Pit No. 1 at the Terracotta Warriors Museum is divided into four areas: excavation area (EX), canopy area (CE), restoration area (RE), and restoration area surface (RS). The excavation area (EX) is the pit in which the Terracotta Warriors archaeological excavations are carried out, in which the soil may come into direct contact with the Terracotta Warriors and other artifacts. The rammed earth walls in the canopy area (CE) effectively supported the huge roof of the trellis, ensuring that it would not collapse, and played an important role in the protection of the Terracotta Warriors. Scaffolding wood filled with green paste mud, with moisture, anticorrosive effect. The restoration area (RE) and the surface of the restoration area (RS) are the areas where the terracotta warriors were restored and exhibited next to the terracotta pits. The samples of RE were collected from the soil about 10 cm below the top soil layer (RS). Gray flocs were found in all four areas, and many flocs accumulated on the restoration area surface. Three parallel samples were collected from each area using sterile spoons and placed in sealed bags. Soil samples were collected from each area at specific locations, times, and temperatures, as listed in Table 1. Some were transported to the laboratory using ice bags, whereas others were transported to the laboratory using dry ice. Those transported in ice bags were stored in a 4°C refrigerator, while those transported with dry ice were stored in a − 80°C refrigerator.

Twelve soil samples were diluted 103 times, and the diluted soil suspension was separately applied to the PDA (Potato dextrose agar medium) medium, with three replicates for each medium of each sample. After 3 days of culture at 28°C, a single fungal colony was selected, inoculated in the new PDA medium, and cultured at 28°C for 3 days again until pure strains were obtained for subsequent experiments. All pure microorganisms were stored at −80°C.

Total microbial genomic DNA was extracted using the DNeasy PowerSoil Kit (QIAGEN, Inc., Netherlands) following the manufacturer’s instructions. The quantity and quality of extracted DNA were measured using a NanoDrop ND-1000 spectrophotometer (Thermo Fisher Scientific, Waltham, MA, United States) and agarose gel electrophoresis, respectively. The extracted microbial DNA was processed to construct metagenome shotgun sequencing libraries with insert sizes of 400 bp using the Illumina TruSeq Nano DNA LT Library Preparation Kit. Each library was sequenced using the Illumina HiSeq X-ten platform (Illumina, United States) using the PE150 strategy at Personal Biotechnology Co., Ltd. (Shanghai, China) (Huang et al., 2021).

The raw sequencing reads were processed to obtain quality-filtered reads for further analysis. Sequencing adapters were removed from sequencing reads using Cutadapt (v1. 2. 1). Low-quality reads were trimmed using a sliding-window algorithm. Sequencing reads were aligned to the host genome using Burrow-Wheeler Aligner (BWA) to remove host contamination. Once quality-filtered reads were obtained, they were assembled de novo to construct a metagenome for each sample using the Iterative De Bruijn graph assembler for sequencing data with highly uneven depth. All coding regions (CDS) of the metagenomic scaffolds longer than 300 bp were predicted using MetaGeneMark. The CDS sequences of all samples were clustered using Cluster Database at High Identity with Tolerance (CD-HIT) with 90% protein sequence identity to obtain a non-redundant gene catalog. The gene abundance in each sample was estimated using soap coverage based on the number of aligned reads. The lowest common ancestor taxonomy of the nonredundant genes was obtained by aligning them against the NCBI NT database using BLASTN (e-value <0. 001). Similarly, functional profiles of non-redundant genes were obtained by annotation against the KEGG, EggNOG, and CAZy databases using the DIAMOND alignment algorithm at Shanghai Personal Biotechnology Co., Ltd. (Shanghai, China) (Huang et al., 2021).

Analysis was performed using GenesCloud, a free online platform for data analysis.¹ QIIME was used to determine the composition and abundance distribution of each sample at each classification level (domain, phylum, class, order, family, genus, and species), integrate the species annotation information of contig sequences with the abundance table in each sample, and obtain the species abundance table of each sample at each classification level. Based on the sequencing results, the composition of the microbial communities in the soil samples was analyzed at the genus level to conclude.

The common and unique functional groups of each sample were analyzed. Based on the composition spectrum of the underlying functional units annotated in each functional database for each sample, the R software was used to calculate and visually present the number of common and unique functional class groups for each sample using Venn diagrams. Principal coordinate analysis (PCoA) was used to analyze the functional beta diversity of each sample. This was done to obtain the spatial distribution characteristics of the sample functional composition and quantify the differences between the samples.

Twelve soil samples of approximately 100 g each were sent to the Institute of Applied Ecology, Chinese Academy of Sciences, for chemical, physical, and bioindicator determination.

