Effects Of Nitrogen Deposition On Carbon Dynamics In The Typical Steppe Of The Loess Plateau

To study the response of grassland ecosystem to increased nitrogen (N) deposition, a field experiment was carried out in a 4-year-exclusion typical steppe of the Loess Plateau. Six levels of N treatments, i.e.0 g N·m-1·a-1,1.15 g N·m-2·a-1,2.3 g N·m-2·a-1,4.6 g N·m-2-a-1,9.2 g N·m-2·a-1 and 13.8 g N·m-2·a-1, were conducted in the experiment. In this study, we examined several key processes associated with carbon dynamics in the steppe ecosystem. The major subjects included vegetation carbon (C) stocks, soil C characteristics, the decay rates and C dynamics in litters and net ecosystem C exchange in the grassland ecosystem. The main conclusions are as follows:(1) The responses of vegetation C storage to the N treatments were not completely consistent in 2010 and 2011 because of different rainfall patterns. In the wet year (2010), the N treatments could significantly increase the total vegetation C storage of grassland within a certain range (4.6-9.2 g N·m-2·a-1 treatments), but the total vegetation C storage decreased with the N treatment of 13.8 g N·m-2·a-1 level. The N treatments significantly increased the aboveground vegetation C storage, and there were hardly obvious changes in belowground vegetation C storage under different N treatments. In the drought year (2011), there were no significant differences between the total and belowground vegetation C storages under different N treatments. The aboveground vegetation C storage under 2.3 g N·m-2·a-1 treatment was significantly higher than those under the other N treatments. N treatments altered the allocation between the aboveground and belowground parts of vegetation in the typical steppe, and reduced the ratio of aboveground C storage to belowground C storage.(2) After the 2-year N treatments, the content of total N in the litter of Stipa bungeana Trin. and Heteropappus altaicus Novopokr. increased, and C/N and lignin/N ratios declined gradually, and lignin first increased and then decreased. After in-situ decomposition for one year, the decomposition rates of the two litters had not changed under different N treatments. Our results indicated that there were a close relationship of litter decomposition rates with lignin/N and lignin content.(3) The N treatments influenced the remaining percentage of N in litter. After in-situ decomposition for one year, the N remaining percentage in Stipa bungeana Trin. under the 1.15,2.3 and 4.6 g N-m-2·a-1 treatments were higher than those under other treatments, and accumulated through the process. The release mode of N in Heteropappus altaicus Novopokr. accorded with the enrichment-release modes under different N treatments. The N treatments had no significant effects on the decomposition of C in the two litters. Therefore, the optimal amount of N treatment could promote the accumulation of C in the grassland ecosystem.(4) Our results showed that the N treatments had no significant effect on the soil total C, total N and SOC (Soil Organic C) contents in the 0-40 cm soil layers in 2010 and 2011. However, the SIC (Soil Inorganic C) content in the 0-10 cm first increased and then decreased in 2011, and the SIC content under 2.3 g N·m-2·a-1 treatment was significantly higher than those under the control and other treatments. The pH values of N treatments in the 0-10 cm were lower in 2011 than that in 2010, and the soil had a trend of acidification.(5) The N treatments significantly influenced the LFOC (Light Fraction Organic C) and MBC (Microbial Biomass C) content in the 0-10 cm soil layer. The LFOC under the N treatment of 2.3 g N·m-2·a-1 was significantly higher than that under the control treatment, and there was no significant difference between 2.3 g N·m-2·a-1 and 4.6-13.8 g N·m-2·a-1 treatments. The optimal amount of N treatment for soil microbial growth and activity occurred at the rate of about 2.3 g N·m-2·a-1 in the two years. The LFOC/SOC and MBC/SOC showed similar seasonal dynamics as the LFOC and MBC under different N treatments, respectively. Therefore, the N treatment of 2.3 g N·m-2·a-1 could promote the accumulation of soil active organic C in the early stage of N treatment in semiarid grassland.(6) The N treatments did not change the seasonal dynamics of ecosystem C exchange, but increased the peak values of GEP (General Ecosystem Production), ER (Ecosystem Respiration), and NEE (Net Ecosystem Exchange). The seasonal accumulation of NEE under the N treatments with 2.3,4.6,9.2 and 13.8 g N·m-2·a-1 increased by 62%,45%,72% and 48%, respectively, while the seasonal accumulation of ER by 66%,69%,78% and 70%, and GEP by 65%,66%,77% and 68%, compared with the control treatment. Therefore, the N treatments could significantly induce C sequestration in the typical steppe of the Loess Plateau.(7) Based on the C dynamics in the vegetation, soil and microbial systems, it was found that the low N deposition level of 2.3 g N·m-2·a-1 could benefit accumulation of vegetation C storage and soil active organic C in the early stage.