The microbial and chemical terroir of agarwood in French Guiana

The microbial and chemical terroir of agarwood in French Guiana | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article The microbial and chemical terroir of agarwood in French Guiana Kenji Maurice, Nicolas Baldovini, Alba Zaresmski, Jérémie Damay, and 3 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-9040724/v1 This work is licensed under a CC BY 4.0 License Status: Posted Version 1 posted You are reading this latest preprint version Abstract Agarwood is a highly valued aromatic resinous wood formed in Aquilaria species following stress or infection, yet the putative microbial drivers of its chemical quality remain poorly understood, particularly outside its native range. In this study, we investigated the bacterial and fungal communities associated with agarwood produced from Aquilaria crassna Pierre ex Lecomte planted in French Guiana and examined their relationships with volatile chemical compounds relevant to agarwood fragrance. Using high‑throughput sequencing and comprehensive chemical profiling, we characterized microbial community composition and agarwood volatile profiles across multiple cultivation plots. Despite spatial variability in microbial assemblages, agarwood samples exhibited a conserved chemical signature dominated by chromone derivatives and sesquiterpenoids, indicating the presence of a stable chemical terroir under Guianese environmental conditions. Network analysis revealed numerous bacterial and fungal taxa significantly associated with key chemical classes, suggesting potential microbial contributions to agarwood chemical complexity through plant–microbe interactions or microbial metabolic activity, although causality remains to be established. Comparative analyses with commercial agarwood samples from South-East Asia and the Middle East revealed a distinct chemical profile for Guianese agarwood, highlighting the influence of geographic origin on agarwood quality and supporting an extension of the terroir concept to woody aromatic products. Overall, this study demonstrates that Aquilaria trees cultivated in French Guiana can produce high‑quality agarwood and provides new insights into the interplay between microbial communities and agarwood chemistry. These findings lay the groundwork for the development of locally adapted, microbiome‑informed strategies for sustainable agarwood production. Agronomy General Microbiology Agroecology Aquilaria Oud microbiome chemistry terroir Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 1. Introduction Agarwood is a resinous wood formed following infection or injury in trees of the family Thymelaeaceae, which originate from the Indomalesian region. Beyond its medicinal uses, agarwood has been burned as incense since antiquity in Asia and the Middle East, where it is regarded as one of the most precious natural fragrant materials. Agarwood is also distilled to produce agarwood oil, a highly valued ingredient in fine perfumery. The principal agarwood‑producing genera, Aquilaria and Gyrinops , are currently listed on the IUCN Red List due to excessive exploitation of wild populations (Lee and Mohamed 2016 ). The planting of these species for industrial agarwood production may alleviate pressure on wild populations, and most current trade now relies on planted trees inoculated with formulations often kept secret that enabled agarwood formation. Although the agarwood formation process in Aquilaria trees is not yet fully understood, it is widely accepted that resin production is triggered by natural or deliberate wounding followed by colonization by environmental microorganisms (Tan et al. 2019 ; Faizal et al. 2022 ). Among these factors, stress plays a central role, as Aquilaria trees produce and accumulate secondary metabolites in response to injury (Rasool and Mohamed 2016 ; Naziz et al. 2019 ). To produce agarwood in managed plantation, induction techniques imply either mechanical wounding of the trunk, the use of chemical inducers (Zhang et al. 2012 ) but also microbial-based techniques using fungal inoculum (Tan et al. 2019 ; Ngadiran et al. 2023 ). These latter methods are often based on inoculation by Aquilaria endophytic or soil fungi, promoting a more natural and environmentally friendly solution to improve productivity and quality. Agarwood microbiome is known to differ from that of non-infected wood, and has been reported to contain Actinomycetes such as Nocardia and Streptomyces , and pathogenic fungi such as Fusarium and Trichoderma (Mohamed et al. 2010 ; Islam et al. 2022 ). Fungal inoculum thus often originates from the specific microbiome of infected Aquilaria trees or from the surrounding soil. However, choosing the appropriate microbial strains to induce a stress response of the tree enabling to produce the molecules of interest to obtain high quality agarwood (Fu et al. 2024 ). Moreover, while the current knowledge on agarwood formation and deliberate induction are progressing (Turjaman et al. 2016 ), major challenges still remain concerning the quality of the resulting agarwood. The chemical composition of agarwood has been studied extensively and is characterized by the occurrence of two main families of specialized metabolites: sesquiterpenoids and 2-phenylethylchromone derivatives (Li et al. 2021 ). Both of these groups contribute to the olfactory properties of agarwood: sesquiterpenoids are generally sufficiently volatile to be intrinsically fragrant, and some of them are well known strong odorants (Baldovini 2022). In contrast, 2‑phenylethylchromones have higher molecular weights and therefore much lower volatility; although essentially odorless, they act as precursors of lighter aromatic compounds produced during thermal degradation when agarwood is burned. Nevertheless, the identity of the most important olfactory contributors remains debated, and no clear relationship has yet been established between odor quality and specific chemical markers. In general, identifying key odorants in aromatic raw materials is challenging due to complex synergistic interactions and the disproportionate influence of trace constituents. It is particularly true in the case of complex mixtures like agarwood oils or extracts containing a very large number of specific constituents with high molecular complexity, and also showing an important variability of composition from one origin to the other. The chemical complexity of agarwood is partly attributable to the contribution of bacteria and fungi during resin development. Microorganisms are well known to influence the flavor profile, nutritional chemistry, and sensory properties of numerous biological products. A well-known example is the grapevine associated soil and endophytic microbiota, which contribute to the taste and quality of wine (Knight et al. 2015a ; Belda et al. 2017 ). The combined influence of biotic and abiotic factors shaping product flavor and quality is increasingly described by the concept of “microbial terroir” (Gilbert et al. 2014 ; Nakano 2024 ). This concept, mainly referring to wine grapes, pertains to the geographic microbial signature of soil, rhizosphere and endophytic fungi in grapes and vineyards (Miura et al. 2017a ) on the fermentation process (Bokulich et al. 2016 ; Mas and Portillo 2022 ) and ultimately on the wine taste. The microbial terroir, beyond wine grapes have also been found to influence the flavor chemistry of mustard seeds, where rhizosphere taxa predicted glucosinolate content, responsible for the desired spicy and bitter flavor of mustard seeds (Walsh et al. 2024 ). In addition to abiotic parameters such as climate, pH, nutrient content and management practices, local microorganisms also have a role on the quality and specificity of food products. While endophytic microbial communities of Aquilaria may have an influence in the process of agarwood formation, their influence on the resulting agarwood chemical quality and signature is yet to be determined, particularly in the concept of microbial terroir. This is particularly important as wood chips are often macerated in water before hydro-distillation to improve oil extraction yield and quality. Therefore, microorganisms are finally involved in two key steps of the process; infection of living trees, and fermentation of wood chips before distillation. Indeed, microbial diversity during maceration (Islam et al. 2022 ) and microbial consortia (Naziz et al. 2024 ) were found to improve agarwood oil quality. Local endophytic microbial communities of agarwood may thus have an influence on the maceration process, and could also be targeted for improving the resulting agarwood oil quality. In French Guiana, the Aquilaria tree is considered a cash crop by farmers in the communities of the villages of Regina and Cacao. These farmers, originally from Laos and with a long agricultural tradition (Vang 2018 ) possess traditional knowledge and expertise in cultivating the Aquilaria tree. In French Guiana, they are working to establish a sustainable Aquilaria production sector. To support high quality agarwood production, it is necessary to take into account local fungal and bacterial communities, as well as to estimate the quality of Aquilaria wood that has been planted outside its natural range. In this study, we aimed at elucidating the microbiome and chemical composition of agarwood in French Guiana, which are expected to differ from those in Aquilaria's geographical area of origin in Asia. Here, we investigated the association between fungi, bacteria, and chemical constituents to identify microbial taxa potentially associated with agarwood scent. Finally, we assessed the effect of geographic location on agarwood quality to test the microbial terroir hypothesis, which we further extend to a broader concept of “chemical terroir.” 2. Material and methods 2.1. Study site Sampling was conducted in two agricultural plots located in the Regina (Farmer A: 4°17'42.3"N 52°10'55.2"W; Farmer B: 4°17'41.8"N 52°10'54.6"W) and one in Cacao (Farmer C: 4°32'22.4"N 52°28'38.7"W) villages in French Guiana. At each of the three sites, two Aquilaria crassna trees previously inoculated in November 2022 with three fungal strains ( Antrodia vaillantii (DC. Ryvarden), Pycnoporus sanguineus (L.) Murril, and Coriolopsis polyzona (Pers.) Ryvarden) were selected. The detail of sampling sites, inoculated tree, and blackwood infection is presented in Fig. 1.A-H. The trunk was sectioned and five cross-sections were made at the inoculation point on each trunk. Two samples (one blackwood and one whitewood) were collected from each cross-section for a total of 60 samples (Fig. 1.I). 2.2. Soil sampling and analysis Five soil samples were sampled at 20 cm deep at each sampling site, for a total of 15 soil samples. To measure the elemental composition of each soil, samples were processed in triplicate using X-ray fluorescence spectroscopy (XRF). For each sample, triplicates were pressed at a 20 tons pressure for 2 min with 1:3 (v:v) of SpectroBlend® (SCP Science). Three technical measures on each triplicate, using the ‘geoexploration’ calibration with a total exposure time of 105 s on the XRF S1 Titan analyzer (Bruker) was used to measure the soil atomic composition of elements from magnesium to uranium, leading to nine measurements for each soil sample. The soil pH was measured at a 1:5 (v:v) ratio of H 2 O using a pH meter (pH Meter Knick 766, Knick International). Soil composition was visualized using composition barplots for the main elements (Al, Ca, Co, Cr, Cu, Fe, K, Mg, Mn, Si and Ti). Principal component analysis (PCA) using centered log-ratio transformed abundances were used to visualize soil composition by farmer plots. 2.3. Library preparation and sequencing Wood chips were crushed in liquid nitrogen using a sterile mortar and pestle. Approximatively 100–150 mg of wood were then used for DNA extraction using the FastDNA Spin Kit for Soil (MP Biomedicals™). DNA concentrations were measured by fluorescence with a Quant-iT™ PicoGreen™ dsDNA Assay Kit (ThermoFisher Scientific™, Waltham, Massachusetts, USA) using a Tecan Spark ® (Tecan Group Ltd., Mannedorf, Switzerland) and normalized to 1 ng.µL − 1 before PCR amplification. Primers 479F (5′ CAGCMGCYGCNGTAANAC 3′) and R888 (5′ CCGYCAATTCMTTTRAGT 3′) were used to amplify the V3-V4 regions of the 16S rRNA gene (Terrat et al. 2015 ). Primers ITS86F (5′ GTGAATCATCGAATCTTTGAA 3′) and ITS4 (5′ TCCTCCGCTTATTGATATGC 3′) were used to amplify the fungal ITS2 (Op De Beeck et al. 2014 ). Reactions were performed in 20 µL and consisted of 10 µL of Buffer Master mix containing Thermo Scientific Phusion™ HighFidelity DNA Polymerase (ThermoFisher Scientific™), 0.5 µL of DMSO, 4.5 µL of DNAse free water, 3 µL of DNA (0.3 ng. µL-1), 1 µL of forward primer and 1 µL of reverse primer tagged with a short nucleotide sequence. The PCR conditions used were identical for the fungal and bacterial amplifications and were performed in a Veriti 96-well Thermal Cycler (ThermoFisher Scientific™) under the following conditions: 10 min initial denaturation, 35 cycles each including denaturation at 94°C for 10 s, annealing at 55°C for 20 s, and extension at 72°C for 20 s, with a final polymerization extension at 72°C for 7 min. All reactions were performed in triplicate and then pooled before deposition on 2% agarose gel, with the corresponding negative control, for amplification quality control. Amplification products of ITS2 region were purified using Agencourt AMPure XP beads (Beckman Coulter Inc., Fullerton, California, USA) and performed on a magnetic rack with a 1:1 (v:v) ratio of AMPure XP per PCR product. The magnetic beads were then rinsed twice with 70% ethanol before final elution in 70 µL of Qiagen EB buffer. For 16S purification, an additional PCR product migration step on 2% agarose gel (80 V, 90 min) was performed to separate and recover the band specific to bacterial 16S (lower band, around 400 bp) and eliminate plant chloroplastic 16S DNA (higher band, > 600 bp, co-amplified during PCR). The recovered bands were purified using the QIAquick gel purification kit (Qiagen, Hilden, Germany). The purified PCR product concentration was then measured with PicoGreen and one equimolar pool (50 ng/sample) for each marker gene (16S and ITS2) was prepared. Each pool was finally purified twice using Agencourt AMPure XP beads, quantified and finally eluted in 40 µL of EB buffer (Qiagen). Library sequencing was performed on a 2 × 250 bp MiSeq system (Fasteris SA, Switzerland). For taxonomic assignment of fungi, OTUs were constructed on a 97% ITS clustering, while ASVs were used for bacteria (16S). We used a pipeline based on VSEARCH 2.18.0 (Rognes et al. 2016 ) and available in GitHub ( https://github.com/BPerezLamarque/Scripts/ ) for data processing. Briefly, only paired-end reads were retained and then merged using the fastq_mergepairs function with default parameters. We applied a quality check and removed merged reads with more than two alignment errors. Then, filtered merged reads were assigned to the respective samples based on tagged primers (demultiplexing) with 0 accepted error regarding the primers and tag sequences using Cutadapt 3.5. We removed chimeras de novo by using the VSEARCH uchime3_denovo option and assigned the taxonomy to each amplicon sequence variant (ASV) and Operational Taxonomic Unit (OTU) using the usearch_global option with default parameters. Taxonomic annotation was performed with SILVA (v.138.2) database for bacteria (16S rRNA) and UNITE database (v.9.0) for fungi (ITS). For bacteria we used ASV (16S rRNA) while 97% OTUs were used for fungi (ITS), as recommended by Tedersoo et al., ( 2022 ). The ASV and OTU tables were then filtered from contaminants based on comparisons with amplified negative controls using the DECONTAM (prevalence method) R package (Davis et al. 2018 ). Only ASVs and OTUs with long sequences (> 200 bp), assigned to the bacterial or fungal kingdom, presenting an acceptable abundance (≥ 10 reads in total dataset) and amplified in at least one sample, were kept for subsequent analyses. ASV and taxonomic tables are provided in Supplemental Tables 1–4, with the associated samples metadata in Supplemental table 5. 2.4. Characterization of volatile constituents in wood samples: sample preparation To determine the chemical composition of the volatile part of the samples used for microbiome analysis, we performed their hydrodistillation using a modified glass made Clevenger apparatus (Modverre, Grasse, France, Supplemental Fig. 1). This sample preparation technique has the advantage to extract the essential oil, which is the second main end product of agarwood industry after wood chips, and which can be analyzed as such by direct injection in a gas chromatography-mass spectrometry (GC-MS) system. Therefore, the analytical results obtained through this study can give an insight in the composition of agarwood oils that would be obtained by the large-scale industrial distillation of similar wood samples. Even if the obtained extract cannot be considered as an essential oil stricto sensu (Bicchi et al. 2018 ), it is nevertheless a good model for analytical investigation on industrial agarwood oil composition. However, in the context of the present work focused on wood amounts lower than 1 g, the classical Clevenger apparatus is not adapted as the low quantity of oil produced (few milligrams) would be lost in the hydrolate. Therefore, we used a modified system specifically designed to isolate such low amounts by trapping the organic volatiles in a solvent (tertbutylmethylether) in a cooled double jacketed collection tube to prevent the evaporation of the solvent during hydrodistillation. After optimization, the following parameters were applied to all hydrodistillations: the powdered wood sample (100–600 mg) and 40 ml water were placed in a 100 ml round bottomed flask equipped with an ovoid stir bar and connected to the Clevenger system. Two ml of tertbutylmethylether were added in the refrigerated collecting tube previously filled with water. The content of the flask was heated in an oil bath (145°C) under stirring for 19 h. At the end of the distillation, the solvent phase was collected, and concentrated until ca. 0.2–0.3 ml by flowing a gentle stream of argon, then diluted in ca . 