Fig. 4: Negative effects of experimental N input on soil respiration (SR) and its components, root respiration (RR) and microbial respiration (MR). | Nature Communications

Fig. 4: Negative effects of experimental N input on soil respiration (SR) and its components, root respiration (RR) and microbial respiration (MR).

From: A general framework for nitrogen deposition effects on soil respiration in global forests

Fig. 4: Negative effects of experimental N input on soil respiration (SR) and its components, root respiration (RR) and microbial respiration (MR).

a Local polynomial regression (LOESS) model on N input rate and the negative response factor (fneg) of decreased SR to N input. Points in different colors represent data from N-limited, N-saturated, and long-term (i.e., N applied for above 3 years) experimental forests. The red line indicates the LOESS model, and the gray shading indicates the standard error of the model. b The relative importance of N input rate and environmental factors for fneg. Error bars show the standard errors of the relative importance values. MAT: mean annual temperature; MAP: mean annual precipitation; Ndep: mean annual N deposition; Sand: soil sand content; Clay: soil clay content; MAT.cv, MAP.cv and Ndep.cv are the corresponding coefficients of temporal variation; pH: soil pH; SOC: soil organic carbon content; TN: soil total nitrogen content; Ninput: N input rates in the experimental plots with artificial N addition. ch Partial dependence of fneg on N input in experimental forests. In each panel, gray curves show partial dependence plots derived from 1000 random forest models built using 1000 different subsets of the experimental data (to avoid the derived pattern being driven by a few observations), and the red curve shows the arithmetic mean of the 1000 gray curves. Blue shading and text show the turning points in the responses of decreased RR (or MR) to N inputs.

Back to article page