Introduction

In the world market, industries are willing to pay for any given price to get the supply, nevertheless of synthetic sources available with cheaper price. Either of its limited sources or artificial production, Aquilaria spp. is also amongst the most expensive agarwood species [1,2,3]. The species itself would not be valuable unless it contains a resinous strip usually found along the inner part of the trunk. Some still unknown mechanism contributes to resin production in agarwood species. Regardless of the fact, countries with available agarwood species seedling have been practising agarwood plantation. Whilst researchers worldwide compete to investigate agarwood species formation, the species keep declining each new year [4,5,6].

Agarwood (Aquilaria malaccensis) is the resinous fragrant infused wood species growing in tropical rainforests in South and Southeast Asian countries [7]. A. malaccensis is the most popular source amongst the Aquilaria species. It is known by various names such as agar, agarwood/sandalwood, aloeswood, eaglewood, gaharu, and kalamabak [7]. It has high commercial demand for its medicines, perfumes, incense, and religious rituals in Asia and the middle east. Resinous agarwood emits a unique soft fragrance when burnt, but its scent is easily detected in super quality agarwood. Distributed in different countries, agarwood is called with a different language and usages relevant to local culture. It is called ‘gaharu’ or ‘karas’ in Malaysia and Indonesia. Whilst vernacular names sound confusing sometimes, their commercial name is widely known as agarwood, aloeswood, and eaglewood [8].

Agarwood is not produced in the wood tissues of Aquilaria species; it is likely formed after healthy tissues have been wounded and infected by a fungus. The formation of volatile agarwood compounds is mainly attributed to the exposure of the agarwood tree to both biotic and abiotic stresses (physically or chemically) that activates the defence mechanism [3]. Various species of fungi have been discovered from agarwood and even applied as inoculants, but the chemical compositions of volatile compounds are still low in quality or frequently inconsistent [9]. The differences between naturally infected and artificially inoculated agarwood may be due to the different fungi diversity in both agarwood trees.

The resin quality is dependent on the resin production process, which relies on types of fungi groups. Premalatha and Kalra [10] found Alternaria sp, Clasdosporium sp, Fusarium sp, Curvularia sp and Clasdosporium sp, and Phaeacremonium sp in resinous wood of Aquilaria malaccensis. Author Mohamed and his co-workers [11] have discovered fungal diversity in agarwood is amongst Cunninghamella, Curvularia, Fusarium, and Trichoderma. Although agarwood microbes in commercial inoculants and nature may not be related [12, 13], certain fungi groups isolated from resinous wood may influence the quality of agarwood scent. Considering the marketable prospect of high-quality agarwood, fungi identification based on its capability and capacity to induce Aquilaria wood into golden agarwood is crucial to large-scale production to fulfil exceeded world demand [4]. Focus is subjected to the relationship between fungi species group and biochemical content in resinous wood compared to the previous study.

There is a lack of prior research studies comparing the fungal community composition in the artificially inoculated and naturally infected agarwood. Thus, comparing these two agarwood samples in terms of quality and fungi diversity is necessary to determine the active fungal group that can effectively induce volatile agarwood compounds. Moreover, the comparison may prove if the quality of artificially inoculated agarwood can match that of naturally infected agarwood based on fungal infection. This present research investigation aims to identify fungi that associate during agarwood formation sourced from inoculants, natural agarwood, and the environment using a fungal culture and molecular approach. Agarwood quality based on agarwood incense smoke can develop chemical profiles expressed by chromatogram.

Materials and Methods

Agarwood Collection

This study used two types of agarwood samples: the artificially inoculated agarwood and naturally infected agarwood. The naturally inoculated agarwood was sourced from the agarwood trees wounded by indigenous people (Kedaik) in Rompin, Pahang, Malaysia, whilst the artificially inoculated samples were collected from Merchang Agarwood Plantation Terengganu, Malaysia. Five agarwood samples from the naturally infected samples were collected and labelled as R1, R2, R3, R4, and R5, whilst healthy woods used as the control were labelled HR1, HR2, HR3, HR4, and HR5. The artificially inoculated agarwood samples were labelled as M1 and M2, whilst healthy artificial woods used as the control were labelled HM1 and HM2. All the collected samples were stored in containers until further use.