1 ml of dichloromethane, dried by filtration through a short pad of magnesium sulfate in a Pasteur pipette, then concentrated under a flow argon and eventually brought to a volume of ca . 50–100 µl by adding dichloromethane. 2.5. Characterization of volatile constituents in wood samples: GC-MS/FID analyses Gas chromatography - Mass spectrometry/Flame Ionisation Detector (GC-MS/FID) analyses were carried out using an Agilent 6890N gas chromatograph equipped a fused silica capillary column DB-1MS (30 m × 0.25 mm i.d., film thickness: 0.25 µm, J&W122-0132). The analytical parameters were the following: the carrier gas was helium at a flow rate of 2 mL.min − 1 , the oven temperature was programmed from 60 to 280°C at 3°C/min and completed by a post run (300°C, 5 min). Five µl samples of the concentrated extract prepared as described above were injected in splitless mode and the injector temperature was 250°C. The column flow was split using a G3184-60065 splitter (Agilent), and transferred to a FID detector (250°C, hydrogen and air flows at 40 and 450 ml/min, respectively) and to an Agilent 5973N mass selective detector working in electron impact (EI) mode at 70 eV (scanning over 40–450 amu range in SCAN mode). The temperatures of the MS ion source and transfer line were 230 and 280°C, respectively. Linear retention indices (LRI) were determined from the retention times of a series of n -alkanes with linear interpolation. The main constituents were identified by comparison of their mass spectra and LRI with those of pure compounds registered in commercial libraries and literature data, and with a laboratory-made database built from authentic compounds, with the help of SearchReview software ( https://waleson.eu/ ). To compare the chemical compositions of the Guianese samples with those of commercial agarwood, samples of commercial blackwood from the Riyadh and AlUla (Saudi Arabia) street market were hydrodistilled and analyzed using the same protocol. The chemical abundance and compounds assignments are provided in tables 6, 7. 2.6. Microbial and chemical diversity To assess if sequencing depth was sufficient to retrieve microbial abundance, rarefaction curves were performed (Supplemental Fig. 2) using the mirlyn package (Cameron et al. 2021 ). Relative abundances of bacterial and fungal community and chemical compositions were then visualized using bubble plots at the genus level according to the wood type and farmer plots. To assess the differences in microbial diversity in black and white wood samples and as a function of sampling sites, Hill diversity index for three orders of diversity (q = 0, 1, and 2) were calculated and the significances of their differences assessed using Anova followed by Tukey HSD tests. 2.7. Microbial and chemical composition To assess the effect of geographical location (farmer plots) on the beta-diversity of microbial composition, and its interaction with the chemistry of wood samples, Hellinger-transformed microbial abundances and clr-transformed chemical abundance were used to perform a Redundancy analysis (RDA). The associated R 2 and its significance were assessed with a Permanova using the Bray-Curtis distance of the Hellinger-transformed microbial abundances using the formula: microbial abundance~Wood type*Farmer. For chemical compounds, beta diversity was visualized using a nMDS with its associated Permanova. To compare the composition of our samples with those of commercial agarwoods, a nMDS of the chemical compositions based on Guianese or foreign origin was also carried out. To identify differentially abundant microbial taxa and chemical compounds between white and black wood, we used DESeq2 (Love et al. 2014 ) on both datasets. The dispersion among individual counts was estimated by the DESeq2 algorithm, and the resulting logarithmic fold changes (Log2FC) with the associated p-values were used to identify differentially abundant features. Analyses were performed separately for the microbial and chemical dataset. To quantify within-group heterogeneity as distance-to-centroid on Aitchison (CLR) distances and compare black vs white woods, betadisper and permutest were used. Then, to relate chemistry to microbiology, linear models of microbiological dispersion (ITS or 16S) on chemical dispersion with wood type and their interaction were fitted; z-scored predictors were also added to interpret group differences at mean chemistry distances. Within-group monotonic concordance was assessed via Spearman ρ, overall sample-to-sample correlations were tested on raw dispersion values, and global multivariate structures between datasets were compared with Mantel tests and Procrustes analyses on cmdscale ordinations. 2.8. Co-occurrence network analysis of microbial-chemistry In order to accurately assess the correlations between microbial taxa and volatile constituents, network analyses based on covariance relationships were carried out. All the constituents were classified according to their putative biosynthetic origin to facilitate the statistical treatment of the chemical information. Among all the volatile constituents, sesquiterpenoids were easily recognized and associated to seven subgroups defined by their sesquiterpenic skeleton (bisabolane, eremophillane, eudesmane, guaiane, humulane, vetispirane, zizaane). Not surprisingly in view of the classical composition of agarwood oils, many aromatic compounds were also identified, and were classified either as “chromone derivatives” or “possible chromone derivatives”, depending on the degree of confidence we had on their suspected biogenetic origin. In addition to these groups, the other families were “linear / fatty acid derivatives”, “adulterants” and “possible pollutants” (to distinguish respectively suspected voluntarily and unvoluntary added compounds, mostly in commercial samples). Moreover, many minor compounds were not present in our database and could not be identified with certainty, therefore they were classified as “unidentified”. Considering microbial and chemical datasets as compositional and the paired samples design of our study, we used the cross-kingdom network inference using SPIEC-EASI (Kurtz et al. 2015 ; Tipton et al. 2018 ), which is robust to the bias of using two compositional datasets (Brunner et al. 2024 ). Multi domain SPIEC-EASI was thus used to build co-occurrence networks for agarwood samples (black wood) using the Meinshausen-Buhlmann's neighborhood selection method using 100 repetitions and 100 subsamples for the Stability Approach to Regularization Selection (StARS) algorithm (Liu et al. 2014 ). The optimal coefficient matrix and symmetric matrix with edge-wise stability was then retrieved after verification of the optimal network stability at the selected sparsity. Sub-networks according to each chemical class were then created using the original ITS-chemistry and 16S-chemistry cross-kingdom networks. To do so, for each chemical class, we retrieved chemical nodes (named hereafter ‘core nodes’) belonging to each chemical class as well as their connections to other chemical compounds and microbial taxa. The internal microbial edges were removed (bacteria-bacteria or fungi-fungi edges) to specifically assess the interaction of microbial taxa and chemical compounds. Then, we calculated the number of positive and negative edges, as well as their taxonomic assignation (chemical or microbial class). Finally, to identify the significant covariance relationships between each microbial class and genus to each chemical class, the ratio of positive to negative edges were calculated for each sub-network specific of each chemical class. Its value (positive, negative or neutral) is an indication of the overall link between microbial taxa and chemical compounds found in agarwood. 3. Results 3.1. A ferralitic tropical soil All plantations were established on ferralitic tropical soils characterized by high iron concentrations (50.5 ± 6.1%; 41.2 ± 7.6%; 60.9 ± 4.3% for farmer A, B and C respectively), followed by Si (20.2 ± 5.0%; 27.4 ± 6.9%; 13.7 ± 2.7%), Al (16.1 ± 4.1%; 16.4 ± 3.6%; 14.9 ± 3.0%), Mg (6.9 ± 4.0%; 6.6 ± 3.6%; 5.1 ± 1.9%) and Ti (5.5 ± 0.4%; 7.0 ± 1.0%; 4.7 ± 0.2% for farmer A, B and C respectively). Soil pH did not differ significantly among plantations (mean pH = 5.3 ± 0.1, p.value = 0.0786). Relative abundances of each element for farmer plot are available in Supplemental table 8. The soil from farmer C plantations had a higher Fe, Cr and Mn content, and a lower K and Si content compared to the other plantations (Supplemental Fig. 3), resulting in a contrasted soil composition compared to the other plantations (Fig. 2.C). 3.2. Agarwood microbial and chemical composition Enterobacter , Kluyvera and Pantoea were the main bacterial genera found in white wood and this assembly did not vary significantly according to sampling sites (Fig. 3.A). Diplodia , Nigrospora , Pestalotiopsis , Gongronella , Fusarium , Trichoderma and Curvularia were the main fungal genera found in white wood (Fig. 3.B). High concentration of fatty acids (hexadecanoic, oleic, octadecanoic and linoleic acids) were found in white wood samples, and their abundance did not significantly vary between samples nor sampling sites (Fig. 3.C). On the other hand, Pantoea , an unknown genus, Acidibacter , Burkholderia and Enrotobacter were the main bacterial genera found in black wood (Fig. 3.A) and their relative abundance significantly vary between samples, and to a higher extent, between sampling sites. Trechispora , Diaporthe , Fusarium , Oliveonia , Calosphaeria and Striaticonidium were the main fungal genera (Fig. 3.B), their abundance also significantly varies according to samples and sampling sites. Black wood volatile constituents were generally more abundant, their number was also higher and their chemical structures were more diverse. Even if the fatty acids abundant in white wood were still present in black wood, we also found high amounts of typical agarwood compounds (Naef 2011 ) such as oxo-agarospirol, 4-methoxyphenyl-butan-2-one, β-agarofurane, jinkoh eremol, agarospirol and dihydrokaranone (Fig. 3.C). These microbial and chemical patterns were confirmed by the differential abundance analysis, where a high share of Pseudomonadota bacteria was found specific to the white wood, while a higher diversity of bacteria, belonging to diverse phyla was specific to the black wood (Supplemental Fig. 3.A; Supplemental Table 9). This was also the case for fungi, where all taxa differentially abundant in white wood belonged to Ascomycota phyla, such as Curvularia , Nigrospora or Chaetomium . In the black wood, both Ascomycota, Oomycota and Basidiomycota were differentially abundant. Interestingly, a high share of unidentified fungal species (Incertidae Sedis = IS) was found specific to the black wood (Supplemental Fig. 3.B). Linear and fatty acids were found highly specific to the white wood, while a more diverse chemical diversity was specific to the black wood (Supplemental Fig. 3.C). We found a specific signature of French Guiana agarwood samples when compared to foreign samples from South-East Asia (Figure S4.A). Chromone derivatives, possible chromone derivatives, and sesquiterpenoids were more abundant in Guiana agarwood samples (Supplemental Fig. 4.B). Furthermore, adulterants such as di-n-butyl phthalate were found specific to foreign samples, testifying to their poor quality. 3.3. Inter- and Intra-sample variability of the microbial and chemical composition Overall, dispersion varied between black and white woods for chemistry and microbial communities (Supplemental Fig. 5.A). Chemical composition dispersion was a strong positive predictor of microbiological communities’ dispersion (ITS: F₁,₄₉ = 21.18, p = < 0.001; 16S: F₁,₄₉ = 206.49, p < 0.001), with an effect of wood type on bacterial dispersion (16S: F₁,₄₉ = 107.58, p < 0.001), but no interaction between chemical dispersion and wood type. In standardized models, the chemistry slopes remained significantly positive (ITS β = 0.410, p = 0.043; 16S β = 0.195, p = 0.034), and white wood showed lower dispersion than black at average chemistry (ITS β = −0.769, p = 0.035; 16S β = −1.578, p = 3.01e − 10; Supplemental Fig. 5.B). Model fit was moderate for ITS (Adj. R² = 0.31) and high for 16S (Adj. R² = 0.86). Overall sample-to-sample correlations between chemical and microbiological dispersion were moderate for fungi (Spearman ρ = 0.434, p = 0.001) and strong for bacteria (ρ = 0.746, p < 2.2e − 16). Mantel tests confirmed significant correspondence between chemical and microbiological distance matrices (chemistry–ITS: r = 0.418, p = 1e − 04; chemistry − 16S: r = 0.390, p = 1e − 04). Ordination structures aligned well between chemistry and microbiology (Procrustes correlations: chem–ITS r = 0.677, p = 1e − 04; chem–16S r = 0.720, p = 1e − 04), supporting substantial shared multivariate patterns (Supplemental Fig. 5.C). Chemical composition strongly structures both bacterial and fungal communities, with bacteria exhibiting more deterministic, chemistry-driven assembly than fungi. 3.4. The terroir effect on microbial and chemical composition Concerning the overall community structure, wood endophytic bacterial beta diversity varied strongly according to the wood type (Fig. 4.A; R 2 = 0.36), while farming plot had a limited effect (R 2 = 0.08). This pattern was similar for fungi, where wood type had the major effect on beta diversity (Fig. 4.B; R 2 = 0.30), while farming plot had a limited effect (R 2 = 0.08). Redundancy analysis (Fig. 4.C) also explained a higher shared of bacterial community composition (Dim 1 = 47.3% and Dim 2 = 12%) than for fungi (Dim 1 = 17.9% and Dim 2 = 9.8%). Concerning the specificity of sequences and taxa found in Aquilaria wood, a high percentage of sequences were shared between black and white wood samples for all farmer plots (Fig. 5.A), ranging from 84.4 to 90.2% of shared sequences. A higher share of unique sequences was found in black wood compared to white wood for both fungi and bacteria, where 9 to 15% of sequences were unique to black wood, while only 0.5 to 1% of sequences were unique to white wood. Conversely, a high number of taxa were found specific to black wood samples, where up to 955 bacterial taxa (out of 1548) were specific to black wood samples for the farmer C plot. This high diversity and high specificity of bacterial taxa in black wood samples was conserved for all plots. For fungi, the share of black wood specific taxa was similar to that of shared taxa. Chemical compounds also had a high proportion of common compounds between black and white woods, but a high number and proportion was specific to black wood (40 to 52 compounds). Concerning the specificity of microbial structure based on sampling sites, while a high share of sequences was found in samples from all locations (98.0% and 81.2% for bacteria and fungi of the white wood; 78.4% and 87.9% for the black wood), a high number of taxa were not found in all plots (Fig. 5.B). For example, the farming plot C had a high number of specific taxa, both in white (113 out of 623 bacterial ASV) and black wood (360 out of 1 730 bacterial ASV). However, between the three sampling sites, the chemistry of white and black wood remained highly consistent (99.8 and 97.3% of share compounds respectively). 