
The underground process of trees is as complicated as the above-ground process. In fact, they are more complex, probably because soil, as solid media, is more dense and richer than atmospheric density as gas media. About 200 billion tons of organic carbon are stored in the soil, which is more than the sum of the carbon reservoirs of the atmosphere and vegetation.
Trees absorb C and assign it to their roots, playing a crucial role in the underground process. Trees can store carbon at the roots, allocate carbon for root growth, release carbon in the form of root secretions, and finally deposit carbon in the form of litter.
Soil microorganisms, including bacteria, archaea, fungi, protozoa and viruses, are an important part of the underground carbon sink in trees. They regulate organic C stored in the soil and released back into the atmosphere according to soil chemistry, temperature and humidity.
The place where tree roots interact with soil microorganisms, called rhizosphere , is the center of the underground carbon process. Rhizosphere processes account for 1/3 of the total amount of C and N mineralization in temperate forest soil.
To quantify the degree of participation of soil microorganisms in the C cycle, we first need to quantify the amount of C allocated underground by evaluating the C flux of the entire tree. A powerful tool for studying ecosystem flux is stable C isotope labeling. By measuring discrete fluxes, the allocation of C can be tracked at a higher time resolution. C labeling includes the enrichment or consumption of δ13CO2 gas, however, labeling can become expensive, especially at the canopy scale. Therefore, the selected method usually uses mildly consumed δ13CO2 gas.
Another method to evaluate C allocation in trees is to use the carbon mass balance method:

Where C source is assimilation (A), C sink is the balance between respiration (R), growth (Gr), waste generation (L), root secretions (Ex), and C storage (S) and reserve consumption (C).
On this issue, Hikino et al. (2022) combined rainwater removal experiments on mature trees in the wild and 13C marking experiments, representing a new achievement in tree ecological physiology research. Furthermore, this large-scale application of experimental methods facilitates the characterization of new mechanisms in tree ecology, such as the current underground allocation of photoassimilars covering only half of the C used for fine root growth.
Another illuminating example in the same study is that despite drought, root secretions have persistent C allocation. This finding confirms what is observed from more drought-invasive forests. Furthermore, the study showed that after the drought release, 90% of C assigned to the root exudation comes from new photoassimilators, compared with 65% under controlled conditions. Increased underground C allocation may increase soil C sequestration, resulting in a net removal of CO2 in the atmosphere.
Soil organic C (SOC) is mainly formed by (1) aboveground litter, (2) root litter and (3) net rhizodeposition of plant rhizodeposition (net rhizodeposition; organic C left after microorganism utilization and decomposition in rhizodes; Figure 1). Mycorrhizal fungi are one of the highest contributors to organic C in soil and may play an important role in adding secretions and residues to net rhizosphere sediments.
SOC can be divided into granular (POC) and mineral-related (MAOC) forms, where POC is more easily decomposed by microorganisms, and MAOC shows higher durability. Rhizosphere sediments have the highest MAOC formation efficiency, while root biomass input has the highest POC formation efficiency.
A recent research institute confirmed that residues from saprophytic and mycorrhizal fungi contribute more to MAOC than plant residues. Although root secretions have "exciting effects" (increased microbial activity leads to the instability of existing C library), the rhizosphere environment of the forest helps stabilize root secretions, resulting in long-term storage due to the accumulation of microbial biomass residues.

▲ Figure | The fate of underground carbon. A tree transfers carbon below the ground through root growth (Gr), carbon storage (St), root secretions (Ex), and leaf and root litter. Microorganisms, such as bacteria, mycorrhizal fungi and saprophytic fungi, utilize C from Ex and litter.Carbon dioxide is being emitted due to the respiration of these microorganisms (Rhetero=heterotrophic respiration) and the respiration of the root system (Rauto=autotrophic respiration). SOC (soil organic C) forms in soil as POC (particle organic C) and MAOC (mineral-related organic C). POC is mainly formed from litter residues, while MAOC is mainly formed from rhizosphere sediments. This diagram is created using BioRender.
C allocation of trees is crucial for C sequestration prediction in soil. Abiotic factors such as increased drought, increased atmospheric CO2 and warming climate are already affecting the carbon flux of trees.
Root secretions increase during seasonal drought in mixed Mediterranean forests. Similarly, in a greenhouse study, oak trees showed an increase in root secretion rate under drought treatment. Under eCO2 conditions, the C assimilation of pine trees is enhanced and the root secretions are increased. In a long-term soil warming experiment conducted in Harvard Forest, C losses at different stages due to warming over 26 years were found. Three-quarters of the total C loss in the period occurred in the first nine years due to frequent microbial activity; after 6 years of microbial community reorganization (the C loss in the warmed plot did not increase), soil respiration increased again until another peak was reached. Overall, the dynamics of SOC are complex because of its sensitivity to warming and dynamic soil microbial communities.
In the precipitation exclusion-13C marking experiment, drought promoted increased growth of fine roots; however, root secretions or ectomycorrhizals did not change. Because more fragile POCs may form first (Figure 1), an increase in root biomass may not contribute to a more stable underground C library.
Although rhizosphere processes are crucial for stable SOC formation, few studies have been conducted on root secretions and their role in SOC formation and stabilization. There are very few field studies on root secretions now, partly due to the technical challenges of collecting root secretions in natural ecosystems. In addition, there is little impact on rhizospheric processes studying global changes. On this issue, Hikino et al. (2022) led the way forward. Nevertheless, research must be accelerated considering the unknown impact of global changes on C allocation in the rhizosphere.
Finally, soil is not the final territory. A large portion of the tree's root system can be found in the rocky layer, where the root system can obtain additional minerals and water resources. Therefore, future research should not only consider the carbon sequestration process in the soil, but also the carbon sequestration process in the rock layer, and the acquisition and research of rock layers in the natural environment is more challenging.