Table 1.
Location of the 19 grassland sites.
Fig 1.
Measurements from May 2002 to December 2003.
(a) daily mean air temperature (Ta, n = 6, from sites where data loggers were installed) and precipitation (P, daily sums of precipitation at the meteorological station Wetzstein); (b) mean manually measured soil moisture (SM), daily mean values of assimilated continuous SM (mean ± SD, n = 19) and (c) mean difference in SM between control and drought plots (ΔSM); (d) mean manually measured soil respiration (SR), daily mean values of the modelled seasonal courses of SR (mean ± SD, n = 19) and e) the difference in SR between control and drought plots (ΔSR). Grey rectangles in the top panel mark the shelter periods (continued by dashed lines into the lower panels). For clarity, SD is shown in one direction only. Numbers on time axis are months of the year. Stars show levels of significance: *p < 0.05, **p < 0.01, ***p < 0.001 (tested by ANOVA).
Fig 2.
Soil respiration and biotic drivers under control and drought conditions.
(a) mean daily soil respiration (SRdaily) and (b) mean annual above-ground productivity (c) standing root biomass and (d) species richness in 2002 (circles) and 2003 (triangles) for control (filled symbols) and drought conditions (open symbols) (mean ± SD, n = 19). Underlying grey symbols show single values. Stars indicate levels of significance: *p < 0.05, **p < 0.01, ***p < 0.001 (tested by ANOVA).
Fig 3.
Correlations between biotic drivers.
Scatterplot matrix for correlations between annual above-ground productivity (under control conditions), species richness and standing root biomass in 2002. Since neither species richness nor standing root biomass showed significant changes in response to the drought treatment, site means from control and drought plots are shown.
Table 2.
Annual above-ground productivity as a covariate.
Fig 4.
Relative change in soil respiration in relation to biotic drivers.
Relative reduction of mean daily sums of soil respiration (SRdaily) in 2002 (circles) and 2003 (triangles) in response to the drought treatment correlated with (a) annual above-ground productivity under control conditions (b) root biomass and (c) species richness. For species richness and root biomass site means from 2002 are shown. Stars indicate levels of significance: *p < 0.05, **p < 0.01 (tested by ANOVA). Note that the interactions terms in our REML approach tested for the same relative effects because the data was log-transformed prior to analysis.
Fig 5.
Interactions of biotic drivers with soil respiration and its response to drought.
Schematic illustration of the interaction of the investigated biotic drivers (i.e. annual above-ground productivity, species richness and root biomass) with soil respiration irrespective of the drought treatment (left) and with the drought response of soil respiration (right). + indicates positive correlations; 0 indicates no significant effect. Note that ΔSR was always calculated as the difference of soil respiration rates between drought and control plots (SR drought—SR control), thus negative ΔSR values depict reduced soil respiration rates due to the drought treatment and the more negative ΔSR, the larger the drought effect on soil respiration. As an example, species richness is positively correlated with ΔSR under drought conditions, which indicates a more positive ΔSR with increasing species richness, i.e. a lower drought effect on SR with increasing species richness.