Putting plant material into soil does not place all of its carbon into long-term storage. Some is respired, some enters living organisms, some remains as fragments, and some becomes associated with minerals. The path matters as much as the input.
This is why soil carbon can be difficult to discuss clearly. A stock is measured in one moment, but that stock emerges from continuing arrivals, transformations, and losses. Microbes participate on both sides of the balance. They release carbon through respiration while also producing biomass and residues that can contribute to organic matter.
The resulting science is more useful than the image of soil as a passive container. It also gives us reasons to be cautious about any proposal that turns one microbial measurement or a map of theoretical capacity into a promise of durable carbon storage.
Efficiency describes a division, not the whole system
Microbial carbon use efficiency describes the proportion of assimilated carbon allocated to growth rather than respiration. In a major 2023 analysis, Tao and colleagues combined published measurements, global soil profiles, and a process-based modeling framework. Their study of carbon use efficiency and soil carbon found a positive relationship and identified efficiency as an influential control within the model.
An essential qualification is that much of the global inference was obtained by fitting model parameters to observed soil carbon profiles. It was not a worldwide experiment in which efficiency was independently changed while everything else remained constant. The model used simplifying assumptions, including steady-state conditions for the profile-level carbon balance.
This does not make the inference uninformative. It makes its role clear: a way of explaining observed patterns and testing how processes might contribute, with empirical measurements providing an additional comparison. The modeled ranking of controls belongs to that framework, its data, and its assumptions.
There is also a basic distinction between a ratio and an amount. Two microbial communities can allocate the same fraction of consumed carbon to growth while processing different total quantities. If the question concerns how much new biomass or residue is produced, efficiency alone cannot supply the answer.
A developing disagreement is useful evidence
A 2026 study comparing microbial growth and efficiency assembled isotope-based observations and compared them with land-surface models. Its reported results found microbial growth rate more strongly associated with soil carbon than efficiency, while soil properties and climate explained comparable or greater variation. The models also differed from observations in their representation of controls.
That result should not be flattened into a contest with one permanent winner. The studies use different observations, model structures, and ways of representing microbial activity. Their disagreement identifies a substantive question: which measurements capture the processes that govern carbon accumulation across scales?
For anyone interpreting a strong headline, the consequence is straightforward. A promising physiological ratio should remain connected to growth, turnover, substrate availability, and the physical environment. Research can establish a useful association without establishing a universally manipulable control knob.
This is especially important when moving from explanation to intervention. Explaining why carbon-rich soils differ from carbon-poor soils is not the same as demonstrating that changing one property in the latter will reproduce the former. Several historical and environmental causes may have produced the observed combination.
Mineral association is protection with a timescale
Minerals can limit microbial access to organic matter, but “mineral-associated” does not mean permanently locked away. A laboratory and field investigation used carbon and nitrogen isotope labels to follow microbial residues in grassland soils. The necromass persistence experiment found rapid association with minerals, followed by substantial loss of the associated labeled carbon during the following months. Carbon and nitrogen did not persist in the same proportions.
The field treatment used added bacterial residues in selected agricultural grasslands. It did not represent every microbial taxon, mineral surface, or natural mode of cell death. The investigators explicitly noted that microbes dying directly on mineral surfaces might follow different pathways from the introduced suspension.
Within that scope, the experiment is a strong warning against equating early capture with long-term retention. A measurement after days and a measurement after months describe different parts of the trajectory. Neither automatically answers what remains after decades.
The distinction also helps explain why residues can be important without being inert. Material can enter a comparatively protected pool while remaining available for later transformations. Persistence is an outcome of conditions and interactions, not simply a label attached to the molecule's origin.
Capacity is not an achievable schedule
Georgiou and colleagues synthesized mineral-associated carbon observations from more than a thousand soil profiles and estimated mineralogical storage capacity. Their global analysis of stocks and capacity showed that mineral type matters and that many soils contain less mineral-associated carbon than the inferred capacity of their minerals.
The authors were careful to distinguish that capacity from what can be achieved in practice. Carbon inputs, nutrient constraints, environmental conditions, management, and time all affect whether additional storage occurs. A deficit on a capacity map is therefore not a timetable for removing carbon dioxide from the atmosphere.
One can appreciate the potential without spending it twice in the imagination. A mineralogical upper estimate helps locate hypotheses and constrain models. A measured increase over an intervention period demonstrates a different kind of result. Evidence of persistence adds another requirement. These quantities should be reported separately so that an attractive theoretical total does not stand in for observed, sustained change.
The common thread through these studies is that carbon storage emerges from a living system embedded in a physical one. Plants supply material, microbes transform it, and minerals influence access and retention. The processes continue after the first measurement.
A better soil-carbon story therefore follows a trajectory: what entered, what was transformed, what remained, and for how long under what conditions. Soil's importance does not depend on making it sound simple. Its continuing work of making and unmaking organic matter is precisely what makes it ecologically consequential.
A thought worth keeping?
