Dust and Delay: Why Years in Storage Can Double a Key Soil Carbon Reading

Every year, agricultural laboratories around the world process millions of soil samples, air-dry them on open trays, seal them in bags, and shelve them for later analysis. The assumption behind this routine is simple: once a soil is dry, it is chemically dormant, a stable snapshot of the field it came from. A new study published in the journal SOIL upends that assumption. Researchers led by Swarnali Mahmood and Yang Lin of the University of Florida report that just three years of storage nearly doubled the concentration of water-extractable organic carbon, or WEOC, one of the most widely used indicators of soil health. The finding suggests that some of the numbers scientists and farmers rely on to judge whether a soil is improving may be quietly distorted by the very act of putting samples on a shelf.

Soil health assessment has become a cornerstone of sustainable agriculture. Laboratories measure a suite of physical, chemical, and biological properties to gauge how well a soil supports nutrient cycling, water regulation, carbon sequestration, and biodiversity. The most sensitive indicators are deliberately chosen because they respond quickly to management changes. Total soil organic carbon, the fundamental currency of soil fertility, moves slowly, often taking years or decades to register the effects of a new cropping practice. Labile carbon pools, by contrast, are designed to be fast. Water-extractable organic carbon reflects the readily available substrates that microbes consume; potentially mineralizable carbon captures how much carbon dioxide microbes release when soil is rewetted; and permanganate-oxidizable carbon, known as POX-C, measures a chemically reactive fraction of organic matter. These rapid-response metrics are attractive precisely because they change quickly, but the new work shows that speed comes with a hidden vulnerability.

The problem begins with water. Even soils classified as air-dried retain a small residue of moisture, typically between 0.3 and 6 percent of their mass, adsorbed to mineral surfaces and trapped in fine pores. That residual water, the researchers explain, can be enough to sustain low levels of microbial activity. Over months and years of storage, slowly metabolizing microbes, cell lysis, and continued enzymatic depolymerization can transform the labile carbon and nitrogen pools that sensitive indicators are supposed to measure. Drying itself may also alter mineral equilibria and the solubility of soil organic matter, releasing carbon that would have stayed bound in a fresh sample. Bulk soil organic carbon and nitrogen pools remain relatively stable during long-term storage, which is why archives have long been trusted for those measurements, but the water-extractable fractions are far more fragile.

To quantify these effects directly, the team designed what they call a Controlled Storage Experiment. They retrieved archived surface soils from a three-year cover crop study at the University of Florida’s Field and Fork Farm in Gainesville, where plots had been planted with pearl millet, sunn hemp, or a mixture of grasses and legumes including sorghum sudangrass, pearl millet, sunn hemp, and cowpea. The samples, collected from 2019 to 2021 before each summer planting season, had been air-dried in a climate-controlled laboratory, sieved, and stored in airtight plastic bags at room temperature. In 2025, the researchers reanalyzed the same archived samples using identical protocols to those applied in 2021 and 2022, isolating the effect of storage from any field-driven changes. Water-extractable carbon and nitrogen were extracted by shaking soil with ultrapure water; mineralizable carbon was measured as carbon dioxide accumulating in sealed jars after rewetting; and POX-C was determined colorimetrically after reaction with potassium permanganate.

The results were striking. Across all sampling years and cover crop treatments, WEOC concentrations increased by 96 percent after three years of storage, a near-doubling that was highly statistically significant. Water-extractable nitrogen moved in the opposite direction, declining by 19 percent, possibly because released organic nitrogen was immobilized by microbes or sorbed onto mineral surfaces. Mineralizable carbon showed a smaller but significant increase of about 6 percent, consistent with the well-known Birch effect, in which rewetting dry soil mobilizes previously protected substrates and triggers a pulse of microbial respiration. Only POX-C remained essentially unchanged, with no significant overall shift, indicating that this indicator is robust to the effects of multi-year storage. Notably, the storage effects were consistent across all three cover crop treatments, suggesting that the composition of plant residues did not buffer or amplify the distortion.

