Introduction

Heavy metals are often added as additives to animal feed (Hashemi 2018). These metal elements cannot be fully utilized during the animal growth, which will increase their contaminative levels in environmental soil (Liu et al. 2020a, b). Moreover, heavy metals may be accumulated in plants or meat (Hashemi 2018; Duan et al. 2019), and pose health risks to the human body through the food chain. Many studies have demonstrated the presence of heavy metals in manure and fertilizing soils. The emissions of heavy metals, Cu and Zn, in the discharged 240 million tons animal manures reached 7153 tons in 2010 (Xu et al. 2019a). The concentration of heavy metals in the animal manure compost ranged from 0.01 to 3692 mg/kg, and the amount of Cr reached up to 6603 mg/kg (Yang et al. 2017). Mean concentration of As in chicken manure from several states in USA was 15.7 mg/kg (Jackson et al. 2003). Therefore, for such a large amount of animal manure and high concentration of heavy metals, it is necessary to monitor the content of heavy metals in animal feces to understand the possible pollution.

Many countries, especially the European Union, have banned the use of antibiotics as growth promoter since 2006, but there are still many antibiotics added to feed in countries dominated by animal-produce (Ben et al. 2017; Carlota et al. 2021). Antibiotics are also often used as the animal feed additives in the livestock breeding, such as cattle, swine, and poultry, to prevent animal disease and promote animal growth (Kumar et al. 2005), which cover more than 16 and 80 million kilograms every year in the USA and China, respectively (Ji et al. 2012). Most antibiotics cannot be metabolized completely by animal body, about 30–90% of which are discharged via the urine or feces excretion (Ajit et al. 2006; Zhang et al. 2021). Large amount of livestock and poultry manures are produced every year around the world, introducing a great deal of antibiotics into the environment via land-applied in rural area (Guo et al. 2018). As we all know, the livestock and poultry manure application can bring antibiotics into the soil system and introduce resistant microorganisms and antibiotic resistance genes (ARGs) (Macedo et al. 2020). Therefore, the potential risk of residual antibiotics in manure to farmland ecosystem should be paid attention to.

The land application of animal manures was considered as the primary original source of antibiotics into the environment (Cui et al. 2016). For example, seven antibiotics were remained in swine manure, with the concentration range of prioritized antibiotics (sulfadiazine and doxycycline) being from 55 to 7744.6 μg/kg (Meersche et al. 2019), and the concentration of antibiotics in manure slurry still presented with the amount of 13.7–685.6 μg/L (Chen et al. 2012). Antibiotic residues were detected in various farm plants, including corn, tomatoes, lettuce, and carrot planted on the farm-applied animal manures (Rahube et al. 2014; Mullen et al. 2019). What’s more, antibiotics also induce the emergence and development of ARGs among soil microorganism (Zhao et al. 2021; Liu et al. 2021a). Previous study reported that the transfer of ARGs in manure was accelerated (Ngigi et al. 2019). Sulfonamide and tetracycline ARGs in bovine manures reached 10−6–10−3 (relative concentration, ARGs/16S rRNA) (Lin et al. 2019). Xu et al. (2019b) found that the abundance of 146 ARGs belonging to four antibiotic types decreased during compost, while the half-life and persistence of ARGs increased. The risk induced by antibiotics and ARGs’ spread was a great challenge for the ecosystem and food chain for human health (Rahube and Yost 2012; Zhang et al. 2015). However, few studies investigate the correlation between heavy metals, antibiotics, and ARGs in the manures as well as their potential risk. Therefore, in this study, we will systematically study it.

In order to discover the original source of heavy metals, antibiotics, and ARGs into the farmland, three different manures were collected to investigate the occurrence and the correlation among those contaminants in this study. The aim of this study is to carry out a comprehensive analysis of the potential pollution source of farmland ecosystem—manure—to provide the basis for blocking pollutants from manure into farmland ecosystem, then make better use of animal manure and reduce its harm to people and the environment. These results will provide basic data for the heavy metal, antibiotics, and ARG pollution in land-applied manures, illuminating the original source of antibiotic risk in soil ecosystem.