Fungi Identification

Agarwood samples, consisting of both resinous and healthy parts, were cleaned from dirt and dust under running water. The samples were surface sterilized by dipping in 10% sodium hypochlorite (NaClO) for 5 min, followed by 2 min in 70% ethanol. The outer part of the samples was peeled off using a sterile blade, whilst the internal part of the samples used for fungal isolation was cut into small pieces and placed onto Potato Dextrose Agar (PDA). The agar plates were incubated at 27 ± 2 °C for three to ten days or until distinct colonies were observed. A single spore culture was implemented after the growth of a uniformed colony on the plate. The fungal culture was scraped using a sterile microbial needle, transferred into sterile distilled water, and mixed. One loop of the spore suspension was observed under a light microscope to determine its concentration. Dilution was made until 10 to 15 spores were seen in one loop suspension. The suspension was streaked onto PDA and incubated for 24 to 72 days to allow spore germination [14].

The incubation was performed at room temperature with 12 h light and dark intervals daily. Using a sterile microbial needle, a single germinated spore was transferred onto a new PDA. The single spore culture was identified through morphological observation using a light microscope (Leica DM500) with an attached camera (Leica ICC50 HD). For molecular determination, the genomic DNA was extracted using I-genomic BYF DNA Extraction Mini Kit (Intron, Belgique) according to the manufacturer's protocol. The identification of the microbes was made based on the similarity of the amplified sequences with the BLAST program of the National Center for Biotechnology Information (NCBI).

Gas Chromatography (GC) Analysis

The chemical analysis of the agarwood was performed using the Solid Phase Micro Extraction (SPME) method to capture volatile compounds released by the wood sample. The agarwood and the healthy wood resinous part was cut into small pieces, heated at 50 °C, and exposed to SPME divinylbenzene-carboxen polydimethylsiloxane (DVB-CAR-PDMS) fibre for 30 min. Analysis of the compounds was conducted by Gas Chromatography-Mass Spectrometry (GC–MS) and Gas Chromatography-Flame Ionization Detector (GC-FID) using column DB-1 [15, 16]. The front inlet was turned into a splitless mode with a heater temperature of 200 °C, the pressure at 12.537 psi, and a septum purge flow of 3 mL/min. The oven was programmed to an initial temperature of 60 °C, which eventually increased to 250 °C with an increment rate of 3 °C/min. The carrier gas flow in the oven was set at 1 mL/min. The front detector heater was set at 250 °C, with a hydrogen gas flow rate of 35 mL/min and an airflow rate of 350 mL/min. The GC-FID results were based on the retention indices or Kovats Index, calculated using linear hydrocarbon C8 to C20.

Results and Discussion

Morphological Identification of Fungi

A total of 55 fungal isolates were grouped into common morphological characteristics based on macroscopic and microscopic observations after seven to ten days of cultivation. The macroscopic observation was based on PDA's fungal colony growth pattern, colour, and shape. Meanwhile, the microscopic observation was done under 10 × , 40 × , and 100 × magnifications to characterize the fungal mycelium, fruiting bodies, and spore structures. The fungal isolates were classified into ten groups based on morphological characterization observed (Table 1).

Table 1 Summary of morphological description of fungal group's presence in agarwood sample
Full size table

Based on the description in this study, fungi from Group 1 (Fig. 1 and Figure a, Supplementary) are more likely to be Fusarium solani. Based on the Universiti Sains Malaysia's Fusarium collection, more than half of the strain obtained from soil and infected plants throughout Malaysia consists of F. solani either as decomposers or pathogens [17]. Therefore, wounded agarwood can be easily infected by F. solani due to the high density and vast population in Malaysia (whether in natural habitats or the plantation).

Fig. 1
figure 1

Macroscopic observation of Group 1 to Group 10 fungi found in healthy wood and agarwood of A. malacensis

Full size image

The similarity of Group 2 with Group 1 and the appearance of pigmentation led to the assumption that Group 2 (Figure b, Supplementary) may be Fusarium oxysporum. Previously, F. oxysporum has been reported to associate with banana wilt [18], which indicates its tendency to exist in moist and colder environments. Hence, it is reasonable that F. oxysporum was the only isolate from the natural wood samples of both agarwood (R1 and R4) and healthy wood (HR3 and HR4). Similar findings were reported by Mohamed [11], who isolated Fusarium spp. from wild species of A.

