For decades, mechanistic soil organic carbon models such as MIMICS, Millennial, and MES-C have represented decomposition as an almost exclusively microbial affair. Litter falling to the ground is split into metabolic and structural pools, digested by microbial functional groups, and routed into dissolved carbon, particulate organic carbon, and mineral-associated organic carbon. Termites, despite being among the most abundant detritivores on Earth and occupying roughly 80 million square kilometers of habitat concentrated in the tropics and subtropics, have been entirely absent from these equations. The omission matters because field evidence has long shown that termites dominate deadwood breakdown in tropical forests, with exclusion experiments suggesting they account for 58 to 64 percent of wood decomposition in tropical rainforests.

The new framework extends the MES-C model by adding an explicit coarse woody debris pool and a parallel, termite-mediated decomposition pathway alongside the familiar microbial one. Three feeding guilds are represented: xylophagous termites that consume wood, fungus-growing termites that harvest plant residues through cultivated fungal combs, and soil-feeding termites that process partially decomposed organic matter. Carbon ingested by termites is partitioned among respiratory carbon dioxide, methane production, growth of termite biomass, and excreted residues. Two of those residues — frass, the finely processed and nutrient-rich waste that decomposes quickly, and necromass, the chitin-rich bodies of dead termites that break down more slowly — re-enter the soil as inputs to microbial and mineral-associated carbon pools.

A crucial subtlety in the model is the fate of termite methane. Between 20 and 80 percent of the methane produced inside termite colonies may be oxidized by methanotrophic microbes within mounds and galleries before it ever reaches the atmosphere, making net emissions highly sensitive to mound architecture, colony density, and soil conditions. The module therefore couples termite metabolism directly to microbial methane cycling, allowing methane that is consumed within mounds to still contribute to microbial biomass and carbon dioxide release. This level of mechanistic detail, the authors argue, is what allows termite processes to be represented without breaking the mass balance that global carbon models depend on.

To run the model globally, the team linked termite biomass to ecosystem productivity, using gross primary productivity as a proxy for the resources available to termite colonies. Potential habitat was constrained by a thermal mask that excludes regions where monthly minimum temperatures fall below minus eight degrees Celsius, consistent with observed limits on termite survival. Simulations across global vegetation classes reveal a striking biogeographic pattern. Tropical evergreen forests, with the highest mean termite biomass density of 8.9 grams of carbon per square meter, dominate every flux: they account for 733.4 teragrams of carbon consumed, 404.1 teragrams released as carbon dioxide, and 3.7 teragrams emitted as methane each year. Savannas, with somewhat lower biomass but vast extent, follow close behind.

Globally, the model estimates that termites release 864.7 teragrams of carbon per year as carbon dioxide — equivalent to roughly 3,171 teragrams of carbon dioxide, or about 3.2 to 4.7 percent of total global soil respiration — and 7.9 teragrams of carbon as methane, a figure consistent with the Global Methane Budget’s estimate that termites contribute around 1.8 percent of global methane emissions. At the same time, termites transfer 689.3 teragrams of carbon per year into labile and mineral-associated soil organic carbon pools, with 209.1 teragrams going to the more stable mineral-associated fraction. The carbon dioxide flux should not be read as an extra source on top of microbial respiration, the authors stress; rather, termites redirect a share of decomposition through a faunal route with different climate sensitivities.

That difference in sensitivity is where the climate implications become sharp. Microbial decomposition rates typically rise by a factor of about 2.2 to 4.6 for every ten degrees of warming, but empirical work on termites suggests their feeding and foraging rates can increase several-fold over the same range, corresponding to an effective behavioral temperature sensitivity close to a Q10 of seven. In other words, warming may disproportionately accelerate termite-driven decomposition and trace-gas fluxes. Model-based reconstructions indicate that termite methane emissions already increased during the twentieth century and may continue rising under warming and land-use change, positioning these insects as a potential amplifier of carbon-climate feedbacks in precisely the tropical and seasonally dry regions most vulnerable to climate shifts.

Sensitivity analysis identified which parameters matter most for taming the uncertainty. Termite ingestion capacity and biomass emerged as the dominant controls on flux magnitude, while the partitioning of consumed carbon among respiration, residues, and biomass governs the ultimate fate of processed carbon. Methane-related parameters, by contrast, had comparatively weak effects. Most of the uncertainty in the headline numbers stems from the wide range of ingestion rates reported in the literature, which points directly to the field measurements the authors say are needed next: biome-resolved termite biomass surveys, guild-specific ingestion rates measured under natural conditions, and better constraints on how termites split carbon between gas, waste, and body tissue.

