Effects of light intensity on the Panax quinquefolius L. growth and soil microbial communities under reduced light spatial heterogeneity: an application to agroforestry systems Authors: Qiaoran Ma, Guobing Tian, Jie Gu, Junwen Chen, Shengchao Yang, Shuhui Zi, Fuseini Issaka, Weizhu Yang, Dehuan Men, Shichao Zheng, Shuran He Published: 2026 ABSTRACT AbstractThe development of under-forest economy within agroforestry systems can enhance the efficiency of land and forest resource utilization, protect the ecological environment, and promote the sustainable development of forests. Forest understory American ginseng (Panax quinquefolius L.) often experiences spatial variability in light availability. However, the impact of light intensity under reduced spatial heterogeneity on soil microenvironments and the regulation of plant physiological traits to promote P. quinquefolius growth remains insufficiently explored. This study adopts field experiments in forest ecosystems to explore how light intensity influences P. quinquefolius physiology, soil biochemical properties, and microbial communities under reduced light spatial heterogeneity. The results indicate that, under reduced spatial heterogeneity with a light transmittance of 13.1%, P. quinquefolius exhibits coordinated physiological responses. The 13.1% light transmittance treatment decreases hydrogen peroxide content, while increasing net photosynthetic rate and chlorophyll content. These changes further improve root dry weight, which is significantly higher than that under 0.7%, 6.4%, and 19.8% light transmittance (with increases of 600.0%, 50.0%, and 5.0%, respectively). Additionally, soil urease activity, pH, and alkali-hydrolyzable nitrogen content are elevated, along with the relative abundance of pH-related microorganisms (e.g., unclassified_Leotiomycetes) and denitrifying bacteria (e.g., unclassified_Gemmatimonadaceae, Anaeromyxobacter, and Bradyrhizobium). The results of the Partial least squares path modeling analysis show that under reduced spatial light heterogeneity in the agroforestry system, light intensity directly promotes the growth of P. quinquefolius by modulating its physiological characteristics. Furthermore, light intensity indirectly fosters the growth of P. quinquefolius by improving soil biological properties (Urease, Sucrase, Catalase, Dehydrogenase, Acid phosphatase) and chemical properties (pH, Alkali-hydrolyzable nitrogen, Available potassium) through the regulation of its physiological characteristics. In conclusion, the optimal light transmittance for P. quinquefolius growth under reduced spatial heterogeneity is 13.1%. These findings offer practical insights for managing P. quinquefolius cultivation in forest ecosystems, contributing to the sustainable development of agroforestry systems.Graphical abstract Supplementary InformationThe online version contains supplementary material available at https://doi.org/10.1186/s12870-026-09315-x.Keywords: Panax quinquefolius L., Light intensity, Spatial heterogeneity, Soil microbial community, Agroforestry Introduction American ginseng derives from the dried roots of Panax quinquefolius L., and is a traditional and precious Chinese herbal medicine [1]. In 2018, the global market for P. quinquefolius was valued at 85 million US dollars [1, 2]. Due to its content of ginsenosides, polysaccharides, volatile oils and other active compounds, it possesses anti-tumor, anti-fatigue and immune-enhancing properties [3, 4]. The main production areas for P. quinquefolius in China include three provinces of Shandong, Jilin and Shaanxi [1, 5]. Currently, the total area under cultivation in China is nearly 3,700 hectares [1]. The mainstream cultivation method for P. quinquefolius is field cultivation, which is a monoculture system requiring the use of shade net to provide the necessary shade [6]. According to the Chinese Pharmacopoeia, only roots that are at least four years old may be used in Chinese medicine. However, long-term cultivation in fields has led to a decline in soil quality, which has exerted a negative impact on the overall quality of P. quinquefolius. To explore and establish sustainable cultivation models, the cultivation of P. quinquefolius within agroforestry systems beneath the forest canopy has rapidly emerged as a new agricultural model in China, Canada and the United States [6–8].This method reduces the use of chemical fertilizers and pesticides by harnessing the benefits of biodiversity and utilizing plant- and microbe-derived materials for P. quinquefolius cultivation [9]. Light heterogeneity is a common phenomenon in agroforestry systems. Light availability for plants varies on a fine scale, such as in forest understory vegetation, intercropping systems in agricultural fields, and aquatic environments [10–12]. Light heterogeneity can be categorized into temporal and spatial variations [13]. Spatial heterogeneity refers to the uneven distribution of light in space [14, 15]. The spatial variability of light is particularly high in forest ecosystems [15, 16]. When plants experience spatially heterogeneous light, photosynthesis in shaded leaves is often limited by insufficient light energy, while non-shaded leaves may experience enhanced photosynthesis to compensate for the shaded areas [17, 18]. This physiological integration, known as photosynthetic compensation, enables plants to balance their resource allocation despite light heterogeneity [17, 19]. As a result, spatial light heterogeneity can drive the accumulation of biomass in light-rich patches, enhancing the efficient use and redistribution of resources within the plant, thus maintaining growth. However, this effect is unlikely to promote substantial growth. Some agricultural studies suggest that reducing light heterogeneity can boost crop yield by providing a more uniform light spectrum and enhancing light energy efficiency [20, 21]. Therefore, understanding how crops adapt to a more uniform light environment after the reduction of spatial heterogeneity holds significant value for optimizing light management in agricultural or agroforestry systems. In most forest ecosystems, the light environment beneath the canopy is typically uneven, characterized by low diffuse light and intermittent direct solar radiation, often referred to as sunflecks [22, 23]. The leaves of understory vegetation absorb substantial solar radiation, creating a scattered light environment, where light levels generally remain below 50 µmol m− 2 s− 1, and the red to far-red light ratio is relatively low [24]. Sunflecks, which intermittently break through diffuse light, create an unpredictable and transient heterogeneous light environment [22, 25]. In agroforestry systems, P. quinquefolius exhibit morphological and growth adaptability to light spatial heterogeneity [8, 22]. After canopy thinning, larger openings enhance light intensity, stimulating photosynthesis in P. quinquefolius, which increases its relative growth rate both in the short and long term, with seed production showing an upward trend up to three years post-thinning [26]. Under uniform light conditions, an appropriate light intensity can increase chlorophyll content and net photosynthetic rate (Pn) in P. quinquefolius, reduce reactive oxygen species (ROS) production, regulate plant signalling molecules (such as gibberellins and auxins) and root exudates (such as organic acids and phenolic compounds), improve the rhizosphere redox environment, and provide a habitat for microorganisms [27, 28]. By regulating the soil environment through the above pathways (e.g. soil mineral nutrients, soil urease (S-UE), transaminase and phosphatase), beneficial bacteria involved in the carbon and nitrogen cycles (such as Micrococcaceae and Xanthobacteraceae) are recruited to the rhizosphere, thereby indirectly influencing the microbial community in the rhizosphere soil. Soil microorganisms promote the growth of P. quinquefolius by releasing bioactive compounds and converting nutrients that are poorly soluble in soil [27, 29].Previous studies have primarily focused on: (1) the relationship between spatial heterogeneity of light and the growth of P. quinquefolius; (2) the relationship