3.5. Association between microbial members and chemical composition The resulting sub-networks of chemical class and microbial composition revealed that each chemical class had specific relationships with different microbial taxa. Linear fatty acid derivatives had the highest connections with bacterial taxa (Supplemental Figure S6, Supplemental table 10); core degree = 68.4), followed by possible chromone derivative (core degree = 66.2), while sesquiterpenoid humulane had the lowest (core degree = 38). The pattern was similar for fungi, where possible chromone derivative had the highest connection with fungal taxa (Supplemental Figure S7; core degree = 15.2), followed by linear fatty acid derivative (core degree = 13.8), and sesquiterpenoid humulane having the lowest (core degree = 9). The most connected bacterial nodes to chemical compositions belonged to the Alphaproteobacteria, Gammaproteobacteria, Actinobacteria and Planctomycetes bacterial class (Supplemental Fig. 7.A; Supplemental table 11). The most connected fungal nodes to chemical compositions belonged to the fungal class Sordariomycetes, Eurotiomycetes, and Dothideomycetes (Supplemental Fig. 7.B). Among these connected microbial nodes to chemical compounds, we were able to identify the significantly positively and negatively associated bacterial (Fig. 6.A) and fungal class (Fig. 6.B) and genus (Supplemental Fig. 9) to each chemical class. Some microbial class exhibited consistent negative relationships with chemical compounds such as members of the Actinobacteria and Vicinamibacteria class, while others had a positive relationship such as Phycisphareae, Spirochaetia, Mucoromycetes and Umbelopsidomycetes. When focusing on chromone derivative, we found 10 bacterial and 5 fungal class positively associated with increased chromone derivative proportion, and 9 bacterial and 2 fungal class negatively associated. For example, Dothideomycetes and Sordariomycetes were negatively associated with chromone and possible chromone derivative proportion in the black wood. 4. Discussion 4.1 French Guiana, a suitable terroir for quality agarwood Endophytic microbial communities of French Guiana Aquilaria resemble those of South-East Asia, where Bacillus and Pantoea were also reported (Bora et al. 2025 ), as well as Aspergillus , Penicillium or Fusarium (Fu et al. 2024 ). Numerous saprotrophs and pathotrophs fungi were identified in black wood. For example, Trechispora , highly abundant in black wood, is a white rot commonly found on decaying wood (Põlme et al. 2020 ). Botryosphaeriaceae were necrotrophic or saprobic in woody tissues and were also found in high abundance in black wood. Calosphaeriaceae , Oliveonia , Coniochaetaceae , Helotiaceae , Kylindria , Stypella were among the most commonly found wood saprobes in black wood. Some fungi may also exhibit a necrotic phase such as the microfungus belonging to Glomerellaceae family. Although overall microbial community composition was conserved across ampling sites, location‑specific differences were observed. Spatial processes have been show to impact the fungal communities of grapevine, where distance decay relationships of microbial community dissimilarity were found on small distances (Miura et al. 2017b ). In our case, the differences in soil composition, where the plot C had more iron and chromium content, may have impacted microbial communities, and thus, agarwood chemical composition. The specificity of Guianese agarwood is further supported by the strong chemical composition differences we found between local and foreign samples (Supplemental Figure S4). Plant genotype, micro-climatic environment and agricultural practices may also drive microbial differences in soil and wood endophytic communities, as demonstrated in the vineyards (Bokulich et al. 2014 ; Knight et al. 2015b ). This is supported by the specific microbial taxa we found between plots, where geographical distance had an impact on soil chemistry and microclimatic conditions and thus, microbial communities. However, the extent to which specific microbial communities drive desirable chemical terroirs and sensory outcomes remains to be determined. While it was not investigated here, the specific microbiome and climate of French Guiana might also influence the maceration process, and further increase the specificity of the resulting essential oil following distillation. While Aquilaria show promises in French Guiana, it originates from South-East Asia, and its implantation should follow strict environmental precautions. Aquilaria spp. can grow from 15 to 25 m in height and have a low number of flowers per inflorescence, obligately outcrossing and the barochoric dispersal of seed do not spread far from the trunk (Soehartono and Newton 2001 ). These trait with the lifespan of the seeds of only a few days, may limit Aquilaria invasiveness in French Guiana (Hamilton et al. 2005 ) but careful monitoring following its plantation should be implemented. No report of invasiveness has been identified as of today, whether within the natural distribution area of ​​ Aquilaria spp. or from plantations within and outside the natural distribution area. 4.2. The interplay of plant and its microbiome for agarwood scent It remains difficult to assess a direct link between chemical compounds and microbial composition. For example, while bacterial and fungal taxa were positively linked to chromone derivative abundance, some biosynthesis pathways exclusively pertain to the plant such as the aromatic ketone benzylacetone or 4-phenylbutan-2-one. This secondary metabolite, present in flowers is an attractant for pollinators (Kessler and Baldwin 2007 ) and the acyltransferase capable of synthesizing benzylbenzoate (D’Auria et al. 2002 ) is only found in Magnolopsida and Pinopsida . The volatile organic compounds belonging to sesquiterpenes are produced by both plant and fungi to attract insect pollinators, defend against pathogens and modulate microbial interactions with antagonistic or synergistic effects on other microbial organisms (Kramer and Abraham 2012 ). Among the fungal kingdom, Basidiomycota are known to produce sesquiterpenes via humulane (Abraham 2001 ), and Ascomycota fungi also produce volatile sesquiterpenes such as members of the genus Penicillium . In our study, we found the Penicillium genus to be strongly positively associated to chromone derivative and sesquiterpenoid eudesmane abundance, which could constitute an interesting inoculum for boosting agarwood quality. Other genera showed promising associations with agarwood chemical quality such as Rhizobium , Solirubrobacter , Acrocalymma and Curvularia . These taxa may be linked to a higher concentration of odorant molecules either through their direct biosynthesis, inducing stronger plant defenses, or modulating microbial competition. Further analysis on the effect of these genera on plant fitness and stress response could provide valuable insights in the production of quality agarwood. Methods such as synthetic communities could be used to produce an optimal inoculum, boosting plant growth and inducing plant response following stress (Xu 2025 ). For the first time, we provide an in-depth analysis of bacterial and fungal taxa associated with volatile constituents of agarwood, and identified positively and negatively associated microbial members to each chemical class. The positively associated taxa with chromone derivative such as members of the Alphaproteobacteria, Baccilli, Thermophilia, Verrucomicrobiia, Agaromycetes, or Mucoromycetes could provide valuable local inoculum resource, for tree infection or maceration starter. Further studies should test the effect of complex communities’ inoculation on agarwood formation throughout plant development and their involvement in the metabolomic induction of biosynthesis pathways relevant to agarwood scent. 5. Conclusion This study provides a comprehensive in-depth characterization of the bacterial and fungal communities associated with agarwood produced from A. crassna cultivated in French Guiana and examines their relationships with volatile chemical constituents relevant to agarwood scent. Our results demonstrate that Aquilaria grown outside its native range in French Guiana is capable of producing high‑quality agarwood, with a conserved chemical profile across distinct sampling sites, supporting the existence of a stable chemical terroir under local environmental conditions. By integrating microbial community profiling with detailed chemical analyses, we identified numerous bacterial and fungal taxa that were positively or negatively associated with major chemical classes, including chromone derivatives and sesquiterpenoids, which are key contributors to agarwood fragrance. Although these relationships remain correlational, they suggest that microbial communities may influence agarwood chemical complexity through interactions with plant stress responses, microbial metabolism, or cross‑kingdom ecological processes. The persistence of a conserved chemical profile despite spatial variability in microbial composition further indicates potential functional redundancy among microbial taxa or a strong host‑mediated regulation of secondary metabolite production. Comparisons with commercial agarwood samples from South-East Asia and the Middle East revealed a distinct chemical signature of Guianese agarwood, reinforcing the influence of geographic origin on agarwood quality and supporting an extension of the terroir concept to woody aromatic products. These findings highlight the importance of considering both microbial communities and environmental context when evaluating agarwood quality and origin. Overall, this work provides a foundation for the development of microbiome‑informed strategies for agarwood induction, maceration, and processing based on local microbial resources. Future research combining experimental inoculation, longitudinal sampling, and metabolic approaches will be essential to disentangle causal relationships between microbial taxa and chemical traits and to optimize sustainable agarwood production. Declarations Funding: This project was founded by the ANR (Agence Nationale de la Recherche) under the project ACQUILASCENT. Conflicts of interest: The authors declare no conflict of interest. Ethics approval: Not applicable Consent to participate: Not applicable Consent for publication: The authors affirm that human research participants provided informed consent for publication of this version of the manuscript, associated metadata, and images for Figure 1. Availability of data and material Sequencing data are available at the NCBI database (Bio-project PRJNA1249911, Biosamples SAMN47917871 to SAMN47917976). All metadata are provided as supplementary tables. Code availability : The datasets generated during and/or analyzed during the current study are provided in supplementary tables and the code is available from the corresponding author on reasonable request. Authors' contributions: Kenji Maurice : Conceptualization, Methodology, Investigation, Data curation, Software, Visualization, Writing – Original draft, Writing – Review and editing. Nicolas Baldovini : Conceptualization, Supervision, Investigation, Data curation, Funding acquisition, Project administration, Writing – Review and editing. Alba Zaresmki : Project administration, Funding acquisition, Writing – Review and editing. Jérémie Damay: Methodology, Investigation, Writing – Review and editing. Romain Lehnebach: Methodology, Investigation, Writing – Review and editing. Yannick Estevez : Methodology, Investigation. Marc Ducousso : Conceptualization, Supervision, Project administration, Funding acquisition, Writing – Review and editing. References Abraham W-R (2001) Bioactive Sesquiterpenes Produced by Fungi are they Useful for Humans as Well. CMC 8:583–606. https://doi.org/10.2174/0929867013373147 Baldovini N (2022) The Chemistry of Agarwood Odorants. In: Kinghorn AD, Falk H, Gibbons S et al (eds) Progress in the Chemistry of Organic Natural Products 118. Springer International Publishing, Cham, pp 47–100 Belda I, Zarraonaindia I, Perisin M et al (2017) From Vineyard Soil to Wine Fermentation: Microbiome Approximations to Explain the terroir Concept. Front Microbiol 8:821. https://doi.org/10.3389/fmicb.2017.00821 Bicchi C, Chaintreau A, Joulain D (2018) Identification of flavour and fragrance constituents. Flavour Fragr J 33:201–202. https://doi.org/10.1002/ffj.3445 Bokulich NA, Collins TS, Masarweh C et al (2016) Associations among Wine Grape Microbiome, Metabolome, and Fermentation Behavior Suggest Microbial Contribution to Regional Wine Characteristics. mBio 7:e00631–e00616. https://doi.org/10.1128/mBio.00631-16 Bokulich NA, Thorngate JH, Richardson PM, Mills DA (2014) Microbial biogeography of wine grapes is conditioned by cultivar, vintage, and climate. Proc Natl Acad Sci USA 111. https://doi.org/10.1073/pnas.1317377110 Bora SS, Ronghang R, Das P et al (2025) Endophytic Microbial Community Structure and Dynamics Influence Agarwood Formation in Aquilaria malaccensis Lam. Curr Microbiol 82:66. https://doi.org/10.1007/s00284-024-04048-2 Brunner JD, Robinson AJ, Chain PSG (2024) Combining compositional data sets introduces error in covariance network reconstruction. ISME Commun 4:ycae057. https://doi.org/10.1093/ismeco/ycae057 Cameron ES, Schmidt PJ, Tremblay BJ-M et al (2021) Enhancing diversity analysis by repeatedly rarefying next generation sequencing data describing microbial communities. Sci Rep 11:22302. https://doi.org/10.1038/s41598-021-01636-1 D’Auria JC, Chen F, Pichersky E (2002) Characterization of an Acyltransferase Capable of Synthesizing Benzylbenzoate and Other Volatile Esters in Flowers and Damaged Leaves of Clarkia breweri . Plant Physiol 130:466–476. https://doi.org/10.1104/pp.006460 Davis NM, Proctor DM, Holmes SP et al (2018) Simple statistical identification and removal of contaminant sequences in marker-gene and metagenomics data. Microbiome 6:226. https://doi.org/10.1186/s40168-018-0605-2 Faizal A, Hermawaty D, Junita E et al (2022) Evaluation of biotic and abiotic stressors to artificially induce agarwood production in Gyrinops versteegii (Gilg.) Domke seedlings. Symbiosis 86:229–239. https://doi.org/10.1007/s13199-022-00835-2 Fu C-C, Huang B-X, Wang S-S et al (2024) Deciphering the roles of bacterial and fungal communities in the formation and quality of agarwood. Stress Biology 4:40. https://doi.org/10.1007/s44154-024-00179-5 Gilbert JA, Van Der Lelie D, Zarraonaindia I (2014) Microbial terroir for wine grapes. Proc Natl Acad Sci USA 111:5–6. https://doi.org/10.1073/pnas.1320471110 Hamilton MA, Murray BR, Cadotte MW et al (2005) Life-history correlates of plant invasiveness at regional and continental scales. Ecol Lett 8:1066–1074. https://doi.org/10.1111/j.1461-0248.2005.00809.x Islam MR, Chakraborty C, Banu S (2022) Isolation and Characterization of Bacteria and Fungi Associated with Agarwood Fermentation. Curr Microbiol 79:313. https://doi.org/10.1007/s00284-022-02999-y Kessler D, Baldwin IT (2007) Making sense of nectar scents: the effects of nectar secondary metabolites on floral visitors of Nicotiana attenuata . Plant J 49:840–854. https://doi.org/10.1111/j.1365-313X.2006.02995.x Knight S, Klaere S, Fedrizzi B, Goddard MR (2015a) Regional microbial signatures positively correlate with differential wine phenotypes: evidence for a microbial aspect to terroir. Sci Rep 5:14233. https://doi.org/10.1038/srep14233 Knight S, Klaere S, Fedrizzi B, Goddard MR (2015b) Regional microbial signatures positively correlate with differential wine phenotypes: evidence for a microbial aspect to terroir. Sci Rep 5:14233. https://doi.org/10.1038/srep14233 Kramer R, Abraham W-R (2012) Volatile sesquiterpenes from fungi: what are they good for? Phytochem Rev 11:15–37. https://doi.org/10.1007/s11101-011-9216-2 Kurtz ZD, Müller CL, Miraldi ER et al (2015) Sparse and Compositionally Robust Inference of Microbial Ecological Networks. PLoS Comput Biol 11:e1004226. https://doi.org/10.1371/journal.pcbi.1004226 Lee SY, Mohamed R (2016) The Origin and Domestication of Aquilaria, an Important Agarwood-Producing Genus. In: Mohamed R (ed) Agarwood. Springer Singapore, Singapore, pp 1–20 Li W, Chen H-Q, Wang H et al (2021) Natural products in agarwood and Aquilaria plants: chemistry, biological activities and biosynthesis. Nat Prod Rep 38:528–565. https://doi.org/10.1039/D0NP00042F Liu H, Roeder K, Wasserman L (2014) Stability Approach to Regularization Selection. StARS) for High Dimensional Graphical Models Love MI, Huber W, Anders S (2014) Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2. Genome Biol 15:1–21 Mas A, Portillo MC (2022) Strategies for microbiological control of the alcoholic fermentation in wines by exploiting the microbial terroir complexity: A mini-review. Int J Food Microbiol 367:109592. https://doi.org/10.1016/j.ijfoodmicro.2022.109592 Miura T, Sánchez R, Castañeda LE et al (2017a) Is microbial terroir related to geographic distance between vineyards? Environ Microbiol Rep 9:742–749. https://doi.org/10.1111/1758-2229.12589 Miura T, Sánchez R, Castañeda LE et al (2017b) Is microbial terroir related to geographic distance between vineyards? Environ Microbiol Rep 9:742–749. https://doi.org/10.1111/1758-2229.12589 Mohamed R, Jong PL, Zali MS (2010) Fungal diversity in wounded stems of Aquilaria malaccensis. Fungal Divers 43:67–74. https://doi.org/10.1007/s13225-010-0039-z Naef R (2011) The volatile and semi-volatile constituents of agarwood, the infected heartwood of Aquilaria species: a review. Flavour Fragr J 26:73–87. https://doi.org/10.1002/ffj.2034 Nakano RT (2024) Taste of microbes: the terroir explained by rhizospheric microbes. New Phytol 243:1639–1641. https://doi.org/10.1111/nph.19796 Naziz PS, Das R, Sen S (2019) The Scent of Stress: Evidence From the Unique Fragrance of Agarwood. Front Plant Sci 10:840. https://doi.org/10.3389/fpls.2019.00840 Naziz PS, Das R, Sen S (2024) Augmentation in fragrant agarwood oil quality by fermentation with a microbial consortium of bacterium (Microbacterium oxydans) and fungus (Penicillium aethiopicum). Eur J Wood Prod 82:1319–1332. https://doi.org/10.1007/s00107-024-02089-9 Ngadiran S, Baba S, Nor NFA et al (2023) The induction techniques of resinous agarwood formation: A review. Bioresource Technol Rep 21:101337. https://doi.org/10.1016/j.biteb.2023.101337 Op De Beeck M, Lievens B, Busschaert P et al (2014) Comparison and Validation of Some ITS Primer Pairs Useful for Fungal Metabarcoding Studies. PLoS ONE 9:e97629. https://doi.org/10.1371/journal.pone.0097629 