Materials and methods

Sample collection

Three kinds of manure samples were collected in April 2019 from a swine farm and a chicken farm in Jinshan and Chongming, Shanghai, China, respectively. The swine farm produced 26000 meat pigs a year, with the annual production of swine manure about 20000 tons. The chicken farm produced 48000 laying hens a year, producing chicken manure about 2100 tons per year. The animal manure produced by the two farms was stored in the cement storage tanks. The manure used in this experiment was fresh manure produced within 2 days. The organic fertilizer (OF) was prepared by mixing the swine manure (SM) and the mushroom growing base material through the composting. The moisture contents of SM, CM, and OF were 68.84%, 66.26%, and 42.28%, respectively, in Table 1. At each point, about 5 L of manure was collected in a plastic container. Each sample was performed in three parallels. All samples were shipped to the laboratory in a portable refrigerator and stored in – 80 °C refrigerator for subsequent experimental analysis.

Table 1 Conventional index and heavy metals in different manures. Units: mg/kg
Full size table

Conventional indices and heavy metals of manure

The tested indices including organic matter, total N, NH4+-N, and NO3−-N. Nine heavy metals (Cd, Pb, Cu, Ni, Zn, As, Ag, Hg, and Cr) in manures were determined. After freeze-dried samples, 0.2 g of the sample was placed in a Teflon crucible with a mixture of HF, HNO3, and HClO3 acid to perform a digestion reaction. The solution obtained from the reaction was diluted to 50 mL in a volume bottle, filtered, and stored at 4 °C for further analysis. The concentration of metal ions was quantified by inductively coupled plasma-mass spectrometry (ICP-MS) (Guo et al. 2018). All measurements were performed three times.

Analysis of antibiotics

Chemicals and reagents

Sulfadiazine (SDZ), sulfamethazine (SMZ), sulfamethoxazole (SMX), chlortetracycline (CTC), oxytetracycline (OTC), tetracycline (TC), doxycycline (DOC), TC-D6, SMX-13C6, and CIP-D8 were obtained from Sigma (USA). Norfloxacin (NOR), ciprofloxacin (CIP), ofloxacin (OFL), and enrofloxacin (ENR) were obtained from Dr. Ehrenstorfer GmbH (Augsburg, Germany). The purity of all drugs was more than 98%. Acetonitrile (99.9%) and methanol (99.9%) were purchased from Shanghai Chemical Reagent Company (Shanghai, China). Ultrapure water was prepared using a Milli-Q water system (Millipore, USA). Na2EDTA-McIlvaine buffer solution was prepared by dissolving 12.9 g of citric acid monohydrate, 27.5 g of Na2HPO4, and 41.18 g of Na2EDTA∙2H2O in 1.0 L of ultrapure water, with the pH adjusted to 4.00 ± 0.05 using HCl or NaOH. This solution was prepared daily and stored in a refrigerator until used.

HLB solid-phase extraction columns (500 mg/6 mL) were obtained from ANPEL Laboratory Technology, Shanghai. The residual antibiotic concentration was analyzed by high-performance liquid chromatographic tandem mass spectrometry (HPLC-MS-MS, Agilent, USA).

Experimental operation

Before the experiment began, all samples were freeze-dried for 48 h. These samples were crushed and refined and then weighed 1 g of samples placed them in a 50-mL centrifugal tube. Add 10 mL of 1:1 mixture of methanol and Na2EDTA-McIlvaine buffer to the tube. Next, concussion of the centrifuge tube for 1 min (2500×g per minute), ultrasonicated for 20 min (50 KHz, 300 W), 5000×g centrifugation for 10 min. After the supernatant was collected, the sample was repeat extracted twice. After three consecutive extracts, a total of 30 mL supernatant was obtained. Dilute it to about 300 mL in a 1000-mL glass bottle with ultrapure water to reduce the methanol content to ≤ 5%.

