Skip to main content
NCBI home page
FEMS Microbiology Ecology logoLink to FEMS Microbiology Ecology
. 2025 Jul 23;101(9):fiaf076. doi: 10.1093/femsec/fiaf076

Polymers and immersion time shape bacterial pathogen and antibiotic resistance profiles in aquaculture facilities

Jeanne Naudet 1,✉, Jean-Christophe Auguet 2, Thierry Bouvier 3,4, Raherimino Rakotovao 5, Tony Motte 6, Loïc Gaumez 7, Tania Crucitti 8, Fabien Rieuvilleneuve 9, Emmanuelle Roque d’Orbcastel 10,✉
PMCID: PMC12396186  PMID: 40699012

Abstract

Most equipment used in aquaculture farms is made of plastic. Plastics-associated biofilms may contain potential human pathogenic bacteria (PHPB) and antibiotic-resistant bacteria (ARB). Understanding the influence of farming practices on the biofouling development and composition is thus essential to control associated microbiological risks. We combined results from metabarcoding analyses, bacterial cultures, and antibiotic susceptibility testing to compare the bacterial pathobiome and resistome associated with plastic aquaculture equipment, including two polyamide nets and a polyester liner, with those associated to a hemp net and a glass control. Over the 3 months of incubation in an aquaculture farm, plastics exhibited neither higher levels of PHPB nor more multiple antibiotic resistance compared to other solid substrates, but they did present specific PHPB and ARB profiles. Bacterial members of the Vibrionaceae and Staphylococcaceae families were more abundant in plastic PHPB communities (respectively 47% and 22% of PHPB reads) than in other substrate ones (4% and 0.22% of PHPB reads). The plastic-associated antibiotic resistance profiles showed higher resistance against quinolones. These results suggest that aquaculture equipment could act as a reservoir for some PHPB and ARB, and that equipment composition and immersion time could be levers to control associated sanitary risks.

Keywords: 16S rRNA gene, antibiotic resistance, aquaculture, pathogen, plastisphere

In situ 3-month-incubation of plastic and nonplastic equipment in an aquaculture farm to characterize potentially human pathogenic and antibiotic-resistant bacteria, and to identify how aquaculture practices could influence their development.

Highlights.

  • PHPB relative abundance decreases with time, suggesting they could be biofilm pioneers.

  • Plastics do not show more PHPB nor antibiotic resistances than other solid substrates.

  • Plastic PHPB communities and resistance profiles differed from other solid substrates.

  • Vibrionaceae bacteria were biomarkers of the PHPB plastisphere.

  • Plastic antibioresistance profiles display higher resistances against quinolones.

List of abbreviations

ARB

Antibiotic-resistant bacteria

ASV

Amplicon sequence variant

MAR

Multiple antibiotic resistant

PA

Polyamide

PHPB

Potential human pathogenic bacteria

PE

Polyethylene

Introduction

Due to its excellent properties, plastic has become pervasive in human society. Its lightweight, durability, and cost-effectiveness make it essential for both long-term applications and single-use items, and aquaculture is no exception, relying heavily on plastic structures and equipment ( Emmerik and Schwarz 2020). Most plastics in aquaculture are of endogenous origin, since all equipment is made of plastic such as nets, ropes, buoys, or pipes (Zhou et al. 2021). Plastics from external origin are also found in aquaculture systems, for example through marine currents and/or from land-based pollution and atmospheric deposition (Chen et al. 2021).

Plastics carry a bacterial biofilm on their surface, called plastisphere, that is composed of various microorganisms such as algae, fungi, viruses, archaea, and bacteria (Zettler et al. 2013, Amaral-Zettler et al. 2021). The presence of potential pathogenic bacteria in the plastisphere has been described by many authors, including potential human pathogenic bacteria (PHPB) (Zettler et al. 2013, Delacuvellerie et al. 2022, Lear et al. 2022, Liang et al. 2023, Naudet et al. 2023) and animal pathogenic bacteria (Oberbeckmann et al. 2016, Frère et al. 2018, Martínez-Campos et al. 2022). Bacterial members of the Vibrionaceae family are the most described PHPB in the marine plastisphere, as they are quite ubiquitous in marine environments, as for instance Vibrio vulnificus, Vibrio parahaemolyticus, and Vibrio cholerae (Metcalf et al. 2022, Junaid et al. 2022). Members of the Enterobacterales order also raised interest in plastisphere studies as they are usually used as fecal indicators in aquatic ecosystems, including Enterococci and Escherichia coli (Liang et al. 2023). However, the enrichment of plastic substrates compared to other solid substrates is still debated (Oberbeckmann et al. 2016, Kelly et al. 2022, Metcalf et al. 2022), and the specific risk that plastic could represent as a fomite or as a reservoir of PHPB is still unclear.

Another concern is antibiotic resistance, known to be promoted in biofilm through diverse mechanisms, such as slow antibiotic penetration, facilitated horizontal gene transfers, and activation of metabolic tolerance mechanisms (Mah 2012, Uruén et al. 2021). High levels of resistance to antibiotics have been detected in the plastisphere, dominated by resistances against sulfonamides, tetracyclines, quinolones, macrolides, and beta-lactams (Sathicq et al. 2021, Yu et al. 2022). Plastics have been suggested to act as a reservoir for antibiotic-resistant bacteria (ARB) and antibiotic resistance genes (Yang et al. 2019, Stenger et al. 2021). But more studies are needed to better understand the risks associated with plastics, and a gap of knowledge still persists in some fields such as in aquaculture contexts.

Microbiologic threats associated with plastics may become a serious sanitary concern, especially in aquaculture facilities, where infectious diseases represent a major cause of mortality in reared species and consequently economic losses (Lafferty et al. 2015), and where workers are exposed to biological hazards (Ngajilo and Jeebhay 2019). This concern is even greater as the number of multiresistant bacteria is known to increase in aquaculture facilities (Cabello 2006, Ge et al. 2023), making it more difficult to find effective treatments. Assessing whether plastic equipment serves as a reservoir for pathogens and antibiotic resistance and how it varies with time is essential to adapt practices in aquaculture and protect reared species, workers, and consumers from potential infectious outbreaks.