Põlme S, Abarenkov K, Henrik Nilsson R et al (2020) FungalTraits: a user-friendly traits database of fungi and fungus-like stramenopiles. Fungal Divers 105:1–16. https://doi.org/10.1007/s13225-020-00466-2 Rasool S, Mohamed R (2016) Understanding Agarwood Formation and Its Challenges. In: Mohamed R (ed) Agarwood. Springer Singapore, Singapore, pp 39–56 Rognes T, Flouri T, Nichols B et al (2016) VSEARCH: a versatile open source tool for metagenomics. PeerJ 4:e2584. https://doi.org/10.7717/peerj.2584 Soehartono T, Newton AC (2001) Reproductive ecology of Aquilaria spp. in Indonesia. Forest Ecology and Management 152:59–71. https://doi.org/10.1016/S0378-1127(00)00610-1 Tan CS, Isa NM, Ismail I, Zainal Z (2019) Agarwood Induction: Current Developments and Future Perspectives. Front Plant Sci 10:122. https://doi.org/10.3389/fpls.2019.00122 Tedersoo L, Bahram M, Zinger L et al (2022) Best practices in metabarcoding of fungi: From experimental design to results. Mol Ecol 31:2769–2795. https://doi.org/10.1111/mec.16460 Terrat S, Dequiedt S, Horrigue W et al (2015) Improving soil bacterial taxa–area relationships assessment using DNA meta-barcoding. Heredity 114:468–475. https://doi.org/10.1038/hdy.2014.91 Tipton L, Müller CL, Kurtz ZD et al (2018) Fungi stabilize connectivity in the lung and skin microbial ecosystems. Microbiome 6:12. https://doi.org/10.1186/s40168-017-0393-0 Turjaman M, Hidayat A, Santoso E (2016) Development of Agarwood Induction Technology Using Endophytic Fungi. In: Mohamed R (ed) Agarwood. Springer Singapore, Singapore, pp 57–71 Vang CY (2018) Thinking Refugee: The Politics of Hmong Place-Making in Argentina and French Guiana. Amerasia J 44:1–21. https://doi.org/10.17953/aj.44.2.1-21 Walsh C, Vanderburgh C, Grant, Lady et al (2024) Microbial terroir: associations between soil microbiomes and the flavor chemistry of mustard ( Brassica juncea ). New Phytol 243:1951–1965. https://doi.org/10.1111/nph.19708 Xu X (2025) Composing a microbial symphony: synthetic communities for promoting plant growth. Trends in Microbiology Zhang XL, Liu YY, Wei JH et al (2012) Production of high-quality agarwood in Aquilaria sinensis trees via whole-tree agarwood-induction technology. Chin Chem Lett 23:727–730. https://doi.org/10.1016/j.cclet.2012.04.019 Additional Declarations The authors declare no competing interests. Supplementary Files SupplementalFiguresACQ.pdf Supplemental Figures Cite Share Download PDF Status: Posted Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-9040724","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":601284674,"identity":"bf1a06cb-411c-4c7d-807d-e49a6473c922","order_by":0,"name":"Kenji Maurice","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA80lEQVRIiWNgGAWjYBACCSA2SAAzGRsYGCqANDNzAylazoC0MBLWggCMbTC9eIBke++BggcM2+QNbh9u/PBxXm00fztQy4+KbTi1SPOcSwA67LbhhnOJzZIztx3PnXGYsYGx58xtnFrkJHIMQFoYN5xhbGPm3XYstwGohZmxDY8W+TdgLfZgLX/nHMudT0iLtAQPWEsiWAtjQ03uBkJaJHtADjO4nTzzDGOzZM+xA7kbgVoO4vOLxPEzZoY/Km7b9p1hf/jhR01d7rzzhw8+AIrg1AIEbAYMBnDOYTB5AJ96IGB+gMSpI6B4FIyCUTAKRiIAAFNOXHNffG93AAAAAElFTkSuQmCC","orcid":"https://orcid.org/0000-0003-4047-7746","institution":"AGAP Institut, Univ Montpellier, CIRAD, INRAE, IRD, SupAgro, Montpellier, France","correspondingAuthor":true,"prefix":"","firstName":"Kenji","middleName":"","lastName":"Maurice","suffix":""},{"id":601284675,"identity":"bed77e74-7248-45f7-a19b-a283bab8afc7","order_by":1,"name":"Nicolas Baldovini","email":"","orcid":"","institution":"Institut de Chimie de Nice, UMR 7272, Université Côte d’Azur, Parc Valrose, 06108 Nice, France","correspondingAuthor":false,"prefix":"","firstName":"Nicolas","middleName":"","lastName":"Baldovini","suffix":""},{"id":601284676,"identity":"661a8a6c-d937-4b35-b396-71a9d8463fcf","order_by":2,"name":"Alba Zaresmski","email":"","orcid":"","institution":"AGAP Institut, Univ Montpellier, CIRAD, INRAE, IRD, SupAgro, Montpellier, France","correspondingAuthor":false,"prefix":"","firstName":"Alba","middleName":"","lastName":"Zaresmski","suffix":""},{"id":601284677,"identity":"6608dc7c-f2a6-44cf-84c5-12e9ff47a71a","order_by":3,"name":"Jérémie Damay","email":"","orcid":"","institution":"CIRAD, UPR BioWooEB, Montpellier, F-34398, France","correspondingAuthor":false,"prefix":"","firstName":"Jérémie","middleName":"","lastName":"Damay","suffix":""},{"id":601284678,"identity":"3d6950d5-0a67-4c81-8313-e90149355405","order_by":4,"name":"Romain Lehnebach","email":"","orcid":"","institution":"ECOFOG, AgroparisTech, Cirad, CNRS, INRAE, Univ. Guyane, Univ. Antilles - Pariacabo - BP 701 97387 Kourou Cedex, Guyane française","correspondingAuthor":false,"prefix":"","firstName":"Romain","middleName":"","lastName":"Lehnebach","suffix":""},{"id":601284679,"identity":"01a1a2f1-3d6c-4d99-97bd-6b9bb62cc941","order_by":5,"name":"Yannick Estevez","email":"","orcid":"","institution":"ECOFOG, AgroparisTech, Cirad, CNRS, INRAE, Univ. Guyane, Univ. Antilles - Pariacabo - BP 701 97387 Kourou Cedex, Guyane française","correspondingAuthor":false,"prefix":"","firstName":"Yannick","middleName":"","lastName":"Estevez","suffix":""},{"id":601284680,"identity":"cfe0d7fa-4fc4-4f2a-9943-6a71867c6190","order_by":6,"name":"Marc Ducousso","email":"","orcid":"","institution":"AGAP Institut, Univ Montpellier, CIRAD, INRAE, IRD, SupAgro, Montpellier, France","correspondingAuthor":false,"prefix":"","firstName":"Marc","middleName":"","lastName":"Ducousso","suffix":""}],"badges":[],"createdAt":"2026-03-05 13:46:36","currentVersionCode":1,"declarations":{"humanSubjects":false,"vertebrateSubjects":false,"conflictsOfInterestStatement":false,"humanSubjectEthicalGuidelines":false,"humanSubjectConsent":false,"humanSubjectClinicalTrial":false,"humanSubjectCaseReport":false,"vertebrateSubjectEthicalGuidelines":false},"doi":"10.21203/rs.3.rs-9040724/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-9040724/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":104222371,"identity":"a6d1b90b-ed73-454f-98ca-754ed1500c83","added_by":"auto","created_at":"2026-03-09 10:33:32","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":583613,"visible":true,"origin":"","legend":"\u003cp\u003eFig 1: Aerial views of sampling sites in experimental \u003cem\u003eAquilaria crassna\u003c/em\u003eplantations. Farmers A (A, B) and B (C, D) are located in the commune of Regina, while farmer C (E, F) is located in the commune of Cacao, in French Guiana. Previous tree inoculation is shown in G, and a cross-section at the inoculation point revealing blackwood is shown in H. The sampling design is described in I, where two \u003cem\u003eAquilaria crassna \u003c/em\u003ewere sampled in each plot. Black wood and White wood samples were collected from five trunk cross-sections of each tree for a total of 60 samples for chemical and microbial analysis (n = 30 white wood, n = 30 black wood).\u003c/p\u003e","description":"","filename":"Figure1.png","url":"https://assets-eu.researchsquare.com/files/rs-9040724/v1/e4201811726ba3a78684a062.png"},{"id":104779450,"identity":"2e6e60a7-363b-4ae8-96d5-64d61b5ac865","added_by":"auto","created_at":"2026-03-17 07:40:27","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":61422,"visible":true,"origin":"","legend":"\u003cp\u003eFig 2: Soil composition for each farmer plot (A) and the associated pH (B). Similar letters represent no significant differences assessed using Anovafollowed by Tukey HSD post-hoc test. Principal component analysis of the soil composition according to the farmer plot, dimension 1 explain 32% of the variance and dimension 2, 24.4%. Vectors represent elements and are colored according to their contribution (C).\u003c/p\u003e","description":"","filename":"Figure2.png","url":"https://assets-eu.researchsquare.com/files/rs-9040724/v1/6060eed4a69924ae352b0835.png"},{"id":104222377,"identity":"b87f7fdb-79f8-4290-874d-5b451730b05c","added_by":"auto","created_at":"2026-03-09 10:33:32","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":162666,"visible":true,"origin":"","legend":"\u003cp\u003eFig 3: Bacteria (A) and Fungi (B) community composition (genus level) for each sample at each site (Farmer A, B and C), ordered by relative abundance and facetted by wood type (white or black). Each column corresponds to a wood sample, and the size of the bubbles is proportional to the relative abundance of the taxonomic level in line. Chemical composition of the wood volatiles compounds samples is also presented (C). Only the top 25 taxa and chemical compounds in relative abundance are presented.\u003c/p\u003e","description":"","filename":"Figure3.png","url":"https://assets-eu.researchsquare.com/files/rs-9040724/v1/58e43d5a0775bde2abbe05ad.png"},{"id":104404176,"identity":"641007d5-207b-4a85-9c45-bc3eafc112a1","added_by":"auto","created_at":"2026-03-11 12:19:46","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":77004,"visible":true,"origin":"","legend":"\u003cp\u003eFig 4: Redundancy Analysis of Bacteria (A) and Fungi (B). Black arrows represent chemical compounds variables, while red arrows denote microbial genus. Points are color-coded by site, while their shape represent wood type. The explained variance for dimensions 1 and 2 is displayed on each panel as percentages. Explained variance of each factor terms and their interaction on the bacterial and fungal community composition are also presented with their associated significance. nMDS of the chemical compounds based on their Hellinger-transformed relative abundances (C). The associated Permanova is also presented in the table D.\u003c/p\u003e","description":"","filename":"Figure4.png","url":"https://assets-eu.researchsquare.com/files/rs-9040724/v1/a9add204f4ebb1107d40660e.png"},{"id":104808335,"identity":"d4b724e0-96d5-4ba2-85ff-1ec426ed6d46","added_by":"auto","created_at":"2026-03-17 12:36:05","extension":"png","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":60545,"visible":true,"origin":"","legend":"\u003cp\u003eFig 5: Venn diagrams showing the unique and shared Bacteria ASVs, fungal OTUs (counts and sequence percentages), and chemical compounds. Comparisons between whitewood and blackwood for each farmer (A). Comparisons between farmers within each wood type (B).\u003c/p\u003e","description":"","filename":"Figure5.png","url":"https://assets-eu.researchsquare.com/files/rs-9040724/v1/8101ab85da4b39e442d1f8ab.png"},{"id":104405048,"identity":"fd817052-0b1b-429c-be59-9df8a11ea3e5","added_by":"auto","created_at":"2026-03-11 12:21:41","extension":"png","order_by":6,"title":"Figure 6","display":"","copyAsset":false,"role":"figure","size":67478,"visible":true,"origin":"","legend":"\u003cp\u003eFig 6: Ratio of positive to negative significant edges in each chemical compound class based on bacterial (A) and fungal (B) class. A negative edge ratio (-, in red), means more negative to positive edges are found for each microbial class to the relative chemical class, a positive edge ratio (+, in green), means more positive to negative edges are found for each microbial class to the relative chemical class. Equal ratio of positive to negative edges are also reported (=, in grey). Only significant edges ratio are reported.\u003c/p\u003e","description":"","filename":"Figure6.png","url":"https://assets-eu.researchsquare.com/files/rs-9040724/v1/b6c2f6c2aa2e4646e67ecc6a.png"},{"id":104809060,"identity":"ad6e54a6-0c4c-48b3-9585-e00c9aa5ff3b","added_by":"auto","created_at":"2026-03-17 12:47:05","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":1946239,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-9040724/v1/7424e2bf-94fb-4176-a58d-aab587e187ac.pdf"},{"id":104222374,"identity":"d17f9d6e-c033-41d4-8403-53019d62130d","added_by":"auto","created_at":"2026-03-09 10:33:32","extension":"pdf","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":3383447,"visible":true,"origin":"","legend":"\u003cp\u003eSupplemental Figures\u003c/p\u003e","description":"","filename":"SupplementalFiguresACQ.pdf","url":"https://assets-eu.researchsquare.com/files/rs-9040724/v1/fd610528fcd0493816ec82e0.pdf"}],"financialInterests":"The authors declare no competing interests.","formattedTitle":"\u003cp\u003e\u003cstrong\u003eThe microbial and chemical terroir of agarwood in French Guiana\u003c/strong\u003e\u003c/p\u003e","fulltext":[{"header":"1. Introduction","content":"\u003cp\u003eAgarwood is a resinous wood formed following infection or injury in trees of the family Thymelaeaceae, which originate from the Indomalesian region. Beyond its medicinal uses, agarwood has been burned as incense since antiquity in Asia and the Middle East, where it is regarded as one of the most precious natural fragrant materials. Agarwood is also distilled to produce agarwood oil, a highly valued ingredient in fine perfumery. The principal agarwood‑producing genera, \u003cem\u003eAquilaria\u003c/em\u003e and \u003cem\u003eGyrinops\u003c/em\u003e, are currently listed on the IUCN Red List due to excessive exploitation of wild populations (Lee and Mohamed \u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e2016\u003c/span\u003e). The planting of these species for industrial agarwood production may alleviate pressure on wild populations, and most current trade now relies on planted trees inoculated with formulations often kept secret that enabled agarwood formation. Although the agarwood formation process in \u003cem\u003eAquilaria\u003c/em\u003e trees is not yet fully understood, it is widely accepted that resin production is triggered by natural or deliberate wounding followed by colonization by environmental microorganisms (Tan et al. \u003cspan citationid=\"CR40\" class=\"CitationRef\"\u003e2019\u003c/span\u003e; Faizal et al. \u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e2022\u003c/span\u003e). Among these factors, stress plays a central role, as \u003cem\u003eAquilaria\u003c/em\u003e trees produce and accumulate secondary metabolites in response to injury (Rasool and Mohamed \u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e2016\u003c/span\u003e; Naziz et al. \u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e2019\u003c/span\u003e). To produce agarwood in managed plantation, induction techniques imply either mechanical wounding of the trunk, the use of chemical inducers (Zhang et al. \u003cspan citationid=\"CR48\" class=\"CitationRef\"\u003e2012\u003c/span\u003e) but also microbial-based techniques using fungal inoculum (Tan et al. \u003cspan citationid=\"CR40\" class=\"CitationRef\"\u003e2019\u003c/span\u003e; Ngadiran et al. \u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e2023\u003c/span\u003e). These latter methods are often based on inoculation by \u003cem\u003eAquilaria\u003c/em\u003e endophytic or soil fungi, promoting a more natural and environmentally friendly solution to improve productivity and quality. Agarwood microbiome is known to differ from that of non-infected wood, and has been reported to contain Actinomycetes such as \u003cem\u003eNocardia\u003c/em\u003e and \u003cem\u003eStreptomyces\u003c/em\u003e, and pathogenic fungi such as \u003cem\u003eFusarium\u003c/em\u003e and \u003cem\u003eTrichoderma\u003c/em\u003e (Mohamed et al. \u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e2010\u003c/span\u003e; Islam et al. \u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e2022\u003c/span\u003e). Fungal inoculum thus often originates from the specific microbiome of infected \u003cem\u003eAquilaria\u003c/em\u003e trees or from the surrounding soil. However, choosing the appropriate microbial strains to induce a stress response of the tree enabling to produce the molecules of interest to obtain high quality agarwood (Fu et al. \u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). Moreover, while the current knowledge on agarwood formation and deliberate induction are progressing (Turjaman et al. \u003cspan citationid=\"CR44\" class=\"CitationRef\"\u003e2016\u003c/span\u003e), major challenges still remain concerning the quality of the resulting agarwood.