The diluted extraction of manure was concentrated and purified by HLB columns. The SPE solid-phase extraction column was activated by adding 5 mL methanol and 5 mL ultrapure water in turn (natural deposition of methanol and ultrapure water in columns). Then connected the diluent to the columns and adjusted the flow rate to about 5 mL/min. Elution of impurities with 5% methanol-water solution after extracting the solution. Finally, 10 mL methanol was used to elute the enriched antibiotics on the columns and collected in the test tube. Blew the eluent close to dry with nitrogen and fixed the volume with 1 mL of 50% methanol-water. Filter the above solution with a 0.22-μm needle filter and transfer to a brown injection bottle to be stored at – 20 °C, waiting to be detected by HPLC/MS/MS system.

The chromatographic separation of antibiotics was carried out by an Agilent Technologies 1260 HPLC system, which was composed of a binary solvent manager and a sample manager. The separation of compounds was performed using an Agilent Zorbax Eclipse Plus C18 column (1.8 m, 2.1 mm × 100 mm). The detection conditions of liquid chromatography are as follows: mobile phase A: acetonitrile solution containing 0.1% formic acid; mobile phase B: pure water containing 0.1% formic acid; flow rate: 0.4 mL/min; injection volume: 10 μL; chromatographic column: Agilent Zorbax Eclipse Plus C18 and column temperature: 35 °C; gradient eluting program: the eluting gradient began with 85% B, decreased to 75% at 0–5 min, and further decreased to 55% at 5–10 min. Finally, the linearity returned to 85% within 4 min and maintained 1 min, until the next injection (Cui et al. 2019). The detailed process of mass spectrometry refers to the literature previously reported (Cheng et al. 2015).

Analysis of ARGs

DNA extraction

DNA was extracted from 0.5 g of fresh manures with three replicates using the FastDNA Spin Kit for Soil (MP Biomedicals, USA). The steps were carried out according to the instruction provided by the manufacturer. The quality and concentration of the purified DNA were measured by gel electrophoresis and spectrophotometer (NanoDrop ND 1000, Thermo, USA) (Hu et al. 2019). The separated DNA was stored at − 20 °C until qPCR analysis was performed.

Quantification of ARGs (qPCR)

According to previous studies (Liu et al. 2021b; Eric et al. 2020; Wang et al. 2019a), the abundances of thirteen ARGs belonging to five different antibiotic classes (tetA, tetW, tetQ, sul1, sul2, qnrD, qnrS, qnrA, qepA, strA, strB, blaOXA-1, and blaTEM-1) were investigated in this study. The 16S rRNA and intI1 gene were also quantified in each sample using qPCR. The construction of standard plasmid, the establishment of standard curve, and other experimental processes were carried out with reference to the previous reports (Ji et al. 2012; Hu et al. 2018). The establishment of standard curve and the quantification of target ARGs were carried out on LightCycler 480 II instrument. The primer sequence, annealing temperature of each ARG subtypes, and the manipulation of qPCR referenced to Hu et al. (Hu et al. 2019). All samples were tested in triplicate and ddH2O was added as the negative control. More information about the qPCR reactions was in Table S4 and Table S5.

Analysis of community structure

Based on high-throughput sequencing, the distribution characteristics of microbial community structure in different manure were analyzed. The raw data were obtained from Sangon Biotech (Shanghai) Co., Ltd. and paired-end sequenced (2 × 300 bp) using the Illumina HiSeq sequencing platform, then follow-up analysis was made. The V3–V4 hypervariable regions were amplified using the primers 341F (5′-CCTACGGGNGGCWGCAG-3′) and 805R (5′-GACTACHVGGGTATCTAATCC-3′). The distribution of community structure at genus level of different samples was analyzed. USEARCH (version 7.1) was employed to assign cluster Operational Taxonomic Units (OTUs) at the 97% similarity. In brief, setting a window with length of 10 bp, and removing the bases with reads tail mass value below 20 in each sample. The sample files were then merged in a fastq file, the primer sequences were removed using cutadapt (version 1.2.1), and the Prinseq (with parameters -lc_method dust -lc_threshold 40 -min_len 200) was used to filter the quality of each sample data. Using RDP for classification analysis, the database is http://rdp.cme.msu.edu/misc/resources.jsp. Chao1 index, Shannon index, and other indexes were obtained to access the alpha diversity. The raw reads were uploaded into the NCBI Sequence Read Archive (SRA) database (Accession Number: PRJNA703060).