We present here the results of a 3-month incubation experiment in an aquaculture farm in the southwest Indian Ocean, where we investigated the dynamics of the bacterial pathobiome and resistome associated with plastic aquaculture equipment, including two polyamide (PA) nets with different mesh sizes and a polyester (PE) liner, and compared them with those associated with a hemp net, a new alternative to plastic net, and a glass control. Our first objective was to assess the role of plastics as a carrier of PHPB and ARB, and secondly to follow the dynamics of PHPB and ARB occurring during the biofouling process on tested substrates. This analysis seeks to determine how aquaculture practices, such as equipment immersion time and composition, can influence bacterial communities in aquaculture, to prevent infectious and/or antibiotic resistance outbreaks.

Materials and methods

Experimental procedures

This study was carried out in Toliara, Madagascar, at the Indian Ocean Trepang farm of holothuria (Holothuria scabra) for the incubation of experimental substrates and at the Halieutic and Marines Sciences Institute for the laboratory analysis, from November 2022 to January 2023. We tested three plastic substrates of interest for this specific farm or more generally in aquaculture farms: a black PE liner (Tianjin Kuifang International Trade Co.,), a white PA net with a small mesh size (mosquito net, 250 µm, Jiexi International Ltd.) and a black PA net with a high mesh size (10 mm, Kerfil®). The liner is the equipment used in all farms to cover the bottom of the ponds, the mosquito PA net was tested by the farmer as a supplementary surface to increase the biofilm formation at the pond bottom, and the high-mesh PA net is the most common net used in fish farming. Two nonplastic substrates were also used: a hemp net (70 mm mesh size, SAS Ciel&Terre), as a promising alternative to plastic nets, already used at some aquaculture farms (Tamburini et al. 2020, Petrocelli et al. 2021) and a borosilicate glass panel as an inorganic control (Metcalf et al. 2022). Seven panels (20 cm × 20 cm; one panel per sampling time-point) of each substrate were immersed 30 cm below the surface, for 3 months, in a 3500 m3 pond (1 m deep, with a 100% water renewal rate per day from the adjacent lagoon), containing a density of 10 holothuria per square meter (Fig. 1). The 3 months duration corresponds to the period that holothuria juveniles spend in the tank for weight gain before sea-ranching.

Figure 1.

Figure 1.

Open in a new tab

Diagram of the experimental design. (A): five substrate types [(1) PE-liner, (2) PA-net, (3) mosquito-PA-net, (4) glass, and (5) hemp net] were incubated for 3 months in an aquaculture tank in which H. scabra were reared, resulting in the formation of a biofilm associated with the substrate. (B) The biofilm was collected in triplicates at seven time-points along the incubation. (C) Biofilm communities were identified using metabarcoding analyses. (D) Culturable bacteria were isolated on selective media and antibiotic resistance was assessed by susceptibility testing. Illustration made using Biorender®.

Panel biofilm triplicates (4 cm2 each on the same panel) and seawater triplicates (1 l each) were sampled after 6 h, 12 h, 24 h, 72 h, 7 days, 21 days, and 3 months (Fig. 1). One glass panel was lost during the experimentation, so there was no data available for the 3 months glass sample. Each panel was carefully rinsed with sterilized 0.2 µm filtered seawater before samplings. Seawater from the pond was sampled to compare the panel’s bacterial biofilm communities to the seawater’s planktonic surrounding ones.

Environmental parameters such as seawater temperature, salinity, and nutrient measures during the experiment were monitored and are presented in Table S1.

Sampling and procedure for bacterial cultures, species identifications, and antibiograms

On-site, a piece of 4 cm2 was cut out from each plastic and hemp panel and stored in a cooler in 50 ml of 0.2 µm filter sterilized seawater (Whattman®) for later biofilm removal according to Trachoo (2004), and culturing. The biofilm from the glass panel (4 cm2) was collected using a sterile swab (SK-2S swabs, Isohelix, UK), which was then processed like the panel pieces. Within 1 h after sampling, each 4 cm2 piece and swab were immersed in 300 ml of sterile seawater with 60 g of sterile and calibrated (300–500 µm size) sand, and then stirred for 5 min to detach the biofilm from the substrates, according to the abrasion protocol described in Trachoo (2004).

Bacterial members of the coliform group were cultured according to the filtration technique (Forster and Pinedo 2015). Seawater (10 ml) and supernatant (10 ml) from plastics, hemp, and glass were filtered through cellulose nitrate filters of 0.45 µm (Sartorius®). Filters were placed on Endo agar (pad set number 14053–47 N Sartorius®), a selective medium for coliforms, making them recognizable via a characteristic pink colouration pattern. Colony forming units of thermotolerant coliforms were counted after a 48 h incubation at 44°C. Individual colonies were collected and inoculated on stock culture agar (BIORAD®) for storage before species identification and antibiogram. Matrix-assisted laser desorption ionization-time of flight (MALDI-TOF) mass spectrometer (Bruker Daltonics, Bremen, Germany) was used to identify the bacteria following the manufacturer’s instruction for identification acceptance (highest score). A representative of each isolates species recovered was tested against a panel of antibiotics using the Kirby–Bauer disk diffusion method on Mueller–Hinton agar (Hudzicki 2009). The antibiotics (n = 24, BIORAD®, Table S2) were selected according to the European Committee on Guidelines for Antimicrobial Susceptibility Testing (EUCAST; v2023). A large selection of antibiotics commonly used in aquaculture and medicine were selected when the EUCAST recommendations did not reference the bacterial genus or family of the isolate. The diameters of the inhibition zones were measured in millimetres using a ruler by the same operator to avoid observation bias, and interpreted according to EUCAST recommendations. Escherichia coli ATCC 25922 was used as a quality control strain. The multiple antibiotic resistance (MAR) index, i.e. the proportion of the tested antibiotics for which the species was resistant, was calculated for each species, following the method described by Krumperman (1983).

