The fungicide azoxystrobin perturbs the gut microbiota community and enriches antibiotic resistance genes in Enchytraeus crypticus

Environment International 131 (2019) 104965

Contents lists available at ScienceDirect

Environment International journal homepage: www.elsevier.com/locate/envint

The fungicide azoxystrobin perturbs the gut microbiota community and enriches antibiotic resistance genes in Enchytraeus crypticus

T

Qi Zhanga, Dong Zhub,c, Jing Dingc,d, Fei Zhengb,c, Shuyidan Zhoub,c, Tao Lua, Yong-Guan Zhub,d, ⁎ Haifeng Qiana,e, a

College of Environment, Zhejiang University of Technology, Hangzhou 310032, PR China Key Laboratory of Urban Environment and Health, Institute of Urban Environment, Chinese Academy of Sciences, Xiamen 361021, PR China c University of the Chinese Academy of Sciences, Beijing 100049, PR China d State Key Laboratory of Urban and Regional Ecology, Research Center for Eco-environmental Sciences, Chinese Academy of Sciences, Beijing 100085, PR China e Xinjiang Institute of Ecology and Geography, Chinese Academy of Sciences, Urumqi 830011, PR China b

A R T I C LE I N FO

A B S T R A C T

Handling Editor: Frederic Coulon

The use of pesticides to ensure global food security is the most important pest control strategy in modern agriculture but causes extensive soil pollution. This pollution involves potential risks to human health and ecosystems. In addition to soil animal growth, the adverse impact of pesticides on the gut microbiomes of nontarget soil fauna remains largely unknown. Here, the effect of the fungicide azoxystrobin (AZ) on soil and the gut microbiota of soil animals (Enchytraeus crypticus) was studied. The tested concentrations of AZ altered the bacterial community in the soil and E. crypticus gut and were slightly toxic with respect to E. crypticus adult mortality and reproduction. The most abundant bacterial phylum, Proteobacteria, significantly increased in response to the 2 and 5 mg/kg AZ treatments, which implied a disordered unhealthy gut bacterial community. Furthermore, bacterial community analysis between the soil and gut showed that the main effect of AZ on the gut microbiota was directly through AZ, not soil microbiota. In addition, AZ exposure significantly enhanced the number and total abundance of antibiotic resistance genes (ARGs) in the E. crypticus gut; these genes may enter the soil food web to affect higher trophic levels and cause a more serious ecological risk. Our study reported the effect of pesticides on the gut of soil animals and on the enrichment of ARGs as global emerging contaminants, revealing unknown potential impacts of fungicides on ecosystem services and sustainable food production.

Keywords: Gut microbiota Antibiotic resistance genes Azoxystrobin Enchytraeus Fungicide

1. Introduction Global food production could fall by as much as 50% without pesticides in modern agricultural production worldwide (Oerke et al., 2004); however, pesticide pollution has become a global concern due to overuse. Generally, the soil ecosystem is the sink of pesticide contamination, and residual pesticides can cause a serious threat to soil fauna that have key functions, from nutrient cycling to energy transfer in the soil ecosystem (Lepage et al., 2006). Fungicides are one of the most commonly used pesticides, and their product registrations have increased more than other types of pesticides in recent years in China (Zhu et al., 2012). In particular, the strobilurin fungicide azoxystrobin (AZ) is the highest selling fungicide, with $1.165 billion in global sales in 2016 (http://cn.agropages.com/, accessed on October 19th, 2017); this fungicide can control four major classes of pathogenic fungi by blocking the transfer of electrons between cytochromes b and c by

⁎

binding to cytochrome b to attenuate mitochondrial respiratory function and inhibiting energy production, ultimately leading to the death of the target microorganism (Bartlett et al., 2002). AZ-containing formulations are extensively used as a foliar fungicide for gutter spraying (Quadris and Uniform) or for seed treatment (Dynasty). AZ can persist in soil for up to 6 months based on average soil affinity according to the predicted environmental distribution (PED) (Bending et al., 2006; Bending et al., 2007). AZ has strong toxicity against single eukaryotic species (Du et al., 2019; Lu et al., 2018b) but displays low toxicity against mammals, birds, bees and other nontarget terrestrial organisms (e.g., enchytraeids, collembolans and earthworms), and the sublethal effects of AZ on terrestrial organism reproduction have been calculated for species such as Folsomia candida (half maximal effective concentration (EC50): 57.9–126.1 mg/ kg), Eisenia andrei (EC50: 23.2–60.8 mg/ kg) and Enchytraeus crypticus (EC50: 73.3–125.7 mg/ kg) (Bartlett et al., 2010; Leitao et al., 2014).

Corresponding author at: College of Environment, Zhejiang University of Technology, Hangzhou 310032, PR China. E-mail address: [email protected] (H. Qian).

https://doi.org/10.1016/j.envint.2019.104965 Received 30 March 2019; Received in revised form 25 June 2019; Accepted 25 June 2019 0160-4120/ © 2019 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/BY-NC-ND/4.0/).

Environment International 131 (2019) 104965

B

C

15

10

a a

a

a

a

a

0 mg/kg mg/kg

5

mg/kg mg/kg

0

Total larva numbers

A

Total adult numbers

Q. Zhang, et al.

mg/kg mg/kg

400 300

a

a a

a

a

a

a a

200 100 0

Shannon index

D

11

a 10

ac

ad b

9 8

0

m g/ kg 0. 1 m g/ kg 0. 3 m g/ kg 0. 8 m g/ kg 2 m g/ kg 5 m g/ kg

7

Fig. 1. The total adult (A) and larval (B) numbers of E. crypticus after exposure to azoxystrobin at different concentrations for 28 days. Different letters indicate significant differences (P < 0.05) between different groups. C: Box plots of the Shannon index from the E. crypticus gut microbiota samples and changes following azoxystrobin exposure. D: Principal coordinate analysis (PCoA) of E. crypticus gut samples using weighted Unifrac distances. Different letters indicate significant differences (P < 0.05, one-way ANOVA) between different groups (mean ± SE, n = 4).

2. Materials and methods

As soil microorganisms affect plant growth, gut microbiota play an important role in host health, including immunity, metabolism and aging (Erkosar and Leulier, 2014; Hachung et al., 2012; Lu et al., 2018a; Motta et al., 2018; Zhan et al., 2018; Zhu et al., 2018a). Soil pollutants, such as glyphosate (herbicide), can interfere with the gut microbiome of bees and reduce host immunity to the pathogen Serratia marcescens (Motta et al., 2018). The reproduction and weight of soil collembolans decreased significantly when their gut microbiota was disturbed by antibiotics (Zhu et al., 2018a). However, few studies have reported the effect of soil pesticide pollution on the gut microbiota of soil fauna. In addition, pesticides in soil combined with other pollutants could produce more serious ecological risks, e.g., antibiotics can increase bioavailability in rat guts (Zhan et al., 2018). Due to the overuse of antibiotics, microorganism community was changed and antibiotic resistance genes (ARGs) have been found in the worldwide (Chen et al., 2019; Martinez, 2009; Lu et al., 2019; Zhu et al., 2017; Y. Zhu et al., 2018, Y.G. Zhu et al., 2018). ARGs are now considered a new soil pollutant, and they can be transferred between bacterial communities via numerous mobile genetic elements (MGEs) (Elena et al., 2015; Teng et al., 2014). Some rare earth oxide nanoparticles can selectively enrich the ARGs in the soil environment (Qi et al. 2019), e.g., graphene (nanoscale) significantly affected the diversity and abundance of resistance genes in the mouse gut (Xie et al., 2016), and antibiotics increased the diversity and abundance of ARGs in the Folsomia candida gut (Zhu et al., 2018a; Zhang et al., 2019). However, the dynamics of ARGs in soil and soil animal gut microbiomes after AZ exposure remain unclarified. In the current study, we selected the representative soil animal E. crypticus as a test model, which plays a key role in reshaping the terrestrial ecosystem through nutrient turnover, organic matter decomposition and soil structure (Vliet et al., 1995). This study aimed to provide new insight into the ecological risk of fungicide related to nontarget soil species, as well as the fungicide contribution to ARG diffusion.

