Forest ecosystems, the largest terrestrial reservoir of carbon (C) in the global C cycle, are crucial in regulating the Earth’s climate (Pan et al., 2011; IPCC, 2022). Forest ecosystems absorb carbon dioxide (CO₂) from the atmosphere through primary production and store the absorbed C as live (plant biomass) and dead (soil) organic material over extended periods (Binkley and Fisher, 2020). The global gross C sink of forests is 3.6 Pg C y⁻¹ (Pan et al., 2024), which represents roughly half of the annual global CO₂ emission from fossil fuel combustion (Harris et al., 2021; Pan et al., 2024). As climate change risks escalate due to human activities, the potential of forest ecosystems to mitigate climate change through C offsets has gained significant global attention (van Kooten and Johnston, 2016). However, this potential of forest C sequestration remains highly uncertain, mainly due to a lack of quantitative evaluation of C dynamics in forest ecosystems (Pugh et al., 2020). An improved understanding of these C dynamics is essential for maintaining and enhancing the C sink function of forests as a nature-based solution for mitigating climate change.

Plant biomass produced through net primary production (NPP) is allocated to both aboveground (stem, branches, and leaves) and belowground (root crown and lateral roots) biomass. This biomass allocation is a key factor in assessing the C sequestration potential of forest ecosystems. Although biomass allocation patterns reflect plant growth, survival, and adaptation strategies, and are largely influenced by climatic, ecological, and nutrient conditions (Qi et al., 2019; Zhou et al., 2022), studies have consistently shown that plants allocate a significant proportion (>30%) of NPP to the production and growth of fine roots (diameter: <2 mm) belowground (Jackson et al., 1997; Matamala et al., 2003; Fernandez-Tschieder et al., 2024). Fine roots, being a dynamic component of the plant root system, play a crucial role in the forest C dynamics through belowground C input, cycling, and the formation of soil organic matter (SOM) (Trumbore and Gaudinski, 2003). However, knowledge about belowground biomass dynamics is still lacking because unlike aboveground processes, belowground processes are hidden and thus difficult to observe directly.

Many studies based on direct observation have used ingrowth cores, sequential soil coring, minirhizotrons, and litter bags to establish fine root dynamics (biomass and neuromas as well as rates of production, turnover, and decomposition after death) in various forest ecosystems (Gill and Jackson, 2000). Fine root lifetimes vary from days to years (Eissenstat and Yanai, 1997; McCormack et al., 2012; Ding et al., 2019), depending on factors such as soil depth (Burton et al., 2000), root diameter size class (Gill and Jackson, 2000), root branch order (Eissenstat et al., 2000; McCormack et al., 2015), and environmental conditions (Kwatcho Kengdo et al., 2023; Conroy et al., 2025). Fine root lifetimes of a few years are typically estimated based on conventional steady-state mass balance calculations (i.e., by dividing the biomass of fine live roots by the annual rate of fine root production determined through direct observation) (Gill and Jackson, 2000; Tierney and Fahey, 2002; Noguchi et al., 2007). While soil microbes rapidly decompose a significant portion of dead fine roots within months, the remainder is decomposed slowly over years (Sun et al., 2013; Wu et al., 2022). However, because of methodological limitations, these direct observation approaches primarily focus on the most dynamic part of fine roots (i.e., typically apical, ephemeral first-order roots) and struggle to capture the full picture of fine root dynamics, especially for larger, more highly branched fine roots (Guo et al., 2008; Pritchard and Strand, 2008). A key limitation is the difficulty in achieving long-term, continuous, in situ monitoring of root growth and death for the whole root system.

Isotopic approaches, utilizing radiocarbon (¹⁴C) measurements of fine root tissues, offer a valuable tool for evaluating the belowground C allocation processes in plants (Gaudinski et al., 2001; Trumbore and Gaudinski, 2003; Pritchard and Strand, 2008). The ¹⁴C content of CO₂ in the atmosphere nearly doubled in the early 1960s due to the production of ¹⁴C (bomb ¹⁴C) from atmospheric nuclear weapons testing and has declined gradually since the Nuclear Test Ban Treaty in 1963 (Hua et al., 2022). Given this temporal change in the ¹⁴C content of atmospheric CO₂, the ¹⁴C content of roots provides an estimate of how long C has been retained in the plant since the photosynthetic fixation of CO₂ from the atmosphere (Gaudinski et al., 2001). Studies based on ¹⁴C measurements have revealed that the presence of C in fine roots exceeding 10 years in age (Gaudinski et al., 2001; Trumbore and Gaudinski, 2003; Trumbore et al., 2006; Solly et al., 2018).

