Phyllostachys edulis (moso bamboo) ages signicantly affect soil nitrogen transformation and endophytic microbes but niche differentiations outweigh ages in shaping microbial communities of moso bamboo-soil system

Background: Phyllostachys edulis (moso bamboo) is a key source of non-wood forest products. Clarifying the linkage among bamboo growth, soil nutrient and core microbe can expand the horizon on nutrient management practices and functional endophytic and rhizospheric microbes. Results: In this study, young (0.5-yr), mature (2.5-yr and 4.5-yr) and old (6.5-yr) bamboo plants were selected in a moso bamboo eld, and above-ground tissues, below-ground roots (rhizomes) and rhizospheric soils were sampled. The bamboo biomass, soil properties and endophytic microbes were determined and quantify their comprehensive relationships. Bamboo ages had negligible impacts on the bamboo height, diameter at breast height and above-ground biomass. Organic matter and nitrogen (N) contents in the rhizospheric soil of the 0.5-yr bamboo were signicantly higher than those of the other three age groups. The rhizospheric soil of the 6.5-yr bamboo had the lowest N mineralization rate and urease activity. Signicant differences in bacterial and fungal communities were found in the aboveground tissues but not in the rhizospheric soils associated with plants of different ages. Bacterial and fungal community structures in the above-ground tissues were signicantly different from their counterparts in the rhizospheric soils. Conclusions: Bamboo ages signicantly affected N transformation rates, functional gene abundances and urease activities of rhizospheric soils and endophytic bacterial community structures. However, niche differentiations outweighed ages in shaping the whole microbial communities of bamboo aboveground tissues, roots and rhizospheric soils. In the future, moso bamboo management should consider balanced applications of ammonium-N and other nutrients and utilisations of Chytridiomycota to stimulate moso bamboo growth. impacts on fungal assemblages including community diversity and structure, which might be evolutionary conservation of endophytic fungi [17, 46, 47].


Introduction
Bamboos are key members of the Gramineae plants and have approximately 90 genera and 1200 species [1,2]. The bamboo output is a key source of non-wood forest products, and bamboo plantations and forests have been continuously increasing in tropical, subtropical and temperate regions [1,2]. After industrial treatment, bamboos have up to 95% glucose and 73% ethanol yield [3]. Bamboo ber has special inherent properties, such as good air permeability and excellent moisture absorbency, and has allegedly better dye-ability than cotton [4,5]. In addition, bamboo shoot is used as healthy food due to the multiple health bene ts including anticancer activity and improvement of digestion [6,7]. Globally, 2.5 billion people are economically associated with bamboo plantations, and the value of bamboo production is about $ 2.5 billion per year [8].
Among all bamboo species in China, the most substantial species is Phyllostachys edulis (moso bamboo), with more than three million hectares planted [9,10]. After emerging, the shoot of moso bamboo grows fast, and under suitable conditions, its growth rate could reach 1 m per day [11,12].
Therefore, different agroforest management practices, especially mineral fertilizer applications, have been used to stimulate moso bamboo growth [10][11][12]. The moso bamboo generally has a rhizomatous clonal growth every second year, and this inherent biological property is markedly different from other plants belonging to the Gramineae [13,14]. The rhizomatous clonal growth signi cantly or markedly changed the physical, chemical and biological properties of rhizospheric soils, relying on root exudates and rhizosphere size [15][16], and therefore ages of moso bamboo might also dominate the rhizospheric effects including nitrogen (N) cycling [15][16][17][18]. However, to the best of our knowledge, few studies have been conducted to reveal and quantify the linkages among bamboo biomass, age and soil nutrient cycling, especially the impacts of bamboo ages on soil N transformations.
Apart from abiotic nutrient supply, key microbes associated with plants also drive the healthand biomass accumulation of host plants [19,20]. A series of taxonomically diverse microbes (viruses, bacteria and fungi) colonizes in plants and serves as the extended secondary genome of host plants [17,21,22]. The relationships between plants and microbes could be classi ed as positive stimulation, neutral interaction and negative pathogenicity. Endophytic microbial colonizations often start with the recognition of speci c root exudate which theoretically 'communicates' with symbiotic, commensalistic and pathogenic microbes [22,23]. However, it has been proven that plants can speci cally attract microorganisms for their own ecological bene ts [24,25]. For instance, endophytic microbes changed membrane lipopolysaccharides, lignins and peroxidase of host plants and improved plant resistance to Fusarium wilt and hazardous compound of invasive species [24][25][26][27]. Much attention has been paid to the roles of endophytic bacteria and fungi to stimulate host plant growth. Furthermore, each niche in plant-soil system, including above-ground tissue, root and rhizospheric soil, has a unique biotope for microbial community and hosts different microbial assemblies [17,28]. The niche effects should be thereby considered, but studies disentangling impacts of niche differentiations on the key microbes in moso bamboo plantation or forestry have rarely been conducted.
The objectives of this study were to: (1) quantify the age impacts of moso bamboo on soil N transformations; (2) examine the selectivity of bamboo age and niche for key microbes; (3) identify potential endophytic and rhizospheric microbes positively affecting moso bamboo growth; and (4) link moso bamboo biomass yield with soil nutrients and key microbes. In this study, moso bamboo plants of different ages (0.5, 2.5, 4.5 and 6.5 years) were selected to serve as different treatments. The bamboo biomass, soil properties and endophytic microbes were determined to reveal and quantify the comprehensive linkages among moso bamboo biomass yield, soil nutrients and endophytic and rhizospheric microbes. Our study could help to develop strategies on nutrient management of bamboo plantations and expand the horizon in utilizing functional microbes to stimulate bamboo growth.