2.1. Test animal and fungicides The model E. crypticus used originated from Aarhus University in Denmark and has been cultured in our laboratory for over two years following the OECD (Organization for Economic Cooperation and Development) guidelines. Ten sexually mature E. crypticus of similar length were randomly selected from the stock culture and transferred to a glass beaker. AZ was purchased from Aladdin (Shanghai, China), and stock solutions (1 g L−1) were diluted in methanol/water (1:5). The final concentration of methanol was 0.003 mg/kg in the soil of the treatment and control groups, and this amount of methanol was assumed to have a negligible influence on soil E. crypticus growth. In the current experiment, six concentrations of AZ in the soil-fungicide mixture were used: 0, 0.1, 0.3, 0.8, 2, and 5 mg AZ kg−1 dry soil, which were close to predicted environmental concentrations (0.01–11 mg AZ kg−1 dry soil) (Adetutu et al., 2010; Gajbhiye et al. 2011; Kim et al. 2017; Liu et al. 2010). Five replicates of each concentration were prepared; each replicate was covered with 30 g soil (clay loam, waterholding capacity (WHC) 46.8%, pH (CaCl2) 4.76, cation exchange capacity (CEC) 13.86 cmol kg−1, organic matter (OM) content 24.6 g kg−1, and total N content 3.8 g kg−1) from Ningbo, Zhejiang, China, which was not contaminated with AZ; ten E. crypticus individuals were stored in a thermostatic box (Safe Co., Ningbo, China) at 18.5 ± 2 °C and 75% relative humidity under cool-white fluorescent light (800 Lux) with a 16:8 h dark: light diurnal regime for 28 days. To maintain the starting weight of the container, we added distilled water twice a week.

2.2. Measurement of E. crypticus reproduction, survival data and AZ residue After 28 d of AZ exposure, the surviving E. crypticus individuals were recorded, and some of them were stored in pure ethanol at −20 °C for 2

Environment International 131 (2019) 104965

Q. Zhang, et al.

Relative abundance (%)

100

Other Bacteroidetes Verrucomicrobia Chloroflexi Planctomycetes

50

Firmicutes Acidobacteria Actinobacteria Proteobacteria

a

20

a

15

c

5

5

m g/ kg

m g/ kg

m g/ kg

2

m g/ kg

a 1

ac

ac

c

m g/ kg 5

m g/ kg 2

m g/ kg

0. 8

m g/ kg

0 m g/ kg

m g/ kg 0. 8 m g/ kg 2 m g/ kg 5 m g/ kg

0. 1

m g/ kg

0

b 2

0. 1

a

ab

3

m g/ kg

a

a 1

0

m g/ kg 5

b

4

0

b

0. 8

m g/ kg

0. 1

m g/ kg

0

Relative abundance (%)

2

b

Bacteroidetes

b

m g/ kg

Relative abundance (%)

3

b

b

10

0

m g/ kg 5

m g/ kg

m g/ kg 2

0. 8

m g/ kg

0. 1

0

m g/ kg

0

25

0. 3

5

5

m g/ kg 2

m g/ kg

0 m g/ kg

Relative abundance (%)

ac ac ac

10

m g/ kg

Relative abundance (%)

a

m g/ kg

m g/ kg

m g/ kg

2

m g/ kg

b 1

0. 8

m g/ kg 5

m g/ kg 2

m g/ kg

0. 8

m g/ kg

0. 3

0. 1

0

m g/ kg

0

a

a

m g/ kg

1

a

Firmicutes

ab ab

Verrucomicrobia

a

2

0. 1

2

Relative abundance (%)

ab

a

m g/ kg

ab ab

3

0

b

a

3

m g/ kg

Relative abundance (%)

5 4

15

Chloroflexi

Planctomycetes ab

0. 8

0

m g/ kg

0

m g/ kg

m g/ kg

5

5

m g/ kg

2

m g/ kg

0. 8

0. 3

0

0. 1

m g/ kg

0

a

a

a

a

0. 3

20

a

10

m g/ kg

40

a

5 mg/kg

Actinobacteria

15

0. 1

b

Relative abundance (%)

c

a

a

a

m g/ kg

Relative abundance (%)

c

60

2 mg/kg

Actinobacteria

Proteobacteria 100 80

0.8 mg/kg

0. 3

0.3 mg/kg

0. 3

0.1 mg/kg

0. 3

0 mg/kg

0. 3

0

Fig. 2. Changes in gut microbiota composition following azoxystrobin exposure of E. crypticus with established gut communities. Relative abundance (> 1%) of gut microbiota after azoxystrobin exposure at phylum levels. Different letters indicate significant differences (P < 0.05, one-way ANOVA) between different groups (mean ± SE, n = 4).

cyclohexane (1:1) at a 1.5-ml/min rate. The eluates were blown dry in air at 40 °C and dissolved in 2 mL acetonitrile. AZ was finally quantified by HPLC using a 225-nm UV detection wavelength at 25 °C with acetonitrile: water (75:25) as the mobile phase and a 1-mL min−1 flow rate.

E. crypticus DNA extraction. To determine E. crypticus reproduction (larval numbers), all samples were fixed with 10 mL of 96% ethanol for 2 min and were then transferred to a plastic container containing 100 mL water mixed with 300 μL of Bengal rose solution (1% in ethanol) after which they were stirred vigorously for 10 s and incubated overnight at 4 °C to achieve optimal staining of the animals. Subsequently, these samples were sieved through a 150 mesh to isolate the larvae, and surplus E. crypticus adults were transferred to a white tray (80 × 50 cm2) to optimize counting of the pink-stained enchytraeids under a magnifying glass. The residual AZ concentration in E. crypticus was determined by solid-phase extraction-HPLC (High Performance Liquid Chromatography) according to the following method. Six worms were crushed by a tissue crusher in the presence of 10 mL of ethyl acetate-cyclohexane (1:1). Then, 10 mL of the tissue mixture was ultrasonically extracted for 10 min and filtered three times through anhydrous sodium sulfate. Subsequently, 2 mL chromatographic grade acetonitrile was added to dissolve the extracted AZ before passing through a solid-phase extraction apparatus WondaSep C18 column, and then the column was eluted with 4 mL of acetate-

2.3. DNA extraction from E. crypticus gut and surrounding soil Two E. crypticus from the control or treated groups were immersed in 2% sodium hypochlorite solution for 10 s for sterilization purposes and were then washed with sterile phosphate-buffered saline solution four times. A FastDNA Spin Kit for soil (MP Biomedicals, Illkirch, France) was used to isolate soil and E. crypticus DNA and E. crypticus whole-body DNA from each glass beaker. E. crypticus and soil DNA extraction steps were described by Zhu et al. (2018c). After DNA extraction, the quality and concentration of the isolated DNA were evaluated by 0.8% agarose gel electrophoresis and spectrophotometric analysis (Nanodrop ND-1000, Thermo Fisher) (Qian et al., 2007).