Sampling and procedure for DNA extraction and sequencing

On site, the biofilm from a 4 cm2 surface of each panel was directly sampled using a sterile swab (SK-2S swabs, Isohelix®), and 400 ml of seawater was collected in a sterile glass bottle. Swabs and bottles were stored on ice in a cooler for 30 min between the site and the lab. Seawater was filtered on 0.2 µm GTTP 47 mm filters (Whattman®) under gentle vacuum pressure (10 mm-Hg max). Swabs and GTTP filters were stored at −80°C in the laboratory until DNA extraction.

Total DNA was extracted from the swabs (for the panels) and filters (for seawater) using the MagAttract PowerSoil® DNA Kit according to the manufacturer’s instructions (Qiagen, Courtaboeuf, France) with automated processing and the liquid handling system KingFisher FlexTM (ThermoScientific®, Waltam, MA, USA). DNA extracts were stored at −20°C until PCR (Polymerase Chain Reaction) amplification.

The V4–V5 region of the 16S rRNA gene was amplified using the primers 515F-Y (5′-GTGYCAGCMGCCGCGGTAA-3′) and 926R (5′-CCGYCAATTYMTTTRAGTTT-3′) coupled with platform specific Illumina adaptor sequences on the 5′ ends (Parada et al. 2016). Each 50 µl PCR reaction was prepared with 25 µl Taq Polymerase Phusion® High-Fidelity PCR Master Mix with GC Buffer (New England Biolabs®, Inc., Ipswich, MA, USA), 1 µl forward primer (10 µM), 1 µl reverse primer (10 µM), 2 µl template DNA, 1.5 µl DMSO, and 19.5 µl molecular water. PCR amplifications involved the following protocol: an initial 98°C denaturing step for 30 s followed by 35 cycles of amplification (10 s denaturation at 98°C; 1 min at 58°C annealing; and 1.5 min extension at 72°C), and a final extension of 10 min at 72°C. Amplification and primer specificity were verified by electrophoresis on a 2.0% agarose gel for confirmation of ∼450 bp amplicon size. To evaluate the quality of our sample processing pipeline, the extraction and 16S rRNA gene amplification of standard mock community (Zymo- BIOMICS Microbial Community DNA Standards II, Zymo Research) and blank samples (no DNA) were performed. Sequencing was performed on an Illumina Miseq by GeT-Biopuces (INSA, Toulouse, France). Raw reads were deposited into the NCBI database under Bioproject number: PRJNA1116056.

In order to follow bacterial abundance over time, relative quantification of the total 16S rRNA gene concentration in each sample was used as a proxy and assessed by quantitative PCR (qPCR), using the same primers and PCR amplification conditions as before. We used a seawater sample collected in the pond at the beginning of the experiment as a reference, and expressed sample concentrations as relative concentrations compared to this reference sample.

Sequence processing

All sequences were analyzed using R (version 4.2.2) with the dada2 package (Callahan et al. 2016). Sequences were first trimmed and filtered based on read-quality profiles (maxN = 0; maxEE = [2,2]; truncQ = 2; and truncLen = [250 250]), and amplicon sequence variants (ASVs) were inferred after pooling dereplicated reads from all samples. Forward and reverse reads were merged and chimeric sequences were removed. Prokaryotic taxonomy was assigned using the naive Bayesian classifier method against the Silva Database nr_V138.1. The ASV abundance table, taxonomy, and sequences were further processed using the phyloseq package (v.1.28.0; McMurdie and Holmes 2013) in R.

To remove contaminants from our dataset, we used the R package decontam (Davis et al. 2018), based on the ‘prevalence method’. Overall, 267 ASVs corresponding to 2.4% of the total reads were removed. Sample read sums were randomly equalized at 11 100 reads per sample, which was the smallest sample read number of our dataset, using the phyloseq package (McMurdie and Holmes 2013). After standardization, our final dataset consisted of 1 587 300 sequences belonging to 26 542 ASVs.

PHPB were identified using a homemade full-length 16S rRNA gene database derived from the enhanced infectious disease database (EID2, Wardeh et al. 2015). This EID2 database gathers data concerning bacteria (i.e. cargos) described to have had interactions with humans. From the 878 human bacterial cargos found in the EID2 database, we constructed a full-length 16S rRNA gene database with 87 405 sequences. Only ASVs matching a 16S rRNA sequence with 100% similarity, 100% coverage, and more than 250 bp were included in our pathobiome dataset. The potential pathogenicity of each ASV was subsequently checked in the literature, and only ASVs matching a bacteria described at least once as pathogenic for humans were kept. Among the 103 PHPB ASVs detected in our study, 13 were discarded. The remaining PHPB are listed in Table S3.

Statistics workflow

Richness (number of ASVs) and the Shannon diversity index H were measured to assess the taxonomic diversity of each microbial community. The change in alpha diversity over time was statistically analyzed using a linear model. Beta diversity was assessed using Bray–Curtis distances with the ‘vegan package (v2.6–4)’ in R (Oksanen et al. 2022). Statistical analyses were performed using permutational analyses of variance (PERMANOVA), and a linear model to assess the changes in Bray–Curtis distances over time. Bray–Curtis dissimilarities were shown in principal coordinate analysis (PCoA) plots. PHPB presence was assessed and compared between samples using relative abundances of PHPB reads. The identification of bacterial families that are plastic biomarkers was performed using the analysis of bacteriome composition with bias correction (ANCOM-BC) in the microbiomeMarker package (Cao et al. 2022).