Bamboo and soil properties
There were negligible differences in the heights, DBH and above-ground biomass among the moso bamboo plants of four different ages (Fig. 1). The average height, DBH and above-ground biomass of moso bamboos ranged from 13.9 to 14.9 m, 10.4 to 10.6 cm and from 16.25 to 23.33 kg, respectively.
The organic matter and total N contents in rhizospheric soil of the 0.5-yr bamboo plant were signi cantly (P < 0.05) higher than those of the other three counterparts, with the lowest organic matter and total N contents being in the rhizospheric soils of the 6.5-yr bamboo plant (Figs. 2a, b). The NH 4 + -N contents did not signi cantly differ among the four treatments (Fig. 2c), and the highest and lowest NO 3 --N contents were present in the rhizospheric soils of 0.5-yr and 4.5-yr bamboo plants, respectively (Fig. 2d).
Biological properties related to soil N transformations The N transformation rates, functional gene abundances and urease activities of the rhizospheric soils were signi cantly but differently affected by moso bamboo ages (Fig. 3). The lowest N mineralization rate was detected in the rhizospheric soil of the 6.5-yr bamboo plant, which was signi cantly lower than those of the 2.5-yr and 4.5-yr bamboo plants (Fig. 3a). The nprA gene abundances in the four different rhizospheric soils ranged from 8.16×10 7 to 1.57×10 8 copies g -1 dry soil, and nprA gene abundances in the rhizospheric soils decreased with bamboo ages (Fig. 3b). The trend of chiA gene abundances among the four different rhizospheric soils was similar to that of N mineralization rates, and chiA gene abundance in the rhizospheric soils of the 6.5-yr bamboo plant was signi cantly (P < 0.05) lower than those of the other three treatments (Fig. 3c). The highest nitri cation rate was in the rhizospheric soil of 4.5-yr bamboo, while the rhizospheric soil of the 2.5-yr bamboo had the lowest nitri cation rate (Fig. 3d). Soil AOA amoA gene abundances decreased in order of 4.5-yr > 0.5-yr > 6.5-yr > 2.5-yr (Fig. 3e). However, the AOB amoA gene abundance in the rhizospheric soil of 0.5-yr bamboo plant was the highest among the four different treatments (Fig. 3f). Relative to that of 0.5-yr bamboo plant, average urease activities in rhizospheric soils of 2.5-yr, 4.5-yr and 6.5-yr bamboos decreased by 81.4%, 81.1% and 88.2%, respectively (P < 0.05; Fig. 3g).

