Ecosystem responses to warming and watering in typical and desert steppes

Abstract

Global warming is projected to continue, leading to intense fluctuations in precipitation and heat waves and thereby affecting the productivity and the relevant biological processes of grassland ecosystems. Here, we determined the functional responses to warming and altered precipitation in both typical and desert steppes. The results showed that watering markedly increased the aboveground net primary productivity (ANPP) in a typical steppe during a drier year and in a desert steppe over two years, whereas warming manipulation had no significant effect. The soil microbial biomass carbon (MBC) and the soil respiration (SR) were increased by watering in both steppes, but the SR was significantly decreased by warming in the desert steppe only. The inorganic nitrogen components varied irregularly, with generally lower levels in the desert steppe. The belowground traits of soil total organic carbon (TOC) and the MBC were more closely associated with the ANPP in the desert than in the typical steppes. The results showed that the desert steppe with lower productivity may respond strongly to precipitation changes, particularly with warming, highlighting the positive effect of adding water with warming. Our study implies that the habitat- and year-specific responses to warming and watering should be considered when predicting an ecosystem's functional responses under climate change scenarios.

Figure 1

Effects of warming and watering on the annual aboveground net primary productivity (ANPP) in typical (a,b) and desert steppes (c,d) in 2011 (a,c) and 2012 (b,d). The dark and light bars represent warming and no warming treatments, respectively. Based on the one-way ANOVA, different lower case letters indicate differences between water treatments at the same temperature with an LSD multiple comparison test, whereas * indicates differences between warming and no warming within a watering treatment at P < 0.05. A three-way ANOVA between temperature, precipitation, and ecosystem type is shown in Tables S1 and S8. T₀W₀, T₀W₁₅, and T₀W₃₀ denote ambient temperature (T₀) with normal precipitation (W₀), plus 15% precipitation relative to average annual precipitation over the past 30 years (1978–2007, W₁₅), and plus 30% precipitation (W₃₀), respectively, whereas T₂W₀, T₂W₁₅, and T₂W₃₀ denote warming (T₂) with normal precipitation (W₀), plus 15% precipitation (W₁₅), and plus 30% precipitation (W₃₀), respectively. Vertical bars represent the SE of the mean (n = 3–4).

Figure 2

Relationships of the aboveground net primary productivity (ANPP) with annual precipitation in typical (a) and desert steppes (b).

Figure 3

Effects of warming and watering on the maximum photochemical efficiency of photosystem II (F_v/F_m) in typical (a) and desert steppes (b) at the growth peak during 2011. The dark and light bars represent warming and no warming treatments, respectively. A three-way ANOVA between temperature, precipitation, and ecosystem type is shown in Tables S2 and S8. For abbreviations of the treatments see Fig. 1. Vertical bars represent the SE of the mean (n = 24–56).

Figure 4

Effects of warming and watering on the soil respiration rate normalized at 20 °C (SR_t20) in typical (a,b) and desert steppes (c,d) at the growth peak in 2011 (a,c) and 2012 (b,d). The dark and light bars represent warming and no warming treatments, respectively. Based on the one-way ANOVA, different lower case letters indicate differences between water treatments at the same temperature with an LSD multiple comparison test, whereas * indicates differences between warming and no warming within a watering treatment at P < 0.05. A three-way ANOVA between temperature, precipitation, and ecosystem type is shown in Tables S3 and S8. Vertical bars represent the SE of the mean (n = 15–30). For abbreviations of the treatments see Fig. 1. Note the differences in the y-axis scales.

Figure 5

Relationships of ANPP with soil TOC and MBC in the typical (a,c) and desert (b,d) steppes. For abbreviated details, see Table 1. Note the differences in the y-axis scales of the upper panels.

Figure 6

Relationships of soil TOC and MBC in typical (a) and desert (b) steppes. For abbreviated details, see Table 1. Note the differences in the x-axis scales.

Figure 7

Loadings from the first two principal components (PCs) derived from principle component analysis (PCA) for all parameters (a); the relationships of annual aboveground net primary productivity (ANPP) with secondary principal component scores (PC2)—only summarizing several belowground process features—in typical (b) and desert steppes (c). F_v/F_m, maximum photochemical efficiency of photosystem II; MAP, mean annual precipitation; MBC, microbial biomass carbon; Q₁₀, soil respiration rate (SR) sensitivity to temperature; SP, growth seasonal precipitation; SR_T20, SR at 20 °C; TOC, soil total organic carbon.