Effects of rock phosphate on arbuscular mycorrhizal fungi-associated bacterial communities and their contribution to phosphorus acquisition in leek

Abstract

Background

Inter-kingdom interactions between arbuscular mycorrhizal fungi (AMF) and bacteria are increasingly recognized for their potential to enhance fertilizer use efficiency in agroecosystems. Here, we investigated the effects of rock phosphate amendment and AMF inoculation on phosphorus (P) nutrition in leek (Allium porrum L.), as well as on bacterial communities associated with AMF extraradical mycelium. A bi-compartmental microcosm was used to disentangle root-derived effects from those mediated by AMF mycelium.

Results

Inoculation with Rhizophagus irregularis significantly increased total plant biomass (p < 0.001), while rock phosphate amendment enhanced arbuscule abundance in roots (p = 0.03), leading to higher shoot P content (p = 0.013) and photosynthetic activity (p < 0.0001). Because rock phosphate was the sole P source, these results indicate that P solubilized in the soil was translocated to the host plant via the mycorrhizal pathway. Rock phosphate amendment also significantly altered the composition of bacterial communities associated with AMF mycelium (p = 0.01). Across treatments, bacterial assemblages were dominated by Planctomycetota, Pseudomonadota, Chloroflexota, and Bacillota, with enrichment of Planctomyces and Gemmata in AMF mycelium, and Planctomyces, Gemmata, and Bacillus in soil. The core bacteriome associated with R. irregularis was primarily composed of Planctomycetota and Bacillota, taxa known to form biofilms on AMF extraradical hyphae.

Conclusion

These findings demonstrate the pivotal role of mycorrhizal symbiosis in enhancing P acquisition from rock phosphate and provide new insights into AMF–bacteria interactions that are relevant for developing sustainable fertilization strategies.

Supplementary Information

The online version contains supplementary material available at https://doi.org/10.1186/s40793-026-00934-3.

Introduction

Phosphorus (P) is a critical but often limiting nutrient in agricultural soils due to its strong fixation in soil matrices and low bioavailability for plant uptake. As a result, modern agroecosystems rely heavily on phosphate-based fertilizers to sustain crop productivity, raising concerns about long-term soil fertility, resource depletion, and environmental impacts [1]. These challenges highlight the need for more sustainable strategies that improve P acquisition efficiency rather than increasing fertilizer inputs.

Plant-associated microbial communities represent a promising but underexploited component of nutrient acquisition systems. In particular, the root microbiome plays a central role in regulating nutrient cycling, plant productivity, and stress resilience [2, 3]. Harnessing beneficial plant–microbe interactions has therefore become a key objective in the development of sustainable agricultural practices [4, 5].

Among soil symbionts, arbuscular mycorrhizal fungi (AMF) are widespread root-associated organisms forming symbioses with approximately 72% of vascular plant species [6]. AMF enhance plant nutrition, particularly P uptake, by extending the soil exploration capacity of roots through their hyphal network in exchange for plant-derived carbon [7]. Beyond nutrient acquisition, AMF contribute to plant resistance against pathogens [8–10], improve tolerance to abiotic stresses [11], and enhance soil structure and stability [12, 13]. These multifunctional benefits position AMF as key biological drivers of sustainable agroecosystem functioning.

However, increasing evidence indicates that the functional role of AMF cannot be fully understood independently of their associated microbial communities. A growing body of work shows that AMF hyphae and spores harbor diverse bacterial assemblages that may contribute to nutrient mobilization, fungal fitness, and plant performance [14–18]. These “mycorrhizosphere” and “hyphosphere” interactions suggest that AMF act not only as plant symbionts but also as ecological platforms shaping bacterial assembly and activity in soil.

Previous studies demonstrated that bacteria can be physically associated with AMF structures, including spore surfaces and internal compartments [19, 20]. In addition, AMF establish close interactions with myriad of bacteria and fungi that can be exploited to enhance mineral mobilization for plants. Subsequent work revealed diverse and taxonomically structured bacterial communities associated with AMF spores and hyphae, including taxa such as Pseudomonas, Burkholderia, and members of the Oxalobacteraceae [21–24]. Functional analyses further suggest that AMF-associated bacteria may contribute to enzymatic activities involved in organic matter turnover and nutrient mobilization [25], while hyphosphere bacteria have been implicated in processes such as phosphorus solubilization, nitrogen fixation, and enhancement of plant growth and AMF development [26–28].

Despite these advances, the ecological drivers shaping AMF-associated bacterial communities remain poorly understood. In particular, it is still unclear how soil nutrient context influences bacterial assembly on AMF hyphae and whether changes in phosphorus availability can modify these interactions. This represents an important gap because nutrient availability is a major determinant of microbial competition, resource exchange, and plant–microbe signalling in the rhizosphere.

Rock phosphate (RP), a low-solubility mineral phosphorus source, provides a relevant system to address this question. Although RP is widely considered agronomically inefficient in the absence of biological assistance, its solubilization can be enhanced by microbial activity and plant–microbe interactions [29–31]. AMF and their associated microbiota may play a central role in this process by altering phosphorus mobilization dynamics and shaping bacterial communities capable of mineral dissolution. However, whether RP amendment modifies AMF-associated bacterial assembly and how this interacts with AMF symbiosis remains largely unexplored.

Leek (Allium porrum L.) is a well-established model host for arbuscular mycorrhizal research because of its high mycorrhizal responsiveness and relatively simple root architecture, which facilitates the assessment of root colonization and AMF-mediated nutrient uptake. Moreover, compartmented microcosm systems using leek have been widely employed to separate root and hyphal compartments, allowing direct investigation of nutrient acquisition through the extraradical mycelium and its associated microbial communities [32]. These characteristics make leek a suitable model for investigating AMF–bacteria interactions involved in phosphorus mobilization.

In this study, we investigated how RP amendment influences AMF–bacteria interactions and phosphorus acquisition by leek. We hypothesized that i) RP amendment influences the composition and structure of bacterial communities associated with AMF mycelium, and ii) AMF inoculation combined with RP application enhances the P nutrition of leek plants through microbial interaction.

Materials and methods

Microcosm setup and substrate preparation

The investigation utilized microcosm units to partition the soil and ensure the physical separation of AMF mycelia in the root-free compartment (RFC) from the plant roots in the root compartment (RC). Each unit consisted of two pots, measuring 6.5 × 6.5 × 8.5 cm (L×W×H), which were bonded together using silicon glue. One pot served as the plant growth compartment (RC), and the other facilitated hyphal propagation (RFC). The pots had a cut side with a 44 µm membrane (manufactured by SEFAR Incorporation, USA) that enabled interconnection without root passage. Before initiating the experiment, the microcosms were sanitized with 70% ethanol and exposed to UV light under laminar flow for 30 min.

