Sheryl L. Bell

When dry soil is rewet, there is a near-immediate pulse of CO2 release that exceeds baseline respiration and can persist for days. This Birch effect is commonly observed across diverse environments and soil types, yet the underlying cellular mechanism of recovery from desiccation remains elusive because most studies are conducted over timescales of hours to days. We used metatranscriptomics, metabolomics, high-resolution mass spectrometry, and modeling to investigate the microbial processes driving the immediate (0-90 minutes) CO2 pulse in response to rewetting soil after 1 and 2 weeks of drought. Post-rewetting CO2 efflux profiles of live versus sterilized soil support persistent low levels of biological activity during drought and the physical flushing of accumulated CO2 from the soil in the first minutes of rewetting. CO2 efflux was 77±12% higher under two-week compared to one-week drought exposure, consistent with continual CO2 accumulation in the pore space throughout drought. Expression profiles for the two-week versus one-week drought/rewet samples were nearly indistinguishable, supporting that the observed difference in CO2 efflux is not driven by a genetically regulated alteration of cellular function. Our metabolic model, MEMPIS, was used to integrate metabolomic and metatranscriptomic data to identify pathways that may contribute to drought responses and post-rewetting respiration activity. The algorithm highlighted trehalose metabolism, polyhydroxybutyrate metabolism, the glyoxylate shunt, and the gamma-aminobutyric acid shunt, implicating metabolites detected from these pathways as carbon sources supporting respiration. Extracellular trehalose was observed to be consumed upon rewetting; however, metatranscriptome profiles at 90 minutes after rewetting largely show a return to pre-drought patterns. This result suggests that genes required for recovery from drought must be transcribed during drought response, priming the cells for recovery. Together, these results improve our understanding of soil microbiome response to drought by clarifying temporal connections between cellular regulatory response, metabolic activities and soil respiration.

Deciphering viral ecology in soils is challenging due to soil’s high physicochemical and microbial community complexity. To enhance detection of DNA and RNA viruses, we applied different preparation methods to soils collected from a grassland field experiment. Analyses included metagenomics and metatranscriptomics of size-fractionated extracellular viruses, total soil metagenomics and metatranscriptomics, total soil metatranscriptomics with polyadenylation enrichment, and metagenomics of bacteria/archaea as well as eukaryote-enriched samples. DNA viromes outperformed total soil metagenomes in viral detection and quality. Contrastingly, RNA viromes and total soil metatranscriptomes performed similarly for viral recovery, though RNA viromes yielded higher-quality genomes. Together, our results highlight how different preparation methods can influence the recovery and quality of DNA and RNA vOTUs. Further, we demonstrate the power of different methods in identifying distinct viral communities with unique host predictions, which in turn can have significant implications for ecological investigations related to interkingdom interactions.

Soil organic matter decomposition is a complex process reflecting microbial composition and environmental conditions. Moisture can modulate the connectivity and interactions of microbes. Due to heterogeneity, a deeper understanding of the influence of soil moisture on the dynamics of organic matter decomposition and resultant phenotypes remains a challenge. Soils from a long-term field experiment exposed to high and low moisture treatments were incubated in the laboratory to investigate organic matter decomposition using chitin as a model substrate. By combining enzymatic assays, biomass measurements, and microbial enrichment via activity-based probes, we determined the microbial functional response to chitin amendments and field moisture treatments at both the community and cell scales. Chitinolytic activities showed significant responses to the amendment of chitin, independent of differences in field moisture treatments. However, for other measurements of carbon metabolism and cellular functions, soils from high moisture field treatments had greater potential enzyme activity than soils from low moisture field treatments. A cell tagging approach was used to enrich and quantify bacterial taxa that are actively producing chitin-degrading enzymes. By integrating organism, community, and soil core measurements we show that (i) a small subset of taxa compose the majority (>50%) of chitinase production despite broad functional redundancy, (ii) the identity of key chitin degraders varies with moisture level, and (iii) extracellular enzymes that are not cell-associated account for most potential chitinase activity measured in field soil.

The rhizosphere, where plant roots meet soil, is a hub of biogeochemical activity with ecosystem impacts on carbon stocks. Root derived carbon has been found to contribute more to soil carbon stocks than aboveground litter. Nonetheless, the molecular chemodiversity of root exudates remains poorly understood due to limited characterization and annotation. Here our goal was to discover the molecular chemodiversity of metabolites and lipids in root exudates to advance our understanding of plant root inputs belowground. We worked with mature, field-grown tall wheatgrass ( Thinopyrum ponticum ) and optimized exudate collection protocols to enable the capture of non-polar lipids in addition to polar and semi-polar metabolites. Rates of carbon input via hydrophobic exudates were approximately double that of aqueous exudates and carbon/nitrogen ratios were markedly higher in hydrophobic compared to aqueous exudates, emphasizing the importance of lipids, due to their high carbon content. To maximize molecular coverage of exudate chemodiversity, we used liquid chromatography coupled tandem mass-spectrometry for paired untargeted metabolomics and lipidomics or ‘metabo-lipidomics’. We substantially increased the characterization of exudate chemodiversity by employing both tandem mass spectral library searching and deep learning-based chemical class assignment. Notably, in this unprecedented characterization of intact lipids in root exudates, we discovered a diverse variety of lipids, including substantial levels of triacylglycerols (∼19 μg/g fresh root per min), fatty acyls, sphingolipids, sterol lipids, and glycerophospholipids. Comparison of the root exudate and tissue lipidomes revealed minimum glycerophospholipids in exudates, suggesting the exudate protocol did not extract lipids from root cell membranes.

Microorganisms drive soil carbon mineralization and changes in their activity with increased temperature could feedback to climate change. Variation in microbial biodiversity and the temperature sensitivities (Q₁₀) of individual taxa may explain differences in the Q₁₀ of soil respiration, a possibility not previously examined due to methodological limitations. Here, we show phylogenetic and taxonomic variation in the Q₁₀ of growth (5–35 °C) among soil bacteria from four sites, one from each of Arctic, boreal, temperate, and tropical biomes. Differences in the temperature sensitivities of taxa and the taxonomic composition of communities determined community-assembled bacterial growth Q₁₀, which was strongly predictive of soil respiration Q₁₀ within and across biomes. Our results suggest community-assembled traits of microbial taxa may enable enhanced prediction of carbon cycling feedbacks to climate change in ecosystems across the globe.

To enable an in-depth survey of the metabolic potential of complex soil microbiomes, we performed ultra-deep metagenome sequencing, collecting >1 Tb of sequence data from three grassland soils representing different precipitation regimes.