The sampling points in this study included two locations: Tiger Ridge (LHL) and South Wall (NCQ). The Tiger Ridge sampling site is located at the Tiger Ridge Dam site of PengGong Village, Pingyao Town, Yuhang District, Hangzhou City, Zhejiang Province (Figure 1a), and the South Wall site is located in Liangzhu Ancient City Site Park, Pingyao Town, Yuhang District, Hangzhou City, Zhejiang Province (Figure 1b).

To isolate microorganisms from the soil samples from the Archaeological Ruins of Liangzhu City, we prepared two kinds of media: malt extract agar (MEA) medium supplemented with 50 mg/mL chloramphenicol to avoid bacterial growth and potato dextrose agar (PDA) medium. The potato dextrose agar (PDA) medium was used for susceptibility testing. Potato dextrose (PD) medium supplemented was used for the degradation testing of protective materials.

(I) Samples for microscopic analysis were obtained using double-sided carbon adhesive tape. In brief, a strip of carbon adhesive tape was gently placed on surfaces at the two sampling locations (Tiger Hill and South Wall). Three samples were taken from different locations at Tiger Ridge and South Wall, resulting in six samples that were finally taken to the laboratory. The collected samples were used for scanning electron microscopy (SEM).

(II) At the same sampling points, we used a sterilization key to collect 500 g soil samples for macro-genome sequencing and isolation of fungi, which were transported to the laboratory in an ice box for subsequent analysis.

(III) Samples for microscopic analysis were obtained using Scotch tape. The collected samples were used for optical microscopy.

Scanning electron microscopy (SEM) was used to observe the microbial samples that adhered to the conductive adhesive and analyze the microbial species on the surface of the sites. The microbial samples were observed by optical microscopy, and the microbial species on the surfaces of the sites were analyzed.

Pure culture of isolate was fostered on a PDA plate, lasting 5 days at 28°C before DNA extraction, which was conducted through CTAB (Abdellaoui et al., 2011). Fungal 28S rRNA gene sequences were augmented by means of LR0R/LR7 primers (Kruys et al., 2015). ITS1-5.8S rRNA-ITS2 fungal gene sequences were augmented by means of ITS1/ITS4 primers. Sequencing of products of PCR were performed by GENEWIZ (Beijing, China), and the National Center for Biotechnology Information (NCBI) made analysis for sequential identity in BLAST program and the GenBank database.

Extraction of the gross genomic DNA was carried out in soil through the MoBio PowerSoil^® DNA Isolation Kit (MO BIO Laboratories, Inc., CA, United States) in compliance with the manufacturer’s protocol.

Bacteria of 16S rRNA gene V4 area and fungi ITS1 area got augmented adopting consensus primers 515F/806R and ITS5–1737F/ITS2-2043R for prokaryotes and fungi, with a distinctive bar code placed on each sample (Lu et al., 2013; Ding et al., 2015). All PCR was conducted by means of Phusion^® High-Fidelity PCR Master Mix with GC Buffer (New England Biolabs, United Kingdom). The reaction mixture of 50 μL was amplified, containing 25 μL of master mix (2X), a 0.5 μM final concentration of forward and reverse primers, 10 ng of template DNA and nuclease-free water to 50 μL. The condition of PCR was set to 98°C for 1 min, and thirty cycles of 10 s at 98°C later, 30 s at 50°C for 16S rRNA genetic augmentation or 55°C for ITS regional augmentation, and 30 s at 72°C, extending 5 min at 72°C in the end. To visually verify the amplification of PCR, a volume equaling 1X the mounting buffer (consisting of SYBR green) and the PCR product was mounted onto a sepharose gel of 2%. A specimen, with an amplifying band from 400 to 450 bp, was selected to carry out further analysis. PCR products from various samples were gathered in equimolar amounts. Afterward, purification of the mixture of PCR product was proceeded by a Qiagen Gel Extraction Kit (Qiagen, Germany).

