Hadal trench sediments are hotspots of biogeochemical activity in the deep sea, but the biogeochemical and ecological factors that shape benthic hadal microbial communities remain unknown. Here, we sampled ten hadal sites from two trench regions with a vertical resolution of down to 1 cm. We sequenced 16S rRNA gene amplicons using universal and archaea-specific primer sets and compared the results to biogeochemical parameters. Despite bathymetric and depositional heterogeneity we found a high similarity of microbial communities within each of the two trench axes, while composition at the phylum level varied strongly with sediment depth in conjunction with the redox stratification into oxic, nitrogenous, and ferruginous zones. As a result, communities of a given sediment horizon were more similar to each other across a distance of hundreds of kilometers within each trench, than to those of adjacent horizons from the same sites separated only by centimeters. Total organic carbon content statistically only explained a small part of the variation within and between trenches, and did not explain the community differences observed between the hadal and adjacent shallower sites. Anaerobic taxa increased in abundance at the top of the ferruginous zone, seeded by organisms deposited at the sediment surface and surviving burial through the upper redox zones. While an influence of other potential factors such as geographic isolation, hydrostatic pressure, and non-steady state depositional regimes could not be discerned, redox stratification and diagenesis appear to be the main selective forces that structure community composition in hadal sediments.

The COVID-19 pandemic has demonstrated the need for massively-parallel, cost-effective tests monitoring viral spread. Here we present SARSeq, saliva analysis by RNA sequencing, a method to detect SARS-CoV-2 and other respiratory viruses on tens of thousands of samples in parallel. SARSeq relies on next generation sequencing of multiple amplicons generated in a multiplexed RT-PCR reaction. Two-dimensional, unique dual indexing, using four indices per sample, enables unambiguous and scalable assignment of reads to individual samples. We calibrate SARSeq on SARS-CoV-2 synthetic RNA, virions, and hundreds of human samples of various types. Robustness and sensitivity were virtually identical to quantitative RT-PCR. Double-blinded benchmarking to gold standard quantitative-RT-PCR performed by human diagnostics laboratories confirms this high sensitivity. SARSeq can be used to detect Influenza A and B viruses and human rhinovirus in parallel, and can be expanded for detection of other pathogens. Thus, SARSeq is ideally suited for differential diagnostic of infections during a pandemic.