Above and belowground net primary productivity of grassland influenced by supplemental water and nitrogen in Inner Mongolia

Abstract

While water availability determines grass- land productivity in semiarid regions, nutrient avail- ability is the main limiting factor under wet conditions. An experiment was conducted in 2008 at two sites in Inner Mongolia with histories of heavy grazing (HG) and moderate grazing (MG) to study the interactive effects of water and nitrogen on above- and belowground net primary productivity (ANPP and BNPP), biomass partitioning, and plant species composition. The study comprises two water treat- ments (no irrigation and irrigated when soil water content was below 70% of the field capacity), and two nitrogen (N) levels (0 and 100 kg N ha �1 ). Mean values of ANPP at the peak biomass time reached 1,028±95 SD g m �2 at the HG site and 568±32 SD gm �2 at the MG site in irrigated and fertilized treatment. Nitrogen use efficiency (NUE) was signif- icantly higher at irrigated plots compared to rain-fed plots at both HG and MG sites. Water use efficiency (WUEt) based on total water input and ANPP decreased with irrigation at the HG site. Meanwhile, N application significantly increased WUEt, WUEp (based on precipitation), and WUEi (based on irrigation water) at both sites. BNPP was significantly higher at irrigated plots compared to rain-fed plots at both HG and MG sites, and it tended to decrease with N addition. However, the fraction of belowground to total biomass (fBNPP = BNPP/(ANPP+BNPP) de- creased with the addition of supplemental resources and exhibited a negative correlation with ANPP. Species diversity remained lower at the HG site compared to the MG site; it decreased with the addition of supplemental resources at the latter site. The annual Salsola collina contributed the most to the total biomass under irrigation. Based on global climate models, more frequent extreme climates are predicted in the future, which can result in changes in resource availabilities. Therefore, our research results have important implications for predicting the pro- duction and other properties of grassland ecosystems.