From the macroscopic observation and microscopic examination, as shown in Figure c, Supplementary, Group 3 fungi were assumed to belong to the genus Botryosphaeria. Macroscopic observation of Group 3 (Figure c, Supplementary) presented a similar morphology as Botryosphaeria rhodina isolated from grapevine and Vitis vinifera from the kenaf plant [19]. However, the microscopic examination revealed many elongated oval structures growing from the hyphae. This has never been previously observed in B. rhodina. In this study, elongated oval conidia were observed around the asci structure after ten days of incubation. Genus Botryosphaeria in Group 3 was isolated from the natural agarwood (R2 and R3) and never from the healthy wood. B. rhodina has been found as an endophyte in Aquilaria sinensis in China [20]. In the endophytic fungi study of the medicinal plant Tinospora cordifolia, B. rhodina was frequently found in the plant's stem. However, it was highly detected during winter compared to other seasons [21]. Considering that B. rhodina prefers a lower temperature as an endophyte, it cannot survive in A. malaccensis healthy wood due to the hot climate in Malaysia compared to China. The existence of Botryosphaeria spp. in A. malaccensis was assumed to be in teleomorph form such as Lasiodiplodia theobromae, where it has been discovered to induce the formation of volatile agarwood compounds in A. malaccensis in Malaysia [22].

Group 4 (Figure d, Supplementary) was isolated from the natural agarwood of R2 and R5 and was never found in the artificially inoculated agarwood. However, according to Sales and Yoshizawa [23], the climate of Peninsular Malaysia (high temperature and high relative humidity all the year) promotes the growth and expansion of Aspergillus spp. Therefore, Aspergillus spp. Infection on agarwood is possible, especially in natural habitats.

Group 5 (Figure e, Supplementary) fungi, isolated from the natural agarwood (R3 and R4), had spiky hyphae, with a round shape clump growing on every septum. The structure of the hyphae was similar to that of the Schizophyllum commune previously characterized by Singh and his co-workers [24] during their medical pathogenic fungi study. The genus Schizophyllum is a novel discovery in agarwood studies. S. commune is a wild edible fungus found in Malaysia as an endophyte of oil palm and is also involved in the antifungal study to control wood-degrading fungi in rubber trees [25]. Since S. commune commonly infects softwood, R3 and R4 must be younger trees with softer wood than R1, R2, and R5. Although the agarwood tree is classified as non-timber, the wood is hardened upon ageing. This may be the reason S. commune was not found in R1, R2, and R5 and never isolated from the wild matured agarwood in previous studies regardless of the conducive natural environments. It was currently discovered that the relationship between agarwood and S. commune might be mutual symbiosis. The interaction of S. commune with the agarwood tree probably inhibits certain fungal strains, which remained a gap of its antifungal capacity. However, due to some other fast-growing fungi like Aspergillus spp. and Lasiodiplodia spp. and the antifungal property of agarwood, S. commune may not entirely survive in agarwood.

Figure f, Supplementary Group 6 isolates were obtained from R5 and M1 agarwood samples alone. Some of the culture colonies were orange in colour. The thin layer colony presented similar characteristics as Phanerochaete spp. Previously, no study had reported the isolation of genus Phanerochaete in agarwood of A. malaccensis. Phanerochaete spp. Played a vital role in the industrial regulation of aromatic carbon degradation, biopulping, brown coal degradation, and lignin degradation [26].

Fungal isolates from agarwood R1 and R3 displayed in Figure g, Supplementary (Group 7) were distinctly noticed due to their tremendous fast growth. Both fungi were fully observed on the culture plates as early as three days of incubation. The fungi appeared dark-brown with greyish cottony mycelium, with black pigmentation observed on the reverse side of the plate. Their morphology conformed with those of L. theobromae reported by [19]. These fungi do not exhibit any conidia or unique structures except a flat and separated mycelium. In this study, the isolate was observed for seven to ten days. Black shiny pycnidia, which hardened at maturity, was observed. The number of pycnidia was found to increase when the isolate was incubated for longer than ten days.