Endophytic and rhizospheric microbial community diversities
After ltering, average sequence length ranged from 374 to 377 bp and from 229 to 270 bp for bacterial and fungal high-quality reads in different samples, respectively (Table S2), and the measured rarefactions of bacterial and fungal sequences all reached saturation plateaus (Fig. S1). The OTU similarities and differences of the above-ground tissues, roots and rhizospheric soils of different ages were present in four-set Venn diagrams (Fig. S2). The 0.5-yr and 6.5-yr above-ground tissues had the highest bacterial Both bamboo plant age and niche differentiation signi cantly (P < 0.05) in uenced the diversity indices and richness estimators of the bacterial community (Fig. 4). Among the above-ground tissues of different ages, the 0.5-yr above-ground tissue always had the highest diversity indices and richness estimators of the bacterial community, which were signi cantly higher than those in the 4.5-yr tissues (Figs. 4a-d).
However, for the roots, the highest diversity indices and richness estimators were in the 6.5-yr plants, and signi cantly (P < 0.05) different diversity indices were observed only between the 0.5-yr and 6.5-yr bamboo plants. Rhizospheric soils had signi cantly (P < 0.05) higher diversity indices and richness estimators than their above-ground tissue counterparts with the same age, although the four different rhizospheric soils shared similar values of community diversity. In the roots, bacterial richness estimators increased with the bamboo root ages (6.5-yr > 4.5-yr > 2.5-yr > 0.5-yr). Bamboo plant ages only signi cantly (P < 0.05) affected fungal community diversity in the above-ground tissues, with negligible impacts being observed in the roots and rhizospheric soils (Figs. 4e, f). Meanwhile, there were no signi cant differences in fungal insimpson diversity among the above-ground tissues, roots and rhizospheric soils in the same age group. In contrast, apart from the above-ground tissue, fungal richness estimators in the roots were also signi cantly in uenced by the bamboo ages, with the lowest richness estimators being in the 0.5-yr roots (Figs. 4g, h).
Bamboo plant ages signi cantly (P < 0.05) affected the relative abundances of Actinobacteria and Firmicutes in the above-ground tissues, Bacteroidetes, Chloro exi, Elusimicrobia, Nitrospirae and Planctomycetes in the roots, and Gemmatimonadetes, Latescibacteria and Verrucomicrobia in the rhizospheric soils (Fig. S4). The relative abundances of Acidobacteria, Chloro exi, Dependentiae, and Nitrospirae in rhizospheric soils were the highest among the three niches and signi cantly (P < 0.05) higher than those in the above-ground tissues. Notably, there were no Latescibacteria detected in any above-ground tissues (Fig. S4j), and phyla Elusimicrobia, Nitrospirae, Planctomycetes and Verrucomicrobia were not detected in the 2.5-yr or 4.5-yr above-ground tissues, either.
The predominant phyla of fungal community were: Ascomycota, Basidiomycota, Chytridiomycota, Glomeromycota, Mortierellomycota, Mucoromycota and Rozellomycota (Fig. S5a). The genus Mortierella were detected in all samples, and 70% of the total Mortierella were from the soil samples (Fig. S5b). The relative abundances of endophyte fungi in the 48 samples ranged from 2.90% to 28.8% (Fig. S5c). In the above-ground tissues, roots and rhizospheric soil, the largest phylum was Ascomycota among all the phyla, (Fig. S6a). The relative abundances of Basidiomycota, Chytridiomycota, Glomeromycota and Mortierellomycota in the above-ground tissues and the relative abundances of Ascomycota and Basidiomycota in the roots were signi cantly (P < 0.05) affected by bamboo ages (Figs. S6b-e).

