GL Hu-nongke 1, widely cultivated in Longquan County, Zhejiang Province, China, was used in this study for GL production by the short wood-log cultivation method as described previously (Tong et al., 2020). The wood logs from a local sawtooth oak (Quercus acutissima) were used for GL cultivation. The GL-inoculated wood logs were embedded in the soil of mushroom houses (10 m width × 50 m length), as shown in Figure 1.

The fungal samples were scraped and labeled from Trichoderma colonies on the surface of GL-cultivated soils in two trails, Lanju (N27^°57′55″, E119^°02′31″) and Chengbei (N28^°13′22″, E119 ^°04′34″), in Longquan County, southeast China. Field sampling was conducted in May and August of 2021 and 2022 (Figure 1). Trichoderma strains were isolated and maintained as described previously (Hatvani et al., 2007). All isolates were deposited in the culture collection of the Department of Forest Protection, Zhejiang A&F University, China. The Trichoderma isolates were inoculated on 9 cm-diameter Petri dishes containing potato dextrose agar (PDA; 200 g potato, 20 g glucose, 18 g agar, 1 l H₂O) or cornmeal dextrose agar (CMD; Oxoid, Hampshire, United Kingdom) at 26°C. Morphological characteristics of the colonies were recorded. Microscopic characteristics were observed under a Motic M200 microscope (Motic, China).

Fungal DNA was extracted using the Ezup Column Fungi Genomic DNA Purification Kit (Sangon Biotech, Shanghai, China), following the manufacturer’s instructions. The primers used to amplify the tef-1α and rpb2 genes from Trichoderma were: tef1-F (5′-CATCGAGAAGTTCGAGAAGG-3′), tef1-R (5′-AACTTGCAGGCAATGTGG-3′), rpb2-F (5′-TGGGGWGAYCARAARAAGG-3′), and rpb2-R (5′-CATRATGACSGAATCTTCCTGGT-3′) (Cai and Druzhinina, 2021). The PCR amplification products were separated using electrophoresis in 1.0% (w/v) agarose gels. The purified PCR products were sent to Tsingke Biotechnology Co., Ltd. for Sanger sequencing. The obtained sequences were deposited in the GenBank database at the National Center for Biotechnology Information (NCBI), and their accession numbers are listed in Supplementary Table S1.

Nucleotide quality and contig assembly were performed using Staden Package v.2.0.0b11 (Staden et al., 2000). The DNA sequence dataset was constructed using MEGA v.7.0 (Kumar et al., 2016). The aligned tef1-1α and rpb2 gene sequences of all Trichoderma isolates were used for BLAST interface analysis and alignment in the NCBI database.¹ After identification, the aligned sequences were submitted to the NCBI GenBank to obtain accession numbers. Molecular phylogenetic analysis between these selected fungi with our strains was performed using the maximum-likelihood method. Nodal robustness was tested using the bootstrap method, and phylogenetic robustness was determined using MEGA v.7.0 with 1,000 replications (Kumar et al., 2016). Protocrea pallida CBS299.78 was used as an outgroup to root the tree.

Trichoderma population assays were conducted in the two trails: Lanju and Chengbei. Trichoderma colonies grown on the surface of GL-cultivated soil from three randomly chosen mushroom houses in each trail were collected and counted from 2021 to 2022. The numbers of Trichoderma colonies were individually recorded at the time points in April and August each year. The species affiliation of each Trichoderma colony was determined by molecular identification, as described above.

One representative strain for each of the eight Trichoderma species was selected to test the effect of different temperatures on fungal mycelial growth: T. atroviride LZ043, T. hamatum LZ007, T. harzianum LZ013, T. koningiopsis LZ033, T. pleuroticola LZ004, T. guizhouense LZ056, T. spirale LZ023, and T. virens LZ045. A 5 mm-diameter mycelial plug was inoculated in a PDA plate and cultured separately at 13, 23, 31, and 35°C. Colony diameter was measured every day using a Vernier caliper. Each treatment was replicated three times. The assays were conducted using two biological replicates.

The eight representative strains were used to test the effect of different pH levels on mycelial growth. Before inoculation, the pH values of PDA were individually adjusted to 4.5, 5.5, 6.5, and 7.5 by adding HCl or NaOH solution filtered through a biofilter after sterilization. A 5 mm-diameter mycelial plug was inoculated into the PDA plate and cultured at 26°C. Colony diameter was measured using a Vernier caliper. Each treatment was replicated three times. The assays were conducted using two biological replicates.