In 2014, L. theobromae was revealed as an endophyte in the root tissue of Mapania kurzii (Cyperaceaa) from the Malaysian rainforest [27]. Therefore, L. theobromae was confirmed as a fungal diversity distributed across peninsular Malaysia and could be found mainly in natural habitats such as in agarwood R3 and R4. This finding was supported by discovering L. theobromae in wild A. malaccensis, where more isolates came from resinous woods than the healthy woods. As a soil-borne fungus, L. theobromae tend to induce injury on the agarwood stem closer to the soil. The existence of L. theobromae from R1 and R3 rather than R2, R4, and R5 might be due to their lower wounded stem area, which made them prone to L. theobromae attack. Unfortunately, the exact distance of the wounded area of A. malaccensis in the forest was not recorded in this study.

Fungi from Group 8 (Figure h, Supplementary) were isolated from the natural agarwood R1 and R4 and are characterized by a white dispersed colony with obvious aerial mycelium. No pigmentation was observed at the back of the agar plate (data not shown). The morphology on the plate matched the description of Rhizopus stolonifer grown on an agar medium [28]. However, the microscopic examination differed as the culture conidia existed as globose and rod shapes. The globose-sized conidia were larger and principally engaged with dense and compact mycelium. Some of the rod-shaped conidia were elongated, slightly curved, and some were septate. Thus, Group 8 cannot be categorized as R. stolonifer. Fungal isolate from Group 9 and 10 appeared as white hyaline mycelium. The difference between both fungi is that Group 9 presented a ring-form colony whilst Group 10 had dense brownish, curled concentric ring colonies.

As displayed in Figure h, Supplementary (Group 9), fungi were found in artificially inoculated agarwood (M2), whilst Group 10 was isolated from the naturally infected agarwood (R1) and healthy woods of artificially inoculated agarwood tree (HM1). Like Group 8, these two fungal groups (Group 9 and Group 10) were unclassifiable in the taxonomical classification due to the lack of unique morphological characteristics; thus, they need further molecular identification. Moreover, the microscopic structure of Group 9 fungi showed non-septate hyphae and irregular fusiform conidia different from those of Group 10 isolates, which showed terminal and intercalary chlamydospore on hyphae.

Molecular Identification of Fungi

Apart from morphological identification, fungal DNA identification was executed to determine the fungal genus and species compared with previous identified fungal sequences in the NCBI database. Morphological observation alone could be erroneous as fungi characteristics vary depending on media, temperature, and environment [29, 30]. Additionally, fungi have diverse genus and species that their physical variation could be confusing, especially to non-experts, despite the fungi atlas that is handy to individual species. Thus, fungal DNA identification has become a favourable approach to support fungal identification with or without prior morphological examination, especially fungi recalcitrant to culture in the laboratory [31].

The employment of a fungal-specific primer is crucial in fungal identification studies as previous studies have used protein-coding markers and ribosomal RNA. Nowadays, the most popular approach to fungal identification is based on ribosomal RNA, and the process is called the Internal Transcribed Spacer (ITS). ITS has shown superior performance by contributing to efficient PCR amplification with lower species discrimination and a successful sequencing rate than other ribosomal RNA such as 18S and 28S [32]. Although it has a competent capability with 28S in barcoding, the probability of correct identification (PCI) rate of ITS is moderately higher. From lichen, basidiomycota, ascomycota, and yeast, a range of fungal species have been successfully identified through ITS [31]. Since the proposal of ITS as primary fungal barcoding was accepted by the Consortium for the Barcode of Life [32], ITS has become a favourable approach that its utilization has increased over the years [31]. The primers (ITS1 and ITS4) employed in this study were expected to amplify a PCR product with a size range from 500 bp to 750 bp. The same pair of primers was used to study fungi associated with A. malaccensis by Premalatha and Kalra [10].