A sandy loam soil of low P content was collected from the organic farm of the IRDA research station in St-Bruno in Québec (45°32′59.6″ N, 73°21′08.0″ W). This soil was characterized by a pH of 6.01 and a Mehlich-3 extractable phosphorus concentration of 0.41 mg/kg of soil [33]. The soil analysis was performed using a commercial service provided by EnvironeX (Longueuil, QC). The chemical properties were described previously [33]. The soil was air-dried, sieved at 2 mm, and then mixed with sand (Bomix company, Laval, QC) at a ratio of 1:1 (v/v) to produce a growth substrate with depleted phosphorus level. To eradicate the indigenous microorganisms, both the substrate and RP were sterilized using γ-irradiation with a minimum dose of 22.3 kGy and a maximum of 46.2 kGy (Nordion’s Gamma Centre of Excellence, Laval, QC). The sterilization protocol is based on standard microbiological quality-control procedures commonly used to assess the effectiveness of substrate sterilization. Sterility was confirmed via the serial dilution method and inoculation of the tryptic soy agar (TSA) plates. Each microcosm compartment was filled by approximately 333 g of substrate. Where required, 0.16 g of RP was thoroughly mixed with the RFC substrate, corresponding to an application rate of approximately 100 kg P₂O₅ ha⁻¹. This rate was selected because it is within the range commonly used for P fertilization of P-deficient agricultural soils and is representative of field application rates used for evaluating RP as a P source. Despite the very low initial Mehlich-3 extractable P concentration of the substrate (0.41 mg kg⁻¹), a moderate RP dose was intentionally chosen to maintain P-limiting conditions and thereby enable the assessment of AMF- and microbially-mediated RP solubilization and phosphorus acquisition. The RP used in this study as the sole P source was the sedimentary RP with a bone phosphate of lime content ≥ 54%, kindly supplied by OCP Group, Morocco.

Host plant and microbial inocula

Seeds from leek (Allium porrum) (McKenzie Seeds, Canada), a highly mycotrophic plant, were used as a host. Under sterile conditions in a laminar flow, the seeds were surface sterilized with ethanol (70%) for 30 s and then bleached (15%) for 5 min, and were then rinsed abundantly with sterile distilled water. After soaking in the final rinse water for 2 h, three surface-sterilized seeds were sown at a depth of 5 mm in each RC. Mycorrhizal treatments were inoculated at sowing with spores of Rhizophagus irregularis DAOM 197198 (Premier Tech, Canada). To enhance the chance of root colonization by the AMF, a second application was performed one month after the first, directly to the roots after thinning the plants to one per microcosm. The AMF species was chosen based on its in vitro production process, its application as a commercial inoculant in agriculture, and its large distribution [34]. In each AMF inoculation, the seeds/seedlings received 200 spores diluted in 300 µl of sterile distilled water.

To enrich the soil microbial community, a microbial suspension was prepared from natural soil collected from the rhizosphere of organically cultivated crops within the Botanical Garden of Montreal. The preparation protocol was developed in-house based on established soil microbial extraction procedures and optimized to recover the indigenous bacterial community while excluding AMF propagules [35, 36]. Soil subsamples were taken separately from potato, leek, tomato, eggplant, and maize during the flowering season and transferred to the lab. The rationale behind using many crop soils was to diversify microbial communities. A composite sample was prepared by pooling equal amounts from each of the five rhizosphere soil samples and homogenizing the mixture. Six kilograms of the composite soil were sieved through a 2 mm mesh to remove roots and rock fragments, and then divided into 3 portions. Firstly, we mixed 1/3 of the soil (2 kg) with 4 L of sterilized distilled water, and the soil was brought into suspension via stirring for 30 min and then kept for at least 2 h at room temperature. The liquid phase of the soil suspension was filtered through autoclaved sieves with mesh widths of 53 µm and 20 µm. This filtration excludes AMF spores in the soil suspension. The same operation was used for the second and the third soil portions using the soil suspension obtained from the first and the second filtration, respectively. The final filtrate was collected in a sterile Erlenmeyer flask and considered as the microbial suspension.The microbial suspension was introduced into the RFC substrate two months after the second AMF inoculation by applying 50 mL per microcosm. This delayed inoculation was intentionally designed to allow the establishment of a functional AM symbiosis, including root colonization and the development of an extensive extraradical mycelial network capable of crossing the membrane and colonizing the distal compartment. Preliminary experiments indicated that approximately two months were required to obtain sufficient fungal biomass and hyphal development before introducing the soil microbial community. This approach ensured that the microbial suspension encountered an already established AMF network, allowing the assessment of AMF-driven microbial recruitment and community assembly within the hyphosphere and mycorrhizosphere while minimizing confounding effects associated with the simultaneous establishment of AMF and bacterial communities.

Experimental design

This study represents a subset of an experiment on the recruitment profile of microbes by AMF carried out in a greenhouse using the microcosm units described earlier. The study examined the effect of sedimentary RP on the community structure of bacteria associated with the mycelium of the AMF R. irregularis DAOM 197198. Two factors were considered: (1) two levels of RP, with or without RP application, and (2) two AMF levels, inoculated or uninoculated. Consequently, the experiment involved four treatment combinations: M0P0, M0P1, M1P0, and M1P1, where M refers to AMF, P to RP, 0 to absence and 1 indicates presence. The trial was organized using a randomized block design, comprising nine replicates on a mesh table to prevent the risk of cross-contamination through water. Three replicates were established each week over a period of three weeks. Leek plants were grown under conditions of photoperiod (16/8 h day/night) and temperature (21/18 °C day/night). An additional set of microcosm units was prepared and inoculated with AMF to monitor plant colonization and the presence of mycelia in the RFCs. A layer of calcinated clay of granules (Turface®), approximately 1 cm thick, was added on top of the microcosm pots to cover the soil surface, and it was renewed weekly to minimize or prevent surface contamination by microalgae. The microcosms were watered to maintain soil moisture at holding capacity. Once a week, the plant compartments received 50 mL of modified half-strength Long Ashton nutrient solution containing 22 ppm (0.71 mM) of P. This concentration of P is known to enhance AMF-induced root colonization [37]. Two months after the second inoculation with the AMF spores (a period allowing the establishment of extraradical AMF mycelia in the RFCs), all RFCs were treated with 50 mL of microbial suspension. Afterwards, the irrigation regime was switched and the diluted nutrient solution without P was supplied once a week to the RFCs. The microcosms were kept watered with tap water as needed throughout the week.