The soil texture and structure, water content, trace element content, heavy metal content, organic carbon content, microbial carbon content, phosphorus-related hydrolase, and carbon-related hydrolase were mainly measured as environmental factors for the correlation analysis. Soil pH (water) was determined with a glass electrode (pH 700 Bench Meter, Eutech Instruments) at a ratio of 1:2. 5. Soil total carbon (C), nitrogen (N), and sulfur (S) were measured using an automatic element analyzer (Analyzer vario MICRO cube, Elementar, Germany). The trace metals were extracted using 0. 1 M BaCl2 and determined using Inductively Coupled Plasma Optical Emission Spectrometer (ICP-OES). Microbial biomass carbon(MBC) was obtained using the chloroform fumigation direct extraction method. α-glucosidase activity (α) and β-glucosidase activity (β) were obtained by the p-nitrophenol method. Cellobiohydrolase activity (CBH), xylosidase activity, phosphomonoesterase I (ALP)and II (ACP) were determinedis obtained by the p-nitrophenylphosphate method, and phosphodiesterase (PDEs) by the bisnitrophenylphosphate method.

An environmental factor association analysis heat map was plotted using heat map tools on the GenesCloud platform.² The tool was developed using the pheatmap package (V1. 0. 8), which was slightly modified to improve the layout style. The data were normalized by z-scores. The package uses popular clustering distances and methods implemented in the dist and hclust functions in R. Based on the species annotations of each sample in the database at the genus level and the obtained soil physicochemical property analysis results, the correlation between each sample species and the environment was visually presented using a heatmap.

The antimicrobial effect of allicin volatile gas was determined using a dichotomy plate with culturable fungi that were isolated and purified as the experimental object (Huang et al., 2019). A PDA medium was poured on one side of a split plate. After medium solidification, a fungal coating with a diameter of approximately 8 mm was applied to the culture medium. The air concentration of allicin was 625 μL/L when 50 μL allicin was added to the other side; the volume of the 90 mm culture dish was about 80 mL. An equal amount of sterile distilled water was used as a negative control to observe the inhibitory effect.

During the excavation of the Terracotta Warriors, many copper weapons were unearthed, so there are high levels of copper in the soil, especially in the excavated areas(the average value is 48.760 mg/kg). Macrogenomic sequencing analysis showed that Cupriavidus were found in the excavation area where a large number of copper artifacts had been buried. Therefore, we carried out further analyses of the Cupriavidus alkaliphilus.

For the particular bacterium C. alkaliphilus, which was isolated and obtained from the excavation area and account for 0. 15% of species diversity, even in the excavation areas the percentage reaches 6. 75%. We performed the growth curve of this strain. One ring of activated C. alkaliphilus plaque was inoculated into the 50 mL LB liquid medium, incubated at 37°C, 180 rpm for 16 h, and reserved for use. The sterilized LB liquid medium was divided into 18 100 mL sterilized triangular bottles, each filled with 20 mL of liquid, sealed with air-permeable sealing film, and labeled and numbered; one bottle was labeled as the control and stored at 4°C, and the other 17 bottles were labeled as 0, 2, 4, 6, 8, 10, 12, 14, 16, 18, 20, 22, 24, 26, 28, 30, and 32 h. Each bottle was inoculated with the above-mentioned bacterial seed solution at an inoculum of 10% (V/V), and shake cultivation was performed for 16 h at 37°C and 180 rpm and the corresponding bottles were taken out according to the labels at the corresponding time and then preserved in the refrigerator at 4°C for the measurement of the OD₆₀₀ nm values by photoelectrophoresis turbidimetry. Three parallel groups were established at each time point. The measurements were performed until a plateau was reached.

The LB liquid medium was prepared by adding CuSO₄-5H₂O so that the Cu²⁺ concentration gradient was 5, 7. 5, 10, 20, 25, and 30 mg/L (in terms of pure Cu²⁺), and 17 triangular flasks of each concentration were labeled as 0, 2, 4, 6, 8, 10, 12, 14, 16, 18, 20, 22, 24, 26, 28, 30, and 32 h. In addition, the LB liquid medium without any heavy metal ions was used as the control, and the pre-cultured bacterial solution was inoculated into the above medium at an inoculum of 10% (V/V) and placed on a shaker (180 rpm) at a constant temperature (37°C) for the shake flask culture. The corresponding flasks were taken out at the corresponding times according to their labels and stored in the refrigerator at 4°C. Three parallel groups were established at each time point. Measurements were taken until a plateau was reached.