3

Environment International 131 (2019) 104965

Q. Zhang, et al.

0 mg/kg

0.1 mg/kg

0.3 mg/kg

0.8 mg/kg

2 mg/kg

5 mg/kg

8.67

16.22

5.99

15.43

3.83

3.37

3.27

2.37

8.64

8.78

19.46

6.94

5.74

4.40

7.69

5.41

4.19

12.99

13.93

17.77

11.66

22.62

19.93

37.19

f__Xanthomonadaceae_OTU_616909

6.24

4.04

7.48

5.42

6.67

5.59

6.65

7.61

4.51

6.60

4.70

14.42

10.58

7.90

8.19

14.59

7.44

10.01

9.80

10.41

7.58

9.86

8.38

13.36

Burkholderia

11.07

5.69

16.60

8.52

8.58

6.45

13.88

9.01

13.81

14.68

4.98

10.74

15.41

14.78

13.59

17.93

12.14

16.46

17.82

13.64

6.70

9.92

8.16

7.79

Kaistobacter

4.20

4.15

1.56

2.06

1.34

1.53

1.71

1.19

3.82

3.89

2.90

2.30

1.27

3.17

3.80

2.77

2.54

3.54

4.14

2.56

4.13

6.16

4.81

3.60

Actinoallomurus

1.47

0.71

0.56

0.92

1.06

0.38

0.40

0.44

0.67

2.12

0.63

6.00

1.20

0.96

1.43

2.26

0.97

2.57

3.06

2.53

2.14

7.24

3.07

3.95

f__Burkholderiaceae;Other

1.88

1.74

2.38

2.48

0.48

0.60

0.73

0.57

1.42

1.88

1.62

1.46

1.03

1.17

2.08

1.15

1.13

2.58

2.60

1.42

2.47

3.31

3.19

3.39

f__Bradyrhizobiaceae_OTU_635366

4.01

3.27

5.01

4.24

2.96

3.06

4.03

3.03

4.47

3.42

2.65

2.33

2.97

3.58

3.13

3.21

3.20

3.51

4.81

3.05

1.98

2.69

2.99

1.81

Rhodoplanes

2.84

1.98

3.05

0.89

3.42

3.74

1.89

3.97

4.43

2.78

3.48

2.91

1.46

2.78

2.51

1.95

2.12

1.38

1.92

1.90

4.94

1.60

2.22

0.58

f__Koribacteraceae_OTU_429733

0.07

0.84

0.02

0.05

0.38

0.59

0.39

0.96

0.09

0.65

0.17

1.59

0.02

0.04

0.08

0.02

0.14

0.08

0.03

0.03

7.65

0.18

0.64

0.36

Flavobacterium

2.59

1.95

3.19

0.94

2.79

2.43

2.33

2.92

4.26

2.61

3.45

2.52

1.52

2.44

2.05

1.59

1.78

1.23

1.59

2.06

3.90

1.41

2.35

0.63

Candidatus

2.18

2.19

1.58

1.23

1.71

2.32

4.34

2.82

2.10

3.55

3.92

2.71

4.99

2.73

3.86

3.42

3.65

1.91

1.61

5.12

2.24

2.06

2.14

1.27

Other

2.12

2.97

1.43

0.99

1.39

1.24

1.27

1.56

1.41

1.50

2.43

1.37

1.23

1.40

1.29

1.75

1.84

2.81

1.80

2.63

1.26

2.12

2.23

1.49

o__Actinomycetales_OTU_464222

3.08

1.96

3.12

2.41

2.10

1.92

2.79

2.31

2.57

2.76

2.45

1.94

1.98

2.50

2.57

1.91

2.34

2.89

2.56

2.16

1.65

1.75

2.02

1.53

f__Rhodospirillaceae_OTU_306381

2.24

1.04

1.70

0.50

2.11

1.78

1.60

2.58

2.31

2.46

2.62

3.05

1.47

2.66

2.25

1.90

1.57

1.10

1.61

1.78

3.78

1.05

1.59

0.37

f__Acidobacteriaceae_OTU_109940

3.38

3.47

5.85

6.44

5.09

3.46

5.00

4.76

2.54

3.10

2.41

2.10

3.43

2.55

3.06

2.94

2.98

3.22

2.12

1.73

1.53

1.70

2.17

1.31

Bacillus

0.75

0.32

0.90

0.17

2.07

2.04

1.91

2.80

0.76

0.69

0.46

0.75

0.62

1.01

0.52

0.62

0.92

0.32

0.43

0.47

0.81

0.22

0.36

0.12

o__Ellin6513_OTU_460038

1.08

4.03

4.08

6.66

3.03

2.47

1.38

0.42

2.87

1.68

2.35

1.12

1.31

0.84

1.57

1.03

1.67

1.29

0.92

1.00

0.31

0.38

0.49

0.30

Alicyclobacillus

1.51

5.66

1.00

3.48

1.04

0.46

0.79

0.79

1.09

0.60

3.45

1.84

0.97

0.63

1.38

0.79

1.20

0.79

0.78

1.40

0.36

0.24

0.37

0.48

Rhodanobacter

0.69

0.68

1.21

1.14

1.53

1.46

2.54

1.71

0.49

0.50

0.62

0.39

1.34

1.09

0.58

0.83

1.17

0.70

0.53

0.46

0.24

0.27

0.33

0.20

f__Gaiellaceae_OTU_264578

Fig. 3. Heatmap describing the genus level or higher possible taxonomic classification based on the 19 most abundant OTUs (> 1% of the total number of reads) detected from all gut samples (mean, n = 4).

sequencing data were analyzed using Quantitative Insights into Microbial Ecology (QIIME, version 1.9.1) based on online instructions (Caporaso et al., 2010). After eliminating uncertain nucleotides, primer sequences and low-quality reads, we obtained clean reads that were assembled into operational taxonomic units (OTUs) using de novo clusters, and OTU sequences with 97% similarity were identified by UCLUST clustering (Edgar, 2010). Then, we selected the most abundant OTU sequence as the representative sequence. PyNAST was used to align the representative sequence (Gregory et al., 2010), and taxonomic status was assigned using RDP Classifier 2.2 based on the Greengenes16S rRNA gene database (Gregory et al., 2010; Thirup et al., 2001). The bacterial alpha and beta diversities were calculated by comparing the level of bacterial OTU diversity in the different samples, which was shown by the alpha diversity index (Shannon, PD whole tree, Chao1 and rarefaction curves) and beta diversity analysis followed by principal coordinate analysis (PCoA) based on the weighted Unifrac distances and Adonis test. All the sequences of this experiment have been uploaded to the National Center for Biotechnology Information Sequence Read Archive and the accession number is PRJNA526791 (SRP188328).

