Soil organic carbon (SOC) is the largest terrestrial organic carbon pool. Plant litter is an important source of SOC, but the knowledge gap between SOC fractions and plant litter input remains inconsistent. Here, a litter input control experiment was conducted at three subalpine forest types (coniferous forest, mixed forest, and broadleaved forest). We assessed the variations of total organic C, active organic C (easily oxidizable C, labile organic C), recalcitrant organic C, and microbial biomass C under litter input or removal. The results showed that soil total organic C decreased greatly under litter input. It was mainly caused by the change of easily oxidizable C and labile C, while the influence of recalcitrant C was small. At the same time, this effect varied among different forest types. Among them, the effect of litter input on SOC was weak and slow in the coniferous forest with low-quality litter input, while a quick effect was observed in the mixed and broadleaved forests with high-quality litter input. Microbial biomass C declined under litter input in most cases, and its variation was strongly controlled by soil temperature and freeze-thaw events. Overall, our results provide new evidence that forest type would strongly control SOC dynamics, in concert with litter quality shifts, with potential consequences for long-term C sequestration. We highlighted that litter input could reduce microbial biomass carbon which might limit the native SOC decomposition, but the loss of active C ultimately changed the SOC in the subalpine forests. It suggested that the interaction of multiple mechanisms should be considered in the study of SOC in this region.

Understory vegetation is a critical component of forest ecosystems. Its removal can substantially alter litter decomposition processes, with cascading effects on carbon (C) and nutrient cycling in terrestrial ecosystems. However, the global response patterns of litter decomposition to understory removal and underlying controlling factors remain unclear. We conducted a meta-analysis of 330 observations from 29 peer-reviewed field litterbag studies to assess the effects of understory removal on litter decomposition. We evaluated the changes in decomposition rate, mass loss, and nutrient dynamics to quantify the impacts of understory removal on litter decomposition. We assessed the associated shifts in soil microbial communities, measured using phospholipid fatty acids (PLFAs), to examine how microbial responses mediate decomposition during understory removal. We examined whether canopy type moderated these responses and explored the key predictors of decomposition for understory removal. Understory removal significantly reduced litter decomposition rate and mass loss by an average of 29.6% and 14.8%, respectively, while increasing lignin remaining by 30.1%. Soil microbial biomass also declined, with total, fungal, and actinomycete PLFAs decreasing by 12.0%, 30.8%, and 27.5%, respectively. Across canopy types, understory removal decreased litter mass loss in both broadleaved and coniferous forests. However, the remaining N and P increased significantly in broadleaved forests but changed only marginally in coniferous forests. Random forest analysis showed that initial litter quality and variations in fungal biomass were the primary predictors of decomposition responses. Understory vegetation removal significantly suppresses litter decomposition by reducing fungal biomass, and interacting with litter quality constraints and canopy type strongly moderates these effects. This highlights the essential role of understory vegetation in sustaining nutrient cycling and microbial functioning in forest ecosystems and underscores its critical role in guiding sustainable forest management.

Trees can resorb nutrients to preserve and reuse them before leaves fall, which could efficiently adapt to environmental changes. However, the nutrient requirements of trees in different months with seasonal climate changes are often neglected. In this study, we selected plantations of an evergreen broadleaf tree (Castanopsis carlesii (Hemsl.) Hayata) and a coniferous tree (Cunninghamia lanceolate (Lamb.) Hook.) in the subtropics. The monthly dynamics of leaf nitrogen (N) and phosphorus (P) resorption efficiencies and C:N:P stoichiometric ratios were checked along a growing season from April to October 2021. Trees in both plantations exhibited efficient N and P resorption but with significant monthly variations. The N and P resorption efficiencies in the Cunninghamia lanceolata plantation ranged from 34.26% to 56.28% and 41.01% to 54.85%, respectively, and were highest in September. In contrast, N and P resorption efficiencies in the Castanopsis carlesii plantation ranged from 11.25% to 34.23% and 49.22% to 58.72%, respectively, and were highest in July. Compared with the Cunninghamia lanceolata, the C:N of the Castanopsis carlesii plantation was significantly lower, while its C:P was significantly higher in May and September. The Castanopsis carlesii plantation was strongly limited by P (the N:P ratios in mature leaves were higher than 20), whereas the Cunninghamia lanceolata plantation might be limited by both N and P (the N:P ratios in mature leaves were between 10 and 20). In addition, the statistical analyses revealed that temperature and precipitation were significantly associated with N and P resorption efficiencies, but the relationships were controlled by forest types. These findings highlight that efficient resorption of N and P may be beneficial in regulating nutrient limitation and balance in subtropical forest ecosystems. These results contribute to the understanding of N and P utilization strategies of trees and provide a theoretical basis for vegetation management in the subtropics.

The influence of forest type on soil aggregates distribution, stability, and the contribution of aggregate-associated carbon (C) to bulk soil organic carbon (SOC) remains poorly understood. This may be crucial for the accumulation and persistence of SOC in subtropical ecosystems. In this study, we examined soil aggregate distribution and stability at two depths (0–15 and 15–30 cm) in 10-, 20-, and 30-year-old Cryptomeria japonica (Japanese cedar) and Chimonobambusa quadrangularis (square bamboo) plantations. We further assessed the contribution of carbon (C) associated with distinct aggregate fractions to bulk SOC. Across all stand ages and soil depths, macroaggregates accounted for 19%–56% of total soil aggregates in Japanese cedar plantations, whereas their proportion was 30%–337% higher in square bamboo plantations. In contrast, fine aggregates constituted 3%–67% of total aggregates in Japanese cedar plantations, but were 29%–94% lower in square bamboo plantations than in Japanese cedar plantations. Compared with Japanese cedar plantations, aggregate mean weight diameter (MWD) and geometric mean diameter (GMD) increased by 17%–88% and 35%–152%, respectively, in square bamboo plantations. In Japanese cedar soils, C and nitrogen (N) were primarily concentrated in coarse macroaggregates and fine macroaggregates, whereas in square bamboo plantations, C and N were mainly associated with coarse macroaggregates only. Both aggregate-associated soil C and N varied significantly with aggregate size and forest type, and Japanese cedar soils exhibited higher aggregate C/N ratios, particularly in older stands. Bulk SOC was positively correlated with macroaggregate-associated C in both forest types and with the silt and clay fractions in Japanese cedar plantations. MWD increased with higher macroaggregate C content and declined as the proportion of C in smaller aggregate fractions increased. These findings indicate that forest type plays a critical role in regulating soil aggregation and SOC stabilization pathways, with square bamboo plantations enhancing C sequestration by promoting macroaggregate formation and stability.