Comparisons of endophytic and rhizospheric microbial structures
The PCoA1 and PCoA2 of two-dimensional plots explained 27.0%-60.0% of the total variances in endophytic and rhizospheric microbial structures, and as elicited in Fig. 5, divergences in endophytic and rhizospheric microbial community structures among the same niches with different ages occurred, although to various extents. There were signi cant (P = 0.01 and P = 0.001) differences in bacterial and fungal communities among the above-ground tissues of different ages (Figs. 5a, d). Signi cant differences in the bacterial communities occurred between the 6.5-yr and 0.5-yr or 2.5-yr above-ground tissues (Fig. 5a). For the fungal communities in the above-ground tissues, signi cant differences were found between the 6.5-yr and the other three age groups (Fig. 5d). In the roots, bacteria rather than fungi had signi cantly (P = 0.024) different community structures among different age groups (Figs. 5b, e).
Signi cant differences in bacterial communities occurred between the 6.5-yr and 0.5-yr or 2.5-yr roots, which was consistent with the results of the bacteria in above-ground tissues. Compared with the microbial community in the above-ground tissues and roots, microbes (both bacteria and fungi) in the rhizospheric soils markedly overlapped (P > 0.05) among different age groups (Figs. 5c, f).

Niche selection of key microbes
On the whole (48 samples), there were no signi cant differences in the bacterial communities detected in the above-ground tissues, roots or rhizospheric soils of different age bamboos (Fig. 6). However, bacterial communities were signi cantly different between the above-ground tissues and their corresponding rhizospheric soils, but identical in the above-ground tissues and roots (Fig. 6a). The fungal community had the same trend with its bacterial counterpart, with the microbes in the above-ground tissues and rhizospheric soils being signi cantly different in the PCoA1 value (Fig. 6b). The same results were also evident in the unweighted UniFrac clustering analyses (Figs. 6c, d). Regardless of bamboo ages, fungal communities in the above-ground tissues were rstly separated, and the fungi in the roots and rhizospheric soils tended to be grouped together (Fig. 6d). The microbes (both bacteria and fungi) from the above-ground tissues had longer distance from their counterparts in the rhizospheric soils, relative to the microbes in the roots. Based on the above, sample niches outweighed bamboo ages in shaping key microbial structures in the moso bamboo plant-soil system. Random forest analysis revealed that the phylum Nitrospirae displayed the most variable importance in the bacterial composition in the aboveground tissues, roots and rhizospheric soils (Fig. 6e). For the fungal community, the phylum Glomeromycota had the largest decrease accuracy (Fig. 6f).
Comprehensive linkages among moso bamboo growth, soil nutrient and microbes The R 2 values in stepwise regression models of N mineralization and nitri cation rates to the regulating factors were 0.264 and 0.483, respectively ( Table 1). The N mineralization rate was best described and positively correlated with chiA gene abundances. The potential nitri cation rate was positively correlated with both AOB amoA gene abundance and bamboo age. The moso bamboo biomass had positive relationships with the total N and NH 4 + -N contents of rhizospheric soils (Table S3). Among all the predominant bacteria and fungi, only the relative abundances of phylum Chytridiomycota in the aboveground tissues, roots and soils were all positively correlated with the moso bamboo biomass (Table S3).