2.4. High-throughput quantitative PCR (HT-qPCR) for analysis of antibiotic resistance genes (ARGs) The abundance and composition of ARGs in all samples were investigated by performing high-throughput quantitative PCR on the DNA from each sample using the SmartChip Real-Time PCR System (Warfergen Inc., Fremont, CA). A total of 296 primer pairs were used to construct the 16S rRNA gene, 285 major ARGs and 10 MGEs (1 clinical class 1 integron, 1 class 1 integron and 8 transposases). A 100-nL reaction system (LightCycler 480 SYBR Green I Master Mix) consisting of each primer, E. crypticus gut DNA template and nuclease-free PCR grade water was used to amplify all primers. The standard conditions of amplification were 95 °C for 10 min and then 40 cycles consisting of 95 °C for 30 s and 60 °C for 30 s. SmartChip qPCR software was used to analyze the data obtained from HT-qPCR. According to the collected ARG data from tests conducted on three technical repetitions, we selected 31 as the threshold cycle (CT). ARG abundance per cell was obtained from the formulation of the normalized copy abundances of ARGs as described in a previous study (Zhu et al., 2018a, 2018d). 2.5. Amplification, Illumina sequencing, and bioinformatic analysis

2.6. Statistical analysis We selected the primer pairs 515F (GTGCCAGCMGCCGCGG) and 907R (CCGTCAATCMTTTRAGTTT) to target the V4-V5 hypervariable region of the 16S rRNA gene to amplify the soil and E. crypticus gut DNA with the unique barcodes of reverse primers (Ma et al., 2017). The PCR conditions and purification process followed a previous study (Zhu et al., 2018a). A NanoDrop spectrophotometer was used to detect the concentration of PCR purified products, and 30 different barcode samples were constructed with a sequencing library of equal quality. The combined library was sequenced on an Illumina Hiseq2500 platform (Novogene, Tianjin, China). The obtained high-throughput 16S

The mean values and standard errors of all data were calculated using Microsoft Excel. PCoA based on the weighted Unifrac metric was conducted with OriginPro 9.1. Histograms and box plot error bars were produced using GraphPad Prism 7.00. The Adonis test was performed in R version 3.4.1 with vegan 2.4–3. In addition, using the heatmap package of R version 3.4.1 was used to draw heatmaps, and the Procrustes test was used to explain the correlation between ARGs and bacterial communities in R version 3.4.1 with vegan 2.4–3. Statistically significant differences (P < 0.05) were determined by one-way 4

Environment International 131 (2019) 104965

Q. Zhang, et al.

B Relative abundance (%)

0.4

0

g

g

m g/ k

m g/ k

0.2

0.13 0.04

0.01 0.02 0.02

0.31 0.22 0.12 0.23 0.05 0.06 0.13 0.13 0.13 0.26 0.07 0.10 0.07 0.07 0.15 0.12 0.18 0.10 0.18

0.01 0.00 0.06 0.04 0.02

0.01 0.05 0.00 0.10 0.05 0.18 0.12 0.19 0.09 0.18 0.15 0.06 0.33 0.22 0.28

0.15 0.00 0.03 0.04 0.03 0.05 0.02 0.07 0.59 0.07 0.02 0.03 0.02 0.03 0.00 0.03 0.03 0.04 0.00

0.32 0.06

0.01 0.08 0.06 0.02

0.01 0.13 0.06 0.06 0.05 0.08 0.02 0.08 0.10 0.04 0.00

0.01 0.04

0.11 0.05 0.07 0.17 0.16

0.01 0.01 0.04 0.06

0.11 0.02 0.04 0.09 0.04 0.02

0.11 0.17 0.08

0.01 0.05 0.09 0.03

0.01

0.01 0.05 0.00 0.05 0.07 0.04 0.02

Aminoglycoside

Beta_Lactamase

Multidrug

0.00 0.00 0.00 0.00 0.00 0.00

Sulfonamide

0.01 0.00 0.00 0.00 0.00 0.00 0.03 0.06

0.01 0.01

0.0

-0.2

Others

0.01 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00

0.1

-0.1

MLSB

0.19 0.05 0.02 0.03 0.12 0.02 0.02 0.09 0.24 0.04 0.06 0.08 0.00 0.09 0.08 0.45 0.29 0.47 0.24 0.25 0.25 0.22 0.25 0.37 0.38 0.39 0.19 0.37 0.07 0.28

0.01 0.00 0.00

PCo2 (7.02%)

D 0.19 0.15 0.10 0.12 0.04 0.09 0.13 0.18

0 mg/kg 0.1 mg/kg 0.3 mg/kg 0.8 mg/kg 2 mg/kg 5 mg/kg

0.3

5

m g/ kg

m g/ k

m g/ kg

0.0

g

m g/ kg

C

a

a

a

0.5

5

2

m g/ kg

m g/ kg

0. 8

m g/ kg

0. 3

0. 1

0

m g/ kg

0

b

b

2

10

b

1.0

m g/ kg

b

a

a

a

a

0. 8

20

1.5

m g/ kg

c 30

0. 3

Numbers of ARGs

40

0. 1

A

-0.3 -0.6

-0.4

-0.2

0.0

0.2

0.4

PCo1 (52.7%) 0.03 0.07 0.03 0.05 0.02 0.03 0.02 0.03 0.02 0.02 0.05 0.08 0.00 0.00 0.00 0.03 0.07 0.00 0.03 0.04 0.03 0.08 0.03 0.03 0.05 0.04 0.06 0.04 0.09 0.04

Tetracycline

0.00 0.00 0.00 0.00 0.00

MGEs

0.01 0.04 0.03 0.02

4

5

5-

Vancomycin

5-

3

2

0.01

0.01 0.00 0.00

5-

5-

1

5

5-

4 2-

2-

3

2

0.01 0.00 0.00 0.02 0.00 0.02 0.00

2-

0.01 0.00 0.00 0.00 0.00

0.01 0.04 0.05 0.00

2-

0.01 0.03 0.00 0.00 0.00

1

0.01 0.00 0.00 0.00 0.00 0.00

2-

0.01 0.00 0.00

0.01 0.00 0.00 0.02 0.03 0.00 0.00

co n1 co n2 co n3 co n4 co n5 0. 11 0. 12 0. 13 0. 14 0. 15 0. 31 0. 32 0. 33 0. 34 0. 35 0. 81 0. 82 0. 83 0. 84 0. 85

0.05

0.01 0.01 0.00 0.00 0.00 0.00 0.02 0.00 0.02

Fig. 4. A: Number and B: Total abundance of ARGs in treated and untreated E. crypticus gut. C: Principal coordinate analysis (PCoA) using Bray-Curtis distances based on the relative abundance of ARGs to explain differences in the distribution patterns of ARGs in E. crypticus. Different letters indicate significant differences (P < 0.05, one-way ANOVA) between different groups (mean ± SE, n = 4). D: Heatmap describing enrichment of the nine classes of ARGs in the E. crypticus gut with and without exposure to azoxystrobin.