After sampling the bacterial growth culture solution, the wavelength of the UV–visible spectrophotometer was adjusted to 600 nm. The OD_600nm value of the bacterial solution of C. alkaliphilus at different incubation times was determined, and the average value of each sample was taken thrice. The range of the spectrophotometer was 0. 0000–3. 0000; if the mass concentration of the bacterial suspension was too large, it was diluted at appropriate times, and the incubation time, dilution times, and OD_600nm value were recorded. The incubation time of C. alkaliphilus served as the horizontal coordinate, the average OD_600nm value of the measured bacterial solution was used as the vertical coordinate, and the growth curve of C. alkaliphilus was drawn. Using the Cu²⁺ concentration as the horizontal coordinate and the OD_600nm value as the vertical coordinate, the histogram of the maximum amount of growth of Cu²⁺-C. alkaliphilus was plotted.

In July 2022, we collected three soil samples from the excavation, canopy, restoration area, and restoration area surface in the exhibition hall of the Terra Cotta Warriors Burial Pit No. 1 of the Emperor Qinshihuang’s Mausoleum Site Museum. A total of 12 samples were collected. Following separation and purification, 28 cultivable fungi were identified and classified. Of these, 16 were identified in previous experiments (Qiang et al., 2021) and may be the main pathogenic fungi, as shown in Table 2. The plate photographs and photomicrographs of the fungi isolated from the soil are presented in Figure 2.

The microbial diversity results obtained from the above-mentioned 12 samples were analyzed. The results are shown in Figure 3, where they rank among the top 20 in terms of genus-level abundance. Penicillium, Aspergillus, Lipomyces, Alternaria, and Fusarium accounted for the largest proportion. Among the 20 genera present in the largest proportions, Penicillium, Aspergillus, Alternaria, Fusarium, Talaromyces, and Trichoderma were obtained in isolation and purified. The analysis of the microbial community composition in the 12 soil samples showed that the top ten microorganisms in abundance were all bacteria, as shown in the Figure 3. Among these, Nocardioides, Amycolatopsis, Arthrobacter, Pseudonocardia, and Streptomyces accounted for the largest proportion. The abundance of fungi and bacteria in the soil suggests the potential harm to soil sites and even to the Qin Terracotta Warriors.

Venn plots of the number of common and unique species groups in the collected samples from the four regions were drawn, and PCoA analysis on their beta diversity was performed to quantify the differences between the samples. From Figure 4 shows that the excavation area contains more unique species, whereas the excavation area (EX) and canopy area (CE) contain more of the same species. There were significant differences in species between the excavation area (EX) and surface area (RS) of the restoration area and other groups. The species group differences between the two samples in the canopy area (CE) and restoration area (RE) were relatively small, and the similarity was relatively high.

The KEGG Metabolic Pathway Database is the core database of KEGG. This database systematically categorizes metabolic pathways into seven major categories: metabolism, genetic information processing, environmental information processing, cellular processes, organic systems, human diseases, and drug development. By integrating the annotated the KO database of soil microorganisms with the gene abundance matri, the top three pathways ranked in abundance were found to be related to metabolism, genetic information processing, and cellular processes. The corresponding KO abundance of each protein was obtained, as shown in Figure 5A.

The EggNOG database is managed by the European Molecular Biology Laboratory, which uses the Smith-Waterman comparison algorithm to annotate the functions of orthologous groups of genes. A relative abundance diagram is shown in Figure 5B. The top three are amino acid transportation and metabolism, energy production and conversion, replication, recombination, and restoration area. Using the EggNOG annotation of soil sample microorganisms, the bias and composition of microbial protein production in excavated soil can be analyzed, providing data support for various mechanisms and functions of research.

The CAZy database focuses on carbohydrate-active enzymes that degrade, modify, and generate glycosidic bonds. This is a professional database used to study related enzymes. Based on the similarity of amino acid sequences in protein domains and their corresponding functions, CAZy divides these carbohydrate-active enzymes into six major functional protein modules: glycoside hydrolases, glycosyl transferases, polysaccharide lyases, carbohydrate esterases, auxiliary oxidoreductases, and carbohydrate-binding modules. A relative abundance diagram is shown in Figure 5C. The top three modules were glycoside hydrolase, glycosyltransferase, and carbohydrate-binding modules. By annotating the CAZy database of soil microorganisms in excavation pits, the bias and composition of microbial carbohydrate enzyme systems can be fully analyzed, providing data support for the study of pathways such as microbial degradation and metabolism.

Venn plots of the number of common and unique functional groups in the collected samples from the four regions were drawn, and PCoA analysis on their beta diversity was performed to quantify the differences between the samples, as shown in Figure 6. The excavation area contains a relatively large number of unique species and functional components. The species group differences between the two samples in CE and RE were relatively small, and the similarity was relatively high.