between uniform light and the growth of P. quinquefolius. Current agroforestry practices often use transparent plastic films to create rain-avoidance shades, which homogenize light intensity and reduce light spatial heterogeneity [9, 30, 31]. What is the optimal light transmission for the growth of P. quinquefolius under conditions of reduced light heterogeneity in agroforestry systems? It remains unclear how the P. quinquefolius physiologically adapts to such conditions. Light heterogeneity contributes to the diversity of understory vegetation species, which in turn leads to spatial variation in soil properties [32]. It has been demonstrated that light heterogeneity and clonal integration of understory vegetation promote dissolved organic carbon (DOC) and soil enzyme activities (sucrase (S-SC) and acid phosphatase (S-ACP)) in forest soils, while also increasing the Chao1 richness and Shannon diversity of fungal communities [15, 29, 33]. As a result, light heterogeneity significantly affects plant community composition and soil microbial diversity [34]. However, it remains unclear how soil microbial communities and the physiological characteristics of P. quinquefolius respond to different light transmittance levels under conditions of reduced spatial heterogeneity of light. Furthermore, it is not yet clear whether these responses promote the growth of P. quinquefolius by regulating soil properties. This study aims to explore how light transmittances, by modulating the physiological traits of P. quinquefolius and the composition of soil microbial communities, can enhance the growth of understory P. quinquefolius in artificial forest environments characterized by reduced light spatial heterogeneity. Field experiments were conducted in artificial pine forests in southwestern China, where diffused light polyethylene films (PE film) were installed to reduce light spatial heterogeneity. The experiments were designed with four distinct light transmission environments. We hypothesize that, under conditions of reduced spatial heterogeneity of light, the optimal light intensity is likely to lie between 10% and 20%. This not only improves the structure of the soil microbial community, but also reduces oxidative damage in P. quinquefolius and increases its maximum photosynthetic rate. Changes in these three factors can promote the growth of P. quinquefolius by regulating the biological and chemical properties of the soil. Materials and methods Study location, climatic conditions, and soil characteristics In 2022, an experimental site was established at the P. quinquefolius planting base in Luquan County, Yunnan Province, China (longitude 102°21′, latitude 25°39′), situated at an altitude of 2,550 m. Meteorological and soil environmental monitoring was conducted from January to December 2022 at the understory P. quinquefolius test site using an integrated weather station (ATMOS 41, Beijing Tanghua Science and Technology Co., Ltd., China) and data collectors (TH302 and TH5G, Beijing Tanghua Science and Technology Co., Ltd., China). The site is characterized by a highland mountain monsoon climate within the temperate zone. The average annual air temperature is 10.9 °C, with a minimum of -4.5 °C and a maximum of 25.5 °C. The annual average relative air humidity is 79.51%. Soil water content ranges from 60% to 80% of the soil’s maximum water-holding capacity. The average annual soil temperature is 11.64 °C, with a minimum of 3.5 °C and a maximum of 17.4 °C. The cumulative annual temperature above 10 °C totals 3500 °C, and the annual average precipitation is 1,000 mm. The natural soil, derived from a Huashan pine (Pinus armandii Franch.) forest, was used for planting. Three composite surface soil samples (0–20 cm, 200 g per sample) were randomly collected from each plot prior to planting, air-dried, and sieved through a 2 mm soil sieve for further analysis. The chemical properties of the soil were determined following the methods of Guo et al. [35], with the following results: pH: 5.03 ± 0.19; total organic carbon (TOC): 37.11 ± 1.05 g kg− 1; alkali-hydrolyzable nitrogen (AN): 128.3 ± 14.15 mg kg− 1; available phosphorus (AP): 30.68 ± 2.602 mg kg− 1; and available potassium (AK): 415.3 ± 19.50 mg kg− 1. Plant materials and treatments The experimental materials used in this study were seeds of P. quinquefolius, and these seeds were provided by Jilin Songhao Biotechnology Co., Ltd. In December 2021, P. quinquefoliu seeds were sown directly in 20 experimental fields within the P. quinquefoliu planting base located under the Huashan pine forest operated by Yunnan Linshen Biotechnology Co., Ltd. Since the seeds were derived from cultivated materials rather than collected from wild populations, no specific permits are required to provide certified cultivated plant materials for research use with these seeds. All P. quinquefolius plant materials are preserved at the National-Local Joint Engineering Research Center on Germplasm Innovation & Utilization of Chinese Medicinal Materials in Southwest China, Yunnan Agricultural University. Plant nomenclature was updated following the International Plant Names Index (IPNI) [36]. All plant material collection and experimental procedures were performed in accordance with the institutional and national guidelines for plant research, and complied with local laws. This study complies with the principles of the Convention on International Trade in Endangered Species of Wild Fauna and Flora (CITES). Experimental treatment Each plot measured 6 m in length, 1.2 m in width, with a ridge height of 40 cm. To prevent interference between light sources from different treatments, the plots were spaced at least 300 m apart. These plots were randomly distributed within the flat areas of the P. quinquefolius cultivation base. To reduce interference from abiotic factors such as spatial heterogeneity of light and rainfall, a rain shelter was constructed above each experimental plot using a PE film and fibre rods (1 m high, 1.4 m wide). Based on pine tree density, four light levels were selected within the forest. To achieve the lowest light intensity treatment, a shade net was placed over the PE film and covered with pine needles. Please refer to the supplementary document for details on how to install polyethylene film and shade nets (Figure S1). Five replicates were established for all treatments to ensure data reliability. During sampling, two replicates in one treatment were damaged by mole gnawing. Consequently, each treatment was ultimately replicated three times. Light intensity under the PE film was measured using a Li-190R quantum sensor (LI-COR, Lincoln, USA) and a Li-1500 photon data collector (LI-COR, Lincoln, USA). Three sites under the PE film, 5 m apart and at the height of P. quinquefolius leaves, were selected for measurement. Light intensity was recorded every 10 s from sunrise (7:00 AM) to sunset (8:00 PM) on sunny days, expressed in terms of photosynthetically active radiation (PAR) (µmol·m− 2·s− 1). Light transmittance was defined as the percentage of light intensity under the PE film relative to the full sunlight intensity outside the forest. Mean light transmittance of each treatment across all time points served as the light intensity indicator for P. quinquefolius growth. The average environmental light intensities for the four treatments were 0.7%, 6.4%, 13.1%, and 19.8%. Detailed PAR data (maximum, minimum, and average) for these treatments can be found in Supplemental File Table S1. The edges of polyethylene film in each plot should be positioned 20 cm away from the ridges to ensure ventilation and heat dissipation, thereby maintaining a consistent soil temperature. Install soil temperature and moisture sensors throughout the estate so that irrigation can be carried out promptly when soil moisture levels drop, thereby maintaining consistent soil moisture levels. In this study, a data logger (TH5, Beijing Tanghua Science and Technology Co., Ltd., China) and soil temperature and moisture sensors (TEROS 12, Beijing Tanghua Science and Technology Co., Ltd., China) are used to monitor soil temperature and moisture data for all treatment groups, in order to rule