At the Root of It: Forests, Soils, and Carbon Storage

A deep dive into the vital role forest soils have in storing carbon.

Forests are most often discussed as carbon stores in terms of trees, timber and above-ground biomass. Trunks, branches and canopies are visible, measurable and intuitive symbols of carbon sequestration. Yet this focus misses where the majority of forest carbon is actually held. In temperate forests, including those in the UK, research consistently shows that around 70–80% of total forest carbon is stored below ground in soils, with only the remaining fraction held in living biomass. At the global scale, soils contain an estimated 1,500 petagrams of organic carbon – around two to three times more carbon than is currently present in the atmosphere. These figures alone underline why soils must sit at the centre of credible forest-based climate strategies. 

Soil carbon represents a balance between inputs of organic matter and losses through decomposition, leaching and erosion. Carbon enters forest soils through multiple pathways. Leaf litter is the most obvious route, but it is often rapidly decomposed, particularly in productive systems, returning much of its carbon to the atmosphere within relatively short timeframes. By contrast, root-derived carbon provides a more persistent pathway. Fine roots grow, die and regrow continuously, delivering organic matter directly into the soil profile, whilst roots also release a wide range of carbon-rich compounds known as exudates that fuel soil microbial communities. These below-ground inputs, together with the turnover of soil organisms, are now recognised as major contributors to long-term soil organic carbon. 

Where carbon is stored within the soil profile is just as important as how much enters the system. Carbon held in surface litter and organic horizons is generally more vulnerable to disturbance, fire and rapid decomposition. More stable carbon is typically found deeper in mineral soils, where organic compounds bind to soil minerals or are physically protected within soil aggregates, forming what is known as mineral-associated organic matter. This form of carbon can persist for decades to centuries. Root inputs and microbial processing are particularly effective at building these stable pools, meaning that two forests with similar amounts of above-ground biomass can differ substantially in how much long-lived carbon they store in soils. 

Mycorrhizal fungi play a central role in this hidden carbon pathway. Most terrestrial plants form symbiotic relationships with mycorrhizal fungi, transferring a proportion of their photosynthetically fixed carbon below ground in exchange for nutrients (such as ammonium and phosphates) and water. This carbon is incorporated into fungal biomass, released into the soil through fungal exudates and returned via the turnover of fungal tissues. Experimental studies using carbon isotopes have demonstrated that these fungal networks can even transfer carbon between plants. In British Columbia, for example, carbon isotope studies showed that net carbon transfer from paper birch to Douglas fir seedlings under shaded conditions can improve fir survival. As paper birch is not native to British Columbia, there is the potential for advocation for its removal - yet removal of these trees would likely present greater risks to beneficial below-ground carbon and nutrient flows than benefits.  Such findings therefore challenge conventional management practices that favour removal of all ‘non-target’ species, illustrating how below-ground interactions can influence both carbon dynamics and forest regeneration.  

Underground fungi © RBG Kew

Not all mycorrhizae are equal:

Mycorrhizal fungi, such as those in the paper birch and Douglas fir example above, come in a wide range of forms. The two most prevalent types of mycorrhizae are known as arbuscular mycorrhizae and ectomycorrhizae. Arbuscular mycorrhizae are found in over 80% of known plant species and are characterised by their ability to penetrate directly into root cells, forming branching structures (arbuscules) where nutrients are transferred. Ectomycorrhizae, on the other hand, are primarily associated with temperate forest trees such as pines, and form sheaths around root tips, growing between cells without penetrating them.  

This is important because the type of mycorrhiza a plant is associated with affects how and where carbon is stored in soils. Forests dominated by ectomycorrhizal tree species often accumulate large carbon stocks in surface organic layers, partly due to slower litter decomposition. In contrast, arbuscular mycorrhizal species are more frequently associated with greater carbon storage in deeper mineral soils. Global analyses suggest that mineral soil carbon stocks are on average around 9–10% higher under arbuscular mycorrhizal trees, while forest floor carbon stocks are often higher under ectomycorrhizal species.  

Despite these insights, identifying which tree species or forest types are ‘best’ for soil carbon storage remains challenging. Below-ground processes are difficult to measure, spatially variable and are slow to respond to change. Root turnover, microbial activity and carbon stabilisation can differ markedly between sites, even under similar vegetation. Latitudinal patterns further complicate the picture: in boreal forests, up to 65% of gross primary production may be allocated below ground, compared with around 30% in tropical forests. These differences mean that results from one region cannot be easily transferred to another, and that broad generalisations about species performance are rarely robust. 

