For decades, soil scientists have recognized that soil represents one of Earth’s largest terrestrial carbon reservoirs, holding more carbon than the atmosphere and all vegetation combined. Within soils, carbon exists in two broad functional pools: particulate organic matter, which consists of relatively unfragmented plant debris and decomposes readily, and mineral-associated organic carbon, in which organic molecules are bound to clay and metal oxide surfaces and can persist for decades to millennia. Mineral-associated carbon is the prize in carbon sequestration strategies because of its stability, yet the precise routes by which carbon reaches this pool remain incompletely mapped.

The prevailing narrative in recent years has emphasized microbial cellular residues—the remains of dead microbial cells—as the dominant microbial contributor to stable soil carbon. Biomarkers such as amino sugars have been used to trace necromass, or dead microbial biomass, through soils, and global assessments suggest that microbial residues account for a large share of organic matter in mineral fractions. However, this framing has largely ignored what living microbes shed into the environment while they are still alive. Extracellular polymeric substances are complex mixtures of polysaccharides, proteins, extracellular DNA, and other biopolymers that microbes exude to build protective biofilm matrices, retain water, adhere to surfaces, and mediate exchanges with their surroundings. Chemically, they are abundant, carbon-rich, and reactive with mineral surfaces, making them plausible candidates for long-term stabilization, yet their contribution has been notoriously difficult to quantify.

That quantification challenge is what the new research, led by Peipei Qian, Yichao Wu, and Peng Cai of Huazhong Agricultural University together with an international team of collaborators, set out to overcome. The team developed an isotope-tracing approach based on heavy-oxygen water, H2(18)O. Because microbes incorporate the oxygen atoms of water directly into newly synthesized biomolecules during growth, the degree of (18)O enrichment in specific biomarkers can be used to calculate how fast those molecules are produced. In this case, the researchers measured (18)O incorporation into DNA, which serves as a proxy for cellular growth, and into purified fractions of EPS polysaccharides and EPS proteins, which serve as proxies for extracellular carbon production. By enzymatically digesting, purifying, and dialyzing EPS fractions extracted from soils with a cation exchange resin, the method sidestepped the long-standing problem of separating freshly produced microbial polymers from the pre-existing, heterogeneous pool of soil organic matter.

Applying this technique across a national-scale survey of cropland soils in China, the researchers produced a striking result: the rate at which microbial communities produce extracellular polymeric substances is comparable to the rate at which they produce cellular biomass. In other words, microbes channel roughly as much carbon into their surrounding slime as they do into their own bodies. Statistical analyses, including partial correlations and random forest modeling, showed that both EPS production rates and the amounts of unassociated EPS carbon in soils were significantly associated with the content of mineral-associated organic carbon across the surveyed soils, rivaling or exceeding the associations observed for traditional cellular residue indicators such as microbial biomass carbon and cellular residue carbon.

Correlation alone, however, cannot prove that EPS actually becomes stable mineral-bound carbon. To test that mechanism directly, the team turned to controlled microcosm experiments using carbon-13-labeled substrates. When soils were amended with isotopically labeled EPS, the labeled carbon was incorporated into the mineral-associated organic carbon pool at rates comparable to, and in some cases exceeding, the rates observed for labeled microbial cellular residues. NanoSIMS imaging, which can map isotope enrichment at submicron spatial resolution, visually confirmed that (13)C from EPS amendments accumulated on individual mineral particles. Solid-state (13)C nuclear magnetic resonance spectroscopy further traced changes in the chemical composition of the mineral-associated fraction, showing that EPS-derived carbon contributed measurably to the stabilized pool across three contrasting soil types—an Alfisol, an Inceptisol, and an Ultisol—under both sterile and non-sterile conditions.

The mechanistic picture that emerges is one of a dual pathway into the stable carbon pool. Dead cells leave behind necromass that can bind to minerals, but living cells continuously secrete polymeric material that adsorbs onto clay surfaces and co-precipitates with iron and aluminum oxides, effectively bypassing the death stage altogether. Laboratory studies have previously shown that EPS fractions are selectively retained when they adsorb to or co-precipitate with ferrihydrite, a poorly crystalline iron oxide abundant in many soils, and that these polymers help glue soil particles into aggregates that physically protect organic matter from decomposition. The new study provides quantitative field evidence that these laboratory-scale interactions scale up to landscape and national levels, effectively defining what the researchers frame as an extracellular arm of the soil mineral carbon pump.

The implications extend well into the domain of climate modeling and agricultural policy. Earth system models that attempt to predict how soils will respond to warming currently rely on representations of microbial carbon use efficiency and necromass turnover, but omitting the extracellular pathway could bias estimates of how much carbon soils can store and how quickly it will turn over. If EPS production proves as significant as the new measurements suggest, then factors that stimulate microbial secretion—such as root exudation, certain fertilization regimes, or changes in moisture and temperature—may influence stable carbon accumulation in ways that current models do not capture. Conversely, disturbances that disrupt biofilm matrices or expose mineral surfaces could destabilize a carbon reservoir that has been invisible to prior accounting frameworks.

For practitioners seeking to rebuild soil carbon in degraded farmland, the study offers a reframing of targets. Regenerative agriculture approaches are often designed to increase particulate and mineral-associated carbon pools through plant inputs, but the new evidence suggests that fostering the microbial communities and conditions that maximize extracellular polymer production could be equally or more important for building durable mineral-associated carbon. Practices that promote microbial activity, maintain diverse communities, and supply the carbon and nutrients microbes need for anabolism may enhance this secretion-based pathway, while soil properties such as clay content and iron oxide abundance will determine how much of the secreted material actually sticks.

Questions remain. The current survey focused on cropland soils, and the balance between extracellular and cellular pathways may differ in forests, grasslands, wetlands, and deeper soil horizons. The method, while elegant, involves incubation and extraction steps whose assumptions will need continued validation across varied soil chemistries. And the long-term fate of mineral-bound EPS—whether it resists microbial attack over decades or eventually turns over—will shape its true contribution to carbon persistence. Nevertheless, by attaching hard numbers to a slippery, previously unmeasured flux, the study transforms extracellular microbial residues from a footnote into a headline act in the story of soil carbon sequestration, reminding researchers that some of the planet’s most consequential climate chemistry may depend on the humblest of substances: microbial slime.

Subject of Research: The contribution of microbial extracellular polymeric substances to mineral-associated soil organic carbon

Article Title: Microbial extracellular residues are a major source of mineral-associated soil carbon

Article References: Qian, P., Wu, Y., Ren, J., Cheng, X., Liu, Z., Peacock, C. L., Liang, C., Zhu, X., Feng, X., Redmile-Gordon, M. A., He, X., Xiao, K.-Q., Zhu, Y.-G., Zhang, M., Dai, K., Gao, C., Qu, C., Liu, Y.-R., Tan, W., … Cai, P. (2026). Microbial extracellular residues are a major source of mineral-associated soil carbon. Nature Geoscience. https://doi.org/10.1038/s41561-026-02103-0

Image Credits: AI Generated

DOI: 10.1038/s41561-026-02103-0

Keywords: soil carbon, extracellular polymeric substances, microbial residues, mineral-associated organic carbon, isotope tracing, carbon sequestration, soil organic matter, microbiology, biogeochemistry, Nature Geoscience, carbon cycle, climate