out any interference from the PE film on soil temperature and moisture. The data on average soil moisture and average soil temperature for each treatment can be found in Table S2 in the Supplementary Materials. Four ridges (6 m in length, 1.2 m in width, and 40 cm in height) were randomly selected, and three loci beneath the PE film, positioned 7 cm above each ridge (simulating the height of the understory P. quinquefolius leaves), were used as replicates. Loci were spaced 1.5 m apart. The final light intensities of the four ridges were 4.90%, 5.00%, 5.30%, and 5.70%, respectively. Artificial regulation reduced spatial light heterogeneity and yielded a stable light environment characteristic of agroforestry systems. During the cultivation of P. quinquefolius, forest land utilization stood at 54%, and organic fertilizer (Yunnan Yunye Fertilizer Co., Ltd.) (16.2 tonnes / hectare) was applied as a base fertilizer during land preparation. Irrigation management should be based on soil moisture content, maintaining it at 20%–25%. No pesticides or chemical fertilizers are applied, and manual weeding and pest and disease control are required. The primary diseases currently affecting the crop are root rot and rust rot; under this cultivation system, the incidence of both diseases is kept below 3%. The study was conducted over approximately six months, from March 15, 2022, to September 10, 2022. Plant morphology and biomass measurement After four months of exposure to different light intensities (July 15, 2022), whole P. quinquefolius plants were harvested from the understory and immediately transported to the laboratory. Plant height (measured from the root to the petiole base), leaf length, and leaf width were recorded using a ruler. Each plant was then separated into leaves, stems, and roots, washed five times with chromatographic water, and surface-dried. Root length, diameter, surface area, and volume were assessed using an Epson Perfection V850 Pro root scanner (SEKEY, Nagano, Japan) and WinRHIZO software (Regent Instruments, Quebec City, Canada). The leaves, stems, and roots were dried at 105 °C for 30 min, followed by drying at 65 °C to a constant weight, after which the dry weight was measured. Photosynthetic parameters and chlorophyll content measurement Pn, transpiration rate (Tr), stomatal conductance (Gs), and intercellular CO₂ concentration (Ci) were measured using the LI-6400XT Portable Photosynthesis System and LI-6400-02B Red-Blue Light Source Leaf Chamber (LI-COR, Lincoln, USA). Measurements were conducted on a sunny, windless day (July 30, 2022), between 8:30 a.m. and 11:30 a.m., at the midpoint of the experiment. For each treatment, three replicates were taken, each consisting of four uniformly growing understory P. quinquefolius plants. Healthy, fully expanded compound leaves with middle leaflets were selected for analysis. The LI-6400XT was set as follows: air flow rate 300 mL min− 1, photosynthetic photon flux density (PPFD) 500 µmol m− 2 s− 1, CO2 concentration 400 µmol·mol− 1, leaf chamber relative humidity 65% ± 5%, and ambient temperature 25 °C. Once leaves were placed in the leaf chamber, measurements were recorded when Gs, Ci, and Tr were positive, with Ci ranging from 0 to 1. The ΔCO2 variation was stable within 0.5 ppm, and Pn remained steady at one decimal place with no further fluctuation. At this point, gas exchange parameters were determined, and data were recorded three times at 30-second intervals for each leaf [37]. Fresh P. quinquefolius leaves were collected and immediately transported in liquid nitrogen to the laboratory. Under low light or dark conditions, fresh leaf samples (0.1 g) were finely chopped, and chlorophyll was extracted using 80% (v/v) pre-chilled acetone. The extraction process was conducted under low light to prevent degradation. Absorbance at 645 nm and 663 nm was measured using a UV-2450 spectrophotometer (Shimadzu, Kyoto, Japan) and a chemical test kit (Suzhou Grey Biotech Co., Ltd., Suzhou, China). Total chlorophyll (Tot Chl), chlorophyll a (Chl a), and chlorophyll b (Chl b) contents were calculated using the formula provided by the supplier [38]. Leaf proline (Pro), ROS, and antioxidant enzyme activities On July 30, 2022, at the midpoint of the experiment, fresh P. quinquefolius leaves were collected and transported to the laboratory in liquid nitrogen. Three replicate samples were collected with four uniformly growing understory P. quinquefolius leaves taken from each replicate. The content of Pro, hydrogen peroxide (H2O2), and the activity of peroxidase (POD, EC 1.11.1.7) were measured following the methodology outlined by Sun et al. [39]. The Pro, H2O2 and POD measurement kit was provided by Suzhou Grey Biotech Co., Ltd. (Suzhou, China). 0.1 g tissue was homogenized in 1 mL extraction solution in an ice bath, transferred to an EP tube, and extracted in a 90 °C water bath for 10 min. Pro content was determined using a UV-2450 spectrophotometer (SHIMADZU, Japan) at 520 nm. For H2O2 determination, 0.1 g tissue was homogenized in 1 mL pre-cooled acetone in an ice bath, and the volume was made up to 1 mL with acetone. After centrifugation at 12,000 rpm at 4 °C for 10 min, H2O2 content was measured at 415 nm using a UV-2450 spectrophotometer. For POD activity, 0.1 g of sample was mixed with 1 mL of extract, homogenized in an ice bath, and centrifuged at 12,000 rpm for 10 min at 4 °C. POD activity was measured at 470 nm using the same spectrophotometer. Soil sampling and property analysis Rhizosphere soil samples were collected on September 10, 2022. The rhizosphere soil was obtained along with the understory P. quinquefolius plants using the root-shaking method. Lightly shake the root ball until all the loose soil would be removed. Next, shake the soil adhering to the surface of the P. quinquefolius roots (defining the soil within 2 mm of the roots as rhizosphere soil) into a collection bag [40]. The rhizosphere soil samples were immediately sieved through a 2 mm mesh sieve, and all visible plant debris and impurities were removed using sterile forceps [41]. Each treatment included three replicates, with four sampling points selected per replicate. Each replicate sample consisted of a composite sample formed from the rhizosphere soil of 3–5 P. quinquefolius plants. Part of the rhizosphere soil sample was used for microbiological analysis, whilst the remainder was air-dried, ground and prepared for chemical analysis. Soil (sieved through a 2-mm sieve) was mixed with CO2-depleted water at a ratio of 1:2.5, shaken, and allowed to settle. The pH was then measured using a calibrated pHS-3E meter (Leici, Shanghai, China). The chemical properties of the soil were determined following the methods of Guo et al. [35]. The content of AN, AP, and AK was measured following standard procedures: AN was determined using the 1 M NaOH hydrolysis diffusion method; AP was quantified using the 0.5 M NaHCO3 extraction method, as outlined by Stevenson, and AK was assessed using the ammonium acetate extraction method combined with the FP640 flame photometer (SPSIC, Shanghai, China). DOC in the rhizosphere soil was determined according to the procedure described by Sun et al. [42]. Soil enzyme activities, including S-UE (EC 3.5.1.5), S-SC (EC 3.2.1.26), catalase (S-CAT, EC 1.11.1.6), S-ACP (EC 3.2.1.26), and dehydrogenase (S-DHA, EC 1.1.1.1), were assessed using corresponding enzyme activity kits from Suzhou Grey Biotech Co., Ltd. and a UV-2450 spectrophotometer (Shimadzu, Japan) [43, 44]. Detailed protocols are provided in the Supplemental File Text S1. Analysis of rhizosphere soil microbial community structure and diversity To prevent contamination during sample collection, all tools were sterilized with ethanol. Samples were homogenized and stored in liquid nitrogen. Total microbial DNA was extracted from 0.5 g soil samples using a soil DNA kit (Mobio, Carlsbad, USA). The V3 - V4 region of the bacterial 16SrRNA gene was amplified by polymerase chain reaction (PCR) with forwarding primer 308F (5’- ACTCCTACGGGAGGCAGCA − 3’) and reverse primer 806R (5’- GGACTACHVGGGTWTCTAAT − 3’) [45]. The fungal ITS1 region was amplified by PCR with for warding primer ITS1F (5′-CTTGGTCATTTAGAGGAAGTAA-3′) and reverse primer ITS2R (5′-GCTGCGTTCTTCATCGATGC-3′) [46]. PCR products were purified and sequenced on the Illumina NovaSeq platform (Biomarker Technologies Co., Ltd., Beijing, China). Raw reads were quality-filtered using Trimmomatic (v0.33) with a window size of 50 bp; downstream bases were trimmed if the average quality score within the window was below 20. Use Cutadapt (version 1.9.1) to identify and remove primer sequences, using parameters of a maximum mismatch rate of 20% and a minimum coverage of 80%. The double-end reads for each sample were then assembled using USEARCH (version 10), with a minimum overlap length of 10 bp, a minimum similarity of 90% in the overlap region, and a maximum of the 5 bp mismatch (default). Finally, UCHIME (version 8.1) was used to identify and remove chimeric sequences based on the criterion that at least one of the two parent sequences shared a similarity of 80% or more with the query sequence. Amplicon Sequence Variants (ASVs) were generated using QIIME 2 (version 2020.6). The quality of the 16 S and ITS sequencing data was assessed at each stage by analyzing sequence numbers and lengths. Sequences were aligned with the Silva 138 database (http://www.arb-silva.de) for 16 S rRNA and the UNITE 8.2 database (http://unite.ut.ee/index.php) for ITS sequences. Alpha diversity was assessed using four indices: Chao1, ACE, Simpson, and Shannon, to evaluate microbial community richness and diversity. Principal Coordinates Analysis (PCoA) was performed on the full-length 16 S rRNA and ITS sequences. The relative abundance of microbes was calculated as percentages. Pearson correlation analysis was conducted to explore the relationships between soil chemical and biological properties and the 10 most abundant microbial genera at the genus level. All graphical outputs were generated on BMK Cloud (www.biocloud.net). To identify the biomarker bacterial and fungi taxa, a linear discriminant analysis (LDA) effect size (LEfSe) was performed on the OTU table (Kruskal-Wallis sum-rank test P < 0.05, LDA > 3). The correlations between different topological roles of network nodes and various soil chemical properties were determined with the Mantel test and visualized with the “ggcor” package. Bacterial and fungal OTUs were annotated with functional information using BugBase, Functional Annotation of Prokaryotic Taxa (FAPROTAX), and the Fungi Functional Guild (FUNGuild v1.0) software. Fungal communities were classified into guilds or groups within FUNGuild based on their feeding strategies [47]. Modeling and statistical analysis Partial Least Squares Path Modeling (PLS-PM) PLS-PM analysis was performed using the “plspm” package in R (version 4.3.3, http://www.r-project.org/) [9]. Complete R code for PLS-PM is available in Supplementary File Text S2. The transmittance included 0.7% and 6.4%, 13.1%, and 19.8%; Bacterial and fungal communities each included the first principal coordinate (PCoA1) from Pearson distance-based PCoA, as well as α-diversity indices (ACE, Chao1, Simpson, and Shannon).Soil biological properties included S-UE, S-SC, S-CAT, S-DHA, S-ACP; Soil chemical properties included pH, AN, and AK; Photosynthesis included Pn, Tot Chl content, Chl a and Chl b; Oxidative damage included Pro and H2O2; Plant growth included plant height, leaf length, leaf width, root length, root diameter, root surface area, root volume, leaf dry weight, stem dry weight, and root dry weight. The AP, Tr and Gs indicators were excluded based on the criterion that their loadings were less than 0.7. However, as soil pH and the fungal communities ACE, Chao1, Simpson and Shannon showed statistical significance in the correlation analysis, these were retained despite their loadings being less than 0.7. Subsequently, the model’s overall performance was evaluated using the goodness-of-fit (GOF) index. Statistical analysis A one-way ANOVA was performed using SPSS (version 18.0, IBM) to evaluate significant differences (P < 0.05) in the characteristics of P. quinquefolius under varying light transmittance conditions. Post-hoc comparisons were conducted using the Waller-Duncan a-b test. Bar charts and box plots were created using Origin software (version 2021). Results Morphological characteristics and biomass Under conditions of reduced light spatial heterogeneity, the morphological characteristics of P. quinquefolius exhibited a trend of initially increasing and then decreasing with rising light intensity, with optimal growth and biomass accumulation observed at a light transmittance of 13.1% (Table 1). It is worth noting that, compared with light transmittance levels of 0.7%, 6.4% and 19.8%, the surface area of P. quinquefolius roots increased significantly by 241.57%, 78.82% and 59.16% respectively at a light transmittance of 13.1% (P < 0.05). The root volume of P. quinquefolius at a light transmittance of 13.1% increased by 271.43%, 136.36% and 52.94% respectively, compared with light transmittances of 0.7%, 6.4% and 19.8%. Meanwhile, root dry weight reached its maximum at a light transmittance of 13.1%, representing increases of 600.00%, 50.00% and 5.00% compared to light transmittances of 0.7%, 6.4% and 19.8% respectively (P < 0.05). Table 1Effects of light intensities on growth of forest-cultivated P. quinquefolius under reduced spatial light heterogeneityTreatmentsPlant height (cm)Leaf length (cm)Leaf width (cm)Root length (cm)Root diameter (cm)Root surface area (cm2)Root volume (cm3)Root dry weight (g)Stem dry weight (g)Leaf dry weight (g)0.7%6.47 ± 0.06c4.10 ± 0.00d1.80 ± 0.00d6.88 ± 0.07d0.72 ± 0.01d0.89 ± 0.00d0.07 ± 0.00d0.03 ± 0.00d0.01 ± 0.00d0.02 ± 0.00d6.4%7.17 ± 0.06b4.30 ± 0.00c2.00 ± 0.00c10.70 ± 0.18c0.96 ± 0.01c1.70 ± 0.07c0.11 ± 0.00c0.14 ± 0.00c0.02 ± 0.00c0.04 ± 0.00c13.1%8.13 ± 0.06a5.17 ± 0.06a2.60 ± 0.00a16.57 ± 0.17a1.84 ± 0.01a3.04 ± 0.05a0.26 ± 0.00a0.21 ± 0.00a0.02 ± 0.00a0.05 ± 0.00a19.8%7.57 ± 0.06b4.53 ± 0.06b2.20 ± 0.00b13.41 ± 0.21b1.05 ± 0.02b1.91 ± 0.04b0.17 ± 0.01b0.20 ± 0.01b0.02 ± 0.00b0.05 ± 0.00bResults are presented as the mean ± SD (n = 3). Percentage indicates light intensity. Different lowercase letters denote significant differences between treatments (P < 0.05) Effects of light intensities on growth of forest-cultivated P. quinquefolius under reduced spatial light heterogeneity Results are presented as the mean ± SD (n = 3). Percentage indicates light intensity. Different lowercase letters denote significant differences between treatments (P < 0.05) Photosynthesis In the agroforestry system, under conditions of reduced light spatial heterogeneity, the Pn, Tr, and Gs of P. quinquefolius exhibited a trend of initially increasing and then decreasing as light intensity rose (Fig. 1A-C). Pn reached its maximum value at a light transmittance of 13.1%, representing significant increases of 197.80%, 79.51%, and 33.49% compared to light transmittances of 0.7%, 6.4%, and 19.8%, respectively (P < 0.05). Meanwhile, the content of photosynthetic pigments (Total Chl, Chl a, and Chl b) in P. quinquefolius leaves initially increased and then decreased as light intensity rose, with the most significant increase occurring at a light transmittance of 13.1% (Figure S2 A-C, P < 0.05). As light intensity increased, Ci exhibited a trend of first decreasing and then increasing, reaching a minimum at a light transmittance of 13.1%. Compared with light transmittances of 0.7%, 6.4%, and 19.8%, the values decreased by 30.28%, 22.74%, and 9.56%, respectively (Fig. 1D, P < 0.05). Effects of light intensities on photosynthesis and oxidative defense of forest-cultivated P. quinquefolius under reduced spatial light heterogeneity. A Net photosynthetic rate (Pn); (B) Transpiration rate (Tr); (C) Stomatal conductance (Gs); (D) Intercellular CO2 concentration (Ci); (E) Proline (Pro); (F) Hydrogen peroxide (H2O2); (G) Peroxidase (POD). The data of the above parameters are mean ± SD (n = 3). Different letters above the plots indicate significant differences between treatments (P < 0.05) Antioxidant defense system In the agroforestry system, under conditions of reduced spatial heterogeneity of light, the antioxidant defense system in P. quinquefolius