Effects of grassland-use on soil respiration and litter decomposition

Aims Land use change affects ecosystem carbon dynamics by changing the plant community structure and soil micro-environment in grassland ecosystems. The aims of this study were to determine the effects of land use on soil respiration and litter decomposition in the temperate grasslands of Nei Mongol and to identify the effects of litter quantity, quality and decomposition on soil respiration during growing season.Methods We measured soil respiration during growing season in 2011 and 2012 under three land use types, i.e. grazing, mowing, and grazing exclusion, by using an automatic infrared gas analyzer (LI-8100) that was connected to a multiplexer system (LI-8150). Quadrat surveys and litter bags were utilized to measure litter production and decomposition. Several chemical indicators of litter quality were measured to calculate the litter decay rates. All data were analyzed with ANOVA and Pearson correlation procedures of SPSS.Important findings Soil respiration and litter decomposition differed greatly among the three land-use types. In the drought year, the total soil respiration at the grazing site was 1.5 times greater than at the mowing site and 1.29 times greater than at the grazing-exclusion site. However, in the wet year, the total soil respiration at the mowing site reached 309 g C·m^–2·a^–1 and was greater than at both the grazing site and the grazing-exclusion site. Precipitation increased soil respiration and litter decomposition, indicating that soil water availability was a primary constraint on plant growth and ecosystem C processes. Also, the responses of soil respiration and litter composition to rainfall differed among the land-use types. Further analysis showed that the litter C:N decreased and the litter N content and lignin:N increased after 2-years of decomposition. In addition, soil respiration was significantly correlated to litter production (r = 0.78, p < 0.01), decay rates, C:N (r = –0.84, p < 0.01), and lignin:N (r = 0.62, p < 0.05).