Harvest and plant analysis

Growth parameters and chlorophyll concentration

The harvest was performed five months after setting up the experiment over three weeks to ensure consistent timing across all replicates. Before cutting the shoots, the stem diameter of plants was measured with an electronic caliper at 5 mm above the soil. To verify the effect of the treatments on the plant’s photosynthetic activity, an atLEAF chlorophyll meter (FT Green LLC, USA) was used to evaluate the concentration of chlorophyll (µg/cm²) on the fully expanded third or fourth leaf from the top according to the manufacturer’s instructions. Afterward, each microcosm was disassembled, and the RC was carefully discharged. The roots were rinsed with sterile distilled water, and the shoots were cut from the roots, put in a paper bag, and dried in a hot-air oven at 80 °C for 72 h. The dry biomass was recorded, and the shoots were ground with a clean dried pestle and mortar for further chemical analysis. The fresh roots were weighed, and to evaluate the mycorrhizal colonization, a small amount of fine roots was placed in a separate 15 mL falcon tube containing 8 mL of 50% ethanol. The remaining roots were weighed a second time and then dried in the oven under the same conditions as the shoots. The second fresh root weight helped to calculate the total amount of dry root weight.

Mycorrhizal colonization and phosphorus content

The mycorrhizal colonization ratio was evaluated using the ink and vinegar staining method [38]. The roots preserved in 50% ethanol were washed with tap water and cut into 1 cm long segments. The root segments were cleared in a 10% KOH solution boiling at 90 °C for 3 to 4 min until they became transparent and were then rinsed with distilled water. To remove the KOH residues, the roots were soaked in another bath of 1% acetic acid for 4 min. Cleared roots were placed in a 5% ink–vinegar solution at 90 °C for 3 min and rinsed with distilled water and then stored in a lactoglycerol solution for a minimum of 24 h at room temperature to remove the excess of ink solution. The stained root segments were mounted on microscope slides with fine-point forceps, and the degree of mycorrhization was determined via the intersection method [37]. The numbers of arbuscules, vesicles, and intraradical mycelia were counted and the percentages were calculated.

The ground shoots were used to determine the total phosphorus content via the modified dry ash method [39]. Ground shoot aliquots of 0.5 g were incinerated at 500 °C for 4 h and digested in 100 mL borosilicate Erlenmeyer containing 6 mL and 1 mL of concentrated sulfuric acid and nitric acid, respectively. The resulting solutions were filtered in an acid-washed flask, filled up to 100 mL with demineralized water, and then homogenized before P concentration analysis via the molybdenum blue colorimetric technique [40]. The shoot P content (µg) was subsequently calculated from the P concentration and shoot biomass.

Extraradical mycelia collection

Prior to harvesting the substrate from the hyphal compartment, the Turface layer and the top 1 cm of the substrate were removed. In addition, a 1 cm strip of substrate adjacent to the membrane was discarded to minimize potential effects of surface-associated microbial communities and root-derived exudates near the interface. Approximately 10 mL of the remaining substrate was collected into a 15 mL Falcon tube for parallel analyses. The remaining substrate was then transferred into a 1 L sterile glass jar and covered with cold sterile distilled water to a height of approximately 1 cm. The jar was sealed, vigorously shaken, and allowed to stand for approximately 15 min to facilitate aggregation of AMF extraradical mycelia due to their adhesive properties. The suspension was subsequently poured through a sterile 40 µm mesh sieve, and the procedure was repeated four to five times until no visible mycelial aggregates were observed. The collected extraradical mycelia were further rinsed with cold sterile 0.9% NaCl solution to remove residual particles and then transferred into microcentrifuge tubes using sterile forceps. Samples were gently squeezed to remove excess liquid, and fresh weights were recorded. Fresh mycelial biomass was expressed as mg pot⁻1 by weighing the recovered extraradical mycelia from each individual microcosm using an analytical balance immediately after collection and removal of excess moisture. No extraradical mycelia were recovered from non-mycorrhizal RFC controls. All substrate and mycelia samples were immediately flash-frozen in liquid nitrogen after collection, maintained on dry ice during transport, and stored at − 80 °C until downstream nucleic acid analyses.

This procedure was developed in-house and is based on established principles for AMF extraradical hyphae extraction (wet-sieving and differential separation), adapted and optimized for this bi-compartmentalized microcosm system to reduce contamination from substrate particles and root-derived residues.

DNA extraction and PCR amplification

Just before DNA extraction from the substrate, all the samples were transferred into 50 mL falcon tubes with pierced lids and placed in a lyophilizer for 3 days until dried. Then, the samples were sieved with a 1 mm mesh sieve to facilitate substrate particle separation. Hereinafter, the materials resulting from the M1 and M0 treatments were named, respectively, hyphosphere soil and bulk soil.

The bulk and hyphosphere soil DNA was extracted from 250 mg of homogenized soil of each sample using the DNeasy PowerSoil Pro Kit (QIAGEN, Canada) according to the manufacturer’s protocol. We performed the DNA extraction from mycelia using the DNeasy Plant Mini Kit (QIAGEN, Canada) following the manufacturer’s instructions, starting by grinding about 50 mg of liquid nitrogen-frozen samples in 1.5 mL tubes containing sterile white sand, which facilitates cell lysis. The elution step was performed in 50 µL of elution buffer for bulk and hyphosphere soils and in 20 µL for mycelia and stored at − 20 °C. Therefore, genomic DNA extracts were first visualized via gel electrophoresis on 1% agarose stained with GelRed diluted at a ratio of 1/10000 using the GelDoc System (BioRad, Montreal, QC, Canada). The DNA concentration was determined using the Qubit 2.0 Fluorometer (ThermoFisher, Canada) and the Qubit double-stranded DNA (dsDNA) HS Assay Kit.

