Japanese larch, native to Honshu Island, Japan, is a medium to large deciduous coniferous tree renowned for its resilient and enduring wood, primarily utilized in construction purposes. Its global introduction to forestry plantations in European countries of the northern hemisphere occurred during the late 19th century. China embarked on the introduction of Japanese larch in the mid-20th century for productive afforestation to satisfy the growing wood demand. The sampling sites are situated in Changlinggang Forest Farm in Jianshi County, Enshi Autonomous Prefecture, Hubei Province, China, and Wanbaoshan Forest Farm in Longshan County, Xiangxi Autonomous Prefecture, Hunan Province, representing the central and southern regions of China, respectively. Established in 1957, Changlinggang Forest Farm stands as the largest Japanese larch base in southern China. Meanwhile, Wanbaoshan Forest Farm holds the distinction of being the largest Japanese larch base in Hubei Province. In 2000, Japanese larch was designated as the afforestation tree species for seedling cultivation and afforestation endeavors. The climate, terrain, altitude, and stand density of the two forest farms exhibit similarities and possess certain representativeness. Detailed stand data is illustrated in Figure 1.

This study investigated a chronosequence of Japanese larch plantations. Soil samples were collected four stand-age stages, namely Young (J1, J2, J3), Middle-aged (M1, M2), Near-mature (N1, N2)and Mature stand (R1) to examine the responses of fungal communities to nitrogen and phosphorus metabolism during forest succession. The collection of samples follows the method of Zhang et al. (2022). Sample collection was conducted during the growing season within each forest stand. In each plot, areas with relatively homogeneous stand structure and similar site conditions were selected as sampling locations to minimize the influence of environmental heterogeneity. Plant samples included leaves, stems, and roots, which were collected using a random sampling approach within each plot. Soil samples were collected around the corresponding plants following an “S-shaped” pattern using a soil auger from the 0–20 cm surface layer. Visible stones, roots, and litter were carefully removed, and the soil was thoroughly homogenized to form one composite sample per replicate. We collected eight forest plot sample groups from four different forest ages, with three biological replicates for each sample group. After collection, all samples were immediately placed in pre-prepared sterile, sealed self-sealing bags. Soil samples were placed in insulated boxes with ice packs immediately after collection in the field to maintain a low temperature as much as possible and minimize changes in microbial activity. The samples were transported back to the laboratory on the same day, with transport time generally kept within 24 h. Upon arrival, soil samples were immediately pretreated: one portion was passed through a 2 mm sieve for physicochemical analyses, while another portion was stored at −20 °C for subsequent microbial analyses. Plant samples were washed, separated by tissue type, and either oven-dried or frozen for further analyses. Regarding sample amounts, approximately 10–20 g of soil was used for physicochemical property measurements, and about 5–10 g was used for microbial biomass and enzyme activity assays. For plant biomass and nutrient analyses, sample mass varied by tissue type, with 0.5–1.0 g (dry weight) used per sample. Independent replicates were established for each plot to ensure data representativeness and reliability. The sampling information is shown in Figure 1c. The collected rhizosphere soil samples were divided into three parts. One part was used to determine the basic physical and chemical properties of the soil, the other part was used to measure microbial biomass and soil enzyme activity, and the remaining part was used for Internal Transcribed Spacer (ITS) measurement analysis.

Total porosity, capillary porosity, non-capillary porosity, and associated water-holding properties were assessed using the ring knife method (Yan et al., 2019). Hydrolyzable nitrogen (HN) was quantified employing the alkaline hydrolysis diffusion method, while total nitrogen was determined through Kjeldahl nitrogen determination (Clark et al., 2015). Available phosphorus was evaluated via sodium bicarbonate extraction and the molybdenum blue method, whereas available potassium was detected using ammonium acetate extraction (Zhu et al., 2016). Total phosphorus and total potassium were analyzed after digestion using an acidic mixture of HNO₃-HCl-HF (4:1:1, v/v/v). The dried samples were pulverized and digested in an HNO₃-HClO₄ (3:1, v/v) mixture at 180 °C employing a graphite furnace (Hong et al., 2022). The resulting solutions were subsequently analyzed using inductively coupled plasma emission spectrometry. Microbial biomass nitrogen and microbial biomass phosphorus were determined via chloroform fumigation (Joshi and Garkoti, 2023). Soil acid phosphatase activity was assessed using the p-nitrophenyl phosphate method (Veloso et al., 2023), and urease activity was determined utilizing the sodium phenol-sodium hypochlorite colorimetric method (Guo et al., 2023).

