Abstract:

We present a soil carbon budget for a grazed Leymus chinensis steppe community in the Xilin River basin of Inner Mongolia based on historical data and current field measurements. The study site was situated in the southern part of the basin, approximately 1 265 m above sea level. The climate is temperate and semi-arid with a mean annual rainfall of 350 mm and annual mean temperature of 0.3 ℃. The dark chestnut soil profile is about 1 m in depth with a 10-20 cm thick humus layer, a pH of 7.4 to 8.3, and an organic C content of 1.53%-1.37% in the upper 0-20 cm layer. A grazed stand (43°32′58″ N, 116°40′34″ E) outside the L. chinensis permanent plot of the Inner Mongolian Grassland Ecosystem Research Station was chosen as a sampling plot. Vegetation consisted mainly of L. chinensis, Stipa grandis, Agropyron cristatum, and Cleistogenes squarrosa, which accounted for 80% of the total aboveground biomass. The pasture has been grazed continuously for more than 40 years at a moderate stocking rate with about 2/3 of the above-ground biomass consumed annually.Soil respiration, and aboveground biomass, and root biomass measurements were initiated May 31, 1998 and measured every ten days over two growing seasons until October 14, 1999. The alkali absorption method was used to measure soil respiration. Root respiration was evaluated indirectly by relating the amount of root biomass under the chamber to rates of CO2 evolution. Aboveground biomass was measured in 10-25 cm diameter circular plots using the harvest method. The maximum biomass value measured during the growing season was taken as the annual net primary productivity (NPP). Root biomass was measured to 30 cm depth, and the root productivity was estimated using the annual increment of growth, as defined by Dahlman and Kucera. Food consumption by insects was determined using the cage-feeding method by using a series of cages with different insect population densities.The annual average carbon input from aboveground biomass production was 78.2 gC·m-2·a-1, and inputs from root biomass to 30 cm depth averaged 322.5 gC·m-2·a-1. The summed mean annual carbon input of shoot and root materials was approximately 400.7 gC·m-2·a-1. The annual amount of aboveground biomass consumed by insects averaged 14.7 gC·m-2·a-1, and the carbon output by leaching or light-chemical oxidation was 3.2 gC·m-2·a-1. The annual evolution rate of CO2 from net soil respiration averaged 343.7 gC·m-2·a-1. Cattle consumption was 49.7 gC·m-2·a-1. The summed mean annual output was approximately 411.3 gC·m-2·a-1 and a net carbon release of about 10.6 gC·m-2·a-1 was calculated. Based on the soil organic carbon density of the field, the turnover rate of soil carbon in 0-30 cm depth was 6.2%, with a turnover time of 16 years.Carbon budgets for each of the two years, 1998 and 1999, were quite different. A net carbon release of 75.1 gC·m-2·a-1 was detected in 1999, while a net carbon accumulation of 54.1 gC·m-2·a-1 occurred in 1998. By comparing carbon cycling patterns between 1998 and 1999, it was evident that the difference between years arose primarily from differences in root production and soil respiration between the two years. Severe drought in 1999 and record high rainfall in 1998 caused a difference in root NPP by as much as 220.5 gC·m-2·a-1. The root NPP accounted for about 73%～85% of the total carbon input in the community, indicating the importance of the belowground biomass in estimating the carbon budget for this ecosystem. Soil respiration seemed to be less sensitive to drought, with only about a 22.5% decrease in 1999 whereas root NPP decreased by about one half in 1999. Our results indicated that drought, a frequent occurrence in this region, could change this ecosystem from a sink to a source for atmospheric CO2. The degree to which this grassland could become a source of CO2 appeared to depend also on the stocking rate.