Applying only ¹⁴C measurements to live fine roots potentially overestimates fine root lifetimes because of two reasons. First, trees can utilize stored C to produce and grow roots (Matamala et al., 2003; Vargas et al., 2009; Sah et al., 2011). Second, isotopic approaches, in contrast to direct observation approaches, tend to focus on larger, higher-branch-order roots, which are more readily collected from soil samples for analysis (Guo et al., 2008; Pritchard and Strand, 2008). Consequently, despite employing various approaches, the dynamics of fine roots—particularly the heterogeneity of the fine root population and its lifetime—remains poorly understood (Gaudinski et al., 2010), representing a significant uncertainty in quantifying forest C dynamics (Pugh et al., 2020; Fernandez-Tschieder et al., 2024). A combination of isotope, ingrowth methods, and fine root classification (Joslin et al., 2006; Solly et al., 2018) can provide a more comprehensive and robust assessment (Gaudinski et al., 2009; Sah et al., 2011).

This study aimed to evaluate the age of C in fine roots using a ¹⁴C-based approach to existing fine roots and newly emerged roots. We collected samples of fine root, including branching fine root systems, in a cool-temperate deciduous broadleaved forest in Japan. The root samples were categorized according to diameter size class (linking with branch order) and live/dead status. Additionally, we captured newly emerged fine roots using ingrowth mesh bags installed in the soil for approximately 3 months during the growing season. The root samples were analyzed for ¹⁴C content to estimate the age of C in fine roots with different morphological properties.

The global ¹⁴C spike resulting from atmospheric nuclear weapons testing, which peaked in the early 1960s, provides a valuable tool for assessing the mean age of root C. The Δ¹⁴C values of existing fine roots in the Japanese beech forest consistently exceeded those of contemporary atmospheric CO₂ (Table 1; Figure 2), indicating that the roots were formed mainly from old (noncurrent-year) photosynthetic products. Using the ¹⁴C signatures and modeling approaches, the mean ages of C in fine roots were estimated to be 5–23, 1–34, and 4–25 years (with an occasional value exceeding 50 years) for live, dead, and bulk roots, respectively (Tables 1, 2). These ¹⁴C-derived estimates of fine root C ages align with findings across biomes and plant types. Gaudinski et al. (2001) reported mean ages of 3–18, 10–18, and 3–18 years for live, dead, and bulk fine roots, respectively, in United States temperate forests. Fine root C ages have also been reported as <1 to >40 years in Amazonian tropical forests (Trumbore et al., 2006), 3–12 years in boreal and hemi boreal pine and spruce forests (Sah et al., 2013), and 3–23 years in European temperate, boreal, and subarctic forests (Solly et al., 2018). Recent work by Kwatcho Kengdo et al. (2023) indicates approximately 6-year turnover times for bulk fine roots in the Austrian Alps. Our observation showed no consistent trend with depth in root C ages. This pattern contrasts with other studies showing an increase in root C age with depth (Gaudinski et al., 2001; Joslin et al., 2006). This is likely due to the characteristics of the Appi forest soil (Andosol) where soil bulk density, clay content, organic matter content are uniformly distributed in the upper 20 cm (Koarashi et al., 2009).

In contrast to the C age observed in the standing stock of fine roots, newly produced (99-day-old) fine roots exhibited a Δ¹⁴C value similar to that of contemporary atmospheric CO₂ (Table 3). This indicates that current-year photosynthetic products are used in the production of new roots in the Japanese beech forest, although the reliance of plants on stored C reserves (e.g., nonstructural carbohydrates) for root production is widely recognized. While Gaudinski et al. (2009) suggested that the production of new roots primarily utilizes stored C, their study of a temperate deciduous oak forest indicated reserves less than 1 year old. Our results align with observations from temperate and tropical forests, where newly emerged roots typically comprise photosynthetic products of age <1–2 years (Gaudinski et al., 2001; Tierney and Fahey, 2002; Trumbore et al., 2006; Vargas et al., 2009; Sah et al., 2011). The use of C reserves in the production of new roots also may depend on the season; in this study, the root collection was conducted in the second half of the growing season (July 30 to November 6), a time when freshly assimilated C reserves are potentially abundant in the tree body (Hilman et al., 2024).