The framework is deliberately minimal, and the authors are candid about its limits. Termite biomass is treated as uniform within grid cells even though colonies create intense local hotspots of consumption and gas exchange; interactions with competing or cooperating free-living microbes are not resolved; and the modeled inputs to mineral-associated carbon represent potential additions rather than guaranteed long-term storage, since mineral surface area and oxide content are not simulated. Evaluation also leans on a single global biomass synthesis from 1996 built on sparse observations. Even so, the consistency of the model’s spatial patterns with known termite biogeography, and the agreement of its flux magnitudes with independent empirical estimates, suggest the dominant controls are plausibly captured.

What the study ultimately offers is a template for a more complete theory of the terrestrial carbon cycle — one that moves beyond the microbe-centric paradigm that has shaped soil modeling for a generation. Because termites process about twelve percent of global coarse woody debris, models that ignore them may misjudge how long woody carbon persists in low-latitude forests, and because their activity peaks in warm, water-limited African savannas where microbial decomposition slows, they may buffer ecosystems against drought in ways current models cannot see. The authors propose that future frameworks aggregate soil fauna by shared substrate use and carbon effects, while keeping termites distinct, since no other faunal group combines massive litter processing with a direct biological source of methane. Embedding these six-legged engineers in next-generation Earth system models, the team concludes, is essential for predicting how the planet’s soils will store — or release — carbon as the climate warms.

Subject of Research: Integrating termite-mediated decomposition into global soil organic carbon and greenhouse gas models

Article Title: Ideas and perspectives: Beyond microbes – integrating termites into global soil carbon cycling models

Article References: Farooq, U., Pasut, C., Wang, Y.-P., Zanne, A. E., Flores-Moreno, H., Wijas, B. J., Forrester, D. I., England, J. R., Macdonald, B., Brown, Z. A., & Karunaratne, S. (2026). Ideas and perspectives: Beyond microbes – integrating termites into global soil carbon cycling models. Biogeosciences, 23(19), 7029-7041. https://doi.org/10.5194/bg-23-7029-2026

Image Credits: AI Generated

DOI: 10.5194/bg-23-7029-2026

Keywords: termites, soil organic carbon, carbon cycle, decomposition, methane, carbon dioxide, tropical ecosystems, savannas, Earth system models, climate feedbacks, biogeosciences, detritivores

The research, led by Gaoji Zhang and Hongyi Liu of Nanjing Forestry University together with colleagues at Nanjing Xiaozhuang University, set out to answer a deceptively simple question: what does a soil-dwelling millipede actually eat, and does it choose? Understanding diet is fundamental to understanding how animals adapt to their environments, yet for cryptic soil invertebrates, direct observation is nearly impossible. The team therefore turned to molecular methods, combining DNA barcoding of gut contents with detailed surveys of the surrounding plant communities, soil properties, and the elemental composition of available leaf litter. This multi-pronged approach allowed the researchers to compare what the millipedes ate against what was actually available to them, a crucial step in distinguishing true dietary selection from mere opportunistic consumption.

The results were striking in their breadth. Across the sampled populations, S. bungii was found to have consumed litter from 76 plant genera spanning 48 families and 30 orders. This extraordinary dietary range suggests that the species functions as a generalist at the population level, capable of exploiting a wide variety of plant material. Population-level niche width, a statistical measure of how broad a population’s resource use is, was correspondingly large, registering 15.32 at the family level and 16.46 at the genus level. Such values indicate that collectively, these millipede populations draw on a deep pool of plant diversity, a flexibility that may underpin the species’ success in habitats heavily altered by human activity.

Yet the broad population-level diet masked considerable variation between individual mountain populations. Analysis of similarity, a non-parametric multivariate technique known as ANOSIM, revealed significant differences in dietary composition among millipedes collected from different mountains. In other words, each population has its own distinct dietary signature, shaped largely by the local plant community on offer. Niche overlap analysis reinforced this picture: most mountain populations showed little dietary overlap, with Ojk values below 0.6, the threshold typically interpreted as meaningful separation. The one notable exception involved populations from Jiangjun Mountain and Fang Mountain, which overlapped at Ojk = 0.65, hinting at either similar local vegetation or genuine sharing of preferred resources between these two sites.