Gut

Soil 0.1 mg/kg

0.1 mg/kg

0.8 mg/kg

0.8 mg/kg

150

1228 2 mg/kg 5 mg/kg

0.3 mg/kg 0 162 71 34 45 0

2 mg/kg

193

0

0 31 136 0 37 0 0 18 58

352

43 0

0 0 28 0 40 30 0

27 0

23

0

4

1

8

1

4

113

2

45

114

0

3

3 22

18 81 36

76

7

3

2

1

25

16

22 17

84

18 2

3

3

0 7

3

2 4

16

0

7 3

10

0 0 263

23

1 2 19

1

0

0 63 0 0 28 106

0

83 20

14 0.3 mg/kg

92 185 9

95 32

5 mg/kg

0 0

296 1860 0 55 0 0 0 28 36

63

646

0

0

1

1 9 10

10

26 27

0

182

Con

61

168

17 12

Con

Treatments

Control Gut

1263

Gut

Soil

190

575

334

Soil

19

57

Fig. 5. Venn diagram showing the number of shared bacterial OTUs among the control and azoxystrobin-treated groups in the E. crypticus gut and surrounding soil and between the gut and soil (mean, n = 4).

5

Environment International 131 (2019) 104965

Q. Zhang, et al.

ANOVA with the post hoc test (Fisher's protected least significant difference (PLSD)) of the StatView 5.0 program.

3.4. Effect of azoxystrobin on the diversity and abundance of ARGs in the gut and soil

3. Results

Overall, 18 and 53 unique ARGs were detected in the E. crypticus gut and surrounding soil, respectively, which could be classified into eight categories: aminoglycoside, multidrug, tetracycline, macrolide-lincosamide-streptogramin B (MLSB), sulfonamide, beta-lactamase, vancomycin and others. After 28 d of AZ exposure, the number of ARGs increased significantly in the gut and soil in the 2 and 5 mg AZ/kg dry soil treatments (Fig. 4A and S3A, one-way ANOVA, P < 0.05). The relative abundance of ARGs in the gut after exposure to 0.8, 2 and 5 mg/kg AZ increased to 180%, and 214% and 200%, respectively, relative to the control (Fig. 4B). The PCoA results also revealed that the ARG patterns in these treatments were significantly different from the control along the X-axis (explaining 52.7% of total variance) (Fig. 4C). Furthermore, the heatmap of the relative abundance of ARGs in the gut indicated that the increased ARGs occurred mainly in the unclassified category (Fig. 4D). However, the relative abundance of ARGs was not obviously changed in soil samples (Fig. S3B, one-way ANOVA, P > 0.05), and the PCoA results showed that only the 5 mg/kg AZ treatment was significantly separated from other groups along the first principal component axis (Fig. S3C).

3.1. Azoxystrobin residue in E. crypticus body and its effect on the reproduction and mortality of E. crypticus E. crypticus specimens were treated with different doses of AZ applied to soil for 28 d. Trace residue (μg individual−1) of AZ was detected in E. crypticus as follows: 0 (control), 0 (0.1 mg/kg treatment), 0.034 (0.3 mg/kg treatment), 0.019 (0.8 mg/kg treatment), 0.036 (2 mg/kg treatment) and 0.076 (5 mg/kg treatment) (Fig. S1). The mortality of adult E. crypticus was not significantly affected by AZ treatment (Fig. 1A, one-way ANOVA, P > 0.05), and the reproduction of E. crypticus was only slightly decreased in the AZ-treated groups, but these results were not significant (Fig. 1B, one-way ANOVA, P > 0.05). 3.2. Effect of azoxystrobin on E. crypticus gut microbiome The PCoA of weighted UniFrac distance matrices based on 16S rRNA gene sequencing significantly separated the AZ-treated and control groups in the two-dimensional quadrant. Interestingly, the lowconcentration AZ treatments (0.1, 0.3 and 0.8 mg/kg) and high-concentration treatments (2 and 5 mg/kg) were clustered on the bottom left and upper right, respectively (Fig. 1C). With respect to the alpha diversity, the Shannon index increased slightly in the 0.1 mg/kg AZ treatment (one-way ANOVA, P > 0.05) but decreased significantly (one-way ANOVA, P < 0.05) in the 5 mg/kg AZ treatment compared to the control group (Fig. 1D). At the phylum level, the eight dominant bacterial groups in all treatments were Proteobacteria, Firmicutes, Actinobacteria, Acidobacteria, Planctomycetes, Chloroflexi, Verrucomicrobia and Bacteroidetes, which together accounted for > 98% of the total gut microbiome abundance (Fig. 2). These bacterial groups responded differently to different concentrations of AZ (Fig. 2, one-way ANOVA, P < 0.05). The most abundant, Proteobacteria, decreased by 23.4% in the 0.1 mg/kg AZ treatment but increased by 110.9% and 128.9% in the 2 and 5 mg/kg AZ treatments, respectively. The change in Proteobacteria involved the following main genera: Xanthomonas, Burkholderia and Kaistobacter (Fig. 3). In contrast, the abundance of Acidobacteria, Verrucomicrobia and Bacteroidetes increased initially and then decreased with increasing AZ concentration, especially Verrucomicrobia and Bacteroidetes, which changed significantly (one-way ANOVA, P > 0.05). In addition, Firmicutes and Chloroflexi gradually declined with increasing AZ concentration, while the abundance of Actinobacteria and Planctomycetes did not change in any of the treatments (one-way ANOVA, P > 0.05). The heatmap of dominant genera (> 1%) also revealed the various responses of the gut microbiome to exposure to different AZ concentrations (Fig. 3).

3.5. Relationship between bacterial communities and ARGs in the gut and soil To assess the relationship between bacterial communities and ARG number and abundance, we analyzed the relationship between microbial OTUs and ARG files of the gut and soil samples using the Mantel test and Procrustes analysis. A significant correlation between the ARG abundance data and bacterial 16S rRNA gene OTU data of gut microbiota was observed using the Mantel test (r = 0.499, P = 0.001). Procrustes analysis indicated a significant correlation between OTUs and ARGs in the gut and soil, exhibited by a goodness-of-fit test (M2 = 0.486, P < 0.01, 9999 permutations) based on the Bray−Curtis dissimilarity metrics (Fig. S4). 4. Discussion The strobilurin fungicide azoxystrobin is the most widely sold fungicide worldwide, making it difficult to prevent entry into the soil environment. Generally, residual fungicide in soil has a mild impact on the soil bacterial community (Bending et al., 2007; Thirup et al., 2001). AZ at a concentration of 0.33 mg/ kg had no significant effect on soil bacterial communities due to soil-adsorbed AZ (Adetutu et al., 2010). Our work demonstrated that only a higher concentration of AZ (5 mg/ kg) could significantly change the structure and composition of the soil bacterial community. A small portion of residual AZ in soil can be ingested by soil animals. In a previous study, high concentrations of azoxystrobin (500 mg/ kg) decreased the biomass of adult earthworms, and the EC50 of E. crypticus reproduction ranged from 73.3 to 125.7 mg/kg (Leitao et al., 2014). However, 28 d of AZ soil exposure at a concentration ranging from 0.1 to 5 mg/kg did not significantly change E. crypticus adult mortality and reproduction, implying that the tested concentrations of AZ are not highly toxic to E. crypticus. However, microbiota diversity and composition in the E. crypticus gut changed in response to AZ treatment, similar to the effects of glyphosate herbicide on the bee gut microbiome (Motta et al., 2018) and nano silver on collembolan gut microbiota (Zhu et al., 2018d). Two reasons for the imbalanced gut microbiota of E. crypticus were postulated: one was that AZ ingested from soil directly disturbed the E. crypticus gut microbiota; another was that AZ exposure changed the bacterial composition of the transient soil, which could settle in gut microbes, indirectly affecting the E. crypticus gut bacterial community. To verify which postulation was the main cause of the change in the