The soil macroelement content was the lowest in the remediation area, but no significant differences among the sampling points were found, as shown in Figure 7A. The soil moisture content, pH, and microbial biomass carbon content in the excavation area were significantly higher than those at the other sampling points, indicating that soil conditions, such as humidity and pH, are more conducive to microbial growth in the area, as shown in Figures 7B,C. The activity of soil carbon-related hydrolase is the highest in the surface sampling area. In particular, α- and β-glucosidases indicate that the hydrolysis of soil organic matter in this area increases, which is not conducive to the stability of soil mass, as shown in Figure 7D. This result was consistent with the metagenomic sequencing results from the CAZy database. Similar to carbon-related hydrolase, the activity of soil phosphorus-related hydrolase was the highest at the surface sampling point, followed by the excavation area, which proves that the hydrolysis of soil organic matter in this area increased, which is not conducive to land stability, as shown in Figure 7E. The proportion of silt particles in the soil at each sampling point was highest, followed by those of sand and clay particles. The proportion of silt particles in the soil of the excavation area was the highest, whereas that of clay particles decreased. The risk of soil erosion in this area was high, as shown in Figure 7F. Available nutrients in the soil are crucial for the growth of soil microorganisms. There was no significant difference in the alkali-hydrolyzed nitrogen content of the soil at each sampling point. The soil of the excavation and canopy area had the highest contents of available phosphorus and potassium, which could provide effective nutrients for microbial activities, as shown in Figure 7G. Due to the alkalinity of the soil, there is a carbonation reaction when adding acid, which increases the determination of soil organic carbon content. The results showed that the total carbon content of the soil was 11. 88–15. 27%, whereas the organic carbon content was only 3. 39–11. 32%. In particular, in the excavation area, the organic carbon content was only 21. 9% of the total carbon content, as shown in Figure 7H.

The following environmental factors most closely related to microbial growth activities were screened: α- glucosidase activity (α) and β-glucosidase activity (β). The correlation between water content (H₂O), soil pH (pH), microbial biomass carbon (MBC), available phosphorus (P), and available potassium (K) was analyzed with the genera of dominant fungi and bacteria, and a heat map was drawn (Figure 8). Soil α-glucosidase activity, cellobiose hydrolase, alkaline phosphomonoesterase, microbial biomass carbon, and organic carbon in the total carbon ratio significantly correlated with the content of dominant fungi and were mostly positively correlated. There was a significant and mostly negative correlation between the soil moisture content and dominant fungal content. Penicillium, Aspergillus, Talaromyces, Filamentomyces, and Phialosimplex differed from the other genera. Soil α-glucosidase activity, available phosphorus, available potassium, water content, microbial biomass carbon, and powder and copper content significantly correlated with the content of dominant bacteria, with most of them being negatively correlated. Soil β-glucosidase activity, phosphodiesterase, sand content, cadmium ion concentration, and selenium ion concentration significantly correlated with the content of dominant bacteria, with most of them being positively correlated. The relationship between Nocardioides and Cupriavidus and soil properties was more specific. The correlation analysis of soil property factors with dominant fungi and bacteria provided data support for analyzing the function of microorganisms in the soil of the Emperor Qinshihuang’s Mausoleum Site Museum and their impact on soil sites.

Dichotomy was used to determine the antifungal effect of allicin volatile gas. Sterile distilled water was used as the negative control. The inhibitory effect was observed when the air concentration of allicin was 625 μL/L. As shown in Figure 9, under normal growth conditions of fungal hyphae in the negative control group, allicin gas had a potent inhibitory effect on the growth of the most culturable fungal hyphae. Allicin gas had a potent inhibitory effect on 12 types of fungi.

The culture of C. alkaliphilus was carried out by traditional shaking cultivation, and its growth was measured by the phototurbidimetric method at OD₆₀₀ nm using an uninoculated pure medium as a negative control. The growth curves and amounts of C. alkaliphilus at different Cu²⁺ concentrations are shown in Figure 10. With an increase in Cu²⁺ concentrations, the acclimatization period of C. alkaliphilus was significantly prolonged, and its growth in the culture containing Cu²⁺ was lower than that in the culture without copper ions.

