Jinsha Site Museum is in No.2 Jinsha Site Road (30.68091 N, 104.01362 E), Chengdu, Sichuan, China (Figure 6a). The earthen wall of the Ivory pit, formed during excavating in a sacrificial area with a lot of buried ivory, had undergone different degree of deterioration. Triplicate soil samples were randomly taken from the same cultural deposit layer (15C) in the Ivory pit wall with no obvious, mild, moderate and severe deterioration, and referred to as S1, S2, S3 and S4 sample groups, respectively (Figure 6b, Figure 6c). Samples were taken using minimally invasive sampling techniques and aseptic procedures, and transported on ice to the laboratory.
Physicochemical properties analyses
Due to the minimal intervention principle in sampling in an archaeological site, the sample quantities were too low to meet the requirements of routine soil property analyses. We analyzed pH, moisture, electrical conductivity (EC) and soluble salt contents that are considered to play significant roles in the deterioration of earthen sites [1, 3, 8, 39]. The samples were air-dried, crushed and sieved through a ø 1 mm sieve. The determination of pH and EC were made on 1:1 slurry of air-dried soil and water . Moisture was determined by oven drying method . Soluble salt contents were measured using ion chromatography (IC): 2 g sample was suspended into 20 mL deionized water, the suspension was shaken for 30 minutes at 150 rpm, filtered first through medium pore sized filter paper and finally through 0.2 μm pore size syringe membrane filter. Ions in the extracts were determined using a Dionex ICS-3000 ion chromatography system with an anion suppressor, a cation suppressor and a conductometric detector (Dionex Corporation, Sunnyvale, USA). Anions and cations in 25 μL of extract were analyzed using 4 mm×50 mm guard columns AG11-11C and CG12A, respectively, 4 mm×250 mm analytical columns IonPac AS11-HC and IonPac CS12A, respectively, and 20 mmol L-1 sodium hydroxide solution and methanesulfonic acid solution, respectively, at 1 mL min-1. The main elements in the samples were analyzed using scanning electron microscope EVO18 and energy dispersive spectrometer X-MaxN (SEM-EDS) (Carl Zeiss, Jena, Germany). The main minerals in Jinsha earthen site are quartz, feldspar, illite, montmorillonite and chlorite, with SiO2, Al2O3, Fe2O3, K2O, MgO, and CaO as the main chemical components (Additional File 1: Table S6) .
DNA was extracted from 0.5 g fresh sample using Fast DNA® SPIN for Soil Kit (MP BIO Laboratories, California, USA) according to the manufacturer's instructions. The concentration and purity of DNA were checked by electrophoresis in 1.0% agarose gel and NanoDrop spectrophotometer (Thermo Scientific Inc., USA). DNA samples were stored at -20 °C.
16S rRNA and ITS amplicon sequencing
The DNA samples were sequenced at Novogene Bioinformatics Technology, Co., Ltd. (Beijing, China). The V3-V4 regions of 16S rRNA genes were amplified using the primers 341F (5’-CCT AYG GGR BGC ASC AG-3’) and 806R (5’-GG ACT ACN NGG GTA TCT AAT-3’) . The ITS2 region of ITS was amplified using the primers ITS3-2024F (5’-GCA TCG ATG AAG AAC GCA GC-3’) and ITS4-2409R (5’-TCC TCC GCT TAT TGA TAT GC-3’) . The primers included sequencing specific adaptor sequences. Amplification was done in 30 μL reactions with 15 μL of Phusion®High-Fidelity PCR Master Mix (New England Biolabs), 0.2 μM of forward and reverse primers, and approximately 10 ng of template DNA. Thermal cycling consisted of initial denaturation at 98 °C for 1 min, followed by 30 cycles of denaturation at 98 °C for 10 s, annealing at 50 °C for 30 s, and elongation at 72 °C for 30 s, and final elongation at 72 °C for 5 min and cooling at 4 °C.
Amplification was checked by mixing equal volume of 1X loading buffer with SYBR green and PCR product, and subjecting the mixture to electrophoresis in 2% agarose gel. A bright band at 400-500 bp indicated successful amplification. PCR products were purified with GeneJET Gel Extraction Kit (Thermo Scientific). Ion Plus Fragment Library Kit 48 rxns (Thermo Scientific) was used to generate sequencing libraries following manufacturer's recommendations. The quality of the libraries was assessed on the Qubit 2.0 Fluorometer (Thermo Scientific). Finally, the libraries were sequenced on a Thermo Fisher Scientific Ion S5 XL platform and 600 bp single-end reads were generated.
Bioinformatic and statistical analyses
Single-end reads were assigned to samples according to their unique barcodes and primers and barcodes were cut off. Low-quality sequences and reads with ambiguous nucleotides were removed using Cutadapt V1.9.1 . Chimeric reads (sequences with ambiguous base and average base quality score <30) were filtered out using UCHIME v. 18.104.22.168 . Sequences were assigned to operational taxonomic units (OTUs) at ≥ 97% similarity using UPARSE v7.0.1001 . The representative sequences of the 16S rRNA OTUs were assigned to taxa using Silva 132 database and Mothur v1.36.1 [49, 50]. The representative sequences of the ITS OTUs were assigned to taxa using Unite database v7.2 (https://unite.ut.ee/) and QIIME v1.9.1 . Chao1 and Shannon indices were calculated using the phyloseq package . Venn diagrams were done at VennDiagramWeb . Principal component analysis (PCA) was done using CANOCO 5 . Differential taxa at phylum to species levels were identified using linear discriminant analysis coupled with effect size (LEfSe) . To analyze the bacterial and fungal community distribution and their correlation with environmental factors, redundancy analysis (RDA) was carried out using CANOCO 5 . Statistical differences among groups were analyzed by One-Way ANOVA with repeated measures followed by a post hoc least significance difference test (SPSS 17.0) . Differences were taken statistically significant at p < 0.05.
The amplicon sequencing data were deposited into the NCBI Sequence Read Archive (SRA) under the accession numbers SRR9678166-SRR9678177 and SRR9678184-SRR9678195.