MAR index was calculated on all isolated bacteria. Statistical differences were assessed using an ANOVA test, following a homogeneity test (Levene test) and a normality test (Shapiro–Wilk test), using the rstatix package in R (Kassambara 2023). Antibiotic resistance profiles were visualized using a principal component analysis (PCA) on diameters of inhibition of bacterial taxa that were tested against the full panel of 24 antibiotics (n = 33; see Table S2). As resistance diameter thresholds vary according to the antibiotic, diameters of inhibition were scaled and centred for each antibiotic before PCA analysis, so we could compare resistance patterns of different antibiotics. Statistical analyses were performed using a PERMANOVA test.

Results

The biofilm’s global composition changes and tends to homogenize over time

The total dataset represented 26 542 ASVs. The increase in biofilm biomass on solid substrates was evaluated through qPCR analyses of the 16S rRNA gene. This analysis pointed out the increase in bacterial load in all samples with time, including seawater samples (linear model, P < .05; Fig. S1A). This growth was accompanied by a significant increase in richness and diversity (Shannon index), particularly at 72 h, for all solid substrates but hemp (linear model, P < .05, Fig. S1A and B). Seawater communities had a constant richness and Shannon index throughout the experiment.

All substrates showed a similar global community structure, displaying three main phyla (Fig. S2): Proteobacteria that represented between 46% and 54% of the reads, Bacteroidota (from 14% to 29%), and Cyanobacteria (from 5% to 17%). However, plastic communities were distinguishable from glass and seawater communities as illustrated by beta-diversity analyses (Fig. 2A and B). Indeed, at the early stages of incubation (≤72 h), no differences were found between PA-nets and the hemp net, but glass and seawater communities were significantly different from all substrates (pairwise PERMANOVA, P < .05). At the late stages of incubation (≥7 days), seawater communities were still different from all other substrates, and glass communities were significantly different from all substrates but the liner (pairwise PERMANOVA, P < .05). While some distinguishable communities were observed at the early incubation timepoints according to the substrate, all solid substrate communities homogenized over time, but remained different from seawater communities (Fig. 2B). This was statistically confirmed by a decrease in Bray–Curtis distances between solid substrates over time, while distances between solid substrates and seawater significantly increased (Fig. S3A and B, linear model, P < .001). For all substrates together, statistical differences were identified between global communities from the first 72 h of incubation and the late stages of incubation (≥7 days) (pairwise PERMANOVA, P < .05; Fig. S4). These differences in alpha and beta diversity over time can be explained by the changes within the taxonomical composition of the biofilm communities on the different immersed substrates. Some phyla appeared during late incubation times, such as Acidobacteriota and Verrucomicrobiota in hemp, in the high-mesh and mosquito PA-nets (Fig. S2).

Figure 2.

Figure 2.

Open in a new tab

Beta-diversity of the whole bacteriomes from the different substrates according to the incubation time. PCoA plots showing the Bray–Curtis dissimilarities between substrates (A) and between substrates according to the incubation time (B).

Focus on the pathogenic communities

Overall, 90 ASVs and 14 879 reads were identified as PHPB, representing 0.34% of all ASVs analysed and 0.94% of the whole dataset reads. This PHPB dataset was composed of 67 species belonging to 38 genera (Table S3). The proportion of potential pathogenic reads varied according to the substrate (Fig. 3A), with a mean of 0.29 ± 0.33% in seawater samples, 0.35 ± 0.40% in the mosquito net samples, 0.37 ± 0.62% in the liner samples, 0.39 ± 0.67% in the high-mesh PA-net samples, 0.74 ± 1.31% in glass samples, and 3.5 ± 4.9% in hemp samples. This proportion also varied with incubation time (Fig. 3A). For all substrates except for seawater, the maximum relative and absolute abundance in PHPB was reached in the first 24 h of the incubation, and a significant decrease in relative abundance was detected for hemp and the mosquito net with time (linear model, P < .05; Fig. 3A). At early stages of incubation (≤72 h), hemp samples displayed the significantly highest relative abundance of PHPB of all substrates, and plastic samples pooled together (PA-nets and PE-liner) displayed a significantly higher relative abundance of PHPB than seawater samples. However, at later stages (≥7 days), seawater samples had the highest relative abundance of PHPB, with significant differences noted for the liner and the mosquito net (Dunn post hoc test, P < .05).

Figure 3.

Figure 3.

Open in a new tab

Relative abundance and alpha diversity of the pathobiome communities. (A) Percentage of pathogenic reads identified in the samples from each substrate according to the incubation time. (B) Taxonomical richness (number of different ASVs). (C) Shannon index of diversity. The symbol * indicates a significant decrease over time (P < .05; linear model).

Substrate type and incubation time exerted a significant effect on the PHPB communities’ beta diversity (PERMANOVA, P < .001, Rtime2 = 0.12, Rsubstrate2 = 0.13) with different composition between the early stages (≤72 h) and late-stages (≥7 days; Fig. 4A and B, Fig. S5). At early stages, PHPB from seawater and hemp were significantly different from all other substrate communities, and glass communities differed from all substrates but liner communities (pairwise PERMANOVA, P < .05, Fig. S5). These differences decreased over time, as only seawater communities differed from other communities in later stages (≥7 days) of incubation (pairwise PERMANOVA, P < .05). The alpha diversity slightly but significantly decreased over time, with a decrease in PHPB richness for liner and high-mesh PA-net samples, and a decrease in Shannon index for high-mesh PA-net samples. In contrast, hemp PHPB richness showed a drastic 4-fold decrease during the first 14 days (linear model, P < .05; Fig. 3B and C).

Figure 4.

Figure 4.

Open in a new tab

Beta diversity of pathobiome communities according to the incubation time. PCoA plot showing the Bray–Curtis dissimilarities between incubation times (A) and between incubation times according to the substrates (B).