Samples were sequenced using ITS, and the raw data obtained were subjected to amplicon sequence variants (ASVs) clustering and annotation through sequence splicing and quality control (Fodelianakis et al., 2022). A rarefaction analysis was performed based on the ASVs clustering results to examine the sample sequencing coverage situation.

Alpha diversity calculations included Chao, Shannon, and Observed_species indices (Li et al., 2022). Chao and Observed species indices reflected community richness, Shannon index indicated community diversity. Species composition analysis integrated the community structure analysis of multiple samples, observing the community structure at different taxonomic levels, and was analyzed and visualized using the reshape2 and ggplot2 packages of R (v3.6.0). ASVs were taxonomically annotated at the 97% similarity threshold (García-López et al., 2021) using the software (QIIME v1.8.0 platform and Vsearch 2.7.1) to determine microbial community structure and composition at the phylum, genus, and ASV levels.

Beta diversity analysis assessed the similarity or difference in sample community composition (Crabot et al., 2020). Non-metric multidimensional scaling (NMDS) was employed as a data analysis method to reduce research objects (samples or variables) in multidimensional space for localization, analysis, and categorization while preserving the original relationships among the objects (Khodabandeh et al., 2023). Analysis and visualization were performed using the vegan and ggplot2 packages of R (v3.6.0).

Redundancy analysis (RDA) and variance partitioning analysis (VPA) were conducted using the vegan package of R. Variable inflation factor values (VIF) were calculated before performing redundancy analysis to eliminate redundancy factors with VIF > 12. Molecular ecological network (MENs) construction was based on the relative abundance of ASVs in different treatments (Wang X. et al., 2023). The network was built using Spearman’s test, with the top 20 genus-level results selected for correlation analysis. Critical values for species interactions were Spearman correlation > |0.6| and p < 0.05, used to filter out weakly correlated ASVs and reduce network complexity. All network analyses were conducted via the Molecular Ecological Network Analysis Pipeline (MENAP),¹ with visualization performed using Gephi 0.9.1-beta. Intra-module connectivity (Zi) and inter-module connectivity (Pi) were calculated for each node based on differences in ASV abundance between modules in the coexistence network.

PICRUSt2 (Phylogenetic Investigation of Communities by Reconstruction of Unobserved States, v2) was used to infer potential functional profiles of microbial communities based on marker gene information, following the approach described by Douglas et al. (2018).

Changes in tree growth and stand age are key factors affecting N and P nutrient accumulation (Wu et al., 2020). The biomass produced during tree growth affects soil N input and P turnover (Ehtesham and Bengtson, 2017). In order to evaluate the effects of tree age on N and P nutrients, we studied growth-related soil physical and chemical properties, and microbial metabolic activities at different tree ages. It was found that the soil N and P elements and their effective states reached the peak in the near-mature stage of the forest (N), and gradually decreased and stabilized in the mature stage of the forest (R) (Figure 1). The variation trends of TP and TN showed similar trends in previous studies (Deng et al., 2019; Wu et al., 2020). The TN in the rhizosphere soil of larch in Northeast China showed an increasing trend in different time series, reaching the peak at the mature stage of forest, while TP showed a decreasing trend, reaching the minimum value of 0.49 g/kg at the near mature stage of forest (Yan et al., 2017). Inconsistent trend of nitrogen and phosphorus nutrient content may be caused by regional and tree species differences (Chen et al., 2016). Compared with the young forest stage (J), the values of soil TN, TP, HN, and AP in the near-mature forest stage (N) increased by 43.1, 50.0, 109.3, and 58.5%, respectively in our study. The nutrient demand of different growth stages (J-R) varied (Jiang et al., 2017).