To investigate the total bacterial community, the V3-V4 region of 16S rRNA gene was targeted using Platinum™ Direct PCR Universal Master Mix (ThermoFisher, Canada). The amplicon libraries were constructed using the primers pair of 341F (5’-CACCTACGGGNGGCWGCAG-3’) and 805R (5′-ACTACHVGGGTATCTAATCC-3′) tagged, respectively, with CS1 (5’-ACACTGACGACATGGTTCTA-3’) and CS2 (5′-TACGGTAGCAGAGACTTGGTCTG-3′), which permit barcoding [41]. The PCR amplification was performed in a 25 µL volume reaction mix containing 10 µL of Platinum Direct PCR Universal Master Mix (Thermo Fisher Scientific, Canada), 0.25 µM of each primer, 4 µL of Platinum GC Enhancer, 7 µL of water, and 1 µL of template DNA. Negative controls with only water were included in each PCR run. The thermal cycling conditions were as follows: initial step of activation at 94 °C for 2 min, followed by 35 cycles of denaturation at 94 °C for 15 s, annealing at 60 °C for 30 s and extension at 68 °C for 20 s, and a final extension at 68 °C for 1 min with a hold at 10 °C. PCR reactions were amplified in an Eppendorf Mastercycler Pro thermocycler (Eppendorf, Canada). The amplification was confirmed by running the PCR products on a 1% agarose gel. Amplicons were sent for Illumina MiSeq sequencing using 2 × 300 bp paired-end reads which were demultiplexed on the instrument in the NGS platform at the Centre d'expertise et de services of Genome Quebec (Montreal, QC, Canada).

Bioinformatics and statistical analyses

A bioinformatic analysis was performed using QIIME2 version 2021.4.0 [42]. The pipeline used to process the 16S rRNA gene sequences was DADA2 v1.18.0 [43]. Initially, Cutadapt 3.4 was employed to remove both primer sequences from the 16S rRNA gene amplicons with parameter values “minimum-length = 50”, “times = 2”, “overlap = 6”, and “p-error-rate = 0.1”. Next, we excluded the forward and reverse sequences with less than 220 bp with the command “–p-trunc-len”, as the base quality of the sequences tended to diminish below that threshold in our data. Afterwards, the amplicon sequence variant (ASV) table was calculated, and chimeras were removed. The taxonomy assignment of the ASVs was performed using the naive Bayesian classifier method on the databases SILVA v183.1 and RDP v2.12, and the identities of the ASVs of interest were verified manually using BLASTn on the NCBI (nr/nt) database.

AMF and plant-related parameters were analyzed employing JMP Pro 17 software (SAS Institute Inc., Cary, NC, USA). Shapiro–Wilk test was used to verify the normal distribution of the data, and Levene’s test to determine the homoscedasticity. The data of root dry-weight biomass and root colonization were transformed using Box–Cox method before performing statistical analyses. An analysis of variance (ANOVA) was used to assess the effects of AMF inoculation, RP application, and their interactions, followed by Tukey’s HSD test to compare the mean values, which differed at p < 0.05. AMF inoculation and RP addition were considered as fixed effects, and the random effect was attributed to the block.

Microbial diversity analysis was performed using R (Version 4.3.2); the rare-curve function from the vegan package v2.6-4 [44] was employed for normalizing the dataset through a random subsampling process that aligned each sample’s read data to the minimal read count observed in the dataset. Shannon and Simpson diversity indices were computed to quantify alpha diversity utilizing the vegan package. The influence of RP application on alpha diversity was statistically examined via an ANOVA, followed by post hoc analyses using the agricolae package v1.3-7 [45]. The beta diversity of bacterial ASVs across samples was determined using Bray–Curtis dissimilarity and visualized with principal coordinate analysis (PCoA) plots using the vegan package. The vegan package’s ADONIS function was used for the permutational multivariate analysis of variance to evaluate the effects of AMF and RP application on beta diversity, applying a Hellinger-transformation to the data and executing 999 permutations. An indicator species analysis was conducted via the indicspecies package v1.7-14 [46] to identify the taxa associated with the application of RP. In defining the core microbiome, the designation “core bacteriota” was attributed to the bacterial taxa consistently detected in the mycelial or hyphosphere soil samples associated with the AMF R. irregularis. The core bacteriome was identified using the microbiome package v1.24-0 [47] with 95% prevalence. The co-occurrence network was built using the R packages SpiecEasi [48] and igraph [49] only using ASVs with a total abundance superior to 0.05%, thus retaining only 1567 ASVs. Metabolic pathways associated with the taxonomy of bacteria were calculated using the Picrust2 of Galaxy tool [50]. The taxa associated with treatments were identified using the microbial package in R [51] and its “ldamarker” command, which implements the Linear discriminant analysis effect size (LEFSe) method.

Results

Mycorrhizal colonization and mycelial production

The mycorrhizal colonization was successfully established in the plants inoculated with R. irregularis, whereas no colonization was observed in non-mycorrhizal roots. Means were 94.82% and 89.32% under the presence and absence of RP respectively, which explains the quantity of mycelial biomass extracted from the hyphal compartment (Table 1). While no hyphae were found in the absence of R. irregularis inoculation, they multiplied extensively in the hyphal compartment when the plants were colonized (Table 1). RP did not significantly influence the mycelial biomass, the percentages of mycorrhizal colonization and vesicle formation. Surprisingly, the percentage of arbuscules was significantly affected by the RP application (p = 0.03).

Table 1.

Mycorrhizal colonization and fresh biomass of the hyphae in the RFC

Treatments	Mycelial biomass (mg/pot)	Mycorrhizal colonization (%)	Arbuscules (%)	Vesicles (%)
M1P0	117.95 ± 27.96a	89.32 ± 3.66a	62.86 ± 10.69a	69.49 ± 6.90a
M1P1	109.60 ± 19.95a	94.82 ± 2.37a	90.12 ± 2.23b	74.07 ± 4.64a

The abbreviations M1P0 and M1P1 indicate samples inoculated with R. irregularis without RP and those inoculated with R. irregularis and supplemented with RP, respectively.

Values are presented as means ± standard error (SE) (n = 9)

Different letters indicate significant differences according to ANOVA followed by Tukey’s HSD test (p < 0.05)

Plant productivity

Plant biomass and stem diameter

Agronomic parameters of Allium porrum were affected by the treatments (Fig. 1). The colonization with R. irregularis significantly enhanced both root (p < 0.001) and shoot dry weight biomass (p < 0.001) (Table S1, Fig. 1A, B). However, RP application did not influence shoot (p = 0.373) and root (p = 0.493) dry-weight biomass (Table S1, Fig. 1A, B). A similar trend was exhibited for stem diameter (Fig. 1C). While the application of AMF spores significantly increased the stem diameter (p < 0.001), RP application was statistically insignificant (p = 0.702) (Table S1). However, the combined treatment (M1P1) tended to increase the stem diameter (p = 0.07) (Table S1). M1P1 treatment improved plant biomass and stem diameter by 125.14% and 51.47%, respectively, compared to the control (Table S1).

Fig. 1.