Given the negligible time lag observed between photosynthetic C assimilation and its utilization for root production in the Japanese beech forest, the estimated ages of C in fine roots suggest the following two processes of fine root dynamics. First, plants use current-year photosynthetic products to produce new roots but use older (¹⁴C-enriched) internally stored organic compounds to support root growth throughout their lifespan. Second, some fine roots live for many years. Both of these processes could explain why live fine roots exhibited a higher Δ¹⁴C value than newly emerged roots.

Regarding the first process, individual roots of woody species undergo primary and secondary growth after emergence. During primary root growth, the primary vascular system and cortex develop, forming a structure largely composed of living parenchyma cells that store starch (Hishi, 2007; Gaudinski et al., 2009). Subsequently, secondary growth leads to the development of the secondary vascular system and cork cambium, resulting in a root structure dominated by structural secondary tissues, such as sclerenchyma, rich in lignin and subring (Hishi, 2007). This shift in root tissue composition raises the hypothesis that trees utilize older C reserves during later root development. The observed increase in ¹⁴C-derived age of root C with diameter size class (branch order) in branching root systems (Figure 1 and Table 2) may support this. However, the significant difference in ¹⁴C signature between existing and newly emerged root C implies a disproportionally large allocation of old C reserves after root emergence. Monitoring changes in both ¹⁴C signature and biomass of fine roots throughout the root lifespan is needed to fully understand this allocation. Solly et al. (2018) have recently reported older ¹⁴C ages than chronological (based on annual growth rings) ages of fine roots. In addition, other studies, in contrast to our results, have documented higher C ages even in newly produced roots (up to 10 years), especially for larger-diameter fine roots (Matamala et al., 2003; Sah et al., 2011; Hilman et al., 2024) and those produced following forest disturbance (Vargas et al., 2009). Richardson et al. (2013) identified both rapidly cycling (<1 year) and decade-old nonstructural carbohydrate reserves in temperate forest trees of the northeastern United States. These findings illustrate the capacity of plants to allocate older stored C to root production and growth, especially under specific environmental conditions.

The second process relies on the assumption that plants do not use older stored C for root growth after emergence. If this assumption holds true, the ¹⁴C-derived ages of root C would primarily reflect root lifetimes, suggesting that a proportion of existing fine roots are long-lived. For the Appi forest, the ¹⁴C method estimates fine root turnover times of over 5 years, significantly longer than approximately 1.9 years estimated by a conventional steady-state mass balance calculation using the fine root biomass (438 g m⁻² measured in this study) and production rate (233 ± 35 g m⁻² y⁻¹ evaluated during 2012–2013; Noguchi et al. unpublished data). The apparent discrepancy in fine root lifetime could arise from differences in the fine root populations being assessed between the methods (Trumbore and Gaudinski, 2003; Guo et al., 2008; Pritchard and Strand, 2008; Strand et al., 2008). The ¹⁴C method relies on a mass-based measure (i.e., mean residence time of fine root mass) and can be skewed by the larger, higher-order fine roots, which contain more C and tend to persist longer than smaller, lower-order fine roots. The difficulty in extracting the most ephemeral components of the fine root system from soil samples could also contribute to this bias (Strand et al., 2008). Conversely, the minirhizotron and ingrowth core methods rely on a number-based measure (growing of individual fine roots) and can be biased by the smaller, apical first-order fine roots, which are more numerous and ephemeral (i.e., short-lived) but contain less C compared to higher-order fine roots. Inherent heterogeneity in root longevity has been documented (Tierney and Fahey, 2002; Guo et al., 2008; Gaudinski et al., 2010; Ahrens et al., 2014), likely due to variations in mortality rates and life cycles across different positions on branching root systems (King et al., 2002; Hishi, 2007). Our observation of increasing ¹⁴C-derived age of root C with branch order in branching root systems (Figure 1 and Table 2) may also reflect this inherent variation in root lifespan. Consequently, the ¹⁴C method could tend to overestimate the actual lifetime of fine roots, whereas the minirhizotron and ingrowth core methods could substantially underestimate the actual lifetime (Guo et al., 2008). A conventional steady-state mass balance calculation assuming a 1-year lifetime for the standing stock of live fine roots (equivalent to ~260 g C m⁻²) estimates a belowground NPP share of approximately 32% at the Appi forest. However, this may probably overestimate the belowground C allocation, and consequently, the transfer of C into the soil through fine root dynamics. Again, tracking ¹⁴C signatures of roots throughout root development is crucial to definitely verify the assumption and belowground C allocation processes.