Statistical analysis

Microsoft Excel was used to filter and calculate all raw data. All related graphics were generated using Origin Pro 9.0. The correlation analysis was performed using SPSS, and p values < 0.05 were considered statistically significant.

Results and discussion

Conventional indices of SM, CM, and OF

Table 1 summarized the conventional indices and heavy metals in three different manures (SM, CM, and OF). The amounts of total organic carbon (TOC) in SM and CM were detected with the concentrations being 3.16 × 105 and 4.27 × 105 mg/kg, respectively, which held the same level as those in SM (3.5 × 105 mg/kg) and CM (3.3 × 105 mg/kg) from Guangdong (Zhang et al. 2019). However, the TOC in OF was slightly higher than that in SM, which might be due to the additional material during the composting process, the amounts of NH4+-N (4.77 × 103–1.98 × 104 mg/kg) and NO3−-N (6.13 × 102–5.40 × 103 mg/kg) in SM and CM were also higher than those in the manures from Guangdong (3.50 × 103–1.90 × 104 mg/kg and 103–124 mg/kg) (Zhang et al. 2019). In this study, the concentration of NH4+-N in OF was lower than that in SM, but the content of the NO3−-N was the opposite. This might be due to the addition of mushroom substrate that can increase the great bioactivity of nitrifying microorganism during composting that NH4+-N can be converted to NO3−-N and released NH3 (Meng et al. 2017).

The residual heavy metals in animal manures were also monitored. The concentrations of Cu and Zn in three manures were higher than other heavy metals, with the concentrations 46.5–311 mg/kg and 3.97 × 102–1.82 × 103 mg/kg, respectively, which was greatly higher than the limit values of the compost/organic fertilizer standards from Australia and the EU (70 and 200 mg/kg) (Yang et al. 2017). This result gave a warning of the residual of Cu and Zn. The main reason for the high concentration of Zn and Cu in manure was their misuse in feed (Hasman et al. 2006). For example, the dosage of Cu and Zn in feed from seven provinces and cities in China were 1017–1591 and 7113–8710 mg/kg, respectively (Zhang et al. 2005). Wang et al. (2013) detected 427.2 mg/kg and 132.3 mg/kg Zn, and 113.7 mg/kg and 141.7 mg/kg Cu in pig manure and chicken manure, respectively. The concentrations of other seven heavy metals were between 0.06 and 10.2 mg/kg. Among the nine detected heavy metals, the amounts of As, Cr, Cd, Pb, and Hg were all lower than the limit values of the Standard of Ministry of Agriculture of China (Yang et al. 2017). The level of Hg, Pb, Cd, and Cr in this study (0.06–7.6 mg/kg) were similar to those in a previous study (below 10 mg/kg) (Wang et al. 2013).

As shown in Table 1, the contents of heavy metals in CM were lower than those in SM, especially for the contents of Zn and Cu. This might be due to that the amount of heavy metals added in the feed of chickens and pigs were greatly different (Wang et al. 2013). The concentrations of Zn and Cu detected in a previous report in pig and chicken feeds were 15.9–2041.8 and 0–392.1 mg/kg, and 5.6–296.3 and 3.8–198.7 mg/kg (Wang et al. 2013). The contents of Zn and Cu in OF were lower than that in SM. The morphology of Cu and Zn can be changed during composting. The composting process reduced the activity and mobility of Zn and Cu, resulting lower contents of Zn and Cu in the OF than those in the SM (Li et al. 2019a). The concentration range of Zn, Cu, Cr, Ni, Pb, and Cd in the OF were from 0.07 to 843 mg/kg, which was higher than those in the composition process (0.012 to 6629 mg/kg) (Yang et al. 2017), showing that the heavy metal content of the compost samples in this study was at a lower level.