The list of identified strains and their accession number, along with the percentage similarity to their best match in the NCBI sequence databases, are given in Table 2. The sequences from Group 1 and Group 2 were identified as F. solani with 99% similarity (Table 2). The identification of F. solani from Group 1 was accurate, considering that the morphological observation of the isolate matched that of F. solani earlier described. Moreover, Group 2 was also identified as Fusarium sp. with 95% identity. Group 3 showed 99% similarity with B. rhodina and concordance with the morphological (microscopic and macroscopic) observation. As expected, Group 4 was confirmed as an Aspergillus spp. And, upon sequence analysis, it was discovered to be similar to Aspergillus aculeatus with 100% similarity. Group 5 isolate was similar to S. commune (99% similarity) and supported morphological observation. Group 6 related to Phanerochaete chrysosporum with 94% similarity, whilst Group 7 was identified as L. theobromae with 100% similarity. Polyporales sp. exhibited 100% ITS sequence identity with Group 9, whilst Group 8 presented a 95% identical sequence with the same genus. Group 10 was found to be similar to the genus Ceriporia sp. Polyporales sp. is one of the large orders under the Basidiomycota phylum. Polyporales covers a range of species, including Ceriporia sp. and P. chrysosporum, although some are yet to be identified at the species level. Unfortunately, correct identification to species level may require longer sequences of approximately 600 bp and above than the readily available NCBI data [33]. However, unavailable data for related species may hinder future studies on the species-level identification of isolates.

Table 2 Internal Transcribed Spacer (ITS) identification of fungal groups in agarwood sample
Full size table

The current study observed that Fusarium sp. was found as the most dominant isolate in all artificially inoculated and naturally infected agarwood of A. malaccensis, endorsing the role of endophyte in the formation of volatile agarwood compounds. Thus, it is assumed that agarwood formation could result from a plant defence mechanism towards fungal attacks by generating resinous compounds as a secondary metabolite.

Quality Analysis of Agarwood

Eleven most common compounds previously reported in agarwood were used as quality indicators to compare agarwood and healthy wood samples. Nine compounds (benzaldehyde, 4-phenyl-2butanone, α-guaiene, β-agarofuran, β-selinene, nor-ketoagarofuran, epoxybulnesene, agarospirol, and jinkoh eremol) were found with the highest quality in A. malaccensis oil-based agarwood on SPME-GCMS-GCFID [34]. Meanwhile, aromadendrene, 10-epi-γ-eudesmol, and guaia-1(10), 11-dien-15-ol are commonly found in agarwood samples from A. malaccensis [35]. Table 3 shows the list of compounds in the healthy wood and agarwood samples. There are notable differences in the chemical compounds contained in the naturally inoculated and artificially infected agarwood compared to the healthy control wood (Table 3). Some of the differences may be due to an unknown duration of inducement, especially from natural agarwood samples, and the location of the trees.

Table 3 GC-FID data analysis of the chemical composition of naturally infected and artificially inoculated agarwood from A. malaccensis
Full size table

From the analysis, natural agarwood samples from Rompin showed that R1 produced the highest number of compounds found in high-quality agarwood, such as benzaldehyde, 4-phenyl-2-butanone, β-agarofuran, β-selinene, nor-keto-agarofuran, epoxybulnesene, 10-epi-γ-eudesmol, agarospirol, and jinkoh eremol. It was followed by the R4 agarwood sample, which produced eight compounds. The R5 agarwood sample produced four similar compounds as the R4 sample except for α-guaiene, Aromadendrene, β-agarofuran, and β-selinene. R2 agarwood sample produced two (10-epi-γ-eudesmol and agarospirol) compounds, whilst R3 produced one (agarospirol) compound. The findings showed that the naturally sourced healthy wood samples contained fewer compounds than the data from previously reported studies [36]. The detection of agarospirol in healthy wood has been reported in A. sinensis but has never been found in healthy A. malaccensis [37].