Impacts of moso bamboo ages on nutrient contents and N transformations in rhizospheric soils
The rhizospheric soils of young bamboo plants had higher organic matter and total N contents, compared with those of the old bamboo plants (Fig. 2). Root exudates and dead ne roots were the major sources of organic matter input into the rhizospheric soils. We found that the rhizospheric soil of 0.5-yr bamboo had the highest organic matter content (Fig. 2a), suggesting that higher amounts of exudates from the young roots contributed to soil organic matter accumulation. During the bamboo invasion, organic C chemical properties could also be altered [29]. Young bamboo growth decreased alkyl C (hydrolyzable polysaccharides) but increased O-alkyl C (recalcitrant substances) contents [29]. Furthermore, Li et al. [30] con rmed that in broadleaf forests, the bamboo invasion could stimulate CO 2 xation potential in soils by enhancing soil cbbL gene abundance and RubisCO enzyme activity,. The moso bamboo is bred from its rhizome, and its under-ground rhizome and roots are connected. Apart from nutrient uptakes from the soils, young bamboos might also obtain nutrients from old and mature bamboo rhizomes, resulting in lower organic matters in rhizospheric soils of old and mature bamboos.
Chou and Yang [31] have shown that even the extracts of bamboo leaves could be detected in the rhizospheric soils. The young moso bamboo has a fast growth rate, which might also contribute to increases in organic matter in the rhizospheric soil of the 0.5-yr bamboo plant.
Both Nitrospirae ratio and random forest analysis showed the impacts of moso bamboo age on N transformation, and thus more attention should be paid to the rhizospheric microbes related to N cycling (Figs. S4 and 6). The entire N cycling includes plant N uptake, N-xation, N mineralization, nitri cation and denitri cation. Soil N mineralization is the key source of NH 4 + -N that can be easily oxidated to NO 3 --N [32]. Relative to the NO 3 --N, plant growth prefers to utilize the NH 4 + -N [33]. In this study, the 6.5-yr bamboo plant was the oldest, and its nutrient intake rate might be slow, which resulted in the lowest N mineralization rate in the rhizospheric soil of 6.5-yr bamboo plant (Fig. 3). Soil nitri cation is mainly mediated by the nitrifying microorganisms, and therefore soil nitri cation rate generally has positive correlations with (AOA and AOB) amoA gene abundances [34,35]. The AOA amoA gene was more abundant than its AOB counterpart in the same rhizospheric soils (Fig. 3), but we found that the potential nitri cation rate was best described by the AOB amoA gene abundance (Table 1). To date, there is no consensus regarding the relationship between acidic soil nitri cation and AOA and AOB amoA gene abundances. Jiang et al. [34] found that nitri cation in acid soil was stimulated with increases in both AOB and AOA amoA gene abundances. Some studies suggested that AOA amoA gene controlled soil nitri cation in acid soils [35]. However, Huang et al. [36] suggested that ammonia oxidation in acidic forest soil was driven by the bacteria rather than archaea. Young bamboo plants also had higher total soil N content than mature and old bamboos (Fig. 2b), which might be also explained by the reasons for increasing organic matter as mentioned above. As a key parameter describing soil fertility, urease activity could also affect the soil N supply and cycling [37]. Our results found that the highest total N and urease activity were both detected in the rhizospheric soil of 0.5-yr bamboo plant (Fig. 3g), implying that young bamboo roots favoured soil organic N accumulations.

