*2.2. DNA Extraction and Purification*

According to the manufacturer's instructions and previous studies [4,22], total DNA was extracted from samples A, B and C using the DNeasy PowerWater Kit (QIAGEN Sciences, Germantown, MD, USA), the DNeasy PowerSoil Kit (QIAGEN Sciences, Germantown, MD, USA) and the QIAamp DNA Stool Mini Kit (QIAGEN Sciences, Germantown, MD, USA), respectively. A Nanodrop 1000™ spectrophotometer (Thermo Scientific, Wilmington, DE, USA) was used to analyze the DNA quality and concentration.

## *2.3. High-Throughput Quantitative PCR*

Detection of ARGs was performed using the Wafergen Smartchip ultra-high-throughput fluorescent quantitative PCR system according to the method of Zhou [4], by controlling the total DNA concentration of bacteria in different sampling media identically. In this experiment, 296 pairs of primers were set up, including 285 pairs of primers for ARGs, 9 pairs of primers for MGEs, 1 pair of primers for the clinical class 1 integron-integrase gene and 1 pair of 16s rRNA internal reference primers to normalize the abundance of ARGs in different sample types [22]. The PCR mixture was introduced into a microwell chip using a nanoscale multi-sample spotter (MSND) 296 (assays) × 16 (samples) mode, followed by qPCR on the cycler. One no-template control (NTC) with three repetitions was set for each assay in each chip for each sample. The reaction system included 1 × LightCycler® 480 SYBR Green I Master, 500 nM of each primer and 2 ng/μL DNA template. The total

reaction volume was 100 nL. PCR reaction conditions were as follows: predenaturation at 95 ◦C, 10 min, 40 cycles of 95 ◦C for 30 s; 60 ◦C for 30 s; dissolution curves. The qPCR results were analyzed through the qPCR software of the instrument. CT = 31 was set as the detection domain, and two or more of the three replicates were detected, with the deviation < 20%. All 16s rRNA internal reference genes of all samples with concentration ≥10 ng/μL in this project were detected, and NTC was not amplified, which proved that the experimental results were credible. The absolute copy number of each gene was calculated indirectly based on the relative copy number. The formula used for calculation was as follows: absolute copy number of target gene = 16 s rRNA gene absolute copy numbe <sup>×</sup> <sup>2</sup>−<sup>Δ</sup>CT, which means the absolute abundance of each gene was recorded as the product of the relative copy number and the 16S rRNA's absolute copy number, where in ΔCT = CT (gene)-CT (16 s). It is worth noting that CT (gene) and CT (16 s) were from the same sample.