3.3. Effect of azoxystrobin on soil microbiota The soil microbial community changed only in the 5 mg/kg AZ treatment compared with the control group (Adonis test, R = 0.896, P = 0.028), and the PCoA results (weighted UniFrac distance matrices) indicated that the soil bacterial communities in the 5 mg/kg AZ treatment were significantly isolated from other groups along with the Yaxis (Fig. S2A). The diversity (Shannon index) did not change significantly after AZ exposure (Fig. S2B, one-way ANOVA, P > 0.05). At the phylum level, the major bacterial groups in the soil were similar to those in the E. crypticus gut, and the abundance of these bacterial groups had no significant change after AZ exposure (one-way ANOVA, P > 0.05). The unclassified phylum increased by 143% after exposure to 5 mg/kg AZ (Fig. S2C, one-way ANOVA, P > 0.05). 6

Environment International 131 (2019) 104965

Q. Zhang, et al.

showed different changes in the diversity and abundance of ARGs between the soil and E. crypticus gut, again demonstrating the speculation that AZ ingested from soil directly affected the gut microbiota. Based on these data, the ARGs in the E. crypticus gut and surrounding soil were tightly related to their bacterial data. A strong correlation was also observed between the bacterial community and ARG profiles in various environments, such as groundwater, sewage, sludge and the phyllosphere (Chen et al., 2017a; Chen et al., 2017b; Ma et al., 2017; Su et al., 2015). The changes in bacterial communities may be an important factor affecting the number and abundance of ARGs. Some studies have determined that the microbial changes in the gut of field collembolans can explain 27.77% of the changes in ARGs based on redundancy analysis (Zhu et al., 2018b). Generally, fungicides mainly act against the fungal community; our results indicated that the fungicides could affect the soil and E. crypticus bacteria, which may be the main reason indirectly affecting the change in ARG (Zhu et al., 2018a, 2018b, 2018d). In conclusion, E. crypticus is an important biological indicator for soil ecosystems (Peijnenburg et al., 1999), and this species indirectly ingests pesticide residues in the soil. Our results demonstrated that residual AZ was ingested by E. crypticus, altering the core potential beneficial bacteria and increasing the number and abundance of ARGs in the E. crypticus gut; this AZ may enter the food web, causing a serious ecological risk. Abuse of soil pesticides could perturb the E. crypticus gut microbiome and enrich the ARGs in the E. crypticus gut, promoting the spread of ARGs through the food chain and amplifying the ecological risks of pesticides.

gut bacterial community, we compared gut and soil microbial compositions. With respect to the gut microbial composition, the most abundant phylum, Proteobacteria, was significantly reduced in the 0.1 mg/ kg AZ treatment but increased in the 5 mg/kg AZ treatment compared to the control groups. The increase in Proteobacteria hinted at the dysbiosis of gut microbiota, indicating an unhealthy gut (Na-Ri et al., 2015). Furthermore, the genus that changed the most in Proteobacteria was Xanthomonas, which is a common phytopathogen that can infect plants by releasing adhesins (Mhedbihajri et al., 2011) and may have a negative effect on the E. crypticus gut. Zhu et al. (2018d) also found an obvious increase in Proteobacteria in collembolan gut microbiota after Ag nanoparticle and AgNO3 exposure. The lower diversity and greater change in the E. crypticus gut microbiota than in the control groups also supported our finding, i.e., an unhealthy E. crypticus gut was caused by the 5 mg/kg AZ treatment. The phylum Acidobacteria represents potentially beneficial bacteria related to immunity in the human gut (Lv et al., 2016); Verrucomicrobia is an E. crypticus gut symbiont for essential amino acid supplementation (Larsen et al., 2016), and Bacteroidetes contributes to host heath, including the immune system and resistance to pathogens (Li et al., 2008). The decline of these major potentially beneficial bacteria in the 5 mg/kg AZ treatment implied the formation of an unhealthy gut. Furthermore, the change in the diversity and composition of the gut bacterial communities exhibited two different trends under different concentrations of AZ exposure, and the critical threshold for different responses of AZ may range from 0.8 to 2 mg/ kg dry soil. The composition of these core bacteria in the E. crypticus gut was different from that in the soil. In total, 1453 and 765 OTUs were identified from E. crypticus and soil in the control groups, respectively, including 190 shared OTUs; meanwhile, 353 and 76 were identified from E. crypticus and soil in the AZ-treated groups, respectively, including only 19 shared OTUs (Fig. 5); this contributed to the specific gut environment (e.g., pH and anoxia) and could be selective for colonized bacteria (Zhu et al., 2018c), which caused an obvious difference between the microbiota in the gut and soil (Berg et al., 2016). Based on the above analysis, the changes in gut microbial composition were completely different from those in the soil; thus, we believe that the changes in the gut microbiome were caused by the ingestion of low doses of AZ rather than the indirect effects of soil microbial changes. In addition, the diversity and numbers of ARGs in both the gut and soil were significantly enriched with increasing concentrations of AZ. This finding may be due to the AZ providing an environmental pressure to enrich ARGs, which similar to the usage of antibiotic (benzylisoquinoline), leading to an increase in the synthesis pathway of other antibiotics in the chicken gut, and also increased numbers of ARGs (Huang et al., 2018), or like other soil pollutants have done (e.g., antibiotics, heavy metal and nano-graphene) in various habitat (Cesare et al., 2016; Xie et al., 2016; Zhu et al., 2018a). Additionally, previous studies have confirmed that after exposure to biocides, bacteria follow five principles to produce new ARGs (Paul et al., 2019), which are: 1) cross-resistance, 2) phenotypic change, 3) co-resistance, 4) the enhancement of less susceptible microbial sub-populations, and 5) activating the SOS response in bacteria to repair the damaged DNA. Thus, in current study, the increase in abundance and number of ARGs may be because some bacteria have acquired new antibiotic resistance at low concentrations of AZ based on the pesticide-antibiotic co-resistance theory (Paul et al., 2019); this supports the increase in the number and abundance of ARGs in the category labeled ‘others’ (Fig. S6). Furthermore, related studies have reported that some bacterial phyla more likely carry specific ARGs (Durso et al., 2012), such as beta-lactamase resistance genes, MDR efflux pump genes and fluoroquinolone resistance genes, which are most frequently associated with Proteobacteria (Durso et al., 2012). In the present study, a significant increase in the relative abundance of beta-lactamase resistance genes was observed in the higher-concentration AZ treatment accompanied by more abundant Proteobacteria than in the control groups. These results also