Niche selection of key microbes
Soil hosts a plethora of microbes and has been widely accepted as the "base camp" and ideal habitats for various bacteria, and it is also the origin of plant-associated microbes [38,39]. Generally, microbes in plants origin from two pathways: horizontal and vertical transmissions [40,41]. After seeds germinate, plant roots exudate different chemicals into the soils to attract soil rhizospheric microbes to enter into the roots. Some microbes prefer to colonize and ourish in the rhizospheric soils or roots, and other microbes gradually translocate from the roots and inhabit in the above-ground tissues, ower and fruit, and nally colonize embryo and endosperm of seed [42]. The vertical transmission from seedhas also been reported in plants: endophytic microbes spread into different plant components with the seedling developments and re-enter the next generation seeds [17,42]. However, relative to the horizontal transmission, the vertical transmission e ciency is lower. During plant development and growth, many microbes were excluded or killed by the host plant [42]. There was no seeding management in our current study, and all test moso bamboo plants were produced by vegetative propagation, and thereby, the horizontal transmission was the only pathway of endophytic microbial colonization in the moso bamboos. Furthermore, anthosphere, carposphere, caulosphere and phyllosphere could also contribute to the endophytic microbes [41,42]. However, these moso bamboo plants have not blossomed and born any fruit yet, and therefore only caulosphere and phyllosphere were alternative sources of endophytic microbes in the test bamboo plants.
The highest bacterial diversity indices and richness estimators were observed in the rhizospheric soils (Figs. 4a-d). The above-ground tissue, root or rhizospheric soil each provides a unique ecological niche for key microbes in the moso bamboo-plant system despite these niches were of the same age. Many factors could limit bacterial colonization and ourishment inside the moso bamboo plants: (1) the signi cant differences in oxygen, osmotic pressure and nutrient supply between bamboo plants and soils rstly impose selective pressure for suitable bacteria; (2) bacterial colonization is a kind of invasion to host plants, and the immune system of the host plants could exclude or inhibit some bacteria [40]; and (3) even if successful colonization, limited intercellular space and environmental heterogeneity may be unsuitable for the reproduction of some bacteria [43]. Fungi and moso bamboo both belong to the eukarya domain, but fungi still need to overcome niche barriers to enter into the above-ground tissues and form plant-endophytic fungi association. Interstingly, the highest fungal alpha diversities were not always in the rhizospheric soils. This result was also supported by Chhipa and Kaushik [44] who suggested that soils and above-ground stems of Aquilaria malaccensi shared similar fungi. Gond et al. [45] also showed that relative to soils, a higher Shannon diversity of fungal community was in the stems of Nyctanthes arbortristis. The reason for the differences in bacterial and fungal diversities might be that the host biogeography was the key factor dominating endophytic fungi, while plant tissues primarily shaped endophytic bacteria [46,47]. Simultaneously, phylogenetic pro ling of key fungi further con rmed that plant compartments had minor impacts on fungal assemblages including community diversity and structure, which might be evolutionary conservation of endophytic fungi [17,46,47].
Linking moso bamboo growth with soil nutrient management and endophytic microbes No differences in plant biomass were observed among age groups in our study. Moso bamboo growth is a "slow-fastslow" process, and the maximum growth rate is generally in the second month after shoot emergence, which contributes to more than 50% of biomass accumulation [11]. The growth of moso bamboo involves soil nutrient uptake by root systems, translocation from roots and then accumulation in the above-ground tissues, and therefore, moso bamboo biomass might be stimulated by the increasing availability of soil nutrients, especially N and phosphorus (P) fertilization. Our previous study has shown that P fertilization alone has minor effects on moso bamboo growth in acidic soil [10], but the previous and current studies con rmed that bamboo biomass might be enhanced by increases in NH 4 + -N content in soil (Table S3). Piouceau et al. [12] have also found that photosynthetic activities of bamboos increase after combined applications of N and other nutrients. Endophytic microbial colonization in plants could also increase plant biomass and commercial production and thus have useful applications in agroforestry [19,27,39]. Endophytic microbes in plants confer various bioactive compounds (protease, lipase laccase, and antiviral or insecticidal compounds) and protection against harsh conditions of contamination, drought and pathogen [48][49][50]. The Chytridiomycotal relative abundances in above-ground tissues, roots and rhizospheric soils were all positively correlated with moso bamboo biomass (Table S3). The Chytridiomycota, widely distributed in terrestrial system, could solubilize CaHPO 4 to enhance plant P bioavailability [51]. Some genera of the Chytridiomycota could generate α-tubulin, effectively decompose straws and litters, mineralize cellulosic substrates to enhance available nutrient content, and the Chytridiomycota and arbuscular mycorrhizal fungi had similar molecular systematics and molecular identi cation, which might be utilized as growth-promoting microbes of moso bamboos [52][53][54]. Therefore, in the future, moso bamboo management should consider balance applications of NH 4 + -N and other nutrients, especially at the fast growth stage, and utilisations of endophytic Chytridiomycota to stimulate moso bamboo growth.

Conclusion
There were signi cantly higher organic matter and total N content in the rhizospheric soils of young bamboo, relative to the mature and old moso bamboos, although the height and above-ground biomass remained relatively similar among the four different age groups. Bamboo ages signi cantly in uenced N transformation rates, functional gene abundances and urease activities of the rhizospheric soils, and the rhizospheric soil of old bamboo plant had the lowest N mineralization rate and urease activity. Bacterial diversity indices and richness estimators in the rhizospheric soils were signi cantly higher than those of the above-ground tissues. Bamboo ages also signi cantly affected endophytic bacterial community structures in the above-ground tissues and roots, but niche differentiations outweighed ages in shaping the whole microbial community of the moso bamboo-soil system. In the future, moso bamboo management should consider balanced applications of NH 4 + -N and other nutrients and utilisations of Chytridiomycota to stimulate moso bamboo growth.

