The capacity of soil as a carbon (C) sink is mediated by interactions between organic matter and mineral phases. However, previously proposed layered accumulation of organic matter within aggregate organo–mineral microstructures has not yet been confirmed by direct visualization at the necessary nanometer-scale spatial resolution. Here, we identify disordered micrometer-size organic phases rather than previously reported ordered gradients in C functional groups. Using cryo-electron microscopy with electron energy loss spectroscopy (EELS), we show organo–organic interfaces in contrast to exclusively organo–mineral interfaces. Single-digit nanometer-size layers of C forms were detected at the organo–organic interface, showing alkyl C and nitrogen (N) enrichment (by 4 and 7%, respectively). At the organo–mineral interface, 88% (72–92%) and 33% (16–53%) enrichment of N and oxidized C, respectively, indicate different stabilization processes than at organo–organic interfaces. However, N enrichment at both interface types points towards the importance of N-rich residues for greater C sequestration.

Organic compounds exuded by plant roots can form organo-mineral associations through physico-chemical interactions with soil minerals but can disrupt existing organo-mineral associations by increasing their microbial decomposition and dissolution. The controls on these opposing processes are poorly understood, as are the chemical and spatial characteristics of these associations which may explain gain or loss of organic matter at the root-soil interface termed the rhizosphere. By pulse-labeling with ¹³C-carbon dioxide, we found that maize root exudates increased organic matter in the rhizosphere clay size fraction and decreased organic matter in the silt size fraction, and that organic matter loss was mitigated by dry conditions. Organic matter associated with rhizosphere clay particles was linked to microbial metabolism of exudates and was more spatially and chemically heterogeneous than non-rhizosphere clay particles. Our findings show that root exudates can simultaneously form and disrupt organo-mineral associations, mediated by mineral size and composition, and soil moisture.

Longstanding theories and models classify mineral-associated organic matter as the large ( ~ 60%) but slow-cycling and persistent portion of soil organic matter. Strong physico-chemical interactions and diffusion limitations restrict the turnover of mineral-associated organic matter, allowing carbon and nitrogen bound therein to persist in soil for as long as centuries to millennia. However, mineral-associated organic matter is a chemically and functionally diverse pool with a substantial portion cycling at relatively fast (i.e., minutes to years) timescales. Despite a growing body of evidence for the heterogenous and multi-pool nature of mineral-associated organic matter, we lack consensus on how to conceptualize and directly quantify fast-cycling mineral-associated organic matter and its ecological significance. We demonstrate that the dynamic qualities of fast-cycling mineral-associated organic matter vary based on 1) the chemistry of the mineral particles and organic matter, 2) the complex set of interactions between organic matter and the mineral matrix, and 3) the presence and strength of destabilizing forces that lead to decomposition or loss of mineral-associated organic matter (i.e., plant-microbe interactions, agricultural intensification, and climate change). Finally, we discuss potential implications and research opportunities for how we measure, manage, and model the dynamic subfraction of this otherwise persistent pool of soil organic matter.

Subsoils store at least 50% of soil organic carbon (SOC) globally, but climate change may accelerate subsoil SOC (SOC_sub) decomposition and amplify SOC-climate feedbacks. The climate sensitivity of SOC_sub decomposition varies across systems, but we lack the mechanistic links needed to predict system-specific SOC_sub vulnerability as a function of measurable properties at larger scales. Here, we show that soil chemical properties exert significant control over SOC_sub decomposition under elevated temperature and moisture in subsoils collected across terrestrial National Ecological Observatory Network sites. Compared to a suite of soil and site-level variables, a divalent base cation-to-reactive metal gradient, linked to dominant mechanisms of SOC_sub mineral protection, was the best predictor of the climate sensitivity of SOC decomposition. The response was “U”-shaped, showing higher sensitivity to temperature and moisture when either extractable base cations or reactive metals were highest. However, SOC_sub in base cation-dominated subsoils was more sensitive to moisture than temperature, with the opposite relationship demonstrated in reactive metal-dominated subsoils. These observations highlight the importance of system-specific mechanisms of mineral stabilization in the prediction of SOC_sub vulnerability to climate drivers. Our observations also form the basis for a spatially explicit, scalable, and mechanistically grounded tool for improved prediction of SOC_sub response to climate change.

Freeze‐thaw is a disturbance process in cold regions where permafrost soils are becoming vulnerable to temperature fluctuations above 0°C. Freeze‐thaw alters soil physical and biogeochemical properties with implications for carbon persistence and emissions in Arctic landscapes. We examined whether different freeze‐thaw histories in two soil systems led to contrasting biogeochemical responses under a laboratory‐controlled freeze‐thaw incubation. We investigated controls on carbon composition through Fourier‐transform ion cyclotron resonance mass spectrometry (FT‐ICR‐MS) to identify nominal carbon oxidation states and relative abundances of aliphatic‐type carbon molecules in both surface and subsurface soils. Soil cores (∼60 cm‐depth) were sampled from two sites in Alaskan permafrost landscapes with different in situ freeze‐thaw characteristics: Healy (>40 freeze‐thaw cycles annually) and Toolik (<15 freeze‐thaw cycles annually). FT‐ICR‐MS was coupled with in situ temperature data and soil properties (i.e., soil texture, mineralogy) to assess (a) differences in soil organic matter composition associated with previous freeze‐thaw history and (b) sensitivity to experimental freeze‐thaw in the extracted cores. Control (freeze‐only) samples showed greater carbon oxidation in Healy soils compared with Toolik, even in lower mineral horizons where freeze‐thaw history was comparable across both sites. Healy showed the most loss of carbon compounds following experimental freeze‐thaw in the lower mineral depths, including a decrease in aliphatics. Toolik soils responded more slowly to freeze‐thaw as shown by intermediary carbon oxidation distributed across multiple carbon compound classes. Variations in the response of permafrost carbon chemistry to freeze‐thaw is an important factor for predicting changes in soil function as permafrost thaws in high northern latitudes.