Fig. S1. (A) Weather conditions during the labelling period, temperature, moisture [g (H2O) g(dry soil)−1] and precipitation (mm) were measured by a micrometeorological tower and provided by Zeri M. and Bernacchi C.

(B) 13CO2 released from switchgrass root zone. Gas samples were collected on the soil surface (excluding plants) ∼ 5 cm from the plant by placing plastic tubes adjacent to the root zone (∼ 300 ml) and collection of gas for 15 min. Error bars are standard error of the mean.

Fig. S2. Density gradient obtained after ultracentrifugation. Purified rhizosphere soil DNA (3 μg) of each sample was mixed with CsCl solution to a final density of ∼ 1.725 g ml−1. The mixture was centrifuged at ∼ 77 000 r.p.m. for 40 h at 20°C. Then it was equally partitioned into 15 fractions using a syringe pump. Each fraction was weighed to determine buoyant density.

Fig. S3. Distribution of DNA along the density gradient.

(A) DNA quantified by Qubit dsDNA HS Assay Kit (detection limitation of 0.2 ng).

(B) PCR detection of 16S rRNA gene in each fraction (5–12) with different buoyant densities. H (heavy DNA), fractions with DNA that are significantly labelled by 13C; L (light DNA), fractions with DNA that are not significantly labelled by 13C. These DNAs were used in the further 16S and 18S rRNA gene sequencing. C1, C2 and C3 were the three controls without 13CO2 labelling; L1–L3, three plants labelled with 13CO2 for 5 days; L4–L6, three plants labelled with 13CO2 for 8 days. Plants L2 and L4, without significant detection of heavy DNA, were not shown in Fig. 2A. Density of each fraction was shown in Supporting Information Fig. S4. In control, heavy DNAs in fractions 5 and 6 were not detected, thus no pyrosequencing was conducted.

Fig. S4. δ13C in the leaves, roots, rhizosphere soil/DNA of individual switchgrass plants labelled with 13CO2 for over 8 days (5 h day−1). Each sample was measured in triplicate. *The initial experiment was done in July 2009, using pure 13CO2 for 8 h to label the plant. Grey shading represents mean background levels.

Table S1. Number of valid pyrosequencing reads for 16S and 18S rRNA used for analysis.

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.