Community structure of SM, CM, and OF

Livestock and poultry manures contained various animal intestinal microorganisms, which might further impact the soil ecosystem, so the microorganism community structure of these three manures was analyzed. Alpha diversity indexes of samples were summarized in Table 2 and the Venn analysis for bacterial species of samples was presented in Fig. S1. All samples’ coverages were higher than 94%. SM, CM, and OF samples covered 1533, 4222, and 2695 OTUs, and only 81 OTUs were shared in three manures, only accounting for 1.9–5.3% of all manures, indicating that three manures had great difference. The Shannon index order of SM, CM, and OF ranked as SM (5.64) > OF (5.15) > CM (3.54), which indicated the most abundant species composition of SM. The Simpson index order was as follows: CM (0.09) > SM (0.02) = OF (0.02), showing the lowest sample diversity in CM. During composting, undegradable organic matter was degraded and utilized microorganisms, which might lead to higher species abundance in OF than CM (Wang et al. 2019b).

Table 2 Alpha diversity index of manure samples
Full size table

The composition of microorganisms in different manure at genus level was studied in this study. There were significant differences in the diversity of bacterial genus level in Fig. 1. The highest abundance of genera in CM, SM, and OF was Sphingobacterium (44.81%), Arcobacter (6.60%), and Nonomuraea (4.35%), respectively. Sphingobacterium was a common dominant bacterium in animal manure, compost, or manured soil (Wang et al. 2019b; Wang et al. 2018; Dong et al. 2018). Sphingobacterium showed tolerance to high temperature, converted cellulose and lignin into stable humus (Wang et al. 2019b). Arcobacter was reported to be a dominant pathogen in aquaculture (Liu et al. 2020a, b), indicating that the SM should not be land-applied directly, especially for the rice fields. Nonomuraea, the dominant bacteria of OF in this study, was reported to be significantly related to the degradation rate of aromatic C and olefin C during composting (R2 = 0.976, p < 0.05) (R2 = 0.976, p < 0.05) (Jiang et al. 2019). Myroides (8.24%) and Luteimonas (7.56%) were the other two abundant genera in the CM. Comamonas (5.94%), Kurthia (5.87%), Acinetobacter (5.82%), Clostridium sensu stricto (5.21%), and Ruminococcus (5.15%) had higher abundance in SM, indicating that multiple microorganisms survive together in the SM. After composting, nearly 60% genera in OF were unclassified. Acinetobacter was reported to be the dominant genus in compost (Jiang et al. 2019; Tortosa et al. 2017). Figure 2 shows that the abundance of the same bacterial genus in the three manures were significantly different, and most of the species in SM had higher abundance. Table S1 also demonstrated significant differences in bacterial community composition among the three manures. When animal manure is used as fertilizer to farmland, there is concern about the spread of zoonotic bacteria, which may contain pathogenic microorganisms (Bui et al. 2011). Luteimonas was the dominant genus in the soil of chicken manure improvement and been widely detected in food waste, but the detection of it in the soil of non-livestock manure improvement was rarely reported (Wang et al. 2018). Therefore, attention should be paid to whether Luteimonas can be used as indicator bacteria in the soil of chicken manure improvement in the future, which will be of great significance to the pollution sources of livestock and poultry manure in the soil.