The Proteobacteria phylum represented the main phylum in all substrates except for hemp PHPB communities, which was dominated by the Firmicutes phylum (83% of hemp reads; Fig. S6). The Bacteroidota phylum was an important component of the glass PHPB community (32% of glass reads), but represented less than 4% on all other substrates. At the family level, the Vibrionaceae family was detected as a biomarker of plastic samples by the ANCOM-BC analysis. Vibrionaceae is the main bacterial family in all PHPB communities from plastic substrates, representing respectively 44%, 43%, and 56% of all high-mesh PA-net, liner and mosquito net PHPB reads. The Vibrionaceae family only accounted for 11% of seawater PHPB reads, 7.5% of glass PHPB reads, and 3.2% of hemp PHPB reads. Among this family, the species V. parahaemolyticus, Photobacterium damselae, and V. vulnificus were identified in our samples, although metabarcoding taxonomic identification can be questionable at this resolution. The Staphylococcaceae family also represented a high PHPB proportion in plastic samples compared to other substrates, reaching 22%, 38%, and 7% of all high-mesh PA-net, liner and mosquito net PHPB reads respectively, whereas it accounted for 1% of seawater and glass PHPB reads and 0.05% of hemp PHPB reads.

Cultures and antibiograms

A total of 40 colonies that showed typical morphotypes of thermotolerant coliforms on the ENDO medium were isolated for identification with the MALDI-TOF, 36 were successfully identified. The isolates belonged to five bacterial genera including Shewanella spp. (n = 20), Enterobacter spp. (n = 7), Pseudomonas spp. (n = 5), Klebsiella spp. (n = 2), and Mixta spp. (n = 2). Surprisingly, no E. coli was identified, and Pseudomonas spp. as well as Shewanella spp. were detected even though they do not belong to the Enterobacterales order. Significant differences in richness and diversity between substrates are difficult to evaluate due to the uneven and low number of isolates for some substrates (n = 0 for the liner, n = 4 for the mosquito PA-net, n = 5 for glass, n = 7 for hemp, n = 11 for the high-mesh PA-net, and n = 13 for seawater; Table S2). However, it can be pointed out that the species Shewanella algae was identified on all substrates, except for the liner, and that all identified genera were at least found in seawater samples.

The MAR index was calculated for each isolate (n = 40) and a global mean of 0.24 ± 0.14 was found, indicating that bacteria from this aquaculture farm were in mean resistant to 24% of the tested antibiotics. This result is higher than the threshold of 0.2, above which samples are suggested to be originating from a high-risk source of antibiotic contamination (Krumperman 1983). Plastic substrates showed the lowest MAR index, with significant differences between the high-mesh PA-net and glass isolates. Surprisingly, seawater isolates displayed the highest MAR index, significantly higher than PA-nets (high mesh and mosquito nets) and the hemp net (Fig. 5A, Tukey post hoc test, P < .05).

Figure 5.

Figure 5.

Open in a new tab

Antibotic resistance according to the substrate and the incubation time. (A) MAR index according to the substrate. ANOVA and post hoc Tukey test, ***: P < .001, **: P < .01, and *: P < .05. (B) PCA plot of the antibiotic resistance profiles based on diameters of inhibition, showing the 10 most contributive antibiotics. Amoxiclav: Amoxicillin + clavulanic acid.

The analysis of the antibiotic resistance profiles showed three main profiles according to the substrates from which the bacteria were isolated and were statistically different from one another (pairwise PERMANOVA, P < .05, Fig. 5B): seawater (n = 13), high-mesh PA-net (n = 11), and hemp and glass (n = 12 altogether). Isolates from the mosquito net did not display a significantly different antibiotic resistance profile than isolates from any other substrate, probably due to the low number of isolates from this substrate (n = 4). The antibiotic resistance profile of the seawater isolates antibiotic resistance profile was driven by high levels of resistance against beta-lactams (from the carbapenem, penicillin, and cephalosporin families), whereas the plastic isolates harboured higher resistance against quinolones, and hemp and glass harboured moderate resistance against both beta-lactams and quinolones (Fig. 5B). The bacterial composition at the genus level also affected the resistance profile observed, as the unidentified isolates significantly differed from Shewanella and Enterobacter isolates (pairwise PERMANOVA, P < .05).

The dynamics of MAR index over time could not be investigated due to too low number of isolates obtained for some time points

Discussion

Bacterial biofilms tend to homogenize over time whatever the aquaculture substrate

Solid substrates in marine environments support bacterial biofilms whose composition differs from the surrounding free-living communities. This is due to common ecological traits of attached living bacteria (Dang and Lovell 2015) and emphasizes the recruitment of rare taxa from surrounding seawater to form biofilms on immersed surfaces (Jousset et al. 2017, Song et al. 2022). Our results, obtained in an aquaculture context, are no exception since all communities developed on solid substrates were significantly different from the bacterial communities of seawater (Fig. 2A and B), and this difference increased over time (Fig. S3B). However, there was also a certain community specificity among solid substrates particularly during the early phase e.g. before 72 h of incubation. The two PA-nets (high-mesh and mosquito nets) communities were different from glass communities, which may be explained by the effect of the polymers (Pinto et al. 2019, Vaksmaa et al. 2021). The chemical effect was not the only factor controlling the community structure, as communities from PE-Liner and glass were similar. Such similarity could be explained by similarities in surface properties, particularly in terms of smoothness (Abdalla et al. 2021, Briand et al. 2022), which suggests that the physical properties of the surface may have had a greater impact on bacterial communities than the substrate itself.

Despite some differences in the initial composition of attached communities, their temporal variability tended to converge towards a similar community structure (Fig. 2A and B, Fig. S3A). This shift towards homogeneity was consistent with the hypothesis that environmental parameters and nutrient concentrations more strongly influence the dynamic of biofilm composition than the nature of the substrate itself (Pinto et al. 2019, Basili et al. 2020, Wright et al. 2021, Lemonnier et al. 2022), at least over long periods.