The eight representative strains were chosen to test the antagonistic potential of the Trichoderma species using the dual confrontation method. Mycelial agar plugs (5 mm in diameter) of GL and each Trichoderma strain were cut individually from the actively growing front of 7-day-old colonies. Each Trichoderma plug was paired against GL at equal distances opposite to each in 90 mm diameter Petri plates containing 20 ml PDA. The plates were incubated at 26°C in darkness for 7 days. The GL growth were recorded. The experiments were carried out with three replicates. The assays were conducted using two biological replicates.

The fermentation broths from eight representative Trichoderma strains were used to test the effects of non-volatile metabolites from different Trichoderma spp. on GL growth. The Trichoderma fermentation broth was prepared as previously described (Wang et al., 2016). Briefly, a 5 mm-diameter plug from each Trichoderma culture on PDA was inoculated in 500 ml PD broth, followed by 7 days culturing in darkness at 26°C at 180 rpm. Each Trichoderma culture was filtered sequentially through cheesecloth, Whatman filter paper, and a 0.2 μm nitrocellulose filter. The filtrates were added to the melted PDA to a volume of 10%; the same volume of sterile water was used as a control. A GL mycelium plug (5 mm in diameter) was inoculated into the medium and cultured in darkness at 26°C. After 7 d, the diameters of the colonies were measured to test the inhibition ratios. The percent inhibition ratio (IR) was computed by the formula: IR = [(R1-R2) /R1] × 100, Where R1 is radial growth of GL in control; R2 is radial growth of GL in treatment. Each treatment was replicated three times. The assays were conducted using two biological replicates.

The sandwich Petri plate setup described previously was employed to test the antagonistic activity of the volatiles emitted by the representative Trichoderma strains against GL (Li et al., 2018). After inoculating GL and Trichoderma on PDA plates, the GL plate was placed on top of a Trichoderma plate, sealed with three layers of Parafilm, and then incubated at 26°C. A non-inoculated PDA plate was used as the control. GL colony diameters were measured after 5 days. The percent inhibition ratio (IR) was computed by the formula: IR = [(R1-R2) /R1] × 100, Where R1 is radial growth of GL in control; R2 is radial growth of GL in treatment. Each treatment included two biological replicates and was repeated three times.

Solid-phase microextraction coupled with gas chromatography–tandem mass spectrometry (SPME–GC–MS) was used to determine the composition of the volatiles produced by the representative Trichoderma strains (Stoppacher et al., 2010). The detailed procedure and conditions have been described previously (Rao et al., 2022). Trichoderma was cultured on PDA and incubated at 26°C for 5 days. SPME inserted into the injection port of a GC 2010 gas chromatograph was used to collect the volatiles. SPME fiber was exposed to the vapor phase above strain LZ42 for 45 min in a culture tube. A Hewlett Packard 7890GC/5975MSD gas chromatograph (Agilent Technologies, Santa Clara, CA, United States) equipped with an HP-5MS capillary column was used. The derived data were analyzed and identified based on a comparison with the mass spectrum in the NIST08.L data bank (National Institute of Standards and Technology). Each experiment was conducted in triplicates. Pure volatile 6-pentyl-2H-pyran-2-one (6-PP, Sigma-Aldrich, St. Louis, MO, United States) was used as the standard.

The antifungal activity of volatile 6-PP on GL was tested by fumigation in an isolated container, as described previously (Wu et al., 2020). The EC₅₀ value of 6-PP on GL was calculated as the 50% effective concentration that inhibits GL mycelial growth using the IBM SPSS Statistics v.19 program (SPSS, Inc., Chicago, IL, United States). A 5 mm-diameter mycelial plug was placed in the center of the PDA. Meanwhile, the 6-PP aliquots of 1, 2, 4, 8, 16, and 32 μl were individually loaded on the filter paper (60 mm in diameter) before incubation, resulting in 0.01, 0.03, 0.05, 0.11, 0.21, and 0.43 μl mL⁻¹ headspace, respectively. Subsequently, the loaded plates were immediately sealed with Parafilm and incubated at 26°C for 5 days. The assays were conducted in triplicate, and plates with filter paper alone were used as controls.

The obtained data were analyzed using the IBM SPSS Statistics 20.0 (SPSS Inc.). Statistical significance was identified at 95% confidence interval (p < 0.05) in all tests.

