Extended Data Fig. 9: Candidate phylum Dormibacteraeota (AD3) genus Ca. ‘Changshengia’ at Stordalen Mire.

a, Total relative abundance of the genus Ca. ‘Changshengia’ correlated with the fraction of the concentration of C mineralized to CO₂ versus CH₄ in the bog porewater samples (R² = 0.19, P = 0.001, n = 51 biologically independent samples). Each point represents an individual sample from 2012, with its colour representing the depth in the core from which the sample was taken. b, Metabolic reconstruction of genomes belonging to the candidate phylum AD3 genus Ca. ‘Changshengia’ correlating with the CH₄:CO₂ concentration ratio in porewater from 2012 bog samples. Genomes from four clades within the AD3 were assembled from across Stordalen Mire. Enzyme colour indicates the families that share that metabolic potential, as outlined in the legend on the left. Arrow colouring indicates whether expression was detected (red arrows) or not detected (black arrows) for genes encoding the enzyme in any of the 24 metatranscriptomes. Orange stars indicate detection of protein expression in any of the 22 metaproteomes from the Ca. ‘Changshengia’ and related genomes. All four lineages encode the potential to oxidize glycerol anaerobically through glycerol transporter (glpF), glycerol kinase (glpK) and a membrane-bound glycerol-3-phosphate dehydrogenase (glpABC), entering glycolysis via dihydroxyacetone phosphate processed to glyceraldehyde-3-phosphate by the triosephosphate isomerase (tpiA). Other glycerol derivatives such as glycerol-3-phosphate could be imported (glpT) by this and other family members, and dihydroxyacetone phosphate can also be processed using the PTS-dependent dihydroxyacetone kinase (dhaLMK) complex. Sinks for the electrons generated from the oxidation of glycerol also varied between the different lineages, with Ca. ‘Changshengia’ and clade 1 having a H⁺-translocating complex I NADH:oxidoreductase, while clade 1 also has a high affinity cytochrome oxidase complex IV, clade 2 genomes encode only a nitrate reductase (narGHI) and clade 4 genomes only a fumarate reductase (sdhABCD). These differences are likely to lead to the differentiation of the niches that each lineage occupies across different sites and depths of the mire. Lineages were considered positive for genes or complexes based on the presence of sequences with 80% homology in 50% of the genomes. c, Phylogenetic subtree showing the family groupings of AD3 for the metabolic analysis. Representative genomes from the 97% average nucleotide identity (ANI) dereplication are indicated in red. Bootstrap support is indicated at the nodes for values over 70% or 90% in grey and black, respectively. Blue clade indicates cluster of seven UBA and RefSeq genomes.