According to the literature, the protection of cultural heritage has become increasingly important. A basic aspect of protection work is to protect cultural relics in the later stage and avoid redestruction. In these later protection activities, many polymer protection materials are needed, such as acrylamide, pure acrylic emulsion, synthetic resin, epoxy, PS, B72, and ethyl orthosilicate (Cappitelli et al., 2004). The protective material will age and be damaged during use. Thus, we performed an experiment to investigate whether microorganisms are related to the aging damage of the protective material. We used epoxy resin as the protective material in this experiment (Pangallo et al., 2015). We evenly mixed the epoxy resin with the curing agent in a 1:1 ratio. To make a mold, we waited until the mixture solidified and obtained 1 ^∗ 1 cm pieces of the film; we chose the four fungal isolates, NCQ2-2, NCQ2-3, NCQ3-1, and NCQ1-4, as the experimental strains. We mixed the four strains, and because these strains came from the South Wall site, we selected soil samples from the southwest and northwest corner of the South Wall to prepare a bacterial suspension. We used the method of shaking. Three parallel samples were selected for each group, and a control group was established. The sterilized epoxy resin material was added to the culture medium, and culturing was carried out for 30 or 60 days. Finally, we used Fourier transform infrared spectroscopy (FTIR-ATR) (Bruker TENSOR II) to characterize the changes in the material groups (Zafeiropoulos et al., 2003) (Figure 2).

In the protection activities performed at the Archaeological Ruins of Liangzhu City, many protective organic polymer materials such as ethyl orthosilicate were used, giving rise to the question of whether these materials cause damage to the soil ecosystem when they are used. Based on the above discussion, we collected some soil samples to simulate the environmental state of the Archaeological Ruins of Liangzhu City, and we selected ethyl orthosilicate as a protective material and added it to the soil to observe the changes in various properties of the soil over a 1-month cycle. The methodology was as follows: Collect two samples from the northwest and southwest corner of the South Wall of the Archaeological Ruins of Liangzhu City. From the above experiments, it was found that the isolated fungi mainly came from the samples from the South wall. For further analysis, two samples (XB, XN) were obtained from two sampling points on the South wall. There were two sets of experimental samples, with each group containing three parallel samples, and 250 g of soil per sample was placed in the medium. Then, 50 mL of tetraethyl orthosilicate was added, and 250 g of each of XN and XB was used as a control group. A total of 12 samples were examined. The above samples were placed in an incubator in which the temperature and humidity were controlled as much as possible in the local environment. After 1 month of culture, appropriate instruments (PH: Mettler Toledo XMTD-8222; microbial C, N: VarioTOC Analyzer, Elementar, Germany) (Šlepetieneė et al., 2010); H₂O%: Oven; organic matter: electric furnace were used to detect PH, microbial C and N, H₂O%, and organic matter levels and other properties (Figure 3).

Nucleotide Sequence Accession Number The nucleotide sequences of strains have been deposited in the DDBJ/GenBank/EMBL database under the accession numbers MN509056-MN509073 for fungal ITS sequences. The raw sequencing data could be downloaded at the NCBI Sequence Read Archive (SRA) with the study accession number PRJNA573470.

Optical microscopy observations of the adhesive tape samples revealed the presence of fungal structures and algal structures. Algae can consist of one or a few cells, and many cells aggregate into tissue-like structures. The filaments can be branched or unbranched. Some algae are single-celled flagellates, while others aggregate into colonies. A large number of green tissue structures were observed in the LHL-1 and LHL-3 samples by optical microscopy, with the fungal hyphae separated; insect traps were observed in the LHL-2 samples, which were generally formed by fungal hyphae. These structures are exploited by fungi to capture nematodes; a beaded fungal hyphal structure was observed in NCQ-1, NCQ-2, and NCQ-3 samples (Figure 4).