Growth responses of Allium porrum plants to inoculation with R. irregularis and RP amendment. A Root dry weight, B shoot dry weight, and C stem diameter measured at harvest in compartmented microcosms (n = 9 biological replicates per treatment). Treatments were defined as follows: M0P0, non-inoculated plants without RP amendement (control); M0P1, non-inoculated plants amended with RP; M1P0, plants inoculated with R. irregularis without RP amendement; and M1P1, plants inoculated with R. irregularis and amended with RP. Boxplots show the median, interquartile range, and whiskers representing the data distribution. Different letters above boxplots indicate significant differences among treatments according two-way ANOVA followed by Tukey’s HSD post hoc multiple comparison tests (p < 0.05)

Shoot P concentration and photosynthetic capacity

Although RP application tended to increase shoot P concentration (p = 0.072), mycorrhization significantly increased the total P acquired by plants (p = 0.014) compared to the control (Fig. 2A; Table S2). Moreover, mycorrhizal plants supplied with RP (M1P1) exhibited 135% increase in shoot P content (P concentration × shoot biomass) relative to the control (Fig. 2A; Table S2). Regarding photosynthetic capacity, both RP application (p = 0.017) and AMF inoculation (p < 0.001) significantly increased chlorophyll concentration compared to the control (Fig. 2B; Table S2). The combined treatment (M1P1) resulted in the highest chlorophyll concentration, representing a 60% increase relative to the control (Fig. 2B; Table S2).

Fig. 2.

Effects of inoculation with R. irregularis and RP amendement on A shoot P content (or concentration, as appropriate) and B chlorophyll concentration in Allium porrum grown in compartmented microcosms (n = 9 biological replicates per treatment). Treatments were defined as follows: M0P0, non-inoculated plants without RP amendment (control); M0P1, non-inoculated plants amended with RP; M1P0, plants inoculated with R. irregularis without RP amendment; and M1P1, plants inoculated with R. irregularis and amended with RP. Bars represent means ± standard error (SE). Different letters above bars indicate significant differences among treatments according to two-way ANOVA followed by Tukey’s HSD post hoc multiple comparison tests (p < 0.05)

Taxonomic profiles of the bacteria

Raw Illumina MiSeq 16S rRNA gene sequencing generated a total of 2,513,803 bacterial reads. We retrieved 1,004,950 non-chimeric reads from 54 samples, ranging from 1948 to 91,128 reads per sample (two samples were excluded due to insufficient sequencing quality). An extra cleaning of the dataset was performed to remove unidentified ASVs at the phylum level as well as those with fewer than 10 reads in all samples. The final dataset comprised a total of 7514 ASVs, which were subsequently categorized into 31 phyla, 85 classes, 130 orders, 149 families, and 194 genera. The rarefaction curves indicated that all samples had reached saturation (Fig. S1). In each of the habitats, Planctomycetota was the most abundant phylum, followed by Pseudomonadota, with 57% and 10% in the mycelia, 39% and 12% in the hyphosphere soil, and 44% and 11% in the bulk soil, respectively (Fig. 3A), while Parcubacteria candidate phylum OD1 came in third place with 7% in the mycelia. Bacillota, on the other hand, was in the hyphosphere and bulk soils, with 11% in each of these habitats. Among the mycelia, the prevalent genera were Planctomyces (10%) followed by Gemmata (9%) and Pirellula (4%) (Fig. 3B). In contrast, the predominant genera in the hyphosphere soil were Planctomyces (9%), Bacillus (7%), and Gemmata (5%). Similarly, in bulk soil, the dominant genera were Planctomyces (11%), Bacillus (6%), and Gemmata (5%). An important proportion of bacterial genera remained unidentified, with 62% in the mycelia, 55% in the hyphosphere soil, and 55% in the bulk soil.

Fig. 3.

Relative abundances of the 10 most common bacterial taxa in the mycelia (M1-P0, M1-P1), hyphosphere soil (SM1-P0, SM1-P1), and bulk soil (SM0-P0, SM0-P1) under no P (P0) or RP (P1) amendment at phylum (A) and genus (B) levels, with all remaining taxa grouped under “Others”. The abbreviations M1, SM1, and SM0 refer to mycelium, hyphosphere soil, and bulk soil, respectively. Relative abundances were calculated from rarefied ASV data

Furthermore, the Venn diagram analysis (Fig. 4) of bacterial ASVs shared among the three habitats revealed 672 ASVs, accounting for 15%, 22%, and 24% of the bacterial communities in mycelia, the hyphosphere, and bulk soil, respectively. Additionally, 281 ASVs (6% of the mycelial community and 9% of the hyphosphere community) were found exclusively in these two habitats. Similarly, 232 ASVs were common between mycelia and bulk soil, representing 5% and 8% of their respective bacterial communities.

Fig. 4.

Venn diagram of bacterial ASVs shared between mycelia, hyphosphere, and bulk soil following sequence quality filtering and rarefaction

In contrast, the highest number of shared ASVs (1,216) was observed between the hyphosphere and bulk soil, comprising 40% and 44% of their respective bacterial communities. Notably, 3,404 ASVs were unique to mycelia (47%), 845 were exclusive to the hyphosphere (12%), and 616 were found only in bulk soil (9%).

Effect of RP amendment on the bacterial community assemblage in mycelia, hyphosphere, and bulk soil

Neither RP nor sample type affected the α-diversity as measured with Shannon and Simpson indices (Fig. 5 and Table 2). However, a PCoA showed a clear separation of bacterial communities associated with the AMF mycelia from those in the hyphosphere soil and bulk soil samples (Fig. 6A). Similarly, RP application led to a distinct categorization of bacterial communities associated with AMF mycelia (Fig. 6B). Given that the bacterial communities were separated by habitats, an additional PERMANOVA analysis was conducted to determine if RP application affected the bacterial communities within each habitat. The results showed a significant impact of RP application on the bacterial communities associated with AMF mycelia (p = 0.01) and in the bulk soil (p = 0.015), accounting for 6.8% and 6.1% of the variation, respectively (Table 3). Another PERMANOVA analysis was conducted on bacterial soil communities to evaluate the effect of AMF spore inoculation. The results revealed a significant effect (p = 0.004) on the bacterial assemblage, accounting for 3.5% of the variation (Table 4).

Fig. 5.

α-diversity of bacterial communities assessed using rarefied ASV data via A Shannon and B Simpson indices in the AMF mycelium, hyphosphere, and bulk soils under RP amendment. Different letters indicate significant differences among habitats according to ANOVA followed by post hoc multiple comparison tests (p < 0.05)

Table 2.