Fig. 1
figure 1

Distribution of community structure at genus

Full size image
Fig. 2
figure 2

Heatmap showing the 50 most abundant genera in different manures

Full size image

Occurrence of antibiotics in different manures

Sulfonamides (SAs), tetracyclines (TCs), and quinolones (QNs) are three types of antibiotics frequently used in livestock and poultry industry (Tang et al. 2015; Pan and Chu 2016; Pan and Chu 2017; Louise et al. 2021; Sameer et al. 2020). Figure 3 summarizes the concentration of antibiotics in the three manures. There was a great difference in the residual kinds and concentration of antibiotics in different manures (Qian et al. 2018). The recoveries of the target antibiotics in different manures ranged from 72.82 to 104.70% on average. The detection frequency of the target antibiotics in manures were 33.30%–100.00%. CTC and OTC were detected at a highest detection frequency of 100% in all samples. The total concentrations of detected antibiotics in SM, CM, OF were 2.15–6.61 × 103 μg/kg, 3.53–89.74 μg/kg, and 7.58–79.03 μg/kg, respectively. Eleven antibiotics were detected in SM, ENR was not detected in CM, DOC and QNs were not detected in OF. Generally, the amount of SAs was not so high as that of TCs and QNs. The organic carbon–normalized partition coefficients (Koc) of SAs were lower than TCs and QNs, which indicated that the persistence of SAs in environment was lower than TCs and QNs (Zhang et al. 2015). In addition, SAs are not inclined to be adsorbed into solid-phase medium, so it may be further released into wastewater or leaching (Wegst-Uhrich et al. 2014). Antibiotics, CIP, and OTC, in the SM, were detected with the concentrations being three orders of magnitude higher than other antibiotics, reaching up to milligram per kilogram level (6.61 and 6.53 mg/kg, respectively). However, antibiotics with highest concentrations in the CM and OF were DOC and SDZ, reaching 89.74 and 79.03 μg/kg, respectively.

Fig. 3
figure 3

Concentration of antibiotics in different manure samples

Full size image

Previous studies showed that the concentrations of residual antibiotics in livestock manure were fairly high. For example, the concentration of OTC and CIP in swine manures reached 59.06 and 33.98 mg/kg, respectively (Zhao et al. 2010). In the report of Pan et al. (2011), the maximum concentration of residual antibiotics in swine manure was 764.4 mg/kg (CTC). The OTC, TC, SDZ, and SMZ in manure samples from two chicken farms detected by Ji et al. (2012) were 21.96 and 3.41 mg/kg, 10.31 and 2.45 mg/kg, and 8.62 and 2.21 mg/kg, respectively. In a previous study, the highest residual concentration of common veterinary antibiotics detected in chicken manure was OTC 416.8 mg/kg (Zhang et al. 2015). Furthermore, Qian et al. (2016) detected 3.1–16280 μg/kg of the antibiotic concentration in pig manure-based organic manure. Their highest levels of antibiotics in different manures were much higher than those in this study. Nine fluoroquinolone (FQ) concentrations of antibiotics in various manure, such as horse manure, chicken manure, and swine manure, were detected from 18.20 to 104.55 μg/kg (Gu et al. 2019). The residues of SAs, TCs, and QNs in chicken and swine manure were detected; the concentrations were 3.1–185.9 and 4.6–202.2 μg/kg, respectively (Xie et al. 2016). Zhou et al. (2020) collected the feces of pig, chicken, cattle, and duck, and obtained the concentration of SAs, FQs, and TCs ranging from 0 to 5.66 × 103 μg/kg. Indicating that in recent years, the strict management of antibiotic use in livestock and poultry breeding industry is effective. At least, the concentration of antibiotics in manure from livestock farms in this study was lower than in many previous studies (Zhang et al. 2015; Pan et al. 2011; Zhao et al. 2010).

Abundance of ARGs in different manures

Figure 4 shows the concentrations of ARGs detected in manure samples. Thirteen ARGs showing resistance to five antibiotic classes (tetracyclines, quinolones, sulfonamides, aminoglycosides, and β-lactams) as well as one class I integrase gene intI1 were detected in all three manures. The absolute abundance of 16S rRNA gene was at 1011–1013 copies/g manure. SM had the highest concentration of 16S rRNA, reaching 3.78 × 1013 copies/g. The order of ARGs’ total amount was CM (7.35 × 1011 copies/g) > SM (3.22 × 1010 copies/g) > OF (2.30 × 108 copies/g). The amount of ARGs detected in this study was lower than 3.01 × 108 to 7.18 × 1014 copies/g in a previous report (Wen et al. 2019). The contents of most ARGs in CM were higher than that in SM; this trend was similar to the reported literature (Wang et al. 2016); this might be due to that the high transferability of the ARGs of chicken microbes has led to the popularity of ARGs in chicken manure (Mu et al. 2015). The contents of ARGs in OF were lower than in SM, but the contents of sul2 was higher than that in SM. This might be due to significant changes in bacterial community composition during composting, such as Ureibacillus in OF, which can lead to ARGs enrich during composting (Su et al. 2015; Zhou et al. 2019).