leaves exhibited significant changes as light intensity increased (Fig. 1E-G). The Pro content in P. quinquefolius leaves decreased most significantly at a light transmittance of 0.7%, followed by 13.1% (P < 0.05). POD activity increased most significantly at a light transmittance of 19.8%, followed by 13.1% (P < 0.05). The H2O2 content showed a trend of first decreasing and then increasing with increasing light intensity, and decreased significantly at a light transmittance of 13.1%, with reductions of 25.50%, 13.37%, and 16.95% compared to light transmittances of 0.7%, 6.4%, and 19.8%, respectively. Properties of the rhizosphere soil Biological properties of rhizosphere soil In agroforestry systems, under conditions of reduced spatial heterogeneity of light, light intensity significantly alters the biological properties of the soil in the rhizosphere of P. quinquefolius (Fig. 2). The DOC content in the rhizosphere soil increased significantly at a light transmittance of 13.1%, rising by 134.21%, 54.56%, and 33.33% compared to light transmittances of 0.7%, 6.4%, and 19.8%, respectively (Fig. 2A, P < 0.05). The concentrations of S-UE, S-SC, S-CAT, S-DHA, and S-ACP in the rhizosphere soil initially increased and then decreased as light intensity rose, with the maximum increase observed at a light transmittance of 13.1% (Fig. 2B-F, P < 0.05). In summary, under conditions of reduced light-space heterogeneity, a light transmittance of 13.1% can increase DOC levels and enzyme activity in the rhizosphere soil, thereby providing an adequate carbon source and creating conditions for efficient nutrient transformation to support plant growth. Effects of light intensities on rhizosphere soil biological properties of forest-cultivated P. quinquefolius under reduced spatial light heterogeneity. A Dissolved organic carbon (DOC); (B) Soil urease (S-UE); (C) Soil sucrase (S-SC); (D) Soil catalase (S-CAT); (E) soil dehydrogenase (S-DHA); (F) Soil acid phosphatase (S-ACP). Different letters above the plots indicate significant differences between treatments (P < 0.05) Chemical properties of rhizosphere soil In agroforestry systems, under conditions of reduced spatial heterogeneity of light, light intensity significantly alters the chemical properties of the soil in the rhizosphere of P. quinquefolius (Fig. 3A-D). Soil pH increased with increasing light intensity, reaching a maximum at 19.8%, followed by 13.1%. Compared to light transmittance levels of 0.7%, 6.4%, and 13.1%, pH increased by 8.49%, 9.32%, and 7.13%, respectively, at 19.8%. The concentrations of AN and AK in the soil first increased and then decreased as light intensity rose, reaching their peak at a light transmittance of 13.1% (P < 0.05). AP content decreased significantly at a light transmittance of 6.4% (P < 0.05), but there were no significant differences at light transmittances of 0.7%, 13.1%, and 19.8%. In summary, under conditions of reduced light heterogeneity, a light transmittance of 13.1% can improve the chemical properties of the rhizosphere soil, thereby providing suitable nutrient conditions for the growth of P. quinquefolius. Effects of light intensities on rhizosphere soil chemical properties of forest-cultivated P. quinquefolius under reduced spatial light heterogeneity. A Soil pH value (pH); (B) Soil alkali-hydrolyzable nitrogen (AN); (C) Available phosphorus (AP); (D)Available potassium (AK). Different letters above the plots indicate significant differences between treatments (P < 0.05) Bacterial and fungal communities of rhizosphere soil Fungal and bacterial community diversity of rhizosphere soil We evaluated the effects of light intensity on the richness and diversity of bacterial and fungal communities in the rhizosphere soil of P. quinquefolius under reduced light spatial heterogeneity using four α-diversity indices (ACE, Chao 1, Simpson, and Shannon). The bacterial communities ACE, Chao 1, Simpson, and Shannon showed a trend of first increasing and then decreasing as light intensity increased. There was no significant difference in bacterial α-diversity at light transmittance levels of 0.7%, 6.4%, and 13.1%, but it was significantly reduced at 19.8% (P < 0.05, Fig. 4A-D). There were no significant differences in fungal diversity as measured by the ACE, Chao 1, Simpson, and Shannon indices across the four light transmittance levels. However, the ACE and Chao 1 indices were the highest at a transmittance of 6.4% (Figs. 5A-B), while the Simpson and Shannon indices reached their maximum values at a transmittance of 19.8% (Figs. 5C-D). Effects of light intensities on soil bacterial diversity and relative abundance of forest-cultivated P. quinquefolius under reduced spatial light heterogeneity. A Bacterial community ACE index; (B) Bacterial community Chao1 index; (C) Bacterial community Simpson index; (D) Bacterial community Shannon index; (E) Bacterial community PCoA; (F) The top 10 species with the highest relative abundance on the bacterial genus level. G Bar chart of metabolic functions predicted by FAPROTAX based on the relative abundance of bacterial groups in rhizosphere soil under different light intensities. (n = 3) Effects of light intensities on soil fungal diversity and relative abundance of forest-cultivated P. quinquefolius under reduced spatial light heterogeneity. A Fungal community ACE index; (B) Fungal community Chao1 index; (C) Fungal community Simpson index; (D) Fungal community Shannon index; (E) Fungal community PCoA; (F) The top 10 species with the highest relative abundance on the fungal genus level. G Bar chart of metabolic functions predicted by FUNGuild based on the relative abundance of fungal groups in rhizosphere soil under different light intensities. (n = 3) Bacterial and fungal community structure of rhizosphere soil The structural changes in bacterial and fungal OTUs within the rhizosphere soil of P. quinquefolius were analyzed using PCoA and Analysis of Similarity (ANOSIM). The PCoA results indicated that the rhizosphere bacterial communities (ANOSIM: R = 0.707, P = 0.001) and fungal communities (ANOSIM: R = 0.364, P = 0.02) differed significantly across the four light intensities (Figs. 4E and 5E). PCoA results indicated that PCoA 1 and PCoA 2 explained 74.76% of the variance in bacterial communities (65.22% and 9.54%, respectively) and 51.13% of the variance in fungal communities (30.96% and 20.17%, respectively). The bacterial communities in the rhizosphere soil at 0.7%, 6.4%, and 13.1% light transmittance were significantly separated along axis 2, while bacterial communities at 0.7%, 6.4%, 13.1%, and 19.8% were significantly separated along axis 1. The bacterial communities at 0.7%, 6.4%, and 13.1% were also significantly separated along axis 2, while fungal communities at 0.7%, 6.4%, 13.1%, and 19.8% were significantly separated along axis 1. The 10 most abundant bacterial genera in the rhizosphere soil of P. quinquefolius were Anaeromyxobacter, Bradyrhizobium, Candidatus_Solibacter, unclassified_Acidobacteriales, unclassified_Bacteria, unclassified_ Gemmatimonadaceae, unclassified_Subgroup_2, unclassified_Vicinamibacterales, unclassified_Xanthobacteraceae, and uncultured_Acidobacteria_bacterium (Fig. 4F). Among these, unclassified_Xanthobacteraceae, unclassified_Acidobacteriales, and unclassified_Gemmatimonadaceae dominated the bacterial populations in the rhizosphere soil. The relative abundance of unclassified_Xanthobacteraceae and unclassified_Acidobacteriales peaked at 19.8%, but was lowest at 13.1%. Notably, the relative abundance of uncultured_Acidobacteria_bacterium and unclassified _Subgroup_2 significantly increased at 19.8%. The relative abundances of unclassified_ Vicinamibacterales and unclassified_Gemmatimonadaceae decreased at a light transmittance of 19.8%. The 10 most abundant fungal genera in the rhizosphere soil of P. quinquefolius were Archaeorhizomyces, Humicola, Mortierella, Peziza, Wilcoxina, Conocybe, unclassified_Ascomycota, unclassified_Fungi, unclassified_Helotiales, and unclassified_Leotiomycetes (Fig. 5F). Among these, Mortierella, Humicola, and Peziza were the dominant fungal populations. The relative abundance of Mortierella was highest at 19.8%, followed