α-diversity of bacterial community assessed by Shannon and Simpson indices in different habitats under RP amendment

Habitats	Index	Source of variation	Sum Sq	Mean Sq	NumDF	DenDF	F value	Pr(> F)
AMF mycelia	Shannon	RP	0.305	0.305	1	8.000	0.839	0.387
	Simpson	RP	6.92E-05	6.92E-05	1	8.000	0.155	0.704
Hyphosphere soil	Shannon	RP	0.043	0.043	1	16	0.468	0.504
	Simpson	RP	2.04E-06	2.04E-06	1	16	0.106	0.749
Bulk soil	Shannon	RP	4.02E-04	4.02E-04	1	8.000	0.006	0.942
	Simpson	RP	3.98E-06	3.98E-06	1	16	0.094	0.763

Values are presented as means ± standard error (SE)

Statistical comparisons were performed using ANOVA followed by post hoc multiple comparison tests

Fig. 6.

Principal coordinate analysis based on Bray–Curtis dissimilarity calculated from Hellinger-transformed ASV abundance data showing variations in the bacterial communities among A sample types and on B R. irregularis mycelia under RP amendment. Each point represents one biological replicate

Table 3.

β-diversity variance partitioning of bacterial community’s structure among RP application in permutational multivariate analysis of variance in different habitats

Variable	Source of variation	Df	SumOfSq	R2	F	Pr(> F)
AMF mycelia	RP	1	0.444	0.068	1.165	0.01**
	Residual	16	6.092	0.932
	Total	17	6.536	1
Hyphosphere Soil	RP	1	0.229	0.044	0.732	0.87
	Residual	16	5.001	0.956
	Total	17	5.230	1
Bulk soil	RP	1	0.317	0.061	1.035	0.015*
	Residual	16	4.900	0.939
	Total	17	5.217	1

Bray–Curtis dissimilarity was calculated from Hellinger-transformed ASV abundance data with 999 permutations

Table 4.

β-diversity variance partitioning of bacterial community’s structure among AMF inoculation in permutational multivariate analysis of variance in soil biotopes

Source of variation	Df	SumOfSq	R2	F	Pr(> F)
AMF	1	0.677	0.035	1.884	0.004**
Residual	52	18.684	0.965
Total	53	19.361	1

Bray–Curtis dissimilarity was calculated from Hellinger-transformed ASV abundance data with 999 permutations

Patterns of indicator species and core bacteriome

Several bacterial ASVs exhibited consistent shifts in abundance across treatments and sample compartments. Indicator species analysis identified 35 ASVs (Table 5), revealing a strong compartment- and treatment-specific signal. In the absence of RP application, 26 indicator ASVs (74%) were detected, the majority of which were associated with mycelial samples (20 ASVs). These indicators were dominated by Planctomycetota (16 ASVs), with additional representation from the Parcubacteria candidate phylum OD1 (three ASVs) and Bacillota (one ASV). Five indicator ASVs were linked to bulk soil, spanning Actinobacteriota, Chloroflexota, Acidobacteriota, Bacillota, and Planctomycetota, while only a single indicator ASV, affiliated with Verrucomicrobiota, was detected in the hyphosphere soil. In contrast, RP application yielded fewer indicator taxa (nine ASVs, 26%), most of which were associated with bulk soil, primarily Planctomycetota (six ASVs) and one Parcubacteria (OD1). Only one indicator ASV was detected in each of the mycelial and hyphosphere compartments, affiliated with Planctomycetota and OP3, respectively.

Table 5.

Indicator species analysis of bacterial ASVs associated with mycelia and enriched in hyphosphere and bulk soils under RP application

Sample type	RP treatment	BASV_ID	stat	p-value	Family	Genus
Bulk soil	P0	BASV1784	0.552	0.025*	Gaiellaceae	Unkown
		BASV2355	0.502	0.049*	Unkown	Unkown
		BASV444	0.5	0.016*	Solibacteraceae	Candidatus Solibacter
		BASV301	0.465	0.049*	Bacillaceae	Bacillus
		BASV254	0.452	0.036*	Unkown	Unkown
	P1	BASV1768	0.733	0.003**	Unkown	Unkown
		BASV1571	0.616	0.011*	Pirellulaceae	Unkown
		BASV1779	0.538	0.035*	Pirellulaceae	Unkown
		BASV3523	0.526	0.031*	Unkown	Unkown
		BASV3592	0.497	0.021*	Planctomycetaceae	Planctomyces
		BASV1001	0.459	0.049*	Pirellulaceae	Unkown
		BASV503	0.438	0.049*	Planctomycetaceae	Planctomyces
Mycelia	P0	BASV622	0.61	0.01**	Pirellulaceae	Unkown
		BASV240	0.578	0.016*	Pirellulaceae	Unkown
		BASV1492	0.574	0.034*	Pirellulaceae	Unkown
		BASV524	0.552	0.002**	Unkown	Unkown
		BASV1430	0.548	0.04*	Pirellulaceae	Unkown
		BASV640	0.545	0.04*	Gemmataceae	Gemmata
		BASV2857	0.535	0.027*	Turicibacteraceae	Turicibacter
		BASV281	0.519	0.017*	Unkown	Unkown
		BASV469	0.513	0.034*	Planctomycetaceae	Planctomyces
		BASV432	0.511	0.034*	Unkown	Unkown
		BASV334	0.508	0.033*	Pirellulaceae	Unkown
		BASV915	0.5	0.04*	Pirellulaceae	Unkown
		BASV21	0.499	0.031*	Planctomycetaceae	Planctomyces
		BASV1698	0.495	0.028*	Unkown	Unkown
		BASV395	0.48	0.041*	Gemmataceae	Unkown
		BASV101	0.47	0.035*	Pirellulaceae	Unkown
		BASV229	0.439	0.04*	Unkown	Unkown
		BASV2733	0.418	0.037*	Gemmataceae	Gemmata
		BASV1812	0.411	0.03*	Planctomycetaceae	Planctomyces
		BASV460	0.319	0.045*	Gemmataceae	Gemmata
	P1	BASV170	0.542	0.021*	Gemmataceae	Gemmata
Hyphosphere soil	P0	BASV779	0.436	0.005**	Unkown	Unkown
	P1	BASV226	0.532	0.043*	Unkown	Unkown

Indicator taxa were identified using the IndVal method. RP application (P1); no RP application (P0)

Core bacteriome analysis further highlighted compartment-specific microbial assemblages. Across all samples, 12 core taxa were identified, occurring in at least 95% of samples (Table 6). The mycelial compartment harbored six core taxa, including Planctomyces sp., three members of the Pirellulaceae family (genera A17, Pirellula sp., and one unclassified taxon), Bacillus flexus, and an unclassified Gemmataceae. Similarly, six core taxa were detected in the hyphosphere soil, comprising Bacillus fumarioli, AKIW781 (Chloroflexota), Planctomyces sp., two Pirellulaceae taxa (A17 and Pirellula sp.), and Gemmata sp. In bulk soil, five core taxa were identified, including AKIW781 (Chloroflexota), Bacillus fumarioli, Planctomyces sp., Pirellula sp., and Gemmata sp. Although no core ASVs were shared between mycelial and hyphosphere compartments, several genera overlapped, whereas two and three shared core ASVs were observed between bulk soil and mycelial samples, and bulk soil and hyphosphere samples, respectively.