Fig. 4
figure 4

Abundance of ARGs in different manure samples

Full size image

The results showed that sul, tet, and strB were the dominant ARGs in these three manures. Zhou et al. (2016) reported strB, strA, and sul1 genes are the most abundant ARGs of soil samples, so it is worth paying attention to these ARGs that may accumulate a large amount in the soil after manure returned to the field. In this study, sul1 and sul2 were the dominant ARGs in CM; this result was similar to Wang et al. (2018). In addition, Wang et al. (2018) proposed that manure should be effectively treated to reduce sul1 and sul2 before they enter the soil, rather than waiting for them to decay after entering the environment. The amount of sul2 in CM was significantly higher than that in SM and OF. Previous study showed that the content of sul ARGs in chicken manure was obviously higher than those in pig manure (Cheng et al. 2013). Meersche et al. (2019) reported sul2 was a representative gene in swine manure, but the abundance of sul2 in SM in this study was much lower than sul1, which was due to their different diffusion mechanisms. sul1 is usually located in the conserved region of the intI1 gene cassette, but sul2 is usually located in small IncQ family plasmids, so there is more extensive host bacteria and larger space for sul1 to transfer and diffusive than sul2, and thus, the abundance of sul1 is often higher than that of sul2 (Duan et al. 2019; Wang et al. 2019a). Many studies have shown that the presence of intl1 in manure may promote ARG accumulation and persistence through horizontal gene transfer (Li et al. 2017; Qian et al. 2018; Mazhar et al. 2021). In summary, ARG abundance in different animal manure varies greatly, the addition of different antibiotics to the feed consumed by different animals, and the different digestion and absorption in animal body resulting in different levels of ARGs.

Correlation analysis

To explore the impact of common indices and heavy metals to the occurrence of ARGs in manures, the correlation analysis of ARGs among them were analyzed in Table S2. There was a strong correlation between TOC with sul2 and blaTEM-1 (p < 0.01), which was consistent with a previous study (Wang et al. 2016). This might be due to that the biological effect of TOC in the growth of ARG host bacteria, affecting the occurrence of ARGs (Li et al. 2020). In this study, sul1, sul2, strB, and blaTEM-1 all had a significant correlation (p < 0.01) with Zn and Cu. It had been reported that sul1 and sul2 were closely related (p < 0.01) to the contents of Zn and Cu (Guo et al. 2018). Zhou et al. (2016) also reported that the contents of Zn and Cu were significantly correlated with many ARGs (p < 0.01). Usually, the abundance of ARGs could be increased under long-term co-selection pressures of heavy metals with metal resistance genes (MRGs) and mobile genetic elements (MGEs), etc. (Zhou et al. 2016; Li et al. 2019b; Zhao et al. 2019). It was reported that tetracycline antibiotics were positively correlated with heavy metals. This phenomenon occurs because it is easy to chelate with heavy metals (Zhang et al. 2015). However, some reports showed that there was a correlation between the use of antibiotics and the abundance of ARGs, while others reported that there was no significant correlation between antibiotics and ARGs (Zhang et al. 2015; Wang et al. 2016; Guo et al. 2018). It requires more and more in-depth research in the future to explore the reasons.

Conclusions

The occurrence of heavy metals, antibiotics, and ARGs in SM, CM, and OF were studied in this research. The amount of TCs, SAs, and QNs remaining in three manures was low, while two antibiotics (OTC and CIP) in SM should be paid more attention to. The sul1, sul2, and strB were the dominant ARGs in animal manures. The abundance of most ARGs was significantly correlated with the concentration of Zn and Cu (p < 0.01). The Luteimonas in CM might be used as an indicator of the source of manure in the soil. Overall, to reduce the risk of pollution from manure to the farmland soil and even agricultural products, here are two suggestions: the addition of antibiotics and heavy metals in animal feed must be strictly controlled. The other one is that the animal manure should be effectively treated before applying into the soil. Based on this study, the migration of pollution from livestock and poultry manure in farmland soil should be deeply studied further.