Additionally, α-guaiene was found in both natural agarwood of R4 (0.414) and healthy wood of HR1 (0.202). Natural agarwood oil from A. malaccensis has been repeatedly reported to contain α-guaiene [34]. It has been suggested that healthy wood samples that indicate sesquiterpenes such as agarospirol and α-guaiene could be found near an infected area of an agarwood tree. Considering that the plant defence system network is not the only affected area, damaged cells tend to signal the adjacent healthy cells to either ward off an active infection or induce a preventive strategy [38]. Thus, the healthy woods play an essential role in synthesizing secondary metabolites such as sesquiterpene in response to damaged cells. In this study, although both samples contained agarospirol and α-guaiene, their percentage area was higher in agarwood samples than in healthy woods.

Epoxybulnesene had been reported as an indicator compound in agarwood that can be found in low and high-quality agarwood, distinguishable by its abundance [34]. However, the current finding showed a contrary result as epoxybulnesene was only found in the healthy agarwood. Epoxybulnesene was found in agarwood isolated from Aquilaria agallocha and A. malaccensis but have not been portrayed as a significant compound in agarwood oil [39]. The differences in the results can be due to the different forms of the analysed samples. Epoxybulnesene was found in the naturally infected agarwood oil but not as a volatile compound, as some substance in the agarwood may interfere more with the volatility of epoxybulnesene than in the healthy wood.

The composition of 10-epi-γ-eudesmol was found to be highest in the agarwood sample of R4 (1.908), followed by those of R1 (1.094), R2 (0.465), and R5 (0.277). This finding was supported by the detection of 10-epi-γ-eudesmol in the naturally infected agarwood samples isolated from A. malaccensis, A. sinensis, and A. crassna, but with no trace in the healthy agarwood [40]. 10-epi-γ-eudesmol can be considered as a marker compound in high-quality agarwood [34]. Moreover, jinkoh eremol was previously identified in naturally infected agarwood. Naef [41] was also observed in the agarwood samples of R4 (1.662), R1 (0.824), and R5 (0.464). This compound is usually found in high-quality agarwood and can only be detected using GC-FID. Norketoagarofuran was abundant in R1 (0.652), R5 (0.42), and R4 (0.209) but not found in the healthy wood samples. Norketoagarofuran was first discovered in 1963 and reported by several over the years from naturally infected agarwood [41]. Recently, Norketoagarofuran was detected in A. malaccensis natural agarwood oil sourced from Selangor, Malaysia and agarwood from different grades [42].

β-agarofuran and β-selinene were found in the naturally infected agarwood of R1 (0.438 and 0.643) and R4 (1.116 and 1.179). Both compounds were frequently found in high-quality agarwood’s oil or volatile compounds [42]. Aromadendrene is usually found in agarwood but is not considered as a marker compound as it does not signify agarwood grade [22]. This study detected aromadendrene only in the naturally infected agarwood R4 (0.157). Benzaldehyde and 4-phenyl-2-butanone were specifically detected in the naturally infected agarwood of R1 (0.465 and 2.697) but absent in all healthy woods. These compounds are rare in healthy woods but are assumed as compulsory compounds that contribute to agarwood fragrance [36]. Amongst all the biological samples, the naturally infected agarwood sample of R4 is suggested to be high-quality agarwood due to the presence of seven out of eleven compounds related to high-quality agarwood compounds.

GC-FID Analysis

From the GC-FID data analysis in Table 3, both artificially inoculated agarwood of M1 and M2 samples gave almost similar compounds composition but differed in compound abundance. The agarwood sample of M1 showed a higher level of agarospirol (4.319) than the sample isolated from the healthy wood (0.86). Comparison of M1 with M2 agarwood samples showed a low agarospirol composition (0.301). Previously, agarospirol was found in artificially inoculated agarwood and healthy wood from A. sinensis [37]. Several studies have reported agarospirol in biologically inoculated agarwood [22]. Meanwhile, α-guaiene was found in both agarwood and healthy samples, with a higher percentage area volume of M1 (0.635) in the agarwood sample than the healthy sample (0.177).