Along the process of succession, the most significant change occurred during the early stages (≤24 h) compared to later stages (≥7 days) of development (Fig. S4A and B). The shift in bacterial communities during biofilm development has been described in the literature as resulting from community specialization according to attachment stages. Early stage bacteria are characterized by particle attachment ability, while following colonizers are characterized by their trophic abilities (Datta et al. 2016, Latva et al. 2022). Autotrophic organisms (as Cyanobacteria) are thus expected to be more abundant on inert substrates as plastics compared to organic substrates as hemp, as observed in our results (Fig. S2). Top-down selection, as predation, can also contribute to variability in communities composition, particularly in late stages (Tobias-Hünefeldt et al. 2021).

However, the early hours of biofilm formation on plastics remain poorly documented in the literature. Latva et al. (2022) described the biofilm composition within the first 15 min of colonization and suggested that the Bacteroidetes group could be the first pioneer of the colonization, before being outcompeted by bacteria from the Proteobacteria phylum. Our initial sampling was carried out 6 h following substrate immersion; our observations are consistent with previous studies describing the rapid colonization of surfaces by Proteobacteria group during the first hours of immersion (Lee et al. 2008, Pollet et al. 2018, Latva et al. 2022). Proteobacteria group was indeed the dominant group detected on all our samples from plastic substrates at 6 h of incubation in the aquaculture ponds. These clades were also highly abundant in biofilms on other solid substrates at similar stages of development, such as on hemp or glass, suggesting that they may act as early colonizers across a range of solid substrates—not just plastics.

PHPB relative abundance decreases over time

The most commonly used method to prevent biofouling in aquaculture is to control the exposure time of equipment, typically by cleaning tanks or replacing nets once biofouling reaches a critical level. These decisions are based on the extent of visible fouling, rather than on any assessment of potential microbiological risk. Our results highlighted the influence of incubation time on PHPB communities with significant differences between early and late stages of incubation (Fig. 4A and B), particularly on solid substrates (Fig. 4B). For each solid substrate, the highest PHPB relative abundance was reached in the first 24 h of incubation, and this proportion of PHPB reads, as well as PHPB alpha diversity significantly decreased with time for several substrates (Fig. 3A–C). These potential pathogens seem to be competitive, opportunistic, and rapid colonizers, explaining their occurrence in young biofilm. The observed decrease of PHPB reads after 72 h seemed to confirm the pioneer status of PHPB communities, which get outcompeted over time by other nonpathogenic communities, more frequent in the environment.

PHPB development on aquaculture equipment did not increase significantly over time, suggesting that more frequent cleaning or equipment turnover may not be necessary. Instead, maintenance schedules should be adapted to the specific socio-economic and environmental context of each farm (Bloecher and Floerl 2021).

Aquaculture plastics do not harbour more PHPB than other solid substrates, but a specific PHPB community

The presence of PHPB in the plastisphere has raised a great interest in the scientific community, suggesting that plastics could be fomites or reservoirs of harmful microorganisms (Viršek et al. 2017, Bowley et al. 2021, Junaid et al. 2022, Naudet et al. 2023). We indeed found in this study a rich (57 ASVs from 47 species) and diverse community of PHPB growing on plastic substrates, although no significant differences were found between plastics and the other substrates’ richness and diversity (Fig. 3B and C). In terms of abundance, plastic samples pooled together were enriched in PHPB compared to seawater in the early stages of incubation but not compared to other substrates, and not in the late stages of incubation. This result suggests that the choice of plastic equipment does not seem to enhance the risk of development of PHPB communities in aquaculture farms in the long term, but a marinization of some equipment could be useful to contribute to decreasing PHPB abundance in biofilms.

Moreover, hemp showed the significantly highest proportion of PHPB reads in its biofilm in the first 72 h of incubation (up to 15%; mean: 3.51 ± 4.86%; Fig. 3A). This can be explained by the intrinsic organic nature of this substrate, providing a nutrient source that is favourable for the development of heterotrophic human bacterial pathogens (Metcalf et al. 2022). The fact that hemp can be presented as an alternative to the use of plastics in aquaculture (Tamburini et al. 2020, Petrocelli et al. 2021) raises the question of the health risks associated with this new equipment.

Although we observed a convergence of compositions in late biofilms communities, plastic pathobiomes in early-stage biofilms distinguish themselves by their taxonomical composition. In the early stages of incubation (≤72 h), plastic substrates showed similar PHPB communities whatever their polymer composition (PA or PE), but were significantly different from seawater and hemp substrate (Fig. S5). Differences in organic and mineral composition between seawater, hemp, and plastic substrates may influence PHPB colonization, influencing the growth of heterotrophic and autotrophic bacteria during early stages of colonization. As suggested before, similarities in surface structure could also influence total and PHPB bacterial communities (Abdalla et al. 2021, Briand et al. 2022), explaining similarities between glass and liner PHPB communities. However, in the late stages of incubation (≥7 days), PHPB communities of all substrates converged and only seawater communities differed from the other ones (Fig. S5).

Among the identified taxonomic families within the DNA metabarcoding pathobiome, one had a particularly high relative abundance in plastic samples: the Vibrionaceae family, detected as a biomarker of plastic samples (Fig. S6). Vibrionaceae members are ubiquitous in marine ecosystems (Haldar 2012, Ina-Salwany et al. 2019), and account for many pathogenic species. Some Vibrionaceae, as the members of the Vibrio and Photobacterium genera, identified in plastic samples, can cause vibriosis in aquatic species, which can be lethal and responsible for huge economic losses in aquaculture systems (Austin and Austin 2016, Stentiford et al. 2017, Ina-Salwany et al. 2019, Zhang et al. 2020). Some Vibrionaceae as V. vulnificus and V. parahaemolyticus, identified here in plastic samples, can also be responsible for human illness, especially foodborne disease (Rivas et al. 2013, Mustapha et al. 2013, Letchumanan et al. 2019, Schröttner et al. 2020). In our samples, its abundance is at least four times higher in plastic samples than in seawater, hemp, or glass samples (Fig. S6), suggesting that plastic substrates could be reservoirs for Vibrionaceae communities in the aquaculture pond. Although, in a lesser fraction, other families such as Staphylococcaceae were detected in high proportions in plastic pathobiomes compared to other substrates. By increasing the contact probability of reared animals to potentially pathogenic bacteria, the use of plastic equipment could thus increase the rate of disease outbreaks in the farm and by human consumers associated to plastic-specific taxa.