The mausoleum of Emperor Qin Shi Huang was built more than 2, 000 years ago. The Terra Cotta Warriors were discovered in March 1974, and excavations began in July. The No. 1 Pit has been excavated three times on large scales (Ma et al., 2016). The exhibition hall of the no. 1 Pit is the earliest semi-open exhibition hall built and opened to the public at the Emperor Qinshihuang’s Mausoleum Site Museum. The exhibition hall is a steel-frame building. Apart from the necessary measures to avoid sunlight and rain, there are no other environmental protection measures. Sunlight can directly illuminate the site hall through glass windows, causing significant changes in temperature and humidity in areas exposed to sunlight, may also cause the growth of some photosynthetic microorganisms (Dexing, 2013; Binyaseen, 2022). Owing to the large space in the exhibition hall, effective temperature and humidity controls cannot be implemented. With an increasing number of tourists and the entry of archaeological excavation personnel, many microorganisms and various coarse and fine particles have gathered in the exhibition hall of Pit 1. The aggregation of flocs further accumulates dark sediments on the surface of the earthen site, leading to further microbial contamination. Early scholars conducted targeted research. Zhang et al. speculated that the surface sediment of the site and cultural relics came from clothing fibers worn by many tourists. Li et al. suggested that dust discharge from nearby industries, power plants, and civilian furnaces also adds to the dust from Pit 1. Furthermore, several poplar and willow trees are planted around the South Ring Road of Pit 1 (Li Hua and Du Weisha, 2019). In the canopy and restoration areas, especially on the surface of the restoration area, these flocs can be observed to gather in large quantities, whereas less flocs accumulates in the excavation area, because of its low elevation and state of excavation. Therefore, PCoA revealed that the species composition of the excavation area differed significantly from that of the other three areas.

Using heat maps, the GenesCloud tool was used to visualize the correlation between dominant species and environmental factors in each sample. Because all four areas are located in the same exhibition hall, some of the physical properties of the soil are relatively similar. In our study, α-glucosidase activity (α), β-glucosidase activity (β), the water content, soil pH, microbial biomass carbon, available phosphorus, and available potassium are closely related to microbial growth activities. The relationship between Cupriavidus and soil properties was more specific. Many copper weapons were unearthed during the excavation of the Terra Cotta Warriors; therefore, there was high copper content in the soil, especially in the excavation area. Cupriavidus sp. was identified as a Cupriavidus alkaliphilus in later identification experiments, and it has genes for resistance to nickel, zinc, arsenate, chromate, cobalt, cadmium, and copper, as well as genes for the production of secondary metabolites, such as bacteriocins, non-ribosomal synthesizing peptides, iron carriers, and polyketides (Rojas et al., 2016). A large group of genes essential for resisting and detoxifying several heavy metals has also been identified (Kutralam-Muniasamy and Pérez-Guevara, 2019). Several genes are involved in the resistance to and detoxification of heavy metals in copper-greedy bacteria. Therefore, C. alkaliphilus can potentially harm the metal cultural relics unearthed in the burial pits of the Terra Cotta Warriors.

In our experiments, C. alkaliphilus was cultured using media containing different concentrations of Cu²⁺. When lower Cu²⁺ were present, C. alkaliphilus had a shorter logarithmic period of growth and a relatively faster growth rate. The acclimatization period of C. alkaliphilus appeared to be significantly prolonged with increasing concentrations of Cu²⁺. The growth level of C. alkaliphilus was lower in cultures containing Cu²⁺ stress compared to those without Cu²⁺ ions. The presence of C. alkaliphilus could potentially be harmful to metallic artifacts excavated from the burial pits of the Terracotta Warriors.

Choosing plant-based antifungal agents from actual plants is safer and more environmentally friendly. In recent years, domestic and international research has revealed multiple known and novel structural types of active plant-derived antibacterial ingredients. It is ideal for developing plant-based antibacterial agents and has received increasing attention from scholars. Allicin (C₆H₁₀OS₂) is an organosulfur compound with the scientific name of diallyl thiosulfinate, is the main bioactive component of garlic and exhibits broad-spectrum antibacterial and antifungal activity. In this study, we used the dichotomy method to determine the antifungal effect of volatile allicin gas, which has a potent inhibitory effect on the mycelial growth of most fungi. However, owing to the odor of allicin itself, we still need to further improve this method by reducing the concentration of allicin. However, because allicin has a potent inhibitory effect on the mycelial growth of most fungi, it can be regarded as a reference for the prevention of microbial diseases in soil sites and cultural relics.

In the different areas of the exhibition hall of Pit 1, there are terracotta warriors that are excavated, restored and conserved. Therefore, the dominant fungi found in the earthen ruins may be the cause of microbial damage to the precious cultural relics of the Terra Cotta Warriors, with potential risks. Therefore, fungi discovered at soil sites must be taken seriously by conservationists. This study is of great significance for the protection of the Terra Cotta Warriors and other precious cultural relics in the Emperor Qinshihuang’s Mausoleum Site Museum.