Table 6.

Core bacteriome of ASVs associated with AMF mycelia and in hyphosphere and bulk soils

Sample type	BASV_ID	Family	Genus
Bulk	BASV4	Unkown	Unkown
	BASV6	Bacillaceae	Bacillus
	BASV2	Planctomycetaceae	Planctomyces
	BASV22	Pirellulaceae	Pirellula
	BASV93	Gemmataceae	Gemmata
Mycelia	BASV2	Planctomycetaceae	Planctomyces
	BASV3	Unkown	Unkown
	BASV42	Pirellulaceae	A17
	BASV22	Pirellulaceae	Pirellula
	BASV24	Bacillaceae	Bacillus
	BASV430	Gemmataceae	Unkown
Hyphosphere soil	BASV6	Bacillaceae	Bacillus
	BASV4	Unkown	Unkown
	BASV14	Planctomycetaceae	Planctomyces
	BASV16	Pirellulaceae	A17
	BASV40	Pirellulaceae	Pirellula
	BASV93	Gemmataceae	Gemmata

Core taxa were defined as ASVs detected with ≥ 95% prevalence within each habitat

Consistent with these patterns, LEfSe analysis identified clear treatment-associated biomarkers. The family Anaerolineaceae and one ASV assigned to the genus Planctomyces were significantly enriched under RP amendment (p < 0.001), whereas the order JG30-KF-AS9 within the Chloroflexota was significantly associated with the absence of RP addition (p < 0.001). These discriminant taxa were predominantly detected in soil samples. At a finer resolution, two ASVs were identified as core taxa within the mycelial compartment, being present in more than 90% of mycelial samples: ASV2 (Planctomyces) and ASV3 (Pirellulaceae). In soil samples, two core ASVs affiliated with the order AKIW781 (Chloroflexi) were consistently detected (Tables 5, 6).

Network analysis and functional profiling

The co-occurrence network of bacterial ASVs was structured into two main groups: one predominantly composed of taxa associated with AMF mycelium, and the other composed of soil-associated taxa (bulk soil and hyphosphere). A hub taxon was identified in both soil and mycelial compartments (ASV3100), belonging to the family Thermogemmatisporaceae within the phylum Chloroflexota, with a betweenness centrality of 2.2 × 105. This hub taxon did not show a significant response to RP amendment (p = 0.322) (Fig. 7), and overall the network structure did not reveal clear differences among RP treatments.

Fig. 7.

Pairwise co-occurrence network representation of bacterial ASVs in the different compartments. Nodes represent ASVs, and edges represent significant pairwise associations. Node colours indicate sample compartments, and node size is proportional to node connectivity (degree). The hub taxon is ASV 3100, belonging to the family Thermogemmatisporacea, in the Chloroflexota phylum

Discussion

This study explored how RP amendment shapes bacterial community assembly associated with Rhizophagus irregularis extraradical mycelium and how these interactions influence P mobilization and plant growth. By using a compartmented microcosm system that physically separated roots from the hyphal compartment, we specifically targeted the AMF mycelium as a microbial scaffold for bacterial recruitment and nutrient exchange, minimizing confounding effects of root exudates.

AMF-mediated nutrient acquisition and plant response

Although AMF colonization rates were not significantly affected by RP, the presence of AMF consistently enhanced plant performance, and RP further increased shoot and root biomass when combined with AMF. This confirms that AMF benefits are not necessarily expressed through changes in colonization intensity, but rather through improved functional efficiency in nutrient acquisition, particularly under low-P conditions. Similar patterns have been reported in Acacia gummifera, where P availability did not alter colonization rates but improved plant P uptake and growth [52]. Enhanced arbuscule abundance under RP further supports improved functional exchange capacity, as arbuscules represent the primary interface for nutrient transfer between AMF and host plants [53].

The observed increase in shoot P content and chlorophyll concentration in AMF + RP treatments suggests enhanced P mobilization from an otherwise poorly available source. This likely results from synergistic AMF–microbial processes operating within the hyphal compartment, where RP served as the sole P source. These findings align with a previous study [54], which demonstrated AMF-associated depletion of organic P pools in hyphal-accessible compartments, supporting the role of extraradical hyphae in mobilizing otherwise inaccessible P forms.

AMF hyphae as ecological scaffolds for microbial assembly

Beyond nutrient acquisition, our results reinforce the concept that AMF hyphae function as ecological scaffolds structuring bacterial communities. The hyphal network extends the soil exploration capacity beyond the rhizosphere, accessing fine pores and nutrient patches unavailable to roots [55]. With estimated hyphal densities reaching up to 2000 cm/cm³ [56], this extensive network provides both physical habitat and carbon inputs through exudation and hyphal turnover, thereby shaping microbial recruitment.

Consistent with this, we observed a clear enrichment of bacteria associated with AMF mycelium compared to bulk soil. AMF hyphae release diverse organic compounds, including sugars, organic acids, and amino acids, which can selectively stimulate microbial colonization and biofilm formation [57, 58]. Moreover, AMF-associated microbial assemblages are not random but responsive to environmental conditions and plant nutrient status [56, 59].

Cooperative AMF–bacteria interactions in P mobilization

Our findings support a cooperative model in which AMF and bacteria jointly enhance P mobilization from RP. While AMF alone may have limited ability to directly solubilize RP, their hyphal network facilitates bacterial recruitment and dispersal to nutrient-rich microhabitats [60, 61]. This is consistent with experimental evidence showing that bacterial inoculation enhances AMF-mediated P uptake by stimulating hyphal growth and activity [61, 62].