Other processes that could potentially influence the ¹⁴C signature of roots include the uptake of C by roots from organic C sources in the soil (Jones et al., 2009; Näsholm et al., 2009; Solly et al., 2018) and the transfer of plant-fixed C to mycorrhizal fungi, which form symbiotic relationships with the roots (Solly et al., 2018; Hawkins et al., 2023). Regarding the uptake of C by roots, however, the Δ¹⁴C values and their depth distribution for fine roots did not match those observed in SOM (Table 1). At the Appi forest site, Δ¹⁴C values for water-extractable organic C (WEOC) in the surface soil were measured in 2009 (102‰, 77‰, and 6‰ at depths of 0–5, 5–10, and 10–15 cm, respectively; Nakanishi et al., 2012). The Δ¹⁴C values for live roots <1 mm in diameter (Figure 2 and Supplementary Table S3) showed a similarity to those for WEOC in the uppermost 10 cm of the soil profile, but diverged in the 10–15 cm layer. The high Δ¹⁴C values for live roots >1 mm in diameter did not correspond to the Δ¹⁴C values for WEOC, suggesting a limited contribution of organic C sources in the soil to root growth. Regarding the transfer of C to mycorrhizal fungi, mycorrhizal fungi are estimated to receive 4–20% of total NPP (Hawkins et al., 2023) and can constitute up to 40% of the biomass of ectomycorrhizal roots (Harley and McCready, 1952; Satomura et al., 2003). However, research by Hobbie et al. (2002) demonstrated that mycorrhizal fungi incorporated recent photosynthetic, with an estimated age of 0–2 years, using four fungal species in a mature Douglas-fir forest in the United States.

Root morphology and chemical composition vary depending on diameter size and branching position (Hishi, 2007), which also affects their decomposition rate in soils after deposition (Sun et al., 2013; Zhang and Wang, 2015; Wu et al., 2022). The observed difference (3–19 years, with no discernible trends related to depth or diameter class, Table 2) in the mean C age between live and dead roots could reflect the time required for root decomposition in soils before root C returns to the atmosphere, a process to be quantified for accurately assessing the role of roots in soil C cycling, and ultimately, forest C dynamics. Zhang and Wang (2015) found faster decomposition of fine roots compared to course roots in middle latitude areas. Conversely Sun et al. (2013), reported slower decomposition of very fine roots (<0.5 mm in diameter) compared to 0.5–2.0 mm roots in a temperate forest in northeastern China. Goebel et al. (2011) noted that root decomposition can be influenced by root pigmentation and root order; pigmented first- and second-order roots decompose more slowly than higher-order roots. These findings suggest that root decomposition can be complicated depending on morphological traits such as diameter-size class and branch order, which may result in the inconsistent difference in ¹⁴C signature between live and dead roots. Many studies rely on buried root techniques and laboratory incubation experiments, potentially misrepresenting actual root decomposition processes (Wu et al., 2022). Root decomposition rates are still controversial, necessitating variable approaches, including ¹⁴C methods, for comprehensive assessment.

Overall, our results provide a piece of knowledge about belowground C allocation processes in plants and suggest a further hypothesis. However, the fine root dynamics—encompassing both ephemeral and long-lived roots—remains a critical knowledge gap. Innovative research methods are still needed to reveal the heterogeneity of fine root dynamics in situ. Future studies should include tracking changes in ¹⁴C signatures of fine roots and their nonstructural C pools throughout their development, for example in combination with direct monitoring techniques such as minirhizotrons and flatbed scanners, to fully elucidate the role of plants in C dynamics and sequestration within forest ecosystems.