During the investigation period, Trichoderma colonies began appearing on the soil surface of cultivation ridges in the mushroom houses in April 2021 when the sprouts of GL fruiting bodies developed (Figure 1). From the GL mushroom houses, 701 and 685 Trichoderma isolates were collected after 2-year GL cultivation in the Lanju and Chengbei trails, respectively. These isolates presented diverse morphological characteristics (data not shown). A phylogenetic tree of the 73 representative Trichoderma isolates with different morphological characters were constructed using the maximum-likelihood method based on the combination of tef-1α and rpb2 sequences (Figure 2). The combined sequence matrix contained 2,271 bp (1,118 for tef1-α and 1,031 for rpb2). Eight Trichoderma species were identified among these isolates with those that were previously documented as threshold of species: T. atrioviride, T. guizhouense, T. hamatum, T. harzianum, T. koningiopsis, T. pleuroticola, T. spirale, and T. virens.

Trichoderma propagation on the surfaces of GL-cultivated soils was analyzed individually in the Lanju and Chengbei trails since April 2021, based on colony sampling and identification of Trichoderma isolates. The colonies of Trichoderma spp. in both trails started appearing in April 2021 and increased significantly thereafter (Figure 3). In the Lanju trail, the average number of Trichoderma colonies in each house was 21 in April 2021 but increased by 2.8-fold in August 2021 and 5.8-fold in April 2022. The average number of Trichoderma colonies reached 214 in August 2022, an increase of 9.2-fold. Similarly, dynamic changes in Trichoderma colonies were observed in the Chengbei trail.

Moreover, population and proportion changes of the eight Trichoderma species in both trails were obtained (Figures 3A,B). Significant differences were observed in Trichoderma community during GL cultivation. T. virens was the most dominant species at the end of GL cultivation, accounting for 32.50 and 33.33% of the community in the Lanju and Chengbei trails, respectively. T. harzianum accounted for 20.37 and 14.10% of the community in the Lanju and Chengbei trails, respectively. T. pleuroticola was found in the lowest amount, accounting for 3.89 and 4.65% of the community in the Lanju and Chengbei trails, respectively.

The strains from the eight Trichoderma spp. showed different growth rates at different temperatures (Figure 4). T. hamatum had the highest radium length at 15°C within 7 days post-inoculation, whereas T. pleuroticola had the lowest radium length, followed by T. virens. All Trichoderma species grew rapidly at 23°C and 31°C. The growth rate of T. pleuroticola at 31°C was higher than that at 23°C. T. harzianum grew rapidly compared to other species at 38°C. The eight Trichoderma species grew normally at pH 4.5–7.5 (Figure 5). Among these species, T. koningiopsis exhibited the lowest growth rate at pH 4.5. Overall, the eight Trichoderma species isolated from GL-cultivated soils could survive over a broad range of environmental temperatures and pH values.

In dual confrontation assays, antagonistic zones were observed in the interactions between the eight representative isolates of different Trichoderma species and GL on PDA plates, which indicated that all eight species exhibited antifungal activities against GL (Figure 6). T. atroviride, T. harzianum and T. virens inhibited heavily GL mycelium growth. The growth of GL was also inhibited by T. guizhouense, T. hamatum, or T. koningiopsis. T. pleuroticola and T. spirale showed less inhibitive activity, but overgrew and spread on GL mycelia. Moreover, the inhibitory effects of non-volatile and volatile metabolites on GL varied among different species (Figure 7). The non-volatile metabolites from T. virens showed an inhibition rate of 85.42% for GL, exhibiting the strongest antagonistic activity against GL. T. guizhouense, T. harzianum, and T. koningiopsis showed low activities, accounting for inhibition rates of 67.62, 66.07, and 64.29%, respectively. Meanwhile, the volatiles from T. atroviride inhibited the mycelial growth of GL at the highest level (70.29%), followed by T. harzianum and T. koningiopsis. The inhibition rate of T. hamatum was the lowest (17.4%), followed by T. spirale.

In this study, the pure chemical 6-PP exhibited an inhibitory effect on GL mycelial growth (Table 1). The EC₅₀ of 6-PP against GL was 8.79 μl mL⁻¹. Moreover, the SPME–GC–MS analysis revealed that all the representative strains for eight Trichoderma species produced 6-PP, but with different levels of production (Figure 8). The differences in the peak areas showed that T. atroviride produced the largest amount of 6-PP, followed by T. guizhouense and T. harzianum. The production of 6-PP by T. pleuroticola was the lowest.