The GC-FID data analysis (Table 3) also shows the presence of benzaldehyde and β-selinene in both artificially inoculated agarwood of M1 (0.879 and 2.293, respectively) and M2 (0.769 and 0.643, respectively) samples. Interestingly, these compounds were not detected in healthy wood. Contrarily, Liu and his co-workers [37] reported the absence of benzaldehyde in agarwood samples inoculated with ten commercial agarwood inoculants. Authors proved that a young A. malaccensis could produce benzaldehyde when induced with fungi [22]. Thus, specific agarwood inducers can induce benzaldehyde formation in healthy trees. Meanwhile, β-selinene has been detected in the biologically induced agarwood of A. malaccensis and natural agarwood [35], though it has also been found in the healthy agarwood tree of A. sinensis. The detection of β-selinene in healthy agarwood trees was influenced by the differences in the species used and the differences in climatic conditions.

Agarwood samples from M1 and M2 also contained 10-epi-γ-eudesmol at higher percentage areas of 3.815 and 0.749. However, the presence of norketoagarofuran (0.416), epoxybulnesene (0.472), and jinkoh eremol (1.38) were only detected in the M1 sample. Compounds like norketoagarofuran, epoxybulnesene, 10-epi-γ-eudesmol, and jinkoh eremol had not been detected previously by GC-FID or GC–MS artificially inoculated agarwood, except for 10-epi-γ-eudesmol [34]. Additionally, M2 exclusively contained 4-phenyl-2-butanone, β-agarofuran, and aromadendrene. Similar compounds have been found in artificially inoculated agarwood sourced from A. sinensis and A. malaccensis, except for β-agarofuran [36], M1 and M2 produced different compound compositions after all the factors, including tree age, species, climate, and inoculant types have been ruled out. This is likely due to the diverse plant responses towards environmental and external stresses [43].

The agarwood quality was observed to be consistent in the artificially inoculated agarwood results after one year of inoculation with Ino A (M1 and M2). Eight of the eleven compounds associated with high agarwood quality were identified in the artificially inoculated agarwood, with compounds like α-guaiene, aromadendrene, norketoagarofuran, and 10-epi-γ-eudesmol being found to be more abundant compared with the best naturally infected agarwood sample from R4. However, the agarwood from R4 had a higher composition of β-agarofuran, β-selinene, agarospirol, and jinkoh eremol than the artificially inoculated agarwood tree maturity may affect compound production in both artificially inoculated and naturally infected agarwood [22]. Artificially induced younger agarwood trees usually produce agarwood close to natural agarwood (as shown by the agarwood samples from M1 and M2). From another perspective, the result may be influenced by the dry and hot conditions of the artificial agarwood plantation. According to Gouinguené and Turlings [44, 45], plants that endure abiotic stress from dry soil may produce more monoterpenes and sesquiterpenes, especially from physical inducement. This means that agarwood compounds produced by agarwood trees in plantations may be influenced by the plant's response to the hot and dry conditions of its habitat, in addition to the physical and biological inducement. Naturally, infected agarwood trees respond more to persistent pathogen infection. Cold and moist environments contribute to the conduciveness of natural habitats for microbial growth. The production of compounds related to high-quality agarwood such as 4-phenyl-2-butanone, β-selinene, α-bulnesene, and agarospirol in R1, R4 and M2; present study is believed to be associated with Fusarium sp. and Polyporales sp. It is also suggested that S. commune interaction might influence the production of β-selinene and β-agarofuran in R4. In this context, it is possible to develop Fusarium sp., Polyporales sp., and S. commune as an artificial agent to inoculate the agarwood.

Conclusion

The current study results demonstrated the formation of volatile agarwood compounds by plant defence mechanism towards fungal attack. This defence mechanism includes the production of resinous compounds (agarwood) as secondary metabolites. Several important compounds related to high-quality agarwood were detected in natural and artificially inoculated agarwood, respectively collected from a natural rainforest habitat and agarwood tree plantation. This study also anticipated that ten groups of fungal isolates associated with the formation of volatile agarwood compounds were identified in both natural and inoculated agarwood using culturing and fungal ITS regions sequencing. Amongst them, the Fusarium sp. was the most dominant group. Other than that, a new fungal isolate of S. commune was also identified as endophytic fungi in agarwood samples. Thus, this fungus should be tested for its endophytic nature and role in forming volatile agarwood compounds. Further research is needed to understand the detailed relationship of these fungal endophytes to Aquilaria species and investigate the economic and pharmaceutical significance of the bioactive compounds from these producers.