64 Chapter 2 highest at Site 5 (3.6x at 34 cm), followed by Site 3 (3x at 22 cm) and Site 7 (2.6x at 22 cm, Figure 4). Putative methanogen abundances (hypothesized from MAG coverages and 16S rRNA gene-based relative abundances) and potential methane production rates have contrasting profiles and do not positively correlate (Figure 4). An explanation for these results is that sequencing data do not reflect potentially higher methanogen biomass or activity in surface sediments (0-10 cm). Alternatively, larger pools of labile organic substrates generated from organic matter degradation could be available in surface sediments (as observed in other aquatic ecosystems (Dalcin Martins et al., 2017), which are depleted at depth, resulting in decreasing potential methane production rates in deeper sediment layers. This might become apparent at the gene expression level, which could correlate with methane production activity, potentially detectable in future metatranscriptomic studies. In Site 3, where sulfate was detected until ca. 30 cm, sulfate reduction-driven competitive inhibition of methanogenesis was expected (Bethke et al., 2011), and low potential methane production rates at this site indicate that this expectation was fulfilled despite the relatively high MAG coverages and 16S-based relative abundances of methanogens. In Sites 5 and 7, high potential methane production rates in surface sediments might also reflect larger pools of labile organic carbon and the rapid depletion of sulfate. These are conditions that could favor methanogens at potentially lower abundances in top sediments to be more active than in deeper sediments, where they could be more abundant but have less substrate availability. Additionally, the observation that the highest potential methane production rates were measured in surface sediments concomitant with relatively high sulfate concentrations suggests cryptic methane cycling, in which methane is consumed as soon as it is produced within the SMTZ, detectable via radiotracer or stable isotope studies (Krause et al., 2023). Methanogenesis from non-competitive substrates is supported by our data, since genomic potential for methanol and methylamine-driven methanogenesis was identified in three out of four methanogen MAGs. This could partially fuel AOM in Site 5, where the ANME-2 MAG coverage was significant (3.85x) within the SMTZ. Methylotrophic methanogenesis has been previously implicated in cryptic methane cycling in similar ecosystems (Krause & Treude, 2021; Xiao et al., 2018). Alternatively, competitive methanogenesis could co-exist with sulfate
RkJQdWJsaXNoZXIy MTk4NDMw