The timing of the changes offered additional clues. Storage effects on WEOC were stronger in soils collected in 2020 and 2021 than in those collected in 2019. The researchers speculate that air-drying itself introduces a strong immediate artifact on WEOC, and that this artifact gradually diminishes over time. Samples collected in 2019 were first analyzed in 2022, after an initial three-year storage period, whereas the 2021 samples experienced only a short interval between collection and first analysis, making them more susceptible to the immediate effects of drying. For mineralizable carbon, the pattern reversed: recent samples showed positive storage effects while older samples showed slight declines, hinting that prolonged storage may alter microbial communities in ways that reduce their capacity to metabolize organic substrates upon rewetting, a phenomenon previously observed in soils recovering from extended drought.

To test whether these effects were unique to one sandy Florida soil, the team conducted a Multi-site Validation Experiment using archived soils from four long-term agricultural studies across the Midwest: the W.K. Kellogg Biological Station Long-Term Ecological Research site and the Great Lakes Bioenergy Research Center in Michigan, the Northern Great Plains Long-Term Agroecological Research site in North Dakota, and the Eastern Nebraska Extension and Education Center in Nebraska. These sites spanned certified organic cropping systems, perennial switchgrass monocultures, fallow treatments following spring wheat, and no-till corn systems with stover retention, on soils ranging from loamy Hapludalfs to silt loam Mollisols. The comparison revealed a consistent signature. At every one of the five studies, WEOC declined significantly over time, with correlation coefficients between minus 0.60 and minus 0.84, while the corresponding trends in bulk soil organic carbon or soil organic matter varied independently, increasing at one site, declining marginally at another, and remaining stable elsewhere.

That divergence is the smoking gun. Because bulk carbon is known to be robust against storage, a labile pool whose trajectory departs from bulk carbon across multiple independent sites points to an additional process beyond genuine soil change. The Controlled Storage Experiment supplies the explanation: prolonged storage inflates WEOC, so the oldest archived samples, which waited longest before analysis, carry the largest storage-induced increases. When those inflated values are plotted against sampling date, the result is an apparent decline in WEOC over time that reflects shelf time rather than soil biology. Although the observational nature of the multi-site data cannot fully separate storage artifacts from every other factor, the consistency of the pattern across such different climates, textures, and management systems strongly supports the conclusion that storage history can substantially bias retrospective analyses of archived soils.

The implications ripple outward. Soil archives are treasured resources, allowing scientists to apply newly developed indicators to samples collected decades ago and to track long-term impacts of management practices. But if labile carbon indicators are systematically inflated by storage, temporal comparisons built on archived samples may confuse artifacts with real management effects, undermining the very decisions soil health testing is meant to inform. The authors recommend that storage-sensitive pools like WEOC be analyzed immediately upon collection, and suggest that controlled-temperature drying or cold storage deserve exploration for labile indicators. They also caution against using WEOC to infer long-term trajectories in carbon storage or composition. In contrast, the stability of POX-C across multi-year storage underscores its potential for retrospective studies, offering new input into ongoing debates about that indicator’s interpretability. As soil health datasets increasingly span years and decades, the study delivers a clear message: a soil sample is not a frozen moment in time, and the shelf it sits on can rewrite its story.

Subject of Research: Effects of long-term air-dried storage on soil health indicators, particularly water-extractable organic carbon

Article Title: Long-term storage of air-dried samples compromises water-extractable organic carbon as a soil health indicator

Article References: Long-term storage of air-dried samples compromises water-extractable organic carbon as a soil health indicator. (n.d.). https://doi.org/10.5194/soil-12-937-2026

Image Credits: AI Generated

DOI: 10.5194/soil-12-937-2026

Keywords: soil health, water-extractable organic carbon, soil archives, soil organic carbon, cover crops, permanganate-oxidizable carbon, mineralizable carbon, sample storage, agricultural research, soil testing, Long-term, storage