Declaration of Competing Interest The authors declare no conflict of interests. Acknowledgments This work was financially supported by the National Natural Science Foundation of China (21777144), the Strategic Priority Research Program of the Chinese Academy of Sciences (XDB15020302 and XDB15020402), CAS Pioneer Hundred Talents Program to H.F. Qian, and Xinjiang Uighur Autonomous Region Talent Project to H.F. Qian. Appendix A. Supplementary data Supplementary data to this article can be found online at https:// doi.org/10.1016/j.envint.2019.104965. References Adetutu, E.M., Ball, A.S., Osborn, A.M., 2010. Azoxystrobin and soil interactions: degradation and impact on soil bacterial and fungal communities. J. Appl. Microbiol. 105, 1777–1790. Bartlett, D.W., Clough, J.M., Godwin, J.R., Hall, A.A., Mick, H., Bob, P.D., 2002. The strobilurin fungicides. Pest Manag. Sci. 58, 649. Bartlett, D.W., Clough, J.M., Godwin, J.R., Hall, A.A., Hamer, M., Parr-Dobrzanski, B., 2010. Review: the strobilurin fungicides. Pest Manag. Sci. 60, 309. Bending, G.D., Lincoln, S.D., Edmondson, R.N., 2006. Spatial variation in the degradation rate of the pesticides isoproturon, azoxystrobin and diflufenican in soil and its relationship with chemical and microbial properties. Environ. Pollut. 139, 279–287. Bending, G.D., Rodríguez-Cruz, M.S., Lincoln, S.D., 2007. Fungicide impacts on microbial communities in soils with contrasting management histories. Chemosphere 69, 82–88. Berg, M., Stenuit, B., Ho, J., Wang, A., Parke, C., Knight, M., Alvarezcohen, L., Shapira, M., 2016. Assembly of the Caenorhabditis elegans gut microbiota from diverse soil microbial environments. ISME J. 10, 1998–2009. Caporaso, J.G., Kuczynski, J., Stombaugh, J., Bittinger, K., Bushman, F.D., Costello, E.K., Knight, R., 2010. QIIME allows analysis of highthroughput community sequencing data. Nat. Methods 7, 335–336. Cesare, A.D., Eckert, E.M., Corno, G., 2016. Co-selection of antibiotic and heavy metal resistance in freshwater bacteria. J. Limnol. 75, 59–66. Chen, Q.L., An, X.L., Zhu, Y.-G., Su, J.Q., Gillings, M.R., Ye, Z.L., Cui, L., 2017a. Application of struvite alters the antibiotic resistome in soil, rhizosphere, and phyllosphere. Environ. Sci. Technol. 51, 8149–8157.

7

Environment International 131 (2019) 104965

Q. Zhang, et al.

determinants. Environ. Pollut. 157, 2893–2902. Mhedbihajri, N., Jacques, M.A., Koebnik, R., 2011. Adhesion mechanisms of plant-pathogenic xanthomonadaceae. Adv. Exp. Med. Biol. 715, 71–89. Motta, E.V.S., Raymann, K., Moran, N.A., 2018. Glyphosate perturbs the gut microbiota of honey bees. Proc. Natl. Acad. Sci. 115, 10305–10310. Na-Ri, S., Tae Woong, W., Jin-Woo, B., 2015. Proteobacteria: microbial signature of dysbiosis in gut microbiota. Trends Biotechnol. 33, 496–503. Oerke, E.C., Dehne, H.W., 2004. Safeguarding production-losses in major crops and the role of crop protection. Crop Protection 23, 275–285. Paul, D., Chakraborty, R., Mandal, S.M., 2019. Biocides and health-care agents are more than just antibiotics: inducing cross to co-resistance in microbes. Ecotoxicol. Environ. Saf. 174, 601–610. Peijnenburg, W.J., Posthuma, L., Zweers, P.G., Baerselman, R., De Groot, A.C., Van Veen, R.P., Jager, T., 1999. Prediction of metal bioavailability in Dutch field soils for the oligochaete Enchytraeus crypticus. Ecotoxicol. Environ. Saf. 43, 170. Qi, L., Ge, Y., Xia, T., He, J.Z., Shen, C.C., et al., 2019. Rare earth oxide nanoparticles promote soil microbial antibiotic resistance by selectively enriching antibiotic resistance genes. Environ. Sci-Nano. 6, 456–466. Qian, H.F., Hu, B.L., Cao, D., Chen, W., Xu, X.Y., Lu, Y.C., 2007. Bio-safety assessment of validamycin formulation on bacterial and fungal biomass in soil monitored by realtime PCR. Bull. Environ. Contam. Toxicol. 78, 239–244. Su, J.Q., Wei, B., Ou-Yang, W.Y., Huang, F.Y., Zhao, Y., Xu, H.J., Zhu, Y.-G., 2015. Antibiotic resistome and its association with bacterial communities during sewage sludge composting. Environ. Sci. Technol. 49, 7356–7363. Teng, Y., Wu, J., Lu, S., Wang, Y., Jiao, X., Song, L., 2014. Soil and soil environmental quality monitoring in China: a review. Environ. Int. 69, 177–199. Thirup, L., Johnsen, K., Torsvik, V., Spliid, N.H., Jacobsen, C.S., 2001. Effects of fenpropimorph on bacteria and fungi during decomposition of barley roots. Soil Biol. Biochem. 33, 1517–1524. Vliet, P.C.J.V., Beare, M.H., Coleman, D.C., 1995. Population dynamics and functional roles of Enchytraeidae (Oligochaeta) in hardwood forest and agricultural ecosystems. Plant Soil 170, 199–207. Xie, Y., Wu, B., Zhang, X.X., Yin, J., Mao, L., Hu, M., 2016. Influences of graphene on microbial community and antibiotic resistance genes in mouse gut as determined by high-throughput sequencing. Chemosphere 144, 1306–1312. Zhan, J., Liang, Y., Liu, D., Ma, X., Li, P., Liu, C., Liu, X., Wang, P., Zhou, Z., 2018. Antibiotics may increase triazine herbicide exposure risk via disturbing gut microbiota. Microbiome 6, 224. Zhang, Q., Zhu, D., Ding, J., Zhou, S., Sun, L., Qian, H., 2019. Species-specific response of the soil collembolan gut microbiome and resistome to soil oxytetracycline pollution. Sci. Total Environ. 668, 1183–1190. Zhu, C.Y., Zhang, N., Bai, X., 2012. Status and analysis of the registration of fungicide products in China. Pest. Sci. Manag 33, 7–11. Zhu, Y.-G., Gillings, M., Simonet, P., Stekel, D., Banwart, S., Penuelas, J., 2017. Microbial mass movements. Science 357, 1099. Zhu, Y., Zhang, Q., Xu, J., Qu, Q., Lu, T., Du, B., Ke, M., Zhang, M., Qian, H., 2018. Changes in bacterial community structure and antibiotic resistance genes in soil in the vicinity of a pharmaceutical factory. Ecotoxicol. Environ. Saf. 158, 87–93. Zhu, Y.-G., Gillings, M., Simonet, P., Stekel, D., Banwart, S., Penuelas, J., 2018. Human dissemination of genes and microorganisms in earth's critical zone. Glob. Chang. Biol. 24, 1488–1499. Zhu, D., An, X.L., Chen, Q.L., Yang, X.R., Christie, P., Ke, X., Wu, L.H., Zhu, Y.-G., 2018a. Antibiotics disturb the microbiome and increase the incidence of resistance genes in the gut of a common soil collembolan. Environ. Sci. Technol. 52, 3081–3090. Zhu, D., Chen, Q.L., Li, H., Yang, X.R., Christie, P., Ke, X., Zhu, Y.-G., 2018b. Land use influences antibiotic resistance in the microbiome of soil collembolans Orchesellides sinensis. Environ. Sci. Technol. 52, 14088–14098. Zhu, D., Chen, Q.L., An, X.L., Yang, X.R., Christie, P., Ke, X., Wu, L.H., Zhu, Y.-G., 2018c. Exposure of soil collembolans to microplastics perturbs their gut microbiota and alters their isotopic composition. Soil Biol. Biochem. 116, 302–310. Zhu, D., Zheng, F., Chen, Q.L., Yang, X.R., Christie, P., Ke, X., Zhu, Y.-G., 2018d. Exposure of a soil collembolan to Ag nanoparticles and AgNO3 disturbs its associated microbiota and lowers the incidence of antibiotic resistance genes in the gut. Environ. Sci. Technol. 52, 12748–12756.