Importantly, AMF hyphae may act as “fungal highways,” enabling bacterial movement and colonization of RP particles [17], while simultaneously supplying labile carbon that fuels microbial metabolism and P solubilization [59]. Such synergistic interactions have been experimentally demonstrated in systems involving Rahnella aquatilis, where AMF-derived carbon exudation triggered bacterial phosphatase activity and enhanced P mineralization [63]. Likewise, AMF–bacteria interactions have been shown to mediate carbon–phosphorus exchange at the peri-arbuscular interface [64].

Novel insight: planctomycetota as a core AMF-associated taxon

A key novel finding of this study is the consistent identification of Planctomycetota as a dominant core bacterial group associated with AMF mycelium under RP amendment. This is particularly significant given the limited knowledge of this phylum in mycorrhizal systems.

Planctomycetota, together with Bacillota, formed the core microbiome of AMF mycelium, while Chloroflexota was additionally enriched in the hyphosphere soil. Although often underrepresented due to cultivation challenges [65, 66], Planctomycetota are increasingly recognized as important contributors to nutrient cycling in soil ecosystems. Their ecological success may be linked to symbiotic lifestyles, slow growth strategies, and production of antimicrobial compounds such as stieleriacines [67, 68].

Importantly, recent studies suggest a functional role of Planctomycetota in P cycling. It has been reported that gcd-containing genomes (coding for the quinoprotein glucose dehydrogenase and crucial for phosphate solubilization by phosphate-solubilizing bacteria such as Rhizobium and Pseudomonas [69]) are enriched in Planctomycetota and are positively correlated with bioavailable soil P [70]. Similarly, Planctomycetota have been identified as dominant taxa contributing to nutrient cycling under RP amendment [71]. In our system, the selective enrichment of Planctomyces sp. following RP addition suggests that AMF hyphae may preferentially recruit taxa involved in P transformation processes, potentially forming co-occurring networks with Bacillota and Chloroflexota that contribute to RP solubilization.

This finding positions Planctomycetota as a previously underappreciated functional component of the AMF hyphosphere microbiome, potentially involved in microbial consortia driving P mobilization from RP.

Study limitations

Despite the insights provided by this study, several limitations should be acknowledged. First, the use of a single AMF species (Rhizophagus irregularis) restricts the generalizability of the observed microbial recruitment patterns, as different AMF taxa are known to vary considerably in their exudation profiles, hyphal traits, and microbial selectivity [32]. Second, the experiment was conducted in gamma-irradiated soil, which, while effective for minimizing background microbial variability, likely reduced native microbial diversity and complexity, potentially amplifying AMF-driven selection effects.

In addition, the study was performed under controlled greenhouse conditions that do not fully reflect the environmental variability encountered in field ecosystems, including fluctuations in moisture, plant diversity, soil heterogeneity, and long-term microbial succession dynamics. These factors are known to strongly influence AMF–bacteria interactions under natural conditions. Finally, the interpretation of microbial functions is based primarily on taxonomic inference, without direct functional validation, meaning that the involvement of key taxa such as Planctomycetota in phosphorus solubilization remains putative.

Future research directions

Future work should extend beyond controlled microcosm systems to address both mechanistic and ecological uncertainties in AMF–bacteria–phosphate interactions. A first priority is to assess AMF diversity effects by comparing multiple AMF species (e.g., Rhizophagus, Funneliformis, Claroideoglomus) to determine whether the recruitment of Planctomycetota is a conserved trait or specific to certain AMF lineages.

In parallel, functional validation of key bacterial taxa is required. This includes isolating Planctomycetota and Bacillota strains from AMF hyphal compartments and experimentally testing their capacity to solubilize RP, both in vitro and in controlled AMF co-culture systems. To complement this, multi-omics approaches integrating metagenomics and metatranscriptomics should be employed to identify functional genes involved in phosphorus cycling, such as gcd, phosphatases, and organic acid biosynthesis pathways, within AMF-recruited microbial communities.

Further mechanistic insight could be gained through stable isotope tracing, particularly using ¹³C-labeled carbon flow from AMF to directly quantify carbon transfer to Planctomycetota and associated P-mobilizing bacteria. Finally, field validation across contrasting soil types is essential to evaluate the ecological stability and agronomic relevance of AMF–RP–bacteria consortia, while network-based analyses should be used to determine whether Planctomycetota function as keystone taxa within AMF hyphal microbiomes.

Conclusions

This study demonstrates that inoculation with Rhizophagus irregularis combined with RP amendment significantly enhances P content in leeks, thereby supporting our first hypothesis that AMF inoculation together with RP improves plant P acquisition. In addition, RP amendment significantly altered the bacterial community structure associated with R. irregularis extraradical mycelium, in agreement with our second hypothesis that AMF inoculation and RP jointly influence hyphosphere bacterial community composition.

Overall, our results highlight the interactive role of abiotic (RP availability) and biotic (AMF activity) factors in shaping bacterial assemblages within the AMF hyphosphere. They further reveal a selective enrichment of specific bacterial taxa under RP amendment, particularly genera such as Planctomyces, Gemmata, and Bacillus, which are known to be involved in P mobilization processes. These findings suggest that AMF hyphae act as ecological filters that promote functionally relevant microbial consortia capable of contributing to RP transformation.

The increased P content observed in leek plants is consistent with the hypothesis that AMF-associated bacterial communities contribute to RP mobilization, thereby enhancing P availability for uptake through the AMF symbiosis. However, direct evidence that P was mobilized by the AMF–bacteria consortium and subsequently translocated to the plant was not provided in this study. Collectively, these results indicate that the combined use of AMF inoculation and rock phosphate amendment represents a promising strategy to enhance RP solubilization efficiency and reduce dependence on chemical phosphorus fertilizers in sustainable agricultural systems.

Supplementary Information

Below is the link to the electronic supplementary material.

Author contributions

ZL designed and conducted the experiments, developed the methodology, performed the data analysis, and drafted the manuscript. JL and BA contributed to the analysis and data interpretation and participated in the manuscript revision. MH conceived the study, supervised the work, and secured the funding. All authors discussed the results and approved the final version of the manuscript.

Funding

This work was funded by the University Mohammed VI Polytechnic (UM6P), which is gratefully acknowledged.

Data availability

The raw reads corresponding to the hyphosphere soil samples and the AMF mycelia samples have been deposited in the NCBI Sequence Read Archive (SRA) (https://www.ncbi.nlm.nih.gov/sra), with accession numbers provided under the BioProject PRJNA1147075.

Declarations

Consent for publication

Not applicable.

Ethics approval and consent to participate

Not applicable.

Competing Interests

The authors declare no competing interests.