Supplementary Figure 1: Comparison of Fe(II) oxidation of 0.5 mm dissolved

Transcription

1 Supplementary Information Supplementary Figure 1: Comparison of Fe(II) oxidation of 0.5 mm dissolved Fe(II) by Rhodobacter ferrooxidans sp. strain SW2 in the absence (a) and presence (b) of 2.0 mm monomeric silica shows no change in Fe(II) oxidation rate. All experiments were conducted in triplicate and with an uninoculated (abiotic) control that showed no Fe(II) oxidation. The left y-axis (labelled in red) shows the concentration of Fe(II) in solution at various time points; ( ) abiotic control, ( ) experimental triplicates. The right y-axis in (b) depicts the concentration of monomeric silica in solution, represented by light (control) and dark grey bars (experimental triplicates).

2 Supplementary Figure 2: Effect of temperature increase on Rhodobacter ferrooxidans sp. strain SW2. The left y-axis displays anoxygenic phototrophic Fe(II) oxidation as a decrease in Fe(II) [mm] over time; ( ) temperature control remaining at 20 C throughout experiment, ( ) mean value of experiment triplicates. The right y-axis shows the change in incubation temperature during the course of the experiment ( ). While microbial Fe(II) oxidation at first continues after temperature change, it is not revived 3 days after exposure to 50 C incubation temperature.

3 Supplementary Figure 3: Geochemical model (Geochemists Workbench Standard 6.0) of silica solubility in a solution of microbial mineral medium at circumneutral ph composed of 0.6 g/l potassium phosphate (KH 2 PO 4 ); 0.3 g/l ammonium chloride (NH 4 Cl); 0.5 g/l magnesium sulphate (MgSO 4 7H 2 O); and 0.1 g/l calcium carbonate (CaCl 2 2H 2 O); and 22 mm carbonate buffer. This model was designed to estimate the precipitation of silica due to temperature decrease in abiotic batch cultures. The batch culture data is presented in the manuscript. The red line separates the two silica phases amorphous silica (precipitate) and dissolved SiO 2 (aq), while the grey arrow marks the change in phase as temperature decreases at the 2 mm silica estimated for the Archean ocean. According to this model, as temperature drops, silica will precipitate as an amorphous silica phase.

4 Supplementary Figure 4: Fe(II) oxidation and silica in solution over time for anoxygenic Fe(II)-oxidizing phototroph Thiodictyon sp. strain F4. All experiments were conducted in triplicate and with an abiotic (uninoculated) control. The left y-axis (labelled in red) shows the concentration of Fe(II) in solution at various time points; ( ) abiotic control, ( ) experimental triplicates. The right y-axis in depicts the concentration of monomeric silica in the culture solution, represented by dark (abiotic control) and light grey bars (experimental triplicates).

5 Supplementary Figure 5: Abiotic oxidation of Fe(II) in the presence of silica. (a) Chemical oxidation of dissolved Fe(II) leading to a much stronger precipitation of silica in comparison to a less pronounced precipitation of silica during Fe(II) oxidation by the anoxygenic phototroph Thiodictyon sp. F4. Values of dissolved silica in solution shown at t = 0 ( ) and t = end of Fe oxidation ( ). (b) Solid phase Si/Fe molar ratios for ferric hydroxide precipitated during the complete oxidation of 0.18 mm (10 ppm Fe) ferrous iron ammonium sulfate solution (ph 8, 25 C and in 0.5 M NaCl) in the presence of various amounts of dissolved silica (up to 1 mm). The coprecipitation of silica with iron appears to display typical adsorption isotherm behavior, with an approximately linear relationship between solid phase Si/Fe and dissolved Si at lower values, a plateau representing saturation at middle values, and what may represent surface precipitation of silica on the iron mineral at the highest dissolved silica concentration investigated, despite undersaturation with respect to amorphous silica in the bulk solution.

6 Supplementary figure 6: Synchrotron-based computer tomography slice images of Thiodictyon sp. strain F4 cell-mineral aggregates precipitated at the end of 0.5 mm Fe(II) oxidation in the absence (a) and in the presence of 2 mm silica (b). Various elements, including Si and Fe, can be distinguished based on the element specific absorption of the synchrotron X-rays. This leads to a marked difference in brightness and contrast of particles that contain either silica or iron. The distinct particles can be seen in Fig. a and b: Silica shown with the white arrow, cell-mineral aggregates with black arrows. Cell-mineral aggregates were taken up with warm 5% low melt agarose solution into 0.5 µm diameter glass capillaries The agarose solidifies with cooling to room temperature, preserving the orientation and structure of the precipitates. These images show larger aggregates forming in the absence of silica but also the separation of the cell-fe(iii) mineral aggregates from the amorphous silica precipitate. Image dimensions are approximately 200 µm x 200 µm.

7 Supplementary Figure 7: Seasonal temperature variations in the modern ocean at ~80 N, 40 N, 0, 40 S and 80 S. Temperatures compiled from World Ocean Atlas 2005 ( profiling depths of 10, 50 and 100 m for Northern Hemisphere winter (January March) and summer (July-September). Full lines ( ) represent temperatures at 10 m, broad dashed lines ( ) represent temperatures at 50 m, and dotted line (---) represent temperatures at 100 m. Symbols in the top row indicate positions of Earth s axis and sun.