Chen, Q.L., Li, H., Zhou, X.Y., Zhao, Y., Su, J.Q., Zhang, X., Huang, F.Y., 2017b. An underappreciated hotspot of antibiotic resistance: the groundwater near the municipal solid waste landfill. Sci. Total Environ. 609, 966–973. Chen, Q.L., Cui, H.L., Su, J.Q., Penuelas, J., Zhu, Y.-G., 2019. Antibiotic resistomes in plant microbiomes. Trends Plant Sci. https://doi.org/10.1016/j.tplants.2019.02.010. Du, B., Zhang, Z., Liu, W., Ye, Y., Lu, T., Zhou, Z., Li, Y., Fu, Z., Qian, H., 2019. Acute toxicity of the fungicide azoxystrobin on the diatom Phaeodactylum tricornutum. Ecotoxicol. Environ. Saf. 168, 72–79. Durso, L.M., Miller, D.N., Wienhold, B.J., 2012. Distribution and quantification of antibiotic resistant genes and bacteria across agricultural and non-agricultural metagenomes. PLoS One 7, e48325. Edgar, R.C., 2010. Search and clustering orders of magnitude faster than BLAST. Bioinformatics 26, 2460. Elena, G., Claudia, R., Sara, F., Anna, E., Karin, E., Magnus, B., 2015. Bacteria-mediated effects of antibiotics on Daphnia nutrition. Environ. Sci. Technol. 49, 5779–5787. Erkosar, B., Leulier, F., 2014. Transient adult microbiota, gut homeostasis and longevity: novel insights from the Drosophila model. FEBS Lett. 588, 4250–4257. Gajbhiye, V.T., Gupta, S., Mukherjee, I., Singh, S.B., Singh, N., Dureja, P., Kumar, Y., 2011. Persistence of Azoxystrobin in/on Grapes and Soil in Different Grapes Growing Areas of India. B. Environ. Contam. Tox. 86, 90–94. Gregory, J.C., Kyle, B., Bushman, F.D., Desantis, T.Z., Andersen, G.L., Rob, K., 2010. PyNAST: a flexible tool for aligning sequences to a template alignment. Bioinformatics 26, 266–267. Hachung, C., Pamp, S.J., Hill, J.A., Surana, N.K., Edelman, S.M., Troy, E.B., Reading, N.C., Villablanca, E.J., Sen, W., Mora, J.R., 2012. Gut immune maturation depends on colonization with a host-specific microbiota. Cell 149, 1578–1593. Huang, P., Zhang, Y., Xiao, K., Jiang, F., Wang, H., Tang, D., Liu, D., Liu, B., Liu, Y., He, X., Liu, H., Liu, X., Qing, Z., Liu, C., Huang, J., Ren, Y., Yun, L., Yin, L., Lin, Q., Zeng, C., Su, X., Yuan, J., Lin, L., Hu, N., Cao, H., Huang, S., Guo, Y., Fan, W., Zeng, J., 2018. The chicken gut metagenome and the modulatory effects of plant-derived benzylisoquinoline alkaloids. Microbiome 6, 211. Kim, M.-G., Hwang, K.-W., Hwang, E.-J., Yoo, S.-C., Moon, J.K., 2017. Absorption of soil residual azoxystrobin to lettuce. J. Appl. Biol. Chem. 60 (4), 343–349. Larsen, T., Ventura, M., Maraldo, K., Triadó-Margarit, X., Casamayor, E.O., Wang, Y.V., Andersen, N., O'Brien, D.M., 2016. The dominant detritus-feeding invertebrate in arctic peat soils derives its essential amino acids from gut symbionts. J. Anim. Ecol. 85, 1275–1285. Leitao, S., Cerejeira, M.J., Brink, P.J.V.D., Sousa, J.P., 2014. Effects of azoxystrobin, chlorothalonil, and ethoprophos on the reproduction of three terrestrial invertebrates using a natural Mediterranean soil. Appl. Soil Ecol. 76, 124–131. Lepage, M., Abbadie, L., Josens, G., Lavelle, P., 2006. Role of soil fauna in nitrogen cycling. Ecol. Study 179, 299–311. Li, W., Ley, R.E., Pavel Yu, V., Stranges, P.B., Lia, A., Stonebraker, A.C., Changyun, H., Susan, F., 2008. Innate immunity and intestinal microbiota in the development of type 1 diabetes. Nature 455, 1109–1113. Liu, Y.P., Sun, H.B., Zeng, F.J., Liu, J.M., 2010. Study on residual dynamics of azoxystrobin in mango and soil. Guangdong Agr. Sci. 37, 106–107. Lu, T., Ke, M.J., Lavoie, M., Jin, Y.J., Fan, X.J., Zhang, Z.Y., Fu, Z.W., Sun, L.W., Gillings, M., Peñuelas, J., Qian, H.F., Zhu, Y.G., 2018a. Rhizosphere microorganisms can influence the timing of plant flowering. Microbiome 6, 231. Lu, T., Zhu, Y., Xu, J., Ke, M., Zhang, M., Tan, C., Fu, Z., Qian, H., 2018b. Evaluation of the toxic response induced by azoxystrobin in the non-target green alga Chlorella pyrenoidosa. Environ. Pollut. 234, 379–388. Lu, T., Zhu, Y., Ke, M., Peijnenburg, W.J.G.M., Zhang, M., Wang, T., Chen, J., Qian, H., 2019. Evaluation of the taxonomic and functional variation of freshwater plankton communities induced by trace amounts of the antibiotic ciprofloxacin. Environ. Int. 126, 268–278. Lv, L.X., Fang, D.Q., Ding, S., Chen, D.Y., Ren, Y., Zhu, Y.X., Chen, Y.F., Li, S., Guo, F.F., Wu, W.R., 2016. Alterations and correlations of the gut microbiome, metabolism and immunity in patients with primary biliary cirrhosis. Environ. Microbiol. 18, 2272–2286. Ma, L., Li, B., Jiang, X.T., Wang, Y.L., Xia, Y., Li, A.D., Zhang, T., 2017. Catalogue of antibiotic resistome and host-tracking in drinking water deciphered by a large scale survey. Microbiome 5, 154. Martinez, J.L., 2009. Environmental pollution by antibiotics and by antibiotic resistance

8