The microscopic samples on the conductive adhesive were observed by scanning electron microscopy to analyze the microbial species present on the surfaces at the sites. The scanning electron microscopy results for LHL-1, LHL-2, and LHL-3 are shown in Figure 5. The structural features of typical fungal hyphae and algal cells were present in the samples. Electron micrographs of the LHL-1 samples showed a large number of hyphae on the surfaces at the site, with a cell structure similar to algae. In the scanning electron microscopy image from the South wall, we can see that there are a lot of algal cells as well as analogs of fungal hyphae. Images from NCQ-1 show the spores of fungi and some analogs of fungal hyphae. In the images from NCQ-2 and NCQ-3, various analogs of algal cells can be observed (Figure 6).

A total of 431,056 fungal community reads were acquired with the filtration of low quality reads, chimeric sequences and trim adapter, bar codes, as well as primer. Ten fungal phyla were present in the three samples (Figure 7A). In different samples, the relevant richness of phyla varied a lot. Among the samples from Tiger Ridge, Basidiomycota is taken as the phylum of highest richness, with a proportion of 97.70% of the Tiger Ridge reads. A summary of the relevant richness of the fungal community at the genus level is shown in Table 1. The genera of fungi varied in samples. Of the 10 richest fungi taxa, Lepidostroma, Penicillium, Aspergillus, Fusarium existed in total specimens, while Lepidostroma proved to be genus with the largest abundance in the Tiger Ridge sample, accounting for between 1.47 and 97.66% of the Tiger Ridge sample, with a mean abundance of 33.85%. Lepidostroma asianum and Penicillium oxalicum were the species with the largest abundance (Table 1).

In the sample of the South Wall, Ascomycota occupied a dominating position of phylum and accounted for 10.93–68.77% of reads of each sample, and the relevant mean abundance reached 47.50%. In different samples, the relative abundance at the genu level was significantly different. Penicillium was the most abundant genus in the South Wall sample, and accounted for 0.42–25.88% of the Tiger Ridge sample, with an average abundance of 9.09%. From the perspective of species, Acrostalagmus luteoalbus, Penicillium oxalicum were the most abundant species (Table 1). Assessment of sample overlapping was conducted to differentiate diversity shared by the 6 samples (Figures 7A,B). There were 177 bacterial operating taxonomic units (OTUs) in total, were shared in the samples (total abundance of 5712 OTUs). In addition, 6 samples shared 45 fungal OTUs from an overall fungal abundance of 324 OTUs. The outcome suggested that the microbial community of various sampling sites overlapped to a certain degree. This was especially distinctive for the fungal community, where Fusarium is the dominant genus, taking up a large proportion in all communities.

Evaluations of bacterial (XB, XN1, XN2) constituent parts (determined through sequencing of 16S rRNA V4 region) were performed with a sequence of amplicon through Illumina HiSeq2500 PE250 for specimens. The acquisition, with a total of 219934 reads of high quality, was achieved after removing READS of inferior quality, chimeras, trim adapter, bar codes and primer by filtration. Gene sequences of the entire16S rRNA were distributed to 36 bacterial phyla. A summary of ten taxa with the largest number at the level of phylum is shown in Table 2. Proteobacteria was phylum with the largest number, taking up 27.08–45.50% of each sample’s reads, and the relevant mean abundance reached 36.29%. Bacteroidetes were the second most numerous phylum, and the relevant mean abundance reached 14.22%. Firmicutes accounted for 20.00% of all reads in the sample XN.2, compared to only 2.78% to 12.08% in the remaining samples. Other major phyla included Cyanobacteria (6.07–11.47%, average of 9.60%), Chloroflexi (0.72–10.44%, average of 3.40%) and Actinobacteria (4.42–10.15%, average of 0.71%) (Figures 7C,D).

The high-throughput sequencing results revealed dominance of Lepidostroma among the six samples, and we therefore attempted to isolate the dominant fungi to study their characteristics. Incubation on potato dextrose agar (PDA) plates at 28°C for 6 days resulted in the growth of some colonies. Nineteen strains were separated from the samples. We then sequenced the ITS1–5.8S rRNA-ITS2 gene regions of nineteen colonies (Table 3). Penicillium oxalicum (NCQ2-2), Fusarium oxysporum (NCQ2-3), Aspergillus aculeatus (NCQ1-4), and Aspergillus nomius (NCQ3-1) were separated by pure culturing (Figure 8), and the results were consistent with the high-throughput results, while Lepidostroma, which showed the highest content, was not cultured successfully.

