Fig. S1. Photographs of the investigated field soils at the military training range in the North-East of Belgium. Left, the bare 2,4-DNT contaminated soil with bomb remnants and detonation container. On the left side, the adjacent grassland soil colonized by Molinia caerulea and Agrostis tenuis, photograph was taken in November 2009.

Fig. S2. Scatterplot of Biolog EcoPlate responses (average net area under the curve for each of the 31 carbon sources) comparing the grassland soil winter (GS-w), 2,4-DNT contaminated soil summer (DS-s) and winter (DS-w) inoculum metabolic profiles to that obtained for the grassland soil summer (GS-s) inoculum (plotted on x-axis).

Fig. S3. Microscopic images of bacteria containing PHB inclusions using the lipid Sudan stain and saffranin counterstain. The bacteria strains are (A) Rhodanobacter sp. HC50 and (B) Pseudomonas sp. HC94. (C) negative control, bacterium with no lipid granules inclusions. The blue dots in the cytoplasma are the lipid-granules, indicated with the arrows. (Nikon Eclipse 80i, oil-immersion objective)

Fig. S4. Clear halos appearing on the opaque white plates indicate bacterial phosphorous solubilisation. A phosphate-buffer washed suspension of bacteria, pre-grown in rich medium was inoculated into the holes. All strains were tested in triplicate.

Fig. S5. Chemotactic response and flagella stain of Ralstonia sp. HC90. (A) Chemotactic response towards aspartic acid and (B) 2,4-DNT. (C) Negative control, lack of chemotaxis of heat-killed cells towards 2,4-DNT. Red arrows indicate the chemotaxis rings. Plates were scanned with a flatbed scanner. (D) Leifson flagella stain of Ralstonia sp. HC90, black arrow indicates flagella. Photograph was taken with oil-immersion objective (1500×).

Fig. S6. Photographs of the Arabidopsis thaliana plants on VAPS-plates (A) control; (B) UHS3 inoculated control; (C) exposed to 2,4-DNT and (D) exposed to 2,4-DNT with UHS3 inoculation, grown at a density of 5 plants per plate. The photograph was taken 9d after transfer of the seedlings when root length measures were taken, root hair pictures and the shoots were collected for reisolation of the inoculated bacteria.

Fig. S7. UPGMA dendrogram showing the relationships among the bacteria isolated from A. thaliana shoot tissues and the inoculated strains from consortium UHS3.

Table S1. Physicochemical characteristics of the 2,4-DNT contaminated soil and adjacent grassland soil. Nitro-aromatic concentrations in the samples were measured with HPLC and are expressed in mg kg −1 dry soil with standard error. EC: Electric Conductivity; 2,4-dinitrotoluene (2,4-DNT); 2,6-dinitrotoluene (2,6-DNT); 1,3-dinitrobenzene (DNB); 1,3,5-trinitrobenzene (TNB); nitrobenzene (NB); 2-amino-4-nitrotoluene (2-A-4NT); 4-amino-2-nitrotoluene (4-A-2NT). ND = not detected (< 0.01 mg l−1 ). DS-s: 2,4-DNT contaminated soil summer; DS-w: 2,4-DNT contaminated soil winter; GS-s: grassland soil summer; GS-w: grassland soil winter.

Table S2. Fingerprint of the inocula tested with Ecoplates. The absorbance data of the Biolog Ecoplates were blanked to the water containing well and then converted into Boolean ones, considering positive (1) the substrates for which the blanked absorbance data > 0.10 and negative (0) the substrates for which the blanked absorbance data < 0.10. DS-s: 2,4-DNT contaminated soil summer; DS-w: 2,4-DNT contaminated soil winter; GS-s: grassland soil summer; GS-w: grassland soil winter.

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.