by 13.1%. Humicola reached its peak relative abundance at 13.1%, but was absent at 19.8%. The relative abundance of Peziza increased initially and then decreased with rising light transmittance, peaking at 6.4%, but was undetected at 19.8%. Notably, the relative abundance of unclassified_Leotiomycetes and unclassified_Helotiales significantly increased at 19.8%. The relative abundances of Archaeorhizomyces, Wilcoxina, and Conocybe decreased at a light transmittance of 19.8%. Using LEfSe, we analyzed differences in the microbial communities of the rhizosphere soil of P. quinquefolius under varying light intensities. At light transmittance levels of 0.7%, 6.4%, 13.1%, and 19.8%, a total of 58 bacterial genera with LDA thresholds ≥ 3 exhibited statistically significant differences (Figure S3A). Soil samples with light transmittance levels of 13.1% and 19.8% were rich in a wide variety of bacterial species, whereas samples with light transmittance levels of 0.7% and 6.4% contained only 5 and 7 abundant bacterial lineages, respectively. Specifically, the potential biomarkers in rhizosphere soil with a light transmittance of 0.7% were unclassified_Micromonosporaceae, Acidothermaceae, Acidothermus, and JG30_KF_AS9. The potential biomarkers identified in 6.4% of the rhizosphere soil were Solirubrobacteraceae, Solirubrobacterales, unclassified_IMCC26256, and Conexibacter. The potential biomarkers in the rhizosphere soil with a light transmittance of 13.1% were Acidimicrobiia, Holophagae, and Dehalococcoidia. Potential biomarkers in the rhizosphere soil with a light transmittance of 19.8% were Acidobacteria and Gammaproteobacteria. LEfSe analysis revealed that, at light transmittance levels of 0.7%, 6.4%, 13.1%, and 19.8%, a total of 21 fungal genera with LDA thresholds ≥ 3 exhibited statistically significant differences (Figure S3B). At a light transmittance of 0.7%, Lactarius, Chaetosphaeriaceae, and Chaetosphaeriales were potential biomarkers. At a light transmittance of 6.4%, Trechisporales and Archaeosporomycetes were potential biomarkers. The potential biomarkers in rhizosphere soil with a light transmittance of 13.1% were Botryotrichum and Rozellomycotina_cls_Incertae_sedis. The primary potential biomarker in the rhizosphere soil with a light transmittance of 19.8% was Tricholoma. Predicting the functions of bacterial and fungal communities in rhizosphere soil Predicting the functions of soil bacterial communities using the FAPROTAX database (Fig. 4G). The results indicated that, within the agroforestry system, there were differences in key metabolic pathways among the rhizosphere soils of P. quinquefolius under different light intensities. Under optimal light intensity (13.1% light transmittance), the bacterial community in the rhizosphere soil of P. quinquefolius grown under forest canopies underwent a significant functional shift, characterized by increases in Nitrogen_fixation, Ureolysis, Predatory_or_exoparasitic, Phototrophy, Photoheterotrophy, and Nitrate_reduction. Conversely, there were decreases in Animal_parasites_or_symbiontsa and Cellulolysis. Chemoheterotrophy, Aerobic_chemoheterotrophy increased with increasing light intensity and exhibited higher relative abundances at a light transmittance of 19.8%. Nitrogen_fixation initially increased and then decreased as light intensity rose, maintaining a relatively high level at a light transmittance of 6.4%. In summary, the shift in the rhizosphere bacterial community from a saprotrophic metabolism focused on hard-to-degrade organic matter to a biotrophic metabolism characterized by efficient nitrogen cycling, light-assisted energy utilization, and enhanced microbial interactions is beneficial for nutrient uptake by P. quinquefolius roots. Functional prediction of soil fungal communities using the FUNGuild database (Fig. 5G). The results indicated that under appropriate light intensity (13.1% light transmittance), the fungal communities in the rhizosphere soil beneath the P. quinquefolius plants maintained high relative abundances of Undefined Saprotroph, Wood Saprotroph, Soil Saprotroph, Animal Pathogen, Dung Saprotroph and Arbuscular Mycorrhizal fungi. However, they maintained relatively low relative abundances in the Ectomycorrhizal, Fungal Parasite, Plant Pathogen, and Lichenized groups. Under high light intensity (19.8% light transmittance), Ectomycorrhizal, Fungal Parasite, Plant Pathogen, Lichenized, Undefined Saprotroph, Wood Saprotroph, Soil Saprotroph, and Arbuscular Mycorrhizal maintained high abundances. This suggests that high light intensity can lead to an accumulation of pathogenic microorganisms in the soil, a reduction in rhizosphere saprotrophic activity, and a shift in mycorrhizal symbiosis patterns. In summary, an appropriate light intensity (13.1%) is more conducive to the formation of a rhizosphere soil microenvironment dominated by saprotrophic decomposition and arbuscular mycorrhizal symbiosis, thereby reducing the presence of pathogenic microorganisms. Relationship between bacterial and fungal genera and soil biochemical properties Correlations between the 10 most abundant fungal and bacterial genera and soil chemical and biological properties were analyzed (Fig. 6A-D). Bacterial Genera and Soil Biochemical Properties: As shown in Fig. 6A, the relative abundances of unclassified_Bacteria, unclassified_Gemmatimonadaceae, and unclassified_Vicinamibacterales were significantly negatively correlated with soil pH. In contrast, the relative abundances of unclassified_Xanthobacteraceae, uncultured_Acidobacteria_bacterium, and unclassified_Subgroup_2 exhibited significant positive correlations with soil pH. Additionally, the relative abundances of Bradyrhizobium and Anaeromyxobacter were significantly negatively correlated with S-UE activity. The relationships among microbial communities, soil properties, plant physiology and plant growth. A Correlation analysis between environmental factors and relative abundances of the ten dominant bacteria based on pearson correlation coefficients; B Correlation analysis between environmental factors and relative abundances of the ten dominant fungi based on pearson correlation coefficients; C Correlation analysis between environmental factors and diversity of the ten dominant bacteria based on pearson correlation coefficients; D Correlation analysis between environmental factors and diversity of the ten dominant fungi based on pearson correlation coefficients. (*P < 0.05, **P < 0.01, ***P < 0.001, n = 3); E Partial Least Squares Path Model (PLS-PM) illustrates light intensity’s effects on the plant-soil-microbesystem. R2 represents the variance of the dependent variable explained by the model; The blue and red arrows represent positive and negative correlations, respectively; The thickness of the arrow represents the size of the standardized path coefficient, *P < 0.05, **P < 0.01, ***P < 0.001 Fungal Genera and Soil Biochemical Properties: Fig. 6B illustrates the relationship between fungal genera and soil biochemical properties. The relative abundance of unclassified_Leotiomycetes showed a significant positive correlation with soil pH levels. Microbial Diversity and Soil Biochemical Properties: The study also examined the correlations between microbial diversity in the rhizosphere and soil chemical and biological properties. The bacterial community diversity indices (ACE, Chao1, Simpson, and Shannon) showed significant negative correlations with soil pH (Fig. 6C). Regarding fungal community diversity, significant negative correlations were observed between the ACE and Chao1 indices and pH, with the ACE index displaying a significant negative correlation with AN (Fig. 6D). The Simpson index was significantly positively correlated with AN, AK, DOC, S-UE, S-SC, S-CAT, S-DHA, and S-ACP. Additionally, the Shannon index showed significant positive correlations with AK, DOC, S-SC, S-CAT, S-DHA, and S-ACP. Mantel-test analysis indicated that bacterial community structure and α-diversity were significantly correlated with soil pH (Figure S4). In contrast, no significant correlation was found between fungal community structure and α-diversity and soil biochemical properties. Factors