Internationally, some countries, such as Switzerland, Japan, the United States, and Italy, use epoxy resin as a reinforcing agent (Selwitz, 2011). In China, we also use epoxy resins frequently, so we chose epoxy resin as the experimental material in the present study. Fourier infrared spectroscopy is an analytical method for qualitatively and quantitatively analyzing the epoxy value of epoxy resin by using the characteristic absorption peak of the epoxy group in the near-infrared region. This method can not only easily analyze the epoxy group content of the epoxy resin matrix but can also analyze the functional structures and change the law of epoxy resin during curing (Wang et al., 2004).

Infrared spectroscopy usually divides the spectrum into a high-frequency region and a low-frequency region at 1500 CM^–1. The former is the characteristic frequency region of chemical bonds and functional groups, and the absorption peak is minimal. The latter reflects the whole molecule due to vibration and rotation. The characteristic absorption peak, the infrared spectrum characteristic absorption peak of the microscopic component, is mainly distributed in the low-frequency region. According to the principles of infrared spectroscopy and organic chemistry, epoxy resin mainly includes epoxy groups. The infrared spectrum mainly exhibits absorption peaks: the characteristic peak of epoxy groups occurs at approximately 910 cm^–1, and there are slight differences between different types of equipment and detection methods. In the figure, the left panel shows the results of the reaction performed for 15 days, the right panel shows the results of the reaction performed for 30 days, in the 15-day reaction, red indicates the control group (Supplementary Figure S1A), and the other three colors indicate the reaction groups. All of them exist in the ring of 910 cm-1 (Wang et al., 2015). The characteristic peak of oxygen was blue in the 30-day sample (Supplementary Figure S1B); the other three groups were the experimental group, and the control group exhibited the characteristic peak at 910 cm^–1, while the characteristic peak of the experimental group at approximately 910 cm^–1 was not as obvious. It can be observed that under the action of multiple strains, certain substances produced by bacteria and fungi may affect the properties of the epoxy resin itself, and the specific effects on the organic matter require further research.

According to the literature, Portuguese Rodrigues et al. used tetraethyl orthosilicate and other materials in limestone and measured their penetration depth. The silicone reinforcing materials that have been widely used include alkyl silicate, ethyl silicate, siloxane and so on (Tarasov, 2001). For example, tetraethyl orthosilicate was widely used in the protection of stone cultural relics in 1936. However, the stress damage caused by the hydrophobicity and hydrophilicity of organosilicon is obvious, and this side effect will cause damage to cultural relics. Therefore, we performed this experiment to explore the effect at the Archaeological Ruins of Liangzhu City. According to the data in the figure, after 1 month of cultivation, the moisture contents of the two groups of soil samples containing tetraethyl orthosilicate did not change significantly compared with the control group, the PH value decreased slightly, and the organic matter content of the two groups was low. The amplitude increased, but the microbial biomass changed greatly, in which the amount of microorganisms in the XB group increased, and the amount of microorganisms in the XN group decreased. Compared with the sand, powder and stick contents of the control group, the percentages in the two groups of samples were significantly reduced, and the proportion of sand decreased. It can also be seen that the powder to stick ratio was decreasing. The amplitude was large. The grain size of the soil changed significantly compared to the control group. The sand, powder, and viscosity ratio decreased, and the soil became somewhat coagulated, while that in the control group was loose (Wilson and Wilson, 2014). In the experimental group, the ratio of microbial biomass C (Janzen, 2006), to N increased, and after the use of tetraethyl orthosilicate, the microorganisms became more abundant. Therefore, it was clear that tetraethyl orthosilicate has a great influence on the soil sites. If ethyl orthosilicate is used at earthen sites, the microbial biomass at the soil sites will increase, and the soil size will change, causing the soil to clot (Figure 9).