Soil samples were collected around the four buildings in Seonamsa temple (Figure 1A). From September 29, 2016 to October 31, 2019, 61 pieces of timber (5 × 5 × 30 cm) were installed at interval of 2 m around the seven buildings, checked and replaced periodically. A sampling point was selected at the location where termites were repeatedly detected in the timbers. The sampling location for each building is shown in Figure 1 (B-E). Termites were observed in the timbers installed around JSD, PSJ and SJ but not in the timbers around DUJ. When the timbers were removed, 50 g of soil samples were taken in triplicates from the Seonamsa temple at a depth of about 30 cm; these were refrigerated for sequencing analysis. For humus soil test, after putting 250 g of humus soil in 1 L culture bottles and adjusting the moisture to 70%, 200 termites (180 workers and 20 soldiers) were added and cultivated in an incubator at 25℃ for four weeks. Each group had five replicates. After four weeks, NGS analysis was performed on the humus soil without termites. The humus soil (Taeheung F&G, Korea) was purchased for gardening, and termite adaptability and microbial diversity were investigated before this test.

Sequencing samples were prepared according to Illumina 16S Metagenomic sequencing library protocols. Forward primer 341F and reverse primer 805R for 16S rRNA amplification were used to amplify the V3-V4 region of the bacterial 16S rRNA. Gene amplification and sequencing were performed by Macrogen Ltd. (Seoul, Korea), and the final results were sequenced using the Illumina MiSeq platform (Illumina Inc., San Diego, CA, USA) (Caporaso et al., 2012). Operational taxonomic unit (OTU) clustering was used in de novo (CD-HIT) OTU picking method. Taxonomic assignment was conducted using the National Center for Biotechnology Information (NCBI) Database (BLAST).

Good’s coverage prediction for all soil samples was >99%, indicating that the number of sequences was satisfactory to reveal the bacterial diversity in each sample. The sequencing of each group was integrated, yielding 71,942 and 72,429 sequence reads from Seonamsa temple’s soil and humus soil samples, respectively. The diversity of bacterial communities in soil samples was predicted by Shannon’s index. In particular, the Shannon’s diversity index of SJ was the highest, followed by PSJ, DUJ, and JSD. In the termite test using humus soil, the Shannon’s diversity index of humus soil with termites was higher than that of humus soil without termites (Table 1).

Figure 2 shows the microbial composition of soil from Seonamsa temple and humus soil samples at the phylum level. A total of 20 different phyla were observed, which represented the overall microbial diversity. On the other hand, about 10 different phyla were observed in humus soil, which represented less diversity than these of Seonamsa temple’s soil. The soil samples from Seonamsa temple were revealed dominant phyla Proteobacteria (27%), Firmicutes (24%), Actinobacteria (21%), Bacteroidetes/ Acidobacteria (5%), and other less abundant phyla such as Chloroflexi/ Cyanobacteria, Gemmatimonadetes, Verrucomicrobia and Nitrospirae (Manjula et al., 2014). The dominant phylum Proteobacteria play a major role in nitrogen fixation in termites. These results are consistent with the occurrence of nitrogen fixing Gammaproteobacteria in termite gut. The second most abundant phylum Firmicutes present in termite gut are well-known cellulose-fermenters and thrive in alkaline environments. The third dominant phylum Actinobacteria was reported to play major roles in degradation of cellulose, lignin, and chitin (Ohkuma and Kudo, 1996; Egert et al., 2003; Husseneder et al., 2010; Grieco et al., 2019). Humus soil test was revealed representative dominant phyla Bacteroidetes (56%) and Proteobacteria (36%). When the compositions of each soil samples were compared, Firmicutes and Actinobacteria predominated the area around JSD, while Proteobacteria and Actinobacteria predominated the areas around PSJ and SJ. Furthermore, Proteobacteria was the most predominant around DUJ, followed by Cyanobacteria. In humus soil experiments, Proteobacteria increased, while Bacteroidetes decreased in humus soil with termites (Figure 3).

In this study, significant differences were observed among the soil microbiotas around the four buildings in the Seonamsa temple. In a principle coordinate analysis (PCA) of soil microbiotas based on OTUs, three soil microbiotas around the JSD, PSJ and SJ buildings were clustered together, while the soil microbiota around the DUJ building was separated. The PCA produced consistent results, showing that the soil microbiotas around JSD, PSJ and SJ with termites were correlated to each other compared to those around DUJ without termites (Figure 4, red circle). Moreover, the PCA of humus soil test showed that the microbiota of humus soil with termites was closer to microbiota of the Seonamsa temple’s soil than those of humus soil without termites (Figure 4, blue circle).