Moso bamboo plantation site and sampling
This experiment was conducted in the Bamboo Education Base of Jiangxi Agricultural University (28°76′N, 115°83′E), which has been enclosed since 1998 to prevent anthropological disturbance. The moso bamboo was the main vegetation, with a minor proportion of understory weeds. Every second year, new moso bamboo plants emerged in spring (March). In August 2019, young (0.5-yr), mature (2.5-yr and 4.5-yr) and old (6.5-yr) moso bamboo plants were selected to serve as different treatments, with four replications for each treatment. The ages of bamboo plants were identi ed based on the stalk colours and fallen leaf traces on twigs. The moso bamboo plants (4 different ages × 4 replicated bamboo plants for each age) were randomly selected, but the distance between adjacent moso bamboo plants was > 8.0 m.
The bamboo plants were harvested in August 2019, and the above-ground plants were cut off, rinsed with clean water and wiped with absorbent paper, prior to determining heights and biomass. Simultaneously, diameters at breast height (DBH) of the bamboo plants were measured at 20 cm above the ground. Below-ground roots (rhizomes) and rhizospheric soils (0-20 cm) were also sampled. The above-ground tissues (approximately 20 cm from the ground) and below-ground roots (approximately 20 cm from the ground) were used to analyze endophytic microbes. Plant samples were surface sterilized with the

Soil nutrient content analyses
Soil organic matter and total N were determined according to the methods used in previous studies [55,56]. In brief, after digesting with sulfuric acid and potassium dichromate, total carbon (C) and N contents were quanti ed by the titrating method and Discrete Auto Analyzer (SmartChem, USA), respectively, and total C was then converted into organic matter contents. Fresh and sieved soil samples were shaken with 2 M KCl (1:5 soil/water ratio) and ltered. The mineral N contents were analyzed with the Discrete Auto Analyzer, and the same ltered KCl solution was used as a blank.

Soil N transformation rates and urease activity analyses
Net N mineralization and potential nitri cation rates of rhizospheric soils were quanti ed with the methods described by Zhang et al. [56]. The rhizospheric soils (10.0 g dry soil equivalent) were incubated in 100 mL asks at 28 ℃ for 7 days. Net N mineralization rate was calculated with the changes in mineral N contents before and after incubations and expressed with mg N kg -1 dry soil d -1 . For determination of potential nitri cation rate, solution of (NH 4 ) 2 SO 4 was added into test soil (10.0 g dry soil equivalent) at 100 mg N kg −1 dry soil as the substrate of nitri cation, and then soil microcosm was incubated at 28 ℃ for 7 days. The potential nitri cation rate was quanti ed with the changes in nitrite and nitrate (NO 2 − +NO 3 --N) contents before and after incubations and expressed with mg N kg -1 dry soil d -

.
Soil urease activity was analyzed with the method described by Guan [57]. After treated with toluene for 15 min, fresh and sieved soil samples (5.0 g dry weight equivalent) were mixed with the substrate (urea) and citrate buffer (pH = 6.7) and incubated at 37 ℃ for 24 h. Soil ammonium-N (NH 4 + -N) concentrations before and after the incubation were extracted with 2 M KCl and then determined with Discrete Auto Analyzer. Soil urease activity was expressed as mg NH 4 + -N kg -1 dry soil d -1 .

Real-time quantitative PCR (qPCR) analysis
Genomic DNA of rhizospheric soils was extracted with Fast DNA SPIN Kit for Soil (MP, USA) and dissolved in 100.0 μL ddH 2 O. The DNA suspensions were evaluated, sub-packed and stored at -20 ºC prior to molecular analyses. The qPCR was employed to determine abundances of functional nprA and chiA genes related to N mineralization and ammonia-oxidizing archaea and bacteria (AOA and AOB) amoA genes related to nitri cation. The determination methods, including the reaction system, calibration curves and PCR product con rmation, were identical to those reported in our previous study [56]. Primer sequences and ampli cation e ciencies of qPCR were listed in Table S1. Quanti cations of diluted DNA suspension suggested that there was no detectable ampli cation inhibition for the qPCR in this study.