\u003c/p\u003e \u003cp\u003eThe chemical composition of agarwood has been studied extensively and is characterized by the occurrence of two main families of specialized metabolites: sesquiterpenoids and 2-phenylethylchromone derivatives (Li et al. \u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e2021\u003c/span\u003e). Both of these groups contribute to the olfactory properties of agarwood: sesquiterpenoids are generally sufficiently volatile to be intrinsically fragrant, and some of them are well known strong odorants (Baldovini 2022). In contrast, 2‑phenylethylchromones have higher molecular weights and therefore much lower volatility; although essentially odorless, they act as precursors of lighter aromatic compounds produced during thermal degradation when agarwood is burned. Nevertheless, the identity of the most important olfactory contributors remains debated, and no clear relationship has yet been established between odor quality and specific chemical markers. In general, identifying key odorants in aromatic raw materials is challenging due to complex synergistic interactions and the disproportionate influence of trace constituents. It is particularly true in the case of complex mixtures like agarwood oils or extracts containing a very large number of specific constituents with high molecular complexity, and also showing an important variability of composition from one origin to the other. The chemical complexity of agarwood is partly attributable to the contribution of bacteria and fungi during resin development. Microorganisms are well known to influence the flavor profile, nutritional chemistry, and sensory properties of numerous biological products. A well-known example is the grapevine associated soil and endophytic microbiota, which contribute to the taste and quality of wine (Knight et al. \u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e2015a\u003c/span\u003e; Belda et al. \u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e2017\u003c/span\u003e). The combined influence of biotic and abiotic factors shaping product flavor and quality is increasingly described by the concept of \u0026ldquo;microbial terroir\u0026rdquo; (Gilbert et al. \u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e2014\u003c/span\u003e; Nakano \u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). This concept, mainly referring to wine grapes, pertains to the geographic microbial signature of soil, rhizosphere and endophytic fungi in grapes and vineyards (Miura et al. \u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e2017a\u003c/span\u003e) on the fermentation process (Bokulich et al. \u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e2016\u003c/span\u003e; Mas and Portillo \u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e2022\u003c/span\u003e) and ultimately on the wine taste. The microbial terroir, beyond wine grapes have also been found to influence the flavor chemistry of mustard seeds, where rhizosphere taxa predicted glucosinolate content, responsible for the desired spicy and bitter flavor of mustard seeds (Walsh et al. \u003cspan citationid=\"CR46\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). In addition to abiotic parameters such as climate, pH, nutrient content and management practices, local microorganisms also have a role on the quality and specificity of food products. While endophytic microbial communities of \u003cem\u003eAquilaria\u003c/em\u003e may have an influence in the process of agarwood formation, their influence on the resulting agarwood chemical quality and signature is yet to be determined, particularly in the concept of microbial terroir. This is particularly important as wood chips are often macerated in water before hydro-distillation to improve oil extraction yield and quality. Therefore, microorganisms are finally involved in two key steps of the process; infection of living trees, and fermentation of wood chips before distillation. Indeed, microbial diversity during maceration (Islam et al. \u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e2022\u003c/span\u003e) and microbial consortia (Naziz et al. \u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e2024\u003c/span\u003e) were found to improve agarwood oil quality. Local endophytic microbial communities of agarwood may thus have an influence on the maceration process, and could also be targeted for improving the resulting agarwood oil quality.\u003c/p\u003e \u003cp\u003eIn French Guiana, the \u003cem\u003eAquilaria\u003c/em\u003e tree is considered a cash crop by farmers in the communities of the villages of Regina and Cacao. These farmers, originally from Laos and with a long agricultural tradition (Vang \u003cspan citationid=\"CR45\" class=\"CitationRef\"\u003e2018\u003c/span\u003e) possess traditional knowledge and expertise in cultivating the \u003cem\u003eAquilaria\u003c/em\u003e tree. In French Guiana, they are working to establish a sustainable \u003cem\u003eAquilaria\u003c/em\u003e production sector. To support high quality agarwood production, it is necessary to take into account local fungal and bacterial communities, as well as to estimate the quality of \u003cem\u003eAquilaria\u003c/em\u003e wood that has been planted outside its natural range. In this study, we aimed at elucidating the microbiome and chemical composition of agarwood in French Guiana, which are expected to differ from those in Aquilaria's geographical area of origin in Asia. Here, we investigated the association between fungi, bacteria, and chemical constituents to identify microbial taxa potentially associated with agarwood scent. Finally, we assessed the effect of geographic location on agarwood quality to test the microbial terroir hypothesis, which we further extend to a broader concept of \u0026ldquo;chemical terroir.\u0026rdquo;\u003c/p\u003e"},{"header":"2. Material and methods","content":"\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e \u003ch2\u003e2.1. Study site\u003c/h2\u003e \u003cp\u003eSampling was conducted in two agricultural plots located in the Regina (Farmer A: 4\u0026deg;17'42.3\"N 52\u0026deg;10'55.2\"W; Farmer B: 4\u0026deg;17'41.8\"N 52\u0026deg;10'54.6\"W) and one in Cacao (Farmer C: 4\u0026deg;32'22.4\"N 52\u0026deg;28'38.7\"W) villages in French Guiana. At each of the three sites, two \u003cem\u003eAquilaria crassna\u003c/em\u003e trees previously inoculated in November 2022 with three fungal strains (\u003cem\u003eAntrodia vaillantii\u003c/em\u003e (DC. Ryvarden), \u003cem\u003ePycnoporus sanguineus\u003c/em\u003e (L.) Murril, and \u003cem\u003eCoriolopsis polyzona\u003c/em\u003e (Pers.) Ryvarden) were selected. The detail of sampling sites, inoculated tree, and blackwood infection is presented in Fig.\u0026nbsp;1.A-H. The trunk was sectioned and five cross-sections were made at the inoculation point on each trunk. Two samples (one blackwood and one whitewood) were collected from each cross-section for a total of 60 samples (Fig.\u0026nbsp;1.I).\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec4\" class=\"Section2\"\u003e \u003ch2\u003e2.2. Soil sampling and analysis\u003c/h2\u003e \u003cp\u003eFive soil samples were sampled at 20 cm deep at each sampling site, for a total of 15 soil samples. To measure the elemental composition of each soil, samples were processed in triplicate using X-ray fluorescence spectroscopy (XRF). For each sample, triplicates were pressed at a 20 tons pressure for 2 min with 1:3 (v:v) of SpectroBlend\u0026reg; (SCP Science). Three technical measures on each triplicate, using the \u0026lsquo;geoexploration\u0026rsquo; calibration with a total exposure time of 105 s on the XRF S1 Titan analyzer (Bruker) was used to measure the soil atomic composition of elements from magnesium to uranium, leading to nine measurements for each soil sample. The soil pH was measured at a 1:5 (v:v) ratio of H\u003csub\u003e2\u003c/sub\u003eO using a pH meter (pH Meter Knick 766, Knick International). Soil composition was visualized using composition barplots for the main elements (Al, Ca, Co, Cr, Cu, Fe, K, Mg, Mn, Si and Ti). Principal component analysis (PCA) using centered log-ratio transformed abundances were used to visualize soil composition by farmer plots.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec5\" class=\"Section2\"\u003e \u003ch2\u003e2.3. Library preparation and sequencing\u003c/h2\u003e \u003cp\u003eWood chips were crushed in liquid nitrogen using a sterile mortar and pestle. Approximatively 100\u0026ndash;150 mg of wood were then used for DNA extraction using the FastDNA Spin Kit for Soil (MP Biomedicals\u0026trade;). DNA concentrations were measured by fluorescence with a Quant-iT\u0026trade; PicoGreen\u0026trade; dsDNA Assay Kit (ThermoFisher Scientific\u0026trade;, Waltham, Massachusetts, USA) using a Tecan Spark \u0026reg; (Tecan Group Ltd., Mannedorf, Switzerland) and normalized to 1 ng.\u0026micro;L\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e before PCR amplification. Primers 479F (5\u0026prime; CAGCMGCYGCNGTAANAC 3\u0026prime;) and R888 (5\u0026prime; CCGYCAATTCMTTTRAGT 3\u0026prime;) were used to amplify the V3-V4 regions of the 16S rRNA gene (Terrat et al. \u003cspan citationid=\"CR42\" class=\"CitationRef\"\u003e2015\u003c/span\u003e). Primers ITS86F (5\u0026prime; GTGAATCATCGAATCTTTGAA 3\u0026prime;) and ITS4 (5\u0026prime; TCCTCCGCTTATTGATATGC 3\u0026prime;) were used to amplify the fungal ITS2 (Op De Beeck et al. \u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e2014\u003c/span\u003e). Reactions were performed in 20 \u0026micro;L and consisted of 10 \u0026micro;L of Buffer Master mix containing Thermo Scientific Phusion\u0026trade; HighFidelity DNA Polymerase (ThermoFisher Scientific\u0026trade;), 0.5 \u0026micro;L of DMSO, 4.5 \u0026micro;L of DNAse free water, 3 \u0026micro;L of DNA (0.3 ng. \u0026micro;L-1), 1 \u0026micro;L of forward primer and 1 \u0026micro;L of reverse primer tagged with a short nucleotide sequence. The PCR conditions used were identical for the fungal and bacterial amplifications and were performed in a Veriti 96-well Thermal Cycler (ThermoFisher Scientific\u0026trade;) under the following conditions: 10 min initial denaturation, 35 cycles each including denaturation at 94\u0026deg;C for 10 s, annealing at 55\u0026deg;C for 20 s, and extension at 72\u0026deg;C for 20 s, with a final polymerization extension at 72\u0026deg;C for 7 min. All reactions were performed in triplicate and then pooled before deposition on 2% agarose gel, with the corresponding negative control, for amplification quality control. Amplification products of ITS2 region were purified using Agencourt AMPure XP beads (Beckman Coulter Inc., Fullerton, California, USA) and performed on a magnetic rack with a 1:1 (v:v) ratio of AMPure XP per PCR product. The magnetic beads were then rinsed twice with 70% ethanol before final elution in 70 \u0026micro;L of Qiagen EB buffer. For 16S purification, an additional PCR product migration step on 2% agarose gel (80 V, 90 min) was performed to separate and recover the band specific to bacterial 16S (lower band, around 400 bp) and eliminate plant chloroplastic 16S DNA (higher band, \u0026gt; 600 bp, co-amplified during PCR). The recovered bands were purified using the QIAquick gel purification kit (Qiagen, Hilden, Germany). The purified PCR product concentration was then measured with PicoGreen and one equimolar pool (50 ng/sample) for each marker gene (16S and ITS2) was prepared. Each pool was finally purified twice using Agencourt AMPure XP beads, quantified and finally eluted in 40 \u0026micro;L of EB buffer (Qiagen). Library sequencing was performed on a 2 \u0026times; 250 bp MiSeq system (Fasteris SA, Switzerland).\u003c/p\u003e \u003cp\u003eFor taxonomic assignment of fungi, OTUs were constructed on a 97% ITS clustering, while ASVs were used for bacteria (16S). We used a pipeline based on VSEARCH 2.18.0 (Rognes et al. \u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e2016\u003c/span\u003e) and available in GitHub (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://github.com/BPerezLamarque/Scripts/\u003c/span\u003e\u003cspan address=\"https://github.com/BPerezLamarque/Scripts/\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e) for data processing. Briefly, only paired-end reads were retained and then merged using the fastq_mergepairs function with default parameters. We applied a quality check and removed merged reads with more than two alignment errors. Then, filtered merged reads were assigned to the respective samples based on tagged primers (demultiplexing) with 0 accepted error regarding the primers and tag sequences using Cutadapt 3.5. We removed chimeras de novo by using the VSEARCH uchime3_denovo option and assigned the taxonomy to each amplicon sequence variant (ASV) and Operational Taxonomic Unit (OTU) using the usearch_global option with default parameters. Taxonomic annotation was performed with SILVA (v.138.2) database for bacteria (16S rRNA) and UNITE database (v.9.0) for fungi (ITS). For bacteria we used ASV (16S rRNA) while 97% OTUs were used for fungi (ITS), as recommended by Tedersoo et al., (\u003cspan citationid=\"CR41\" class=\"CitationRef\"\u003e2022\u003c/span\u003e). The ASV and OTU tables were then filtered from contaminants based on comparisons with amplified negative controls using the DECONTAM (prevalence method) R package (Davis et al. \u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e2018\u003c/span\u003e). Only ASVs and OTUs with long sequences (\u0026gt;\u0026thinsp;200 bp), assigned to the bacterial or fungal kingdom, presenting an acceptable abundance (\u0026ge;\u0026thinsp;10 reads in total dataset) and amplified in at least one sample, were kept for subsequent analyses. ASV and taxonomic tables are provided in Supplemental Tables\u0026nbsp;1\u0026ndash;4, with the associated samples metadata in Supplemental table 5.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec6\" class=\"Section2\"\u003e \u003ch2\u003e2.4. Characterization of volatile constituents in wood samples: sample preparation\u003c/h2\u003e \u003cp\u003eTo determine the chemical composition of the volatile part of the samples used for microbiome analysis, we performed their hydrodistillation using a modified glass made Clevenger apparatus (Modverre, Grasse, France, Supplemental Fig.\u0026nbsp;1). This sample preparation technique has the advantage to extract the essential oil, which is the second main end product of agarwood industry after wood chips, and which can be analyzed as such by direct injection in a gas chromatography-mass spectrometry (GC-MS) system. Therefore, the analytical results obtained through this study can give an insight in the composition of agarwood oils that would be obtained by the large-scale industrial distillation of similar wood samples. Even if the obtained extract cannot be considered as an essential oil \u003cem\u003estricto sensu\u003c/em\u003e (Bicchi et al. \u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e2018\u003c/span\u003e), it is nevertheless a good model for analytical investigation on industrial agarwood oil composition. However, in the context of the present work focused on wood amounts lower than 1 g, the classical Clevenger apparatus is not adapted as the low quantity of oil produced (few milligrams) would be lost in the hydrolate. Therefore, we used a modified system specifically designed to isolate such low amounts by trapping the organic volatiles in a solvent (tertbutylmethylether) in a cooled double jacketed collection tube to prevent the evaporation of the solvent during hydrodistillation. After optimization, the following parameters were applied to all hydrodistillations: the powdered wood sample (100\u0026ndash;600 mg) and 40 ml water were placed in a 100 ml round bottomed flask equipped with an ovoid stir bar and connected to the Clevenger system. Two ml of tertbutylmethylether were added in the refrigerated collecting tube previously filled with water. The content of the flask was heated in an oil bath (145\u0026deg;C) under stirring for 19 h. At the end of the distillation, the solvent phase was collected, and concentrated until ca. 0.2\u0026ndash;0.3 ml by flowing a gentle stream of argon, then diluted in \u003cem\u003eca\u003c/em\u003e. 1 ml of dichloromethane, dried by filtration through a short pad of magnesium sulfate in a Pasteur pipette, then concentrated under a flow argon and eventually brought to a volume of \u003cem\u003eca\u003c/em\u003e. 50\u0026ndash;100 \u0026micro;l by adding dichloromethane.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec7\" class=\"Section2\"\u003e \u003ch2\u003e2.5. Characterization of volatile constituents in wood samples: GC-MS/FID analyses\u003c/h2\u003e \u003cp\u003eGas chromatography - Mass spectrometry/Flame Ionisation Detector (GC-MS/FID) analyses were carried out using an Agilent 6890N gas chromatograph equipped a fused silica capillary column DB-1MS (30 m \u0026times; 0.25 mm i.d., film thickness: 0.25 \u0026micro;m, J\u0026amp;W122-0132). The analytical parameters were the following: the carrier gas was helium at a flow rate of 2 mL.min\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e, the oven temperature was programmed from 60 to 280\u0026deg;C at 3\u0026deg;C/min and completed by a post run (300\u0026deg;C, 5 min). Five \u0026micro;l samples of the concentrated extract prepared as described above were injected in splitless mode and the injector temperature was 250\u0026deg;C. The column flow was split using a G3184-60065 splitter (Agilent), and transferred to a FID detector (250\u0026deg;C, hydrogen and air flows at 40 and 450 ml/min, respectively) and to an Agilent 5973N mass selective detector working in electron impact (EI) mode at 70 eV (scanning over 40\u0026ndash;450 amu range in SCAN mode). The temperatures of the MS ion source and transfer line were 230 and 280\u0026deg;C, respectively. Linear retention indices (LRI) were determined from the retention times of a series of \u003cem\u003en\u003c/em\u003e-alkanes with linear interpolation. The main constituents were identified by comparison of their mass spectra and LRI with those of pure compounds registered in commercial libraries and literature data, and with a laboratory-made database built from authentic compounds, with the help of SearchReview software (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://waleson.eu/\u003c/span\u003e\u003cspan address=\"https://waleson.eu/\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e). To compare the chemical compositions of the Guianese samples with those of commercial agarwood, samples of commercial blackwood from the Riyadh and AlUla (Saudi Arabia) street market were hydrodistilled and analyzed using the same protocol. The chemical abundance and compounds assignments are provided in tables 6, 7.