Perhaps the most consequential finding concerns habitat fragmentation itself. The researchers measured fragmentation using two standard landscape metrics: patch density (PD), which counts habitat patches per unit area, and edge density (ED), which quantifies the amount of forest edge relative to area. Both were strongly correlated with the class-level trophic niche width of the millipedes, with patch density showing a Spearman correlation of ρ = 0.93 (p = 0.007) and edge density ρ = 0.79 (p = 0.048). In practical terms, as forests become more chopped up and edge-dominated, the dietary repertoire of S. bungii appears to shift. This is one of the clearest demonstrations to date that the spatial configuration of urban habitat, not just its total area, can leave a measurable imprint on the feeding ecology of a soil animal.

The study also uncovered pronounced selectivity. When the researchers compared consumption against availability using null-model approaches, S. bungii exhibited a positive selection tendency toward three plant genera: Cinnamomum, the genus that includes camphor and cinnamon trees; Carya, the hickories; and Ailanthus, the tree of heaven. Conversely, the millipedes showed a negative selection tendency toward Trachelospermum, a genus of woody vines and lianas common in Chinese forest understories. The avoidance of Trachelospermum is intriguing from a biochemical standpoint, as the genus is known to contain a rich arsenal of phytochemical compounds, some of which may render its litter unpalatable or even deterrent to detritivores. Preference for the favored genera, by contrast, may reflect more favorable litter chemistry, softer tissues, or higher nutritional returns per bite.

Elemental analysis added another layer to the story. The team found that S. bungii tended to consume plants with high calcium content, with a Spearman correlation of ρ = 0.83 and a p-value of 0.058, just shy of conventional significance thresholds but strongly suggestive of a pattern. Calcium matters enormously to millipedes. Their calcified exoskeletons, which must be built and repeatedly rebuilt through successive molts, demand substantial calcium reserves drawn directly from their food. A millipede that preferentially grazes calcium-rich litter is, in effect, mining the leaf fall for skeletal raw material. This finding connects the feeding behavior of a humble arthropod to the broader biogeochemistry of the forest floor, where leaf elemental composition varies widely among tree species and shapes the entire decomposer food web.

The implications extend beyond millipede biology. Detritivores such as S. bungii are engine rooms of soil nutrient cycling: by fragmenting litter and converting it into faeces, they accelerate organic matter turnover, release nutrients locked in dead leaves, and create conditions that favor microbial decomposition. Their bioturbation, the mixing of organic and mineral soil layers as they burrow, further alters soil structure and water movement. If urban fragmentation changes what these animals eat, it may also change how efficiently they perform these ecosystem services. A population forced onto a narrower or lower-quality diet could process litter more slowly, with knock-on effects for soil fertility in the very green spaces that cities depend on for cooling, flood regulation, and recreation.

The study also fits into a growing body of work on urban soil biodiversity. Previous research by some of the same authors showed that urbanization affects millipede gut microbiota by impeding host gene flow, and other studies have found that even small urban forest fragments can maintain complex food webs of litter-dwelling arthropods. Taken together, these findings challenge the assumption that fragmented urban habitats are ecological deserts. Instead, they suggest that the dietary flexibility of generalist detritivores, combined with thoughtful conservation of plant diversity within fragments, could sustain functioning decomposer communities even in heavily built-up landscapes. For urban planners, the message is concrete: preserving a diverse mix of native trees, including calcium-rich species that detritivores favor, may be as important for soil health as simply preserving green space.

Methodologically, the study showcases the power of DNA barcoding for diet analysis in organisms that cannot be observed feeding directly. By sequencing plant DNA recovered from millipede gut contents and matching it against reference libraries, the researchers achieved a taxonomic resolution that traditional gut-content microscopy could never match. All raw sequences from the study were deposited in the NCBI Sequence Read Archive under accession number PRJNA1182097, ensuring that other researchers can verify and build upon the work. As cities continue to expand and fragment the forests around them, studies like this one make clear that the fate of above-ground biodiversity and the hidden economy of the soil are tightly intertwined, and that even a millipede’s dinner menu can tell us whether an urban forest is truly thriving.

Subject of Research: Dietary selection and trophic niche of the millipede Spirobolus bungii in fragmented urban forests

Article Title: Dietary selection of millipedes (Spirobolus bungii) in fragmented urban forests

Article References: Dietary selection of millipedes (Spirobolus bungii) in fragmented urban forests. (n.d.). https://doi.org/10.1007/s11104-026-09152-9

Image Credits: AI Generated

DOI: 10.1007/s11104-026-09152-9

Keywords: millipedes, Spirobolus bungii, DNA barcoding, diet selection, habitat fragmentation, urban forests, soil biodiversity, detritivores, leaf litter, calcium, trophic niche, nutrient cycling