PCR ampli cation of endophytic and rhizospheric microbes
The surface-sterilized plant samples were ground and extracted for DNA with the method described by Wang et al. [58]. The V5-V7 region of bacterial 16S rRNA gene was ampli ed with the nested PCR: in the rst-round PCR, the forward and reverse primers were 779F and 1392R, and the ampli ed product fragment was approximately 593 bp; the forward and reverse primers for the second-round PCR were 779F and 1193R, and the nal product fragment was approximately 394 bp. The internal transcribed spacer (ITS) region of fungi was ampli ed, and the forward and reverse primers were ITS1F and ITS2R. All the primers were added with special barcodes, and after ampli cation, the products were estimated by 2.0% agarose gel electrophoresis and puri ed with a gel extraction kit (Qiagen, Germany).
Illumina Miseq sequencing, quality control and data processing The Illumina Miseq platform (Majorbio, China) was employed to analyse puri ed PCR products. The fastp software was used to lter raw reads to obtain high-quality reads: (1) raw reads were truncated at the end side, and quality score < 20 within 10 bp sliding window and truncated reads < 50 bp were removed; (2) raw reads with nitrogenous bases were eliminated; (3) reads were merged to one sequence with at least 10 overlapped bases; and (4) the unassembled reads were discarded [59][60][61]. A series of high-quality reads was classi ed into operational taxonomic units (OTUs) with > 97% sequence similarity. Based on the OTUs, rarefaction curves, diversity indices (Shannon and invsimpson) and richness estimators (Ace and Chao1) of microbial communities were analyzed. Meanwhile, bacterial and fungal OTUs were identi ed with the SILVA and UNITE databases and classi ed into different bacterial and fungal taxonomies, respectively, and presented at phylum and genus levels. Bacterial and fungal functional pro lings were predicted with the PICRUSt and FUNGuild analyses, respectively.

Calculation and statistical analysis
The impacts of bamboo plant ages, sample niches and their interactions on microbial properties were evaluated by the two-way analysis of variance (ANOVA), and signi cant differences (P < 0.05) among the same samples with different ages or among different niches were tested by the one-way ANOVA, followed by a Duncan multiple range test. Bamboo properties, soil nutrient contents and functional gene abundances served as candidate variables in stepwise regression analyses of N transformation rates. The ANOVA and stepwise regression analyses were conducted with the SPSS 24.0 software (IBM SPSS Inc., USA). Principal coordinate analysis (PCoA) and unweighted UniFrac clustering were employed to reveal the divergences in endophytic or rhizospheric microbial communities. The PCoA was based on the Bray-Curtis distances, and statistical analyses of similarity with 999 random permutations were also utilized to compare the mean of ranked dissimilarities among different communities to the mean of ranked dissimilarities within the same community. The rarefaction curves, Venn gures, circos diagrams revealing microbial genus levels, PCoA plots and unweighted UniFrac clustering were completed with the online platform of Majorbio Cloud Platform (www.majorbio.com).

Declarations
Ethics approval and consent to participate Not applicable Consent for publication Not applicable Availability of data and material The sequencing raw data of bacterial and fungal communities have been submitted to the NCBI, and bacterial and fungal BioProject accession numbers were SRP256129 and SRP256169, respectively.
MYZ and WYZ designed the study. MYZ, YZ and WYZ sampled the soils and plants, determined soil nuitrient contents and N transformation rates, extracted DNA and performed bioinformatics and statisticalanalysis;. MYZ wrote themanuscript, and WJW, SHB and ZHX revised the manuscript. All authors read and approved the nal manuscript.  Impacts of bamboo ages on the biological properties related to rhizospheric soil N cycling. (a) net N mineralization rate, (b) nprA gene abundance, (c) chiA gene abundance, (d) potential nitri cation rate, (e) AOA amoA abundance, (f) AOB amoA gene abundance and (g) urease activity. Vertical bars indicate the maximum and minimum values. Lowercase letters show signi cant differences among the different rhizospheric soils  Impacts of bamboo ages on microbial community structures at different niches. Bacterial community structures in (a) above-ground tissues, (b) roots and (c) rhizospheric soils and fungal community structures in (d) above-ground tissues, (e) roots and (f) rhizospheric soils