affecting the growth of under forests P. quinquefolius PLS-PM analysis further clarified the relationships among light intensity, plant photosynthesis, oxidative damage, soil biochemical properties, soil microbial communities, and P. quinquefolius growth under reduced light spatial heterogeneity. The goodness-of-fit (GOF) was 0.6165, meeting the fit requirements for the PLS-PM model in field experiments (Fig. 6E) [48]. Optimal light intensity positively affected plant photosynthesis (0.4749) and negatively affected oxidative damage (− 0.2766). Photosynthesis (0.4356, P < 0.01) and oxidative damage (− 0.1811, P < 0.05) significantly affected the growth of P. quinquefolius. Meanwhile, photosynthesis (0.9645, P < 0.05) and oxidative damage (0.1097, P < 0.05) directly promoted soil biological properties, which in turn indirectly affected soil chemical properties (0.7110, P < 0.01). Soil chemical properties had a direct and significant positive effect on the growth of P. quinquefolius (0.5572, P < 0.001). Soil biological properties directly enhanced soil bacterial (0.0077) and fungal communities (0.5347), which indirectly regulated soil chemical properties (bacteria: −0.3648; fungi: 0.2090) and promoted understory P. quinquefolius growth. In addition, bacterial communities (0.0688) and fungal communities (0.0322) could have a direct positive effect on the growth of P. quinquefolius grown in forest understory. In summary, under reduced light spatial heterogeneity, optimal light intensity promotes P. quinquefolius growth by directly enhancing photosynthesis and antioxidant capacity, and indirectly improving rhizosphere soil biochemical properties and microbial communities. Discussion Under conditions of reduced light spatial heterogeneity, distinct differences were observed in the response of understory P. quinquefolius and soil microbial community structures to light intensity. Notably, the synergistic effects of reduced light spatial heterogeneity and improved soil properties provided direct benefits for selecting the optimal light intensity for P. quinquefolius, thereby facilitating its growth in forested environments. The relationship between light intensity and the growth of under forests P. quinquefolius An appropriate light intensity (13.1% light transmittance) under reduced light spatial heterogeneity conditions promoted the growth of understory P. quinquefolius. Consistent with our findings, a study demonstrated that at 6%, 9%, 13%, and 17% light transmittance, root weight of P. quinquefolius was maximized at 13% light transmittance [49]. However, research conducted at the University of Guelph in Ontario, Canada, reported an optimal light intensity for P. quinquefolius growth at 35.6% light transmittance under greenhouse conditions [50]. This discrepancy may stem from differences in altitude, as the experimental site in our study was situated at an elevation of 2550 m, whereas Guelph University is at 376 m [51]. Higher altitudes are typically associated with stronger ultraviolet radiation. When ultraviolet radiation levels are high, the growth of P. quinquefolius is inhibited, whereas lower light intensities strike a balance between sufficient light availability and reduced UV damage. In 2022, the annual average temperature in Guelph, Ontario, Canada, was 7.27 °C; the annual minimum temperature was − 25.6 °C; the annual maximum temperature was 32.8 °C; and the annual average relative humidity was 72.20% (Source: https://climate.weather.gc.ca/). With the exception of January to March, the monthly average temperatures at this experimental site are broadly similar to those in Guelph, Ontario, Canada, during the remaining months; however, there is a significant difference in the monthly average relative humidity between the two locations (Figure S5). Therefore, humidity may be the primary factor responsible for this difference. Low air humidity can restrict the growth of P. quinquefolius by regulating leaf stomatal characteristics, photosynthetic systems, plant carbon balance and nutrient uptake [52, 53]. The relationship between light intensity and the physiological characteristics of under forests P. quinquefolius Excessive or insufficient light intensity often inhibits plant growth through the integration or imbalance of internal physiological characteristics. One prominent example is the suppression of Pn and chlorophyll content (Fig. 1A and Figure S2). The PLS-PM results indicate that the growth of P. quinquefolius under forest canopies can be enhanced by directly improving photosynthesis at the optimal light intensity (Fig. 6E). Photosynthesis is a fundamental process in plants, directly linked to growth and biomass production. Studies have shown that compared to optimal light conditions, P. quinquefolius experiences a significant decline in photosynthetic performance under high light intensity, accompanied by marked reductions in root dry weight, leaf area, and plant height. Similarly, exposure to low light intensity results in significant reductions in both root and leaf dry weight [50, 54]. Consequently, the suppression of plant growth under unsuitable light intensities could be associated with the photosynthesis process. Under optimal light intensity (13.1% transmittance), decreased H2O2 content (a typical ROS indicator, Fig. 1F) indicated that this light regime alleviated root damage in P. quinquefolius by reducing oxidative stress. PLS-PM analysis revealed that under optimal light intensity, P. quinquefolius growth was directly promoted by reduced oxidative damage (Fig. 6E). Pro accumulation was a common plant adaptive response to alleviate stress. Pro content was higher at 0.7% and 19.8% light transmittance but lower at 13.1% (Fig. 1E). As an osmotic regulator, Pro stabilized the antioxidant system and maintained cell membrane integrity, thereby mitigating ROS damage. Research showed that Pro content peaked under extreme light intensities (27% and 100%), where stress was most pronounced [55]. Reduced Pro accumulation at 13.1% light transmittance indicated alleviated stress and more effective growth. Moreover, under 13.1% light transmittance, activated antioxidant enzyme systems in P. quinquefolius leaves elevated POD activity (Fig. 1G), facilitating ROS scavenging. Excessive ROS disrupted redox homeostasis, triggering irreversible DNA damage and cell death. POD decomposed H2O2 into H2O and O2, neutralizing ROS and sustaining normal plant physiology. The elevated POD activity may arise from the upregulation of POD gene expression [56]. Consistent trends were observed in P. quinquefolius roots, with 11.5% light transmittance yielding greater POD activity relative to 29.8% and 0.2% transmittance [57]. The relationship between light intensity and soil biological properties Soil enzymes are integral to soil energy flow and nutrient cycling, serving as key indicators of soil fertility and quality. PLS-PM analysis indicated that optimal light intensity could improve soil biological properties (DOC, S-UE, S-SC, S-CAT, S-DHA, and S-ACP) by directly influencing plant photosynthesis and oxidative damage (Fig. 6E). Furthermore, existing studies have shown that the changes in soil enzyme activity caused by light exposure have a significant impact on plant growth, which is consistent with our research findings [58]. This may be because the optimal light intensity improves the photosynthetic characteristics of P. quinquefolius, promotes the secretion of root exudates and the release of carbohydrates, and directly regulates soil enzyme activity [58]. Soil enzymes can convert unavailable nutrients into available nutrients. S-UE plays a crucial role in converting organic nitrogen sources, such as urea, into inorganic nitrogen (ammonium ions) that plants can absorb [59]. S-SC, S-CAT, and S-DHA are key enzymes involved in the decomposition of soil organic matter (SOM), playing a pivotal role in its mineralization [60]. The results of this study also indicated that the light transmittance of 13.1% increased not only soil enzyme activities but also AN content (Fig. 6E). Therefore, under conditions of reduced light spatial heterogeneity in agroforestry systems, P. quinquefolius responds