Aquaculture plastics do not harbour more ARB than other solid substrates, but a specific resistome

Regarding antibiotic resistance, the global MAR index measured in the samples was 0.24 ± 0.14, thus higher than the conventional threshold of 0.2 considered in ARB studies. Beyond this value, samples are suggested to be originating from a high-risk source of antibiotic contamination (Krumperman 1983). Our results were also higher than MAR index of aquaculture-related bacteria observed in the South Indian Ocean with 0.19 in South Africa (Reverter et al. 2020) and 0.15 in Mauritius (Naudet et al. 2023) but lower than countries in East Indian Ocean such as Indonesia or India and Sri Lanka (with 0.35 and 0.36 values, respectively; Reverter et al. 2020). The 10 most contributing antibiotics belonged to the beta-lactam family (including carbapenems, penicillins, and cephalosporins; Fig. 5B). We did not perform Combined Disk Test to confirm the presence of the extended-spectrum β-lactamases (ESBL) bacteria, but this result is a first element of interest knowing the growing worry concerning ESBL (Nwafia et al. 2022, Husna et al. 2023). The presence of these multiresistant bacteria has raised concern in aquaculture, including fish and invertebrate culture, and in many countries around the world (Sivaraman et al. 2021, Tolentino et al. 2022, Young et al. 2023). The reared aquatic species may become a zoonotic source of ESBL bacteria and disseminate them in human populations from especially low- and middle-income countries such as Madagascar (Olaru et al. 2023).

Regarding to the resistance profiles, the high-mesh PA-net displayed a significantly different resistance profile compared to seawater, hemp, and glass substrates, with more resistance against quinolones (Fig. 5B). However, plastic-associated PHPB had a lower MAR index than glass and seawater (Fig. 5A), which is unexpected since biofilms typically promote antibiotic resistance, and this contrasts with previous studies on aquaculture plastics and seawater (Lu et al. 2019, Zhang et al. 2020, Naudet et al. 2023). The PHPB and resistance profiles of each substrate showed that seawater communities were distinct from solid ones (Fig. 5B and Fig. S5A and B), suggesting that biofilm colonization may stem from the recruitment of rare taxa from seawater, shaping both PHPB and ARB communities (Jousset et al. 2017, Song et al. 2022), with species intrinsically different in terms of resistance, especially during early colonization stages. The isolated strains used to calculate MAR indexes differed between seawater (mostly Pseudomonas sp. and Enterocbacter sp.) and other substrates (mostly Shewanella sp.). The too low isolate count from some substrates and their taxonomical differences could explain the observed differences between seawater and plastic MAR index.

Although our results did not demonstrate a higher risk of antibiotic resistance development associated with the use of plastic equipment compared to glass and hemp in the context of this aquaculture farm, the specificity of the resistance profile observed in plastic substrates suggests that through specific bacteriome selection, plastics could represent a reservoir for some antibacterial resistances in aquaculture environment.

Conclusion

In this study, the objective was to compare the pathobiome and the resistome of different aquaculture equipment over a 3-month incubation in a farm pond. The results show the high influence of substrate composition and incubation time on the development of PHPB and ARB communities. PHPB did not appear more abundant in plastics, but aquaculture plastics showed a specific PHPB communities’ profile, including Vibrionaceae and Staphylococcaceae bacteria, and a specific antibiotic resistance profile, suggesting that they could represent a nonnegligible reservoir of PHPB and ARB in aquaculture contexts. The incubation time had a great influence on PHPB pointing out the importance of finding the right cleaning or replacement frequency, adapted to each aquaculture context. The potential transfer from these plastics to reared animals should still be investigated to explore the role of fomite that plastic could have in aquaculture.

Supplementary Material

fiaf076_Supplemental_Files
fiaf076_supplemental_files.zip (887.4KB, zip)

Acknowledgements

Special thanks to Moustapha Dieng and Pascal Lacroix, Directors of the Indian Ocean Trepang farm in Madagascar, Loïc Gaumez and all the staff, without whom this experiment would not have been possible. They welcomed us and made their ponds available for our experiments in Tulear as long as we needed. We thank the IHSM and the University of Toliara for hosting the microbiological analyses, and the Pasteur Institute of Madagascar for hosting the pathogens identification analyses. We would like to extend our warmest thanks to Jean François BRIAND and Lenaik BELEC from the University of Toulon, MAPIEM laboratory, for stimulating discussions prior to this experiment. We also thank Loic Le Doaré from KERFIL who kindly provided the PA net with high mesh size (10 mm).

Contributor Information

Jeanne Naudet, MARBEC, University of Montpellier, CNRS, Ifremer, IRD, Montpellier 34000, France.

Jean-Christophe Auguet, MARBEC, University of Montpellier, CNRS, Ifremer, IRD, Montpellier 34000, France.

Thierry Bouvier, MARBEC, University of Montpellier, CNRS, Ifremer, IRD, Montpellier 34000, France; Institut Halieutique et des Sciences Marines (IH.SM), Université de Toliara, Toliara, Madagascar.

Raherimino Rakotovao, Institut Halieutique et des Sciences Marines (IH.SM), Université de Toliara, Toliara, Madagascar.

Tony Motte, MARBEC, University of Montpellier, CNRS, Ifremer, IRD, Montpellier 34000, France.