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec8\" class=\"Section2\"\u003e \u003ch2\u003e2.6. Microbial and chemical diversity\u003c/h2\u003e \u003cp\u003eTo assess if sequencing depth was sufficient to retrieve microbial abundance, rarefaction curves were performed (Supplemental Fig.\u0026nbsp;2) using the mirlyn package (Cameron et al. \u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e2021\u003c/span\u003e). Relative abundances of bacterial and fungal community and chemical compositions were then visualized using bubble plots at the genus level according to the wood type and farmer plots. To assess the differences in microbial diversity in black and white wood samples and as a function of sampling sites, Hill diversity index for three orders of diversity (q\u0026thinsp;=\u0026thinsp;0, 1, and 2) were calculated and the significances of their differences assessed using Anova followed by Tukey HSD tests.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec9\" class=\"Section2\"\u003e \u003ch2\u003e2.7. Microbial and chemical composition\u003c/h2\u003e \u003cp\u003eTo assess the effect of geographical location (farmer plots) on the beta-diversity of microbial composition, and its interaction with the chemistry of wood samples, Hellinger-transformed microbial abundances and clr-transformed chemical abundance were used to perform a Redundancy analysis (RDA). The associated R\u003csup\u003e2\u003c/sup\u003e and its significance were assessed with a Permanova using the Bray-Curtis distance of the Hellinger-transformed microbial abundances using the formula: microbial abundance~Wood type*Farmer. For chemical compounds, beta diversity was visualized using a nMDS with its associated Permanova. To compare the composition of our samples with those of commercial agarwoods, a nMDS of the chemical compositions based on Guianese or foreign origin was also carried out.\u003c/p\u003e \u003cp\u003eTo identify differentially abundant microbial taxa and chemical compounds between white and black wood, we used DESeq2 (Love et al. \u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e2014\u003c/span\u003e) on both datasets. The dispersion among individual counts was estimated by the DESeq2 algorithm, and the resulting logarithmic fold changes (Log2FC) with the associated p-values were used to identify differentially abundant features. Analyses were performed separately for the microbial and chemical dataset.\u003c/p\u003e \u003cp\u003eTo quantify within-group heterogeneity as distance-to-centroid on Aitchison (CLR) distances and compare black \u003cem\u003evs\u003c/em\u003e white woods, betadisper and permutest were used. Then, to relate chemistry to microbiology, linear models of microbiological dispersion (ITS or 16S) on chemical dispersion with wood type and their interaction were fitted; z-scored predictors were also added to interpret group differences at mean chemistry distances. Within-group monotonic concordance was assessed \u003cem\u003evia\u003c/em\u003e Spearman ρ, overall sample-to-sample correlations were tested on raw dispersion values, and global multivariate structures between datasets were compared with Mantel tests and Procrustes analyses on cmdscale ordinations.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec10\" class=\"Section2\"\u003e \u003ch2\u003e2.8. Co-occurrence network analysis of microbial-chemistry\u003c/h2\u003e \u003cp\u003eIn order to accurately assess the correlations between microbial taxa and volatile constituents, network analyses based on covariance relationships were carried out. All the constituents were classified according to their putative biosynthetic origin to facilitate the statistical treatment of the chemical information. Among all the volatile constituents, sesquiterpenoids were easily recognized and associated to seven subgroups defined by their sesquiterpenic skeleton (bisabolane, eremophillane, eudesmane, guaiane, humulane, vetispirane, zizaane). Not surprisingly in view of the classical composition of agarwood oils, many aromatic compounds were also identified, and were classified either as \u0026ldquo;chromone derivatives\u0026rdquo; or \u0026ldquo;possible chromone derivatives\u0026rdquo;, depending on the degree of confidence we had on their suspected biogenetic origin. In addition to these groups, the other families were \u0026ldquo;linear / fatty acid derivatives\u0026rdquo;, \u0026ldquo;adulterants\u0026rdquo; and \u0026ldquo;possible pollutants\u0026rdquo; (to distinguish respectively suspected voluntarily and unvoluntary added compounds, mostly in commercial samples). Moreover, many minor compounds were not present in our database and could not be identified with certainty, therefore they were classified as \u0026ldquo;unidentified\u0026rdquo;. Considering microbial and chemical datasets as compositional and the paired samples design of our study, we used the cross-kingdom network inference using SPIEC-EASI (Kurtz et al. \u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e2015\u003c/span\u003e; Tipton et al. \u003cspan citationid=\"CR43\" class=\"CitationRef\"\u003e2018\u003c/span\u003e), which is robust to the bias of using two compositional datasets (Brunner et al. \u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). Multi domain SPIEC-EASI was thus used to build co-occurrence networks for agarwood samples (black wood) using the Meinshausen-Buhlmann's neighborhood selection method using 100 repetitions and 100 subsamples for the Stability Approach to Regularization Selection (StARS) algorithm (Liu et al. \u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e2014\u003c/span\u003e). The optimal coefficient matrix and symmetric matrix with edge-wise stability was then retrieved after verification of the optimal network stability at the selected sparsity. Sub-networks according to each chemical class were then created using the original ITS-chemistry and 16S-chemistry cross-kingdom networks. To do so, for each chemical class, we retrieved chemical nodes (named hereafter \u0026lsquo;core nodes\u0026rsquo;) belonging to each chemical class as well as their connections to other chemical compounds and microbial taxa. The internal microbial edges were removed (bacteria-bacteria or fungi-fungi edges) to specifically assess the interaction of microbial taxa and chemical compounds. Then, we calculated the number of positive and negative edges, as well as their taxonomic assignation (chemical or microbial class). Finally, to identify the significant covariance relationships between each microbial class and genus to each chemical class, the ratio of positive to negative edges were calculated for each sub-network specific of each chemical class. Its value (positive, negative or neutral) is an indication of the overall link between microbial taxa and chemical compounds found in agarwood.\u003c/p\u003e \u003c/div\u003e"},{"header":"3. Results","content":"\u003cdiv id=\"Sec12\" class=\"Section2\"\u003e \u003ch2\u003e3.1. A ferralitic tropical soil\u003c/h2\u003e \u003cp\u003eAll plantations were established on ferralitic tropical soils characterized by high iron concentrations (50.5\u0026thinsp;\u0026plusmn;\u0026thinsp;6.1%; 41.2\u0026thinsp;\u0026plusmn;\u0026thinsp;7.6%; 60.9\u0026thinsp;\u0026plusmn;\u0026thinsp;4.3% for farmer A, B and C respectively), followed by Si (20.2\u0026thinsp;\u0026plusmn;\u0026thinsp;5.0%; 27.4\u0026thinsp;\u0026plusmn;\u0026thinsp;6.9%; 13.7\u0026thinsp;\u0026plusmn;\u0026thinsp;2.7%), Al (16.1\u0026thinsp;\u0026plusmn;\u0026thinsp;4.1%; 16.4\u0026thinsp;\u0026plusmn;\u0026thinsp;3.6%; 14.9\u0026thinsp;\u0026plusmn;\u0026thinsp;3.0%), Mg (6.9\u0026thinsp;\u0026plusmn;\u0026thinsp;4.0%; 6.6\u0026thinsp;\u0026plusmn;\u0026thinsp;3.6%; 5.1\u0026thinsp;\u0026plusmn;\u0026thinsp;1.9%) and Ti (5.5\u0026thinsp;\u0026plusmn;\u0026thinsp;0.4%; 7.0\u0026thinsp;\u0026plusmn;\u0026thinsp;1.0%; 4.7\u0026thinsp;\u0026plusmn;\u0026thinsp;0.2% for farmer A, B and C respectively). Soil pH did not differ significantly among plantations (mean pH\u0026thinsp;=\u0026thinsp;5.3\u0026thinsp;\u0026plusmn;\u0026thinsp;0.1, p.value\u0026thinsp;=\u0026thinsp;0.0786). Relative abundances of each element for farmer plot are available in Supplemental table 8. The soil from farmer C plantations had a higher Fe, Cr and Mn content, and a lower K and Si content compared to the other plantations (Supplemental Fig.\u0026nbsp;3), resulting in a contrasted soil composition compared to the other plantations (Fig.\u0026nbsp;2.C).\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec13\" class=\"Section2\"\u003e \u003ch2\u003e3.2. Agarwood microbial and chemical composition\u003c/h2\u003e \u003cp\u003e \u003cem\u003eEnterobacter\u003c/em\u003e, \u003cem\u003eKluyvera\u003c/em\u003e and \u003cem\u003ePantoea\u003c/em\u003e were the main bacterial genera found in white wood and this assembly did not vary significantly according to sampling sites (Fig.\u0026nbsp;3.A). \u003cem\u003eDiplodia\u003c/em\u003e, \u003cem\u003eNigrospora\u003c/em\u003e, \u003cem\u003ePestalotiopsis\u003c/em\u003e, \u003cem\u003eGongronella\u003c/em\u003e, \u003cem\u003eFusarium\u003c/em\u003e, \u003cem\u003eTrichoderma\u003c/em\u003e and \u003cem\u003eCurvularia\u003c/em\u003e were the main fungal genera found in white wood (Fig.\u0026nbsp;3.B). High concentration of fatty acids (hexadecanoic, oleic, octadecanoic and linoleic acids) were found in white wood samples, and their abundance did not significantly vary between samples nor sampling sites (Fig.\u0026nbsp;3.C). On the other hand, \u003cem\u003ePantoea\u003c/em\u003e, an unknown genus, \u003cem\u003eAcidibacter\u003c/em\u003e, \u003cem\u003eBurkholderia\u003c/em\u003e and \u003cem\u003eEnrotobacter\u003c/em\u003e were the main bacterial genera found in black wood (Fig.\u0026nbsp;3.A) and their relative abundance significantly vary between samples, and to a higher extent, between sampling sites. \u003cem\u003eTrechispora\u003c/em\u003e, \u003cem\u003eDiaporthe\u003c/em\u003e, \u003cem\u003eFusarium\u003c/em\u003e, \u003cem\u003eOliveonia\u003c/em\u003e, \u003cem\u003eCalosphaeria\u003c/em\u003e and \u003cem\u003eStriaticonidium\u003c/em\u003e were the main fungal genera (Fig.\u0026nbsp;3.B), their abundance also significantly varies according to samples and sampling sites. Black wood volatile constituents were generally more abundant, their number was also higher and their chemical structures were more diverse. Even if the fatty acids abundant in white wood were still present in black wood, we also found high amounts of typical agarwood compounds (Naef \u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e2011\u003c/span\u003e) such as oxo-agarospirol, 4-methoxyphenyl-butan-2-one, β-agarofurane, jinkoh eremol, agarospirol and dihydrokaranone (Fig.\u0026nbsp;3.C). These microbial and chemical patterns were confirmed by the differential abundance analysis, where a high share of \u003cem\u003ePseudomonadota\u003c/em\u003e bacteria was found specific to the white wood, while a higher diversity of bacteria, belonging to diverse phyla was specific to the black wood (Supplemental Fig.\u0026nbsp;3.A; Supplemental Table\u0026nbsp;9). This was also the case for fungi, where all taxa differentially abundant in white wood belonged to Ascomycota phyla, such as \u003cem\u003eCurvularia\u003c/em\u003e, \u003cem\u003eNigrospora\u003c/em\u003e or \u003cem\u003eChaetomium\u003c/em\u003e. In the black wood, both Ascomycota, Oomycota and Basidiomycota were differentially abundant. Interestingly, a high share of unidentified fungal species (Incertidae Sedis\u0026thinsp;=\u0026thinsp;IS) was found specific to the black wood (Supplemental Fig.\u0026nbsp;3.B). Linear and fatty acids were found highly specific to the white wood, while a more diverse chemical diversity was specific to the black wood (Supplemental Fig.\u0026nbsp;3.C). We found a specific signature of French Guiana agarwood samples when compared to foreign samples from South-East Asia (Figure S4.A). Chromone derivatives, possible chromone derivatives, and sesquiterpenoids were more abundant in Guiana agarwood samples (Supplemental Fig.\u0026nbsp;4.B). Furthermore, adulterants such as di-n-butyl phthalate were found specific to foreign samples, testifying to their poor quality.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec14\" class=\"Section2\"\u003e \u003ch2\u003e3.3. Inter- and Intra-sample variability of the microbial and chemical composition\u003c/h2\u003e \u003cp\u003eOverall, dispersion varied between black and white woods for chemistry and microbial communities (Supplemental Fig.\u0026nbsp;5.A). Chemical composition dispersion was a strong positive predictor of microbiological communities\u0026rsquo; dispersion (ITS: F₁,₄₉ = 21.18, p\u0026thinsp;=\u0026thinsp;\u0026lt;\u0026thinsp;0.001; 16S: F₁,₄₉ = 206.49, p\u0026thinsp;\u0026lt;\u0026thinsp;0.001), with an effect of wood type on bacterial dispersion (16S: F₁,₄₉ = 107.58, p\u0026thinsp;\u0026lt;\u0026thinsp;0.001), but no interaction between chemical dispersion and wood type. In standardized models, the chemistry slopes remained significantly positive (ITS β\u0026thinsp;=\u0026thinsp;0.410, p\u0026thinsp;=\u0026thinsp;0.043; 16S β\u0026thinsp;=\u0026thinsp;0.195, p\u0026thinsp;=\u0026thinsp;0.034), and white wood showed lower dispersion than black at average chemistry (ITS β = \u0026minus;0.769, p\u0026thinsp;=\u0026thinsp;0.035; 16S β = \u0026minus;1.578, p\u0026thinsp;=\u0026thinsp;3.01e\u0026thinsp;\u0026minus;\u0026thinsp;10; Supplemental Fig.\u0026nbsp;5.B). Model fit was moderate for ITS (Adj. R\u0026sup2; = 0.31) and high for 16S (Adj. R\u0026sup2; = 0.86).\u003c/p\u003e \u003cp\u003eOverall sample-to-sample correlations between chemical and microbiological dispersion were moderate for fungi (Spearman ρ\u0026thinsp;=\u0026thinsp;0.434, p\u0026thinsp;=\u0026thinsp;0.001) and strong for bacteria (ρ\u0026thinsp;=\u0026thinsp;0.746, p\u0026thinsp;\u0026lt;\u0026thinsp;2.2e\u0026thinsp;\u0026minus;\u0026thinsp;16). Mantel tests confirmed significant correspondence between chemical and microbiological distance matrices (chemistry\u0026ndash;ITS: r\u0026thinsp;=\u0026thinsp;0.418, p\u0026thinsp;=\u0026thinsp;1e\u0026thinsp;\u0026minus;\u0026thinsp;04; chemistry \u0026minus;\u0026thinsp;16S: r\u0026thinsp;=\u0026thinsp;0.390, p\u0026thinsp;=\u0026thinsp;1e\u0026thinsp;\u0026minus;\u0026thinsp;04). Ordination structures aligned well between chemistry and microbiology (Procrustes correlations: chem\u0026ndash;ITS r\u0026thinsp;=\u0026thinsp;0.677, p\u0026thinsp;=\u0026thinsp;1e\u0026thinsp;\u0026minus;\u0026thinsp;04; chem\u0026ndash;16S r\u0026thinsp;=\u0026thinsp;0.720, p\u0026thinsp;=\u0026thinsp;1e\u0026thinsp;\u0026minus;\u0026thinsp;04), supporting substantial shared multivariate patterns (Supplemental Fig.\u0026nbsp;5.C). Chemical composition strongly structures both bacterial and fungal communities, with bacteria exhibiting more deterministic, chemistry-driven assembly than fungi.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec15\" class=\"Section2\"\u003e \u003ch2\u003e3.4. The terroir effect on microbial and chemical composition\u003c/h2\u003e \u003cp\u003eConcerning the overall community structure, wood endophytic bacterial beta diversity varied strongly according to the wood type (Fig.\u0026nbsp;4.A; R\u003csup\u003e2\u003c/sup\u003e\u0026thinsp;=\u0026thinsp;0.36), while farming plot had a limited effect (R\u003csup\u003e2\u003c/sup\u003e\u0026thinsp;=\u0026thinsp;0.08). This pattern was similar for fungi, where wood type had the major effect on beta diversity (Fig.\u0026nbsp;4.B; R\u003csup\u003e2\u003c/sup\u003e\u0026thinsp;=\u0026thinsp;0.30), while farming plot had a limited effect (R\u003csup\u003e2\u003c/sup\u003e\u0026thinsp;=\u0026thinsp;0.08). Redundancy analysis (Fig.\u0026nbsp;4.C) also explained a higher shared of bacterial community composition (Dim 1\u0026thinsp;=\u0026thinsp;47.3% and Dim 2\u0026thinsp;=\u0026thinsp;12%) than for fungi (Dim 1\u0026thinsp;=\u0026thinsp;17.9% and Dim 2\u0026thinsp;=\u0026thinsp;9.8%).\u003c/p\u003e \u003cp\u003eConcerning the specificity of sequences and taxa found in \u003cem\u003eAquilaria\u003c/em\u003e wood, a high percentage of sequences were shared between black and white wood samples for all farmer plots (Fig.\u0026nbsp;5.A), ranging from 84.4 to 90.2% of shared sequences. A higher share of unique sequences was found in black wood compared to white wood for both fungi and bacteria, where 9 to 15% of sequences were unique to black wood, while only 0.5 to 1% of sequences were unique to white wood. Conversely, a high number of taxa were found specific to black wood samples, where up to 955 bacterial taxa (out of 1548) were specific to black wood samples for the farmer C plot. This high diversity and high specificity of bacterial taxa in black wood samples was conserved for all plots. For fungi, the share of black wood specific taxa was similar to that of shared taxa. Chemical compounds also had a high proportion of common compounds between black and white woods, but a high number and proportion was specific to black wood (40 to 52 compounds). Concerning the specificity of microbial structure based on sampling sites, while a high share of sequences was found in samples from all locations (98.0% and 81.2% for bacteria and fungi of the white wood; 78.4% and 87.9% for the black wood), a high number of taxa were not found in all plots (Fig.