to optimal light intensity by physiologically coordinating soil nutrient transformation processes (such as nutrient conversion catalyzed by enzymes), thereby increasing soil nutrient content. Furthermore, the small sample size in this study may have led to bias in the results of the PLS-PM model analysis. It is recommended that future studies of this kind increase the sample size to improve the reliability of the model fit. The impact of light intensity on soil microbial communities and soil chemical properties Optimal light intensity can create a favorable soil environment, which may influence the soil microbial community. In the present study, a reduction in light spatial heterogeneity led to increased bacterial richness and diversity at 6.4% and 13.1% light transmittance, as indicated by indices such as ACE, Chao 1, Simpson, andv Shannon (Figs. 4A-D and 5A-D). Due to the high diversity of soil microorganisms, challenges in cultivation, and limitations of available databases, some taxa’s 16 S rRNA/ITS sequences had poor matches with existing databases, resulting in incomplete or unclassified taxonomic annotations, a common issue in complex environments such as forest soils [61]. Despite the presence of unclassified taxa, PCoA and ANOSIM analyses revealed significant differences in microbial community structure between light treatments (Figs. 4E and 5E). Phenolic compounds in forest soils are the main causes of soil acidity and can be degraded by soil microorganisms [62, 63]. In this study, soil bacteria negatively affected soil chemical properties, whereas soil fungi exerted positive effects (Fig. 6E). Soil pH was maximum at 19.8% light transmittance, followed by 13.1% (Fig. 3A). Interactions between phenolic compounds and bacteria were stronger than those with fungi [64]. In agroforestry systems, the dominant rhizosphere microbial biomarkers under 13.1% light transmittance were Acidimicrobiia and Botryotrichum (Figure S3A and B). Reduced soil pH promoted the enrichment of Acidimicrobiia, thereby facilitating the decomposition and utilization of soil recalcitrant organic carbon [65]. Botryotrichum could synthesize acetic acid by anaerobic pathways [66]. Moreover, under optimal light intensity, the relative abundance of several lignin-degrading microbes (e.g., unclassified_Acidobacteria_bacterium, unclassified_Acidobacteriales, unclassified_Leotiomycetes) increased, enhancing lignin degradation and thus elevating soil phenolic compounds [67–69]. Functional prediction showed that optimal light intensity maintained low relative abundance of rhizosphere fungi in the plant pathogen pathway, whereas high light intensity resulted in high abundance (Fig. 5G). In the soil used for P. quinquefolius cultivation, an increase in phenolic acid content could inhibit certain fungal pathogens [70]. In agroforestry systems, optimal light intensity can increase the relative abundance of Bradyrhizobium and the nitrogen-fixing bacterium unclassified_Gemmatimonadaceae in the rhizosphere soil of P. quinquefolius (Fig. 4F). Bradyrhizobium contains multiple urease gene clusters that catalyze the hydrolysis of urea into CO2 and nitrogen [71, 72]. In this study, the highest levels of AN content and S-UE activity in the rhizosphere soil were observed at the optimal light intensity (13.1%) (Figs. 2B and 3B). This may be due to the promotion of symbiosis between understory P. quinquefolius and Bradyrhizobium at this light intensity. The unclassified_Gemmatimonadaceae was an important group of nitrogen-fixing microorganisms in soil [73]. In this study, the optimal light intensity maintained high levels of Nitrogen_fixation, Ureolysis, Nitrate_reduction and Mycorrhizal relative abundance in the rhizosphere soil (Figs. 4 and 5G). Therefore, in agroforestry systems, changes in light intensity can reshape the microbial communities in the rhizosphere, and regulate soil pH and nitrogen cycling processes, thereby promoting the growth of P. quinquefolius. Conclusion This study reveals that, under conditions of reduced light spatial heterogeneity in agroforestry systems, 13.1% is the optimal light transmission for promoting P. quinquefolius growth. Under these conditions, a light transmittance of 13.1% regulated the following characteristics of the P. quinquefolius–soil system: (1) this reduced oxidative damage to P. quinquefolius and increased its maximum photosynthetic rate and root dry weight; (2) it increased the DOC content in the rhizosphere soil, as well as the activity of S-UE, S-SC, S-CAT, S-DHA, and S-ACP; (3) increased the AN and AK content in the rhizosphere soil; (4) The microbial composition of the rhizosphere soil was improved, with unclassified_ Gemmatimonadaceae,Bradyrhizobium, Anaeromyxobacter, unclassified_Vicinamibacterales, Humicola being the dominant phyla in the soil; LEfSe analysis revealed that the bacterial biomarker in the soil is unclassified_Bacteria, Acidimicrobiia,Holophagae and Dehalococcoidia. The fungal biomarker are Botryotrichum and Rozellomycotina_cls_Incertae_sedis; (5) Microorganisms associated with the soil nitrogen cycle have adapted well within this system. PLS-PM analysis indicates that, under conditions of reduced spatial heterogeneity of light in agroforestry systems, light intensity indirectly promotes the growth of P. quinquefolius by regulating its physiological characteristics and improving the biological properties (S-UE, S-SC, S-CAT, S-DHA, S-ACP) and chemical properties (pH, AN, AK) of the soil. Supplementary Information Supplementary Material 1. Acknowledgements The authors would like to thank all the practitioners who took part in this study for their support. The authors would like to appreciate Yunnan Linshen Biotechnology Co., Ltd. for providing experimental field plots, and Yongli Zhou and Xiongkun Li of this Co. Ltd. for providing suggestions on the control of Panax quinquefolius L. pests and diseases. We would like to thank Ke Li and Tiancai Wang of Yunnan Shenlin Bencao Biotechnology Co., Ltd. for providing suggestions on water management in the field. Abbreviations Panax quinquefolius L. Linear discriminant analysis effect size Soil organic matter Diffused light polyethylene film Principal coordinates analysis Photosynthetically active radiation Photosynthetic photon flux density Net photosynthetic rate Transpiration rate Stomatal conductance Intercellular CO2 concentration Total chlorophyll Chlorophyll a Chlorophyll b Proline Reactive oxygen species Hydrogen peroxide Peroxidase Total organic carbon Alkali-hydrolyzable nitrogen Available phosphorus Available potassium Dissolved organic carbon Soil urease Sucrase Catalase Acid phosphatase Dehydrogenase Partial least squares path modeling Authors’ contributions QRM, and GBT: writing - original draft; JG, SCY, WZY, and SRH: methodology; JWC, SCY, SHZ, FI and SRH: writing - review & editing; JWC, DHM, and SCZ: investigation; SCY, and SRH: resources; SCY, and SRH: supervision; SRH: funding acquisition. All the authors reviewed and approved the manuscript. Funding This work was financially supported by the Central Finance Demonstration Fund for Forest and Grass Science and Technology Promotion (No. YUN [2026]TG01), National Key R&D Program of China (No. 2021YFD1601003), Yunnan Fundamental Research Projects (No.202401BD070001-014), Construction Project for Postgraduate Tutor Team of Yunnan Province in 2024 (Yunnan Academic Degrees Committee Document [2024] No. 19). Data availability The datasets generated and/or analysed during the current study are available in the Figshare repository, https://doi.org/10.6084/m9.figshare.28366283.v1 (Original data). Declarations Ethics approval and consent to participate Not applicable. Consent for publication Not applicable. Competing interests The authors declare no competing interests. Footnotes Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. Qiaoran Ma and Guobing Tian contributed equally to this work. References Associated Data Supplementary Materials Supplementary Material 1. Data Availability Statement The datasets generated and/or analysed during the current study are available in the Figshare repository, https://doi.org/10.6084/m9.figshare.28366283.v1 (Original data).