Loïc Gaumez, Indian Ocean Trepang (IOT), Toliara, Madagascar.

Tania Crucitti, Experimental Bacteriology Unit, Institut Pasteur de Madagascar, Antananarivo, Madagascar.

Fabien Rieuvilleneuve, MARBEC, University of Montpellier, CNRS, Ifremer, IRD, Montpellier 34000, France.

Emmanuelle Roque d’Orbcastel, MARBEC, University of Montpellier, CNRS, Ifremer, IRD, Sète 34200, France.

Author contributions

Jeanne Naudet (Conceptualization, Data curation, Formal analysis, Methodology, Writing – original draft), Jean-Christophe Auguet (Investigation, Methodology, Supervision, Validation, Writing – review & editing), Thierry Bouvier (Conceptualization, Project administration, Supervision, Writing – review & editing), Raherimino Rakotovao (Conceptualization, Data curation, Methodology), Tony Motte (Formal analysis, Investigation), Loïc Gaumez (Resources, Validation), Tania Crucitti (Data curation, Methodology, Writing – review & editing), Fabien Rieuvilleneuve (Data curation, Formal analysis, Methodology), and Emmanuelle Roque (Conceptualization, Data curation, Formal analysis, Funding acquisition, Methodology, Supervision, Validation, Writing – review & editing)

Conflict of interest

The authors declare no competing interests that could have appeared to influence the work reported in this paper.

Funding

This study was supported by funds from the ANR-21-CE34-0020 VECTOPLASTIC and AFD-COI ExPLOI projects (to T.B. and E.R.), and the NEMESIS project (2021-EST-149) funded by the French Agency for Food, Environmental and Occupational Health and Safety (ANSES) (to J.C.A.). J.N. was supported by the Ecole Normale Supérieure (ENS) PhD grant.

References

  1. Abdalla  MM, Ali  IAA, Khan  K  et al.  The influence of surface roughening and polishing on microbial biofilm development on different ceramic materials. J Prosthodont. 2021;30:447–53. 10.1111/jopr.13260. [DOI] [PubMed] [Google Scholar]
  2. Amaral-Zettler  LA, Ballerini  T, Zettler  ER  et al.  Diversity and predicted inter- and intra-domain interactions in the Mediterranean plastisphere. Environ Pollut. 2021;286:117439. 10.1016/j.envpol.2021.117439. [DOI] [PubMed] [Google Scholar]
  3. Austin  B, Austin  DA. Vibrios. In: Bacterial Fish Pathogens: Disease of Farmed and Wild Fish. Austin  B, Austin  DA (eds), Cham: Springer International Publishing, 2016, 499–601. 10.1007/978-3-319-32674-0_10. [DOI] [Google Scholar]
  4. Basili  M, Quero  GM, Giovannelli  D  et al.  Major role of surrounding environment in shaping biofilm community composition on marine plastic debris. Front Mar Sci. 2020;7:262. 10.3389/fmars.2020.00262. [DOI] [Google Scholar]
  5. Bloecher  N, Floerl  O. Towards cost-effective biofouling management in salmon aquaculture: a strategic outlook. Rev Aquacult. 2021;13:783–95. 10.1111/raq.12498. [DOI] [Google Scholar]
  6. Bowley  J, Baker-Austin  C, Porter  A  et al.  Oceanic hitchhikers—assessing pathogen risks from marine microplastic. Trends Microbiol. 2021;29:107–16. 10.1016/j.tim.2020.06.011. [DOI] [PubMed] [Google Scholar]
  7. Briand  J-F, Pollet  T, Misson  B  et al.  Surface characteristics together with environmental conditions shape marine biofilm dynamics in coastal NW Mediterranean locations. Front Mar Sci. 2022;8:746383. 10.3389/fmars.2021.746383. [DOI] [Google Scholar]
  8. Cabello  FC. Heavy use of prophylactic antibiotics in aquaculture: a growing problem for human and animal health and for the environment. Environ Microbiol. 2006;8:1137–44. 10.1111/j.1462-2920.2006.01054.x. [DOI] [PubMed] [Google Scholar]
  9. Callahan  BJ, Paul  J, McMurdie  MJ.  et al.  DADA2: high-resolution sample inference from Illumina amplicon data. Nat Methods. 2016;13:581–3. 10.1038/nmeth.3869. [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Cao  Y, Dong  Q, Wang  D  et al.  microbiomeMarker: an R/bioconductor package for microbiome marker identification and visualization. Bioinformatics. 2022;38:4027–9. 10.1093/bioinformatics/btac438. [DOI] [PubMed] [Google Scholar]
  11. Chen  G, Li  Y, Wang  J. Occurrence and ecological impact of microplastics in aquaculture ecosystems. Chemosphere. 2021;274:129989. 10.1016/j.chemosphere.2021.129989. [DOI] [PubMed] [Google Scholar]
  12. Dang  H, Lovell  CR. Microbial surface colonization and biofilm development in marine environments. Microbiol Mol Biol Rev. 2015;80:91–138. 10.1128/mmbr.00037-15. [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Datta  MS, Sliwerska  E, Gore  J  et al.  Microbial interactions lead to rapid micro-scale successions on model marine particles. Nat Commun. 2016;7:11965. 10.1038/ncomms11965. [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Davis  NM, Proctor  DM, Holmes  SP  et al.  Simple statistical identification and removal of contaminant sequences in marker-gene and metagenomics data. Microbiome. 2018;6:226. 10.1186/s40168-018-0605-2. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Delacuvellerie  A, Ballerini  T, Frère  L  et al.  From rivers to marine environments: a constantly evolving microbial community within the plastisphere. Mar Pollut Bull. 2022;179:113660. 10.1016/j.marpolbul.2022.113660. [DOI] [PubMed] [Google Scholar]
  16. Emmerik  Tv, Schwarz  A. Plastic debris in rivers. WIREs Water. 2020;7:e1398. 10.1002/wat2