Fig. S1. Hindcast models of bottom salinity in the Columbia River estuary during sampling. Models were downloaded and modified from the Model Browser created by the Center for Coastal Margin and Prediction Program (, a multi-institutional National Science Foundation Science and Technology Center. A and B. June samples taken (A) during an ETM with turbidity at 50 ntu and salinity at 19 psu, and (B) after an ETM with turbidity at 8 ntu and salinity at 25 psu. C and D. July samples taken (C) during an ETM with turbidity at 70 ntu and salinity at 5 psu, and (D) after an ETM with turbidity at 6 ntu and salinity at 9 psu. For reference, Astoria (represented by a polygon) and Youngs Bay are labelled in the first panel.

Fig. S2. Changes in total, dissolved and ascorbate-released manganese with depth (A) during and (B) before an ETM event in the South Channel of the Columbia River in July 2007.

Fig. S3. Relationship between turbidity and manganese (A) during and after an ETM event in the South Channel in July 2007, and (B) regression analysis of total manganese versus turbidity and dissolved manganese(II) versus turbidity for all deep water samples (c. 1 m from the bottom, or 10–18 m depth) collected in the South Channel on June, July or August 2007 cruises on the Columbia River.

Fig. S4. Maximum-likelihood tree inferring the phylogenetic relationship between Mn(IV)-reducing and Mn(II)-oxidizing Shewanella sp. cultured strains* (asterisks) and those found in clone libraries† (daggers) in the Columbia River in this study. Top blast hits to the isolates were included, as well as a broad range of described species. Alignments were created using the on-line SILVA aligner and then exported from ARB. Dendrogram was created using PHYLIP. Bootstrapping values are shown for nodes that were supported at least 50% of the time. The number of the sequences from each library that represented a particular sequence/OTU is given in parentheses. The dendrogram was only weakly supported with neighbour-joining analysis, and many branches had low bootstrapping values in both analyses (Fig. 5). Escherichia species were used as the outgroup.

EMI_2360_sm_fS1-4.pdf481KSupporting info item

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.