Tree species identity and diversity can also influence soil carbon outcomes, even if their effects are highly context dependent. Meta-analyses indicate that afforestation of former croplands can increase soil carbon stocks by around 25% under broadleaf species, compared with approximately 2% under conifers over two to three decades. Conversely, coniferous forests often accumulate greater carbon stocks in forest floor layers, particularly in boreal and nutrient-poor environments, due to slower litter decomposition. However, this surface-stored carbon is more vulnerable to disturbance and climate warming. Evidence that higher tree species diversity consistently enhances soil carbon is mixed, with species identity and environmental conditions often exerting a stronger influence than diversity alone. 

Nutrient availability adds a further layer of complexity. Nitrogen fertilisation has been shown to increase mineral soil carbon stocks by approximately 20–25% in temperate forests, largely by suppressing microbial decomposition rather than increasing litter inputs. In some studies, increases in soil carbon under nitrogen fertilisation have far exceeded gains in tree biomass carbon. However, these benefits must be considered alongside increased nitrous oxide emissions and the substantial carbon costs associated with fertiliser manufacture, reinforcing the need to evaluate the full carbon balance of interventions. 

Forest in Madagascar © RBG Kew

Forest Management and Its Role in Soil Carbon  

Forest management practices often exert an even stronger and more immediate influence on soil carbon than species choice alone. Intensive harvesting, whole-tree removal, short rotation lengths and heavy soil disturbance consistently reduce soil carbon stocks, particularly in surface and upper mineral soils. In contrast, management approaches that minimise disturbance – such as longer rotations, retention forestry and continuous cover systems – are more likely to maintain soil structure, root systems and fungal networks that underpin carbon stability. 

Peatland forests are a particularly interesting example and display both the risks and opportunities associated with forest management decisions. Drainage of peatlands lowers the water table, exposing organic matter to oxygen and thus accelerating microbial decomposition. One study, (Joosten, 2009), showed that, globally, carbon dioxide emissions from drained peatlands increased by more than 20% between 1990 and 2008. Rewetting peatland can substantially reduce emissions and help restore these habitats, yet full recovery of peatland carbon sink function can take at least 8 to 10 years, highlighting the long-term consequences of short-term management choices. 

Looking ahead, improving forest-based climate mitigation depends on a better understanding of not only how much carbon enters soils, but what determines its persistence. Soil carbon is only climatically meaningful if it remains stored over long timeframes. Achieving this requires long-term monitoring and improved modelling that link management practices with microbial processes, soil structure and carbon stability. 

Research at Royal Botanic Gardens, Kew’s Wakehurst site is contributing to this effort by integrating above- and below-ground measurements. At Wakehurst, landscape-scale research combines soil carbon stocks, monitoring of soil carbon dioxide emissions, DNA-based characterisation of fungal communities and high-resolution remote sensing (LIDAR) of forest structure. These approaches reflect a growing recognition that understanding soils – and the complex biological processes within them – is essential if forests are to deliver durable and credible climate benefits. 

Ultimately, forestry for climate mitigation is not simply about planting more trees or maximising growth rates. The most secure carbon is stored slowly, below ground, shaped by roots, fungi, soil minerals and management decisions. By recognising and managing forests in ways that integrate above- and below-ground systems we can help boost carbon storage in forest systems – changing soil carbon from an overlooked component of forest carbon budgets to a central pillar of long-term climate strategy. 

Key References: 

Mayer, M. et al. (2020) ‘Tamm Review: Influence of forest management activities on soil organic carbon stocks: A knowledge synthesis’, Forest Ecology and Management, 466, p. 118127. Available at: https://doi.org/10.1016/j.foreco.2020.118127. 

Simard, S. et al. (1997) ‘Net transfer of C between ectomycorrhizal tree species in the field’, Nature, 388, pp. 579–582. Available at: https://doi.org/10.1038/41557. 

Mäkipää, R. et al. (2023) ‘How does management affect soil C sequestration and greenhouse gas fluxes in boreal and temperate forests? – A review’, Forest Ecology and Management, 529, p. 120637. Available at: https://doi.org/10.1016/j.foreco.2022.120637. 

Keller, A.B. et al. (2021) ‘Root-derived inputs are major contributors to soil carbon in temperate forests, but vary by mycorrhizal type’, Ecology Letters, 24(4), pp. 626–635. Available at: https://doi.org/10.1111/ele.13651. 

Peng, Y. et al. (2020) ‘Tree species effects on topsoil carbon stock and concentration are mediated by tree species type, mycorrhizal association, and N-fixing ability at the global scale’, Forest Ecology and Management, 478, p. 118510. Available at: https://doi.org/10.1016/j.foreco.2020.118510. 

Hijri, M. and Bâ, A. (2023) ‘Editorial: Mycorrhizal fungi and plants in terrestrial ecosystems, volume II’, Frontiers in Plant Science, 14. Available at: https://doi.org/10.3389/fpls.2023.1180884.