\u0026nbsp;5.B). For example, the farming plot C had a high number of specific taxa, both in white (113 out of 623 bacterial ASV) and black wood (360 out of 1 730 bacterial ASV). However, between the three sampling sites, the chemistry of white and black wood remained highly consistent (99.8 and 97.3% of share compounds respectively).\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec16\" class=\"Section2\"\u003e \u003ch2\u003e3.5. Association between microbial members and chemical composition\u003c/h2\u003e \u003cp\u003eThe resulting sub-networks of chemical class and microbial composition revealed that each chemical class had specific relationships with different microbial taxa. Linear fatty acid derivatives had the highest connections with bacterial taxa (Supplemental Figure S6, Supplemental table 10); core degree\u0026thinsp;=\u0026thinsp;68.4), followed by possible chromone derivative (core degree\u0026thinsp;=\u0026thinsp;66.2), while sesquiterpenoid humulane had the lowest (core degree\u0026thinsp;=\u0026thinsp;38). The pattern was similar for fungi, where possible chromone derivative had the highest connection with fungal taxa (Supplemental Figure S7; core degree\u0026thinsp;=\u0026thinsp;15.2), followed by linear fatty acid derivative (core degree\u0026thinsp;=\u0026thinsp;13.8), and sesquiterpenoid humulane having the lowest (core degree\u0026thinsp;=\u0026thinsp;9). The most connected bacterial nodes to chemical compositions belonged to the Alphaproteobacteria, Gammaproteobacteria, Actinobacteria and Planctomycetes bacterial class (Supplemental Fig.\u0026nbsp;7.A; Supplemental table 11). The most connected fungal nodes to chemical compositions belonged to the fungal class Sordariomycetes, Eurotiomycetes, and Dothideomycetes (Supplemental Fig.\u0026nbsp;7.B). Among these connected microbial nodes to chemical compounds, we were able to identify the significantly positively and negatively associated bacterial (Fig.\u0026nbsp;6.A) and fungal class (Fig.\u0026nbsp;6.B) and genus (Supplemental Fig.\u0026nbsp;9) to each chemical class. Some microbial class exhibited consistent negative relationships with chemical compounds such as members of the Actinobacteria and Vicinamibacteria class, while others had a positive relationship such as Phycisphareae, Spirochaetia, Mucoromycetes and Umbelopsidomycetes. When focusing on chromone derivative, we found 10 bacterial and 5 fungal class positively associated with increased chromone derivative proportion, and 9 bacterial and 2 fungal class negatively associated. For example, Dothideomycetes and Sordariomycetes were negatively associated with chromone and possible chromone derivative proportion in the black wood.\u003c/p\u003e \u003c/div\u003e"},{"header":"4. Discussion","content":"\u003cdiv id=\"Sec18\" class=\"Section2\"\u003e \u003ch2\u003e4.1 French Guiana, a suitable terroir for quality agarwood\u003c/h2\u003e \u003cp\u003eEndophytic microbial communities of French Guiana \u003cem\u003eAquilaria\u003c/em\u003e resemble those of South-East Asia, where \u003cem\u003eBacillus\u003c/em\u003e and \u003cem\u003ePantoea\u003c/em\u003e were also reported (Bora et al. \u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e2025\u003c/span\u003e), as well as \u003cem\u003eAspergillus\u003c/em\u003e, \u003cem\u003ePenicillium\u003c/em\u003e or \u003cem\u003eFusarium\u003c/em\u003e (Fu et al. \u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). Numerous saprotrophs and pathotrophs fungi were identified in black wood. For example, \u003cem\u003eTrechispora\u003c/em\u003e, highly abundant in black wood, is a white rot commonly found on decaying wood (P\u0026otilde;lme et al. \u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e2020\u003c/span\u003e). \u003cem\u003eBotryosphaeriaceae\u003c/em\u003e were necrotrophic or saprobic in woody tissues and were also found in high abundance in black wood. \u003cem\u003eCalosphaeriaceae\u003c/em\u003e, \u003cem\u003eOliveonia\u003c/em\u003e, \u003cem\u003eConiochaetaceae\u003c/em\u003e, \u003cem\u003eHelotiaceae\u003c/em\u003e, \u003cem\u003eKylindria\u003c/em\u003e, \u003cem\u003eStypella\u003c/em\u003e were among the most commonly found wood saprobes in black wood. Some fungi may also exhibit a necrotic phase such as the microfungus belonging to \u003cem\u003eGlomerellaceae\u003c/em\u003e family.\u003c/p\u003e \u003cp\u003eAlthough overall microbial community composition was conserved across ampling sites, location‑specific differences were observed. Spatial processes have been show to impact the fungal communities of grapevine, where distance decay relationships of microbial community dissimilarity were found on small distances (Miura et al. \u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e2017b\u003c/span\u003e). In our case, the differences in soil composition, where the plot C had more iron and chromium content, may have impacted microbial communities, and thus, agarwood chemical composition. The specificity of Guianese agarwood is further supported by the strong chemical composition differences we found between local and foreign samples (Supplemental Figure S4). Plant genotype, micro-climatic environment and agricultural practices may also drive microbial differences in soil and wood endophytic communities, as demonstrated in the vineyards (Bokulich et al. \u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e2014\u003c/span\u003e; Knight et al. \u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e2015b\u003c/span\u003e). This is supported by the specific microbial taxa we found between plots, where geographical distance had an impact on soil chemistry and microclimatic conditions and thus, microbial communities. However, the extent to which specific microbial communities drive desirable chemical terroirs and sensory outcomes remains to be determined. While it was not investigated here, the specific microbiome and climate of French Guiana might also influence the maceration process, and further increase the specificity of the resulting essential oil following distillation.\u003c/p\u003e \u003cp\u003eWhile \u003cem\u003eAquilaria\u003c/em\u003e show promises in French Guiana, it originates from South-East Asia, and its implantation should follow strict environmental precautions. \u003cem\u003eAquilaria\u003c/em\u003e spp. can grow from 15 to 25 m in height and have a low number of flowers per inflorescence, obligately outcrossing and the barochoric dispersal of seed do not spread far from the trunk (Soehartono and Newton \u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e2001\u003c/span\u003e). These trait with the lifespan of the seeds of only a few days, may limit \u003cem\u003eAquilaria\u003c/em\u003e invasiveness in French Guiana (Hamilton et al. \u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e2005\u003c/span\u003e) but careful monitoring following its plantation should be implemented. No report of invasiveness has been identified as of today, whether within the natural distribution area of ​​\u003cem\u003eAquilaria\u003c/em\u003e spp. or from plantations within and outside the natural distribution area.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec19\" class=\"Section2\"\u003e \u003ch2\u003e4.2. The interplay of plant and its microbiome for agarwood scent\u003c/h2\u003e \u003cp\u003eIt remains difficult to assess a direct link between chemical compounds and microbial composition. For example, while bacterial and fungal taxa were positively linked to chromone derivative abundance, some biosynthesis pathways exclusively pertain to the plant such as the aromatic ketone benzylacetone or 4-phenylbutan-2-one. This secondary metabolite, present in flowers is an attractant for pollinators (Kessler and Baldwin \u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e2007\u003c/span\u003e) and the acyltransferase capable of synthesizing benzylbenzoate (D\u0026rsquo;Auria et al. \u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e2002\u003c/span\u003e) is only found in \u003cem\u003eMagnolopsida\u003c/em\u003e and \u003cem\u003ePinopsida\u003c/em\u003e.\u003c/p\u003e \u003cp\u003eThe volatile organic compounds belonging to sesquiterpenes are produced by both plant and fungi to attract insect pollinators, defend against pathogens and modulate microbial interactions with antagonistic or synergistic effects on other microbial organisms (Kramer and Abraham \u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e2012\u003c/span\u003e). Among the fungal kingdom, Basidiomycota are known to produce sesquiterpenes \u003cem\u003evia\u003c/em\u003e humulane (Abraham \u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e2001\u003c/span\u003e), and Ascomycota fungi also produce volatile sesquiterpenes such as members of the genus \u003cem\u003ePenicillium\u003c/em\u003e. In our study, we found the \u003cem\u003ePenicillium\u003c/em\u003e genus to be strongly positively associated to chromone derivative and sesquiterpenoid eudesmane abundance, which could constitute an interesting inoculum for boosting agarwood quality. Other genera showed promising associations with agarwood chemical quality such as \u003cem\u003eRhizobium\u003c/em\u003e, \u003cem\u003eSolirubrobacter\u003c/em\u003e, \u003cem\u003eAcrocalymma\u003c/em\u003e and \u003cem\u003eCurvularia\u003c/em\u003e. These taxa may be linked to a higher concentration of odorant molecules either through their direct biosynthesis, inducing stronger plant defenses, or modulating microbial competition. Further analysis on the effect of these genera on plant fitness and stress response could provide valuable insights in the production of quality agarwood. Methods such as synthetic communities could be used to produce an optimal inoculum, boosting plant growth and inducing plant response following stress (Xu \u003cspan citationid=\"CR47\" class=\"CitationRef\"\u003e2025\u003c/span\u003e). For the first time, we provide an in-depth analysis of bacterial and fungal taxa associated with volatile constituents of agarwood, and identified positively and negatively associated microbial members to each chemical class. The positively associated taxa with chromone derivative such as members of the Alphaproteobacteria, Baccilli, Thermophilia, Verrucomicrobiia, Agaromycetes, or Mucoromycetes could provide valuable local inoculum resource, for tree infection or maceration starter. Further studies should test the effect of complex communities\u0026rsquo; inoculation on agarwood formation throughout plant development and their involvement in the metabolomic induction of biosynthesis pathways relevant to agarwood scent.\u003c/p\u003e \u003c/div\u003e"},{"header":"5. Conclusion","content":"\u003cp\u003eThis study provides a comprehensive in-depth characterization of the bacterial and fungal communities associated with agarwood produced from \u003cem\u003eA. crassna\u003c/em\u003e cultivated in French Guiana and examines their relationships with volatile chemical constituents relevant to agarwood scent. Our results demonstrate that \u003cem\u003eAquilaria\u003c/em\u003e grown outside its native range in French Guiana is capable of producing high‑quality agarwood, with a conserved chemical profile across distinct sampling sites, supporting the existence of a stable chemical terroir under local environmental conditions. By integrating microbial community profiling with detailed chemical analyses, we identified numerous bacterial and fungal taxa that were positively or negatively associated with major chemical classes, including chromone derivatives and sesquiterpenoids, which are key contributors to agarwood fragrance. Although these relationships remain correlational, they suggest that microbial communities may influence agarwood chemical complexity through interactions with plant stress responses, microbial metabolism, or cross‑kingdom ecological processes. The persistence of a conserved chemical profile despite spatial variability in microbial composition further indicates potential functional redundancy among microbial taxa or a strong host‑mediated regulation of secondary metabolite production. Comparisons with commercial agarwood samples from South-East Asia and the Middle East revealed a distinct chemical signature of Guianese agarwood, reinforcing the influence of geographic origin on agarwood quality and supporting an extension of the terroir concept to woody aromatic products. These findings highlight the importance of considering both microbial communities and environmental context when evaluating agarwood quality and origin. Overall, this work provides a foundation for the development of microbiome‑informed strategies for agarwood induction, maceration, and processing based on local microbial resources. Future research combining experimental inoculation, longitudinal sampling, and metabolic approaches will be essential to disentangle causal relationships between microbial taxa and chemical traits and to optimize sustainable agarwood production.\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eFunding:\u0026nbsp;\u003c/strong\u003eThis project was founded by the ANR (Agence Nationale de la Recherche) under the project ACQUILASCENT.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eConflicts of interest:\u003c/strong\u003e The authors declare no conflict of interest.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eEthics approval:\u003c/strong\u003e Not applicable\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eConsent to participate:\u003c/strong\u003e Not applicable\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eConsent for publication:\u003c/strong\u003e The authors affirm that human research participants provided informed consent for publication of this version of the manuscript, associated metadata, and images for Figure 1.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAvailability of data and material\u0026nbsp;\u003c/strong\u003eSequencing data are available at the NCBI database (Bio-project PRJNA1249911, Biosamples SAMN47917871 to SAMN47917976). All metadata are provided as supplementary tables.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCode availability :\u0026nbsp;\u003c/strong\u003eThe datasets generated during and/or analyzed during the current study are provided in supplementary tables and the code is available from the corresponding author on reasonable request.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAuthors' contributions: Kenji Maurice\u003c/strong\u003e: Conceptualization, Methodology, Investigation, Data curation, Software, Visualization, Writing – Original draft, Writing – Review and editing. \u003cstrong\u003eNicolas Baldovini\u003c/strong\u003e: Conceptualization, Supervision, Investigation, Data curation, Funding acquisition, Project administration, Writing – Review and editing. \u003cstrong\u003eAlba Zaresmki\u003c/strong\u003e: Project administration, Funding acquisition, Writing – Review and editing.\u0026nbsp;\u003cstrong\u003eJérémie Damay:\u0026nbsp;\u003c/strong\u003eMethodology, Investigation,\u0026nbsp;Writing – Review and editing. \u003cstrong\u003eRomain Lehnebach:\u0026nbsp;\u003c/strong\u003eMethodology, Investigation,\u0026nbsp;Writing – Review and editing. \u003cstrong\u003eYannick Estevez\u003c/strong\u003e: Methodology, Investigation.\u0026nbsp;\u003cstrong\u003eMarc Ducousso\u003c/strong\u003e: Conceptualization, Supervision, Project administration, Funding acquisition, Writing – Review and editing.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003eAbraham W-R (2001) Bioactive Sesquiterpenes Produced by Fungi are they Useful for Humans as Well. CMC 8:583\u0026ndash;606. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.2174/0929867013373147\u003c/span\u003e\u003cspan address=\"10.2174/0929867013373147\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBaldovini N (2022) The Chemistry of Agarwood Odorants. In: Kinghorn AD, Falk H, Gibbons S et al (eds) Progress in the Chemistry of Organic Natural Products 118. Springer International Publishing, Cham, pp 47\u0026ndash;100\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBelda I, Zarraonaindia I, Perisin M et al (2017) From Vineyard Soil to Wine Fermentation: Microbiome Approximations to Explain the terroir Concept. Front Microbiol 8:821. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.3389/fmicb.2017.00821\u003c/span\u003e\u003cspan address=\"10.3389/fmicb.2017.00821\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBicchi C, Chaintreau A, Joulain D (2018) Identification of flavour and fragrance constituents. Flavour Fragr J 33:201\u0026ndash;202. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1002/ffj.3445\u003c/span\u003e\u003cspan address=\"10.1002/ffj.3445\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBokulich NA, Collins TS, Masarweh C et al (2016) Associations among Wine Grape Microbiome, Metabolome, and Fermentation Behavior Suggest Microbial Contribution to Regional Wine Characteristics. mBio 7:e00631\u0026ndash;e00616. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1128/mBio.00631-16\u003c/span\u003e\u003cspan address=\"10.1128/mBio.00631-16\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBokulich NA, Thorngate JH, Richardson PM, Mills DA (2014) Microbial biogeography of wine grapes is conditioned by cultivar, vintage, and climate. Proc Natl Acad Sci USA 111. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1073/pnas.1317377110\u003c/span\u003e\u003cspan address=\"10.1073/pnas.1317377110\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBora SS, Ronghang R, Das P et al (2025) Endophytic Microbial Community Structure and Dynamics Influence Agarwood Formation in Aquilaria malaccensis Lam. Curr Microbiol 82:66. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1007/s00284-024-04048-2\u003c/span\u003e\u003cspan address=\"10.1007/s00284-024-04048-2\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBrunner JD, Robinson AJ, Chain PSG (2024) Combining compositional data sets introduces error in covariance network reconstruction. ISME Commun 4:ycae057. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1093/ismeco/ycae057\u003c/span\u003e\u003cspan address=\"10.1093/ismeco/ycae057\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCameron ES, Schmidt PJ, Tremblay BJ-M et al (2021) Enhancing diversity analysis by repeatedly rarefying next generation sequencing data describing microbial communities. Sci Rep 11:22302. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1038/s41598-021-01636-1\u003c/span\u003e\u003cspan address=\"10.1038/s41598-021-01636-1\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eD\u0026rsquo;Auria JC, Chen F, Pichersky E (2002) Characterization of an Acyltransferase Capable of Synthesizing Benzylbenzoate and Other Volatile Esters in Flowers and Damaged Leaves of \u003cem\u003eClarkia breweri\u003c/em\u003e. Plant Physiol 130:466\u0026ndash;476. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1104/pp.006460\u003c/span\u003e\u003cspan address=\"10.1104/pp.006460\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDavis NM, Proctor DM, Holmes SP et al (2018) Simple statistical identification and removal of contaminant sequences in marker-gene and metagenomics data. Microbiome 6:226. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1186/s40168-018-0605-2\u003c/span\u003e\u003cspan address=\"10.1186/s40168-018-0605-2\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFaizal A, Hermawaty D, Junita E et al (2022) Evaluation of biotic and abiotic stressors to artificially induce agarwood production in Gyrinops versteegii (Gilg.) Domke seedlings. Symbiosis 86:229\u0026ndash;239. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1007/s13199-022-00835-2\u003c/span\u003e\u003cspan address=\"10.1007/s13199-022-00835-2\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFu C-C, Huang B-X, Wang S-S et al (2024) Deciphering the roles of bacterial and fungal communities in the formation and quality of agarwood. Stress Biology 4:40. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1007/s44154-024-00179-5\u003c/span\u003e\u003cspan address=\"10.1007/s44154-024-00179-5\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGilbert JA, Van Der Lelie D, Zarraonaindia I (2014) Microbial \u003cem\u003eterroir\u003c/em\u003e for wine grapes. Proc Natl Acad Sci USA 111:5\u0026ndash;6. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1073/pnas.1320471110\u003c/span\u003e\u003cspan address=\"10.1073/pnas.1320471110\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHamilton MA, Murray BR, Cadotte MW et al (2005) Life-history correlates of plant invasiveness at regional and continental scales. Ecol Lett 8:1066\u0026ndash;1074. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1111/j.1461-0248.2005.00809.x\u003c/span\u003e\u003cspan address=\"10.1111/j.1461-0248.2005.00809.x\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eIslam MR, Chakraborty C, Banu S (2022) Isolation and Characterization of Bacteria and Fungi Associated with Agarwood Fermentation. Curr Microbiol 79:313. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1007/s00284-022-02999-y\u003c/span\u003e\u003cspan address=\"10.1007/s00284-022-02999-y\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKessler D, Baldwin IT (2007) Making sense of nectar scents: the effects of nectar secondary metabolites on floral visitors of \u003cem\u003eNicotiana attenuata\u003c/em\u003e. Plant J 49:840\u0026ndash;854. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1111/j.1365-313X.2006.02995.x\u003c/span\u003e\u003cspan address=\"10.1111/j.1365-313X.2006.02995.x\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKnight S, Klaere S, Fedrizzi B, Goddard MR (2015a) Regional microbial signatures positively correlate with differential wine phenotypes: evidence for a microbial aspect to terroir. Sci Rep 5:14233. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1038/srep14233\u003c/span\u003e\u003cspan address=\"10.1038/srep14233\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKnight S, Klaere S, Fedrizzi B, Goddard MR (2015b) Regional microbial signatures positively correlate with differential wine phenotypes: evidence for a microbial aspect to terroir. Sci Rep 5:14233. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1038/srep14233\u003c/span\u003e\u003cspan address=\"10.1038/srep14233\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKramer R, Abraham W-R (2012) Volatile sesquiterpenes from fungi: what are they good for? Phytochem Rev 11:15\u0026ndash;37. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1007/s11101-011-9216-2\u003c/span\u003e\u003cspan address=\"10.1007/s11101-011-9216-2\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKurtz ZD, M\u0026uuml;ller CL, Miraldi ER et al (2015) Sparse and Compositionally Robust Inference of Microbial Ecological Networks. PLoS Comput Biol 11:e1004226. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1371/journal.pcbi.1004226\u003c/span\u003e\u003cspan address=\"10.1371/journal.pcbi.1004226\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLee SY, Mohamed R (2016) The Origin and Domestication of Aquilaria, an Important Agarwood-Producing Genus. In: Mohamed R (ed) Agarwood. Springer Singapore, Singapore, pp 1\u0026ndash;20\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLi W, Chen H-Q, Wang H et al (2021) Natural products in agarwood and \u003cem\u003eAquilaria\u003c/em\u003e plants: chemistry, biological activities and biosynthesis. Nat Prod Rep 38:528\u0026ndash;565. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1039/D0NP00042F\u003c/span\u003e\u003cspan address=\"10.1039/D0NP00042F\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLiu H, Roeder K, Wasserman L (2014) Stability Approach to Regularization Selection. StARS) for High Dimensional Graphical Models\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLove MI, Huber W, Anders S (2014) Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2. Genome Biol 15:1\u0026ndash;21\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMas A, Portillo MC (2022) Strategies for microbiological control of the alcoholic fermentation in wines by exploiting the microbial terroir complexity: A mini-review. Int J Food Microbiol 367:109592. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.ijfoodmicro.2022.109592\u003c/span\u003e\u003cspan address=\"10.1016/j.ijfoodmicro.2022.109592\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMiura T, S\u0026aacute;nchez R, Casta\u0026ntilde;eda LE et al (2017a) Is microbial terroir related to geographic distance between vineyards? Environ Microbiol Rep 9:742\u0026ndash;749. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1111/1758-2229.12589\u003c/span\u003e\u003cspan address=\"10.1111/1758-2229.12589\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMiura T, S\u0026aacute;nchez R, Casta\u0026ntilde;eda LE et al (2017b) Is microbial terroir related to geographic distance between vineyards? Environ Microbiol Rep 9:742\u0026ndash;749. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1111/1758-2229.12589\u003c/span\u003e\u003cspan address=\"10.1111/1758-2229.12589\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMohamed R, Jong PL, Zali MS (2010) Fungal diversity in wounded stems of Aquilaria malaccensis. Fungal Divers 43:67\u0026ndash;74. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1007/s13225-010-0039-z\u003c/span\u003e\u003cspan address=\"10.1007/s13225-010-0039-z\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNaef R (2011) The volatile and semi-volatile constituents of agarwood, the infected heartwood of \u003cem\u003eAquilaria\u003c/em\u003e species: a review. Flavour Fragr J 26:73\u0026ndash;87. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1002/ffj.2034\u003c/span\u003e\u003cspan address=\"10.1002/ffj.2034\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNakano RT (2024) Taste of microbes: the terroir explained by rhizospheric microbes. New Phytol 243:1639\u0026ndash;1641. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1111/nph.19796\u003c/span\u003e\u003cspan address=\"10.1111/nph.19796\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNaziz PS, Das R, Sen S (2019) The Scent of Stress: Evidence From the Unique Fragrance of Agarwood. Front Plant Sci 10:840. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.3389/fpls.2019.00840\u003c/span\u003e\u003cspan address=\"10.3389/fpls.2019.00840\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNaziz PS, Das R, Sen S (2024) Augmentation in fragrant agarwood oil quality by fermentation with a microbial consortium of bacterium (Microbacterium oxydans) and fungus (Penicillium aethiopicum). Eur J Wood Prod 82:1319\u0026ndash;1332. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1007/s00107-024-02089-9\u003c/span\u003e\u003cspan address=\"10.1007/s00107-024-02089-9\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNgadiran S, Baba S, Nor NFA et al (2023) The induction techniques of resinous agarwood formation: A review. Bioresource Technol Rep 21:101337. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.biteb.2023.101337\u003c/span\u003e\u003cspan address=\"10.1016/j.biteb.2023.101337\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eOp De Beeck M, Lievens B, Busschaert P et al (2014) Comparison and Validation of Some ITS Primer Pairs Useful for Fungal Metabarcoding Studies. PLoS ONE 9:e97629. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1371/journal.pone.0097629\u003c/span\u003e\u003cspan address=\"10.1371/journal.pone.0097629\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eP\u0026otilde;lme S, Abarenkov K, Henrik Nilsson R et al (2020) FungalTraits: a user-friendly traits database of fungi and fungus-like stramenopiles. Fungal Divers 105:1\u0026ndash;16. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1007/s13225-020-00466-2\u003c/span\u003e\u003cspan address=\"10.1007/s13225-020-00466-2\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRasool S, Mohamed R (2016) Understanding Agarwood Formation and Its Challenges. In: Mohamed R (ed) Agarwood. Springer Singapore, Singapore, pp 39\u0026ndash;56\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRognes T, Flouri T, Nichols B et al (2016) VSEARCH: a versatile open source tool for metagenomics. PeerJ 4:e2584. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.7717/peerj.2584\u003c/span\u003e\u003cspan address=\"10.7717/peerj.2584\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSoehartono T, Newton AC (2001) Reproductive ecology of Aquilaria spp. in Indonesia. Forest Ecology and Management 152:59\u0026ndash;71. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/S0378-1127(00)00610-1\u003c/span\u003e\u003cspan address=\"10.1016/S0378-1127(00)00610-1\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTan CS, Isa NM, Ismail I, Zainal Z (2019) Agarwood Induction: Current Developments and Future Perspectives. Front Plant Sci 10:122. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.3389/fpls.2019.00122\u003c/span\u003e\u003cspan address=\"10.3389/fpls.2019.00122\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTedersoo L, Bahram M, Zinger L et al (2022) Best practices in metabarcoding of fungi: From experimental design to results. Mol Ecol 31:2769\u0026ndash;2795. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1111/mec.16460\u003c/span\u003e\u003cspan address=\"10.1111/mec.16460\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTerrat S, Dequiedt S, Horrigue W et al (2015) Improving soil bacterial taxa\u0026ndash;area relationships assessment using DNA meta-barcoding. Heredity 114:468\u0026ndash;475. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1038/hdy.2014.91\u003c/span\u003e\u003cspan address=\"10.1038/hdy.2014.91\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTipton L, M\u0026uuml;ller CL, Kurtz ZD et al (2018) Fungi stabilize connectivity in the lung and skin microbial ecosystems. Microbiome 6:12. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1186/s40168-017-0393-0\u003c/span\u003e\u003cspan address=\"10.1186/s40168-017-0393-0\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTurjaman M, Hidayat A, Santoso E (2016) Development of Agarwood Induction Technology Using Endophytic Fungi. In: Mohamed R (ed) Agarwood. Springer Singapore, Singapore, pp 57\u0026ndash;71\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eVang CY (2018) Thinking Refugee: The Politics of Hmong Place-Making in Argentina and French Guiana. Amerasia J 44:1\u0026ndash;21. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.17953/aj.44.2.1-21\u003c/span\u003e\u003cspan address=\"10.17953/aj.44.2.1-21\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWalsh C, Vanderburgh C, Grant, Lady et al (2024) Microbial terroir: associations between soil microbiomes and the flavor chemistry of mustard (\u003cem\u003eBrassica juncea\u003c/em\u003e). New Phytol 243:1951\u0026ndash;1965. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1111/nph.19708\u003c/span\u003e\u003cspan address=\"10.1111/nph.19708\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eXu X (2025) Composing a microbial symphony: synthetic communities for promoting plant growth. Trends in Microbiology\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZhang XL, Liu YY, Wei JH et al (2012) Production of high-quality agarwood in Aquilaria sinensis trees via whole-tree agarwood-induction technology. Chin Chem Lett 23:727\u0026ndash;730. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.cclet.2012.04.019\u003c/span\u003e\u003cspan address=\"10.1016/j.cclet.2012.04.019\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":true,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":true,"hideJournal":true,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true},"keywords":"Aquilaria, Oud, microbiome, chemistry, terroir","lastPublishedDoi":"10.21203/rs.3.rs-9040724/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-9040724/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eAgarwood is a highly valued aromatic resinous wood formed in \u003cem\u003eAquilaria\u003c/em\u003e species following stress or infection, yet the putative microbial drivers of its chemical quality remain poorly understood, particularly outside its native range. In this study, we investigated the bacterial and fungal communities associated with agarwood produced from \u003cem\u003eAquilaria crassna\u003c/em\u003e Pierre ex Lecomte planted in French Guiana and examined their relationships with volatile chemical compounds relevant to agarwood fragrance. Using high‑throughput sequencing and comprehensive chemical profiling, we characterized microbial community composition and agarwood volatile profiles across multiple cultivation plots. Despite spatial variability in microbial assemblages, agarwood samples exhibited a conserved chemical signature dominated by chromone derivatives and sesquiterpenoids, indicating the presence of a stable chemical terroir under Guianese environmental conditions. Network analysis revealed numerous bacterial and fungal taxa significantly associated with key chemical classes, suggesting potential microbial contributions to agarwood chemical complexity through plant\u0026ndash;microbe interactions or microbial metabolic activity, although causality remains to be established. Comparative analyses with commercial agarwood samples from South-East Asia and the Middle East revealed a distinct chemical profile for Guianese agarwood, highlighting the influence of geographic origin on agarwood quality and supporting an extension of the terroir concept to woody aromatic products. Overall, this study demonstrates that \u003cem\u003eAquilaria\u003c/em\u003e trees cultivated in French Guiana can produce high‑quality agarwood and provides new insights into the interplay between microbial communities and agarwood chemistry. These findings lay the groundwork for the development of locally adapted, microbiome‑informed strategies for sustainable agarwood production.\u003c/p\u003e","manuscriptTitle":"The microbial and chemical terroir of agarwood in French Guiana","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2026-03-09 10:33:21","doi":"10.21203/rs.3.rs-9040724/v1","editorialEvents":[{"type":"communityComments","content":0}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"a9ee708c-4489-45a0-a6cf-ef1f9525653d","owner":[],"postedDate":"March 9th, 2026","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[{"id":63991600,"name":"Agronomy"},{"id":63991601,"name":"General Microbiology"},{"id":63991602,"name":"Agroecology"}],"tags":[],"updatedAt":"2026-03-09T10:33:21+00:00","versionOfRecord":[],"versionCreatedAt":"2026-03-09 10:33:21","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-9040724","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-9040724","identity":"rs-9040724","version":["v1"]},"buildId":"XKTyCvWXoU3ODBz1xrDgd","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}