Heterogeneity of grassland soil respiration: Antagonistic effects of grazing and nitrogen addition

The effects of grazing and nitrogen addition on the spatial variability of soil respiration (SR) at fine spatial scales are poorly understood, which can lead to substantial uncertainty in estimations of soil CO2 flux. Our objectives were to examine how changes in grassland structure and ecosystem properties in response to grazing and nitrogen addition affect the spatial variability of SR, and to explore the underlying ecological mechanisms. We conducted a manipulative experiment with control (CK), grazing (G), nitrogen addition (N) and grazing plus nitrogen addition (NG) treatments in a Leymus chinensis meadow steppe, in northeastern China. Simultaneous measurements of SR, along with a suite of above- and below-ground response variables (aboveground biomass, root biomass, soil water content, soil nutrients and enzyme activities), were conducted for 75 sampling points in each of the four 15 x 15 m plots, both under drought and wet conditions. Compared to the CK treatment, the G treatment reduced the degree of spatial dependence of SR. No spatial structure was detected for SR in the N plot. Grazing-induced reductions in the heterogeneity of SR were lower in the fertilized plots than in the unfertilized plots. The autocorrelation range for SR in the CK treatment was higher under wet conditions (4.58 m) than under drought conditions (1.61 m). The degrees of spatial dependence of SR in the G and NG treatments under the wet condition were higher than those under the drought condition. The suite of aboveground and belowground variables explained 26-59% of the spatial variability of SR. The factors driving the spatial variability in SR and their contributions to the models varied among the different treatments. Our results demonstrated that grazing and N addition have antagonistic effects on the heterogeneity of SR. Spatial patterns of SR under different treatments and their related determinant factors are subject to control by soil water content.