In this study, the soil TN concentration was between 2.21 to 3.55 g/kg, and the TP content was between 0.5 to 0.75 g/kg (Figure 1). Soil N: P reflects the balance of N and P during plant growth and development, and can indicate the supply of soil nutrients during plant growth (Zhao et al., 2015). In our study, the N: P ratios of rhizosphere soil at different forest ages were 4.96 (J), 3.81 (M), 4.73 (N) and 5.26 (R), which were lower than the national soil N:p (5.20) level in the early stage of forest maturity (Sun et al., 2024). This indicates that the soil nitrogen content is generally low, and the growth may be inhibited by nitrogen. In the mature stage of the forest (R), the soil phosphorus content decreases, and the trees are limited by phosphorus. Soil N: P showed a trend of decreasing first and then increasing, which was consistent with the trend of N: P ratio in rhizosphere soil of Pinus massoniana Lamb. plantation in southern China (Xu et al., 2022).

Microorganisms are the main driving forces to assist plants in absorbing nitrogen and phosphorus nutrients from the soil (Cao et al., 2022; Peng et al., 2022). In detail, plant N and P uptake is mediated by specific transporters, such as nitrate transporters (NRTs) and phosphate transporters (PHTs), which involved various nutrient signal transduction, gene regulation, and morphological and physiochemical changes that may recruit beneficial microorganisms and reallocate their structures and functions to help the plant for more intelligent and precise nutrient uptake (Oldroyd and Leyser, 2020). The analysis of microbial biomass in this study showed that with the increase of forest age, the content of MBP increased first and then decreased, which reflected the trend of TP and AP in rhizosphere soil of different forest ages. It reached the maximum value of 25.39 mg/kg in near mature forest stage (N), which was 32.1% higher than that in young forest stage (J). MBP is a dynamic phosphorus pool that regulates the availability and turnover of phosphorus (Peng et al., 2022). Similarly, the MBN content of different forest ages reflected the change trend of TN and SUE. The MBN value was the highest in the near-mature forest stage (N), reaching 850.56 mg/kg, which was 36.9% higher than that in the young forest stage (J). Microorganisms can help plants acquire these key nutrients more effectively through nitrogen fixation, phosphorus and potassium solubilization (Chang et al., 2024; Philippot et al., 2023; Zhu et al., 2023). For example, phosphate-solubilizing bacteria can secrete organic acids and phosphatases to improve the bioavailability of phosphorus and regulate the conversion process of phosphorus (Li H.-Z. et al., 2024); nitrogen-fixing bacteria can produce nitrogenase to convert atmospheric nitrogen into ammonia (Sankari et al., 2022). Collectively, these microbial-mediated processes highlight the potential role of soil microorganisms in supporting plant nitrogen and phosphorus nutrition.

Increasing forest age is generally associated with changes in soil organic matter (SOM), soil nutrient availability, and microbial abundance in the rhizosphere, which may be partly related to age-related differences in tree growth and litter inputs. Root exudates are an important form of communication between plants and microorganisms, and can also microorganisms that are potentially involved in soil nutrient transformations relevant to plant nutrient acquisition (Bai et al., 2022). For example, flavonoids in maize root exudates recruit rhizosphere oxalate bacteria that can promote N utilization and maize lateral root growth (Yu et al., 2021). Increased microbial biomass and enzyme activities may reflect enhanced microbial involvement in soil nutrient transformation processes in the rhizosphere (Figure 1).

Tree age was significantly associated with changes in the diversity of plant rhizosphere microbial communities. Alpha diversity analysis showed that the diversity and abundance of fungal communities gradually increased and then decreased with changes in forest age (Figure 3). A similar pattern has been reported in Korean pine forests, where soil fungal diversity exhibited an increase followed by a decrease along forest development stages (Han et al., 2016). Litter input is limited, which may lead to low microbial community diversity. Similar to previous results (Hu et al., 2022), as larch stands develop, soil nutrient availability is improved and soil microbial diversity is also increased.

Tree age was an important factor associated with variations in plant rhizosphere microbial community structure. The structure of the microbial community varied among forest age stages, and the fungal community was mainly composed of Ascomycota, Basidiomycota and Mortierellomycota (Figure 5). Other studies have also confirmed this result (Liu et al., 2020; Sui et al., 2022; Wang et al., 2022a). Most Ascomycota, Basidiomycota and Mortierellaomycota are predominantly saprotrophic fungi. They are key decomposers of forest plant biomass, which can decompose forest litter and increase SOM content (Büttner et al., 2021; Manici et al., 2024; Salo et al., 2019). Microorganisms convert organic nitrogen into inorganic forms by decomposing organic matter available for plant uptake. The input of organic matter can also change the composition of genes related to the phosphorus cycle and promote the transformation of phosphorus (Huang et al., 2024). In addition, NMDS analysis showed that the rhizosphere microbial community of the forest changed greatly in the early stage of development, and tended to gather when the forest was close to maturity (Figure 4c). The soil fungal community structure is related to the change of soil nutrient accumulation (Li T. et al., 2023). The soil nutrient accumulation tends to be stable when the forest is nearly mature, and the number of microbial species is also relatively stable (Figure 3).

Key taxa occupying the core niches of microbial communities play crucial roles in shaping community structure and function. Molecular ecological network (MENs) analysis across different forest age stages revealed that Ascomycota was the dominant phylum shared among all three networks (Supplementary Table 1). Basidiomycota and Mortierellomycota were the dominant phyla in the molecular ecological network of middle-aged forest (M), and the abundance of key taxa in the MENs was higher in the middle-aged forest (M). Ascomycota was not only the key taxa of the important niches of MENs, but also the dominant species of each forest age sample, with an average percentage of 59.18% (Figure 5). Ascomycota is involved in the decomposition of organic matter and the assimilation of root exudates (Hugoni et al., 2018; Wang et al., 2022b). In this study, the abundance of Basidiomycota in J1-J3 decreased and the abundance of Ascomycota increased (Figure 5). It was reported that the increase of litter was beneficial to the transfer of Basidiomycota to Ascomycota (Dang et al., 2017). With the increase of forest age, SOM was accumulated due to the increase of diversity of litter and understory vegetation (Wan et al., 2023), but the ability of Ascomycota to degrade lignin was limited (Osono, 2007). Basidiomycota dominated the decomposition of litter in the middle and late stages. It can produce a variety of extracellular oxidases (Baldrian, 2006) and degrade recalcitrant lignin compounds (Lundell et al., 2010; Osono, 2007). Ascomycota and Basidiomycota are important groups for ectomycorrhiza (ECM) formation. Ectomycorrhizal fungi (ECMF) regulate soil nutrients through a variety of mechanisms, such as enhancing soil phosphorus bioavailability, promoting SOM decomposition and improving soil microbial activity (Nicolás et al., 2019; Qi et al., 2022). Interestingly, Glomeromycota, which forms arbuscular mycorrhiza (AM), became a key species in M (Supplementary Table 1). Arbuscular mycorrhizal fungi (AMF) can recruit microbial communities, stimulate their nitrogen and phosphorus turnover function, and improve N and P utilization (Wang G. et al., 2023; Zhang J. et al., 2024).

Rhizosphere microorganisms are the key driving factors between soil and forest nitrogen and phosphorus cycles (Cui et al., 2018; Zhang Y. et al., 2024). Their composition and functions are affected by the regulation of tree growth and soil properties, and can affect the circulation of nitrogen and phosphorus nutrients in the rhizosphere, which in turn participates in and regulates the turnover of nitrogen and phosphorus in plants and soils. Microbial functional genes are mainly involved in biochemical processes and mediate element cycling by encoding specific functional enzymes and metabolites. Their abundance can predict the biogeochemical cycling potential and metabolic characteristics of soil microorganisms (Tu et al., 2017). The results of this study showed that the microbial community-mediated N and P cycle processes were related to plant age (Figure 8), and the abundance of soil microbial functional genes also changed with forest age (Li Y. et al., 2024).

In this study, there was a significant positive correlation between MBN and TN (p ≤ 0.05) (Figure 9), and MBN at different forest ages also reflected the changing trend of TN and SUE (Figure 1). When the soil TN content was the lowest in the middle-aged forest (Figure 1), we found that in the process of nitrogen cycle, the expressions of nitrogen fixation (nifH, nifD and nifK) and assimilative nitrate reduction (nasA) functional genes were more abundant in the middle-aged forest (M) (Figure 8b). As trees grow from young to middle-aged, tree growth is more dependent on nitrogen uptake to meet their nitrogen needs (Sun et al., 2016). Soil nitrogen storage is reduced, and microorganisms accelerate biological nitrogen fixation and establish symbiotic or reciprocal relationships with plant roots, thereby promoting root expansion and nutrient uptake (Islam et al., 2013; Li P. et al., 2024). Microorganisms are a key link in driving nitrification and denitrification and regulating soil nitrogen availability and N₂O emissions (Song and Niu, 2022). The abundance of functional genes related to nitrification process (nosZ, pmoA-amoA, pmoB-amoB, pmoC-amoC and hao), denitrification process (nirK, nirS, norB and norC) and dissimilatory nitrate reduction processes (narG, narH and nirB) was the highest in the young forest period (Figure 8b). Young trees grow faster and have higher nitrogen requirements, resulting in more nitrogen uptake (Sun et al., 2016). The interaction between plants and rhizosphere microorganisms affects the physical and chemical properties of the soil, shapes the rhizosphere microbial community, activates nitrogen cycle genes, and promotes nitrogen fixation and transfer (Dong et al., 2024).

There was a significant negative correlation between MBP and ACP, as well as between AP and ACP (p ≤ 0.05) (Figure 9), which was consistent with previous observations (Hofmann et al., 2016). The contents of AP and MBP were the lowest, while the activity of ACP was the highest, which was 21.21 umol d⁻¹ g⁻¹ (Figure 1). ACP activity is one of the main factors affecting MBP (Chen et al., 2017). It has been reported that ACP activity is affected by plant phosphorus starvation. The degree of phosphorus limitation in soil and the availability of organic phosphorus compounds (He et al., 2024)could induce microbial ACP secretion to mobilize organic phosphorus in low-phosphorus soils (Ma et al., 2021). AP and MBP showed an increasing trend with the increase of forest age (Figure 1), and the increase of soil available phosphorus could be attributed to the utilization of mineral phosphorus by microorganisms. Microorganisms can dissolve insoluble inorganic phosphorus and mineralize more organic phosphorus than they need to be fixed in biomass and provide soluble inorganic phosphorus to plants (Chen et al., 2021; Hofmann et al., 2016). The increase in soil AP may be due to the increase in gene abundance that controls soil microbial inorganic phosphorus solubilization and organic phosphorus mineralization (Zhi et al., 2023). During the phosphorus cycle, the abundance of functional genes related to microbial phosphorus starvation response and phosphorus mobilization regulation increased with the increase of forest age, reaching a peak in the near-mature forest stage. The abundance of phosphorus starvation genes (PhoU, PhoR and PhoB), phosphorus mobilization genes (gcd, phnP, ppx, ppa) and phosphorus uptake and transport (pstB, pstC, pstA, pstS and Pit) were much higher than other genes in the same group (Figure 8a). Inorganic phosphorus dissolution and organic phosphorus mineralization related genes gcd, ppx, ppa and PhnP dominated the increase of soil phosphorus utilization capacity (Zhi et al., 2023). The genes (phoU, phoR and phoB) involved in the regulation of phosphorus starvation response enable microorganisms to utilize external phosphorus sources. These genes are closely related to genes involved in phosphorus uptake and transport (such as pst) and control the expression of alkaline phosphatase-encoding genes (Hsieh and Wanner, 2010).