinjiang of China, using the Biome-BGC model. The results showed that: 1) the NPP increased by 0.05 g C/(m(2)center dot a) with every increase of 1-m elevation, reached the maximum at the mid-high elevation (1,600 m asl), and then decreased by 0.06 g C/(m(2)center dot a) per 1-m increase in elevation; 2) the grassland NPP was positively correlated with temperature in alpine meadow (AM, 2,700-3,500 m asl), mid-mountain forest meadow (MMFM, 1,650-2,700 m asl) and low-mountain dry grassland (LMDG, 650-1,650 m asl), while positive correlations were found between NPP and annual precipitation in plain desert grassland (PDG, lower than 650 m asl); 3) an increase (from 0.08 to 1.09 g C/(m(2)center dot a)) in mean NPP for the grassland in TMJB under a real climate change scenario was observed from 1959 to 2009; and 4) remarkable differences in WUE were found among different elevations. In general, WUE increased with decreasing elevation, because water availability is lower at lower elevations; however, at elevations lower than 540 m asl, we did observe a decreasing trend of WUE with decreasing elevation, which may be due to the sharp changes in canopy cover over this gradient. Our research suggests that the NPP simulated by Biome-BGC is consistent with field data, and the modeling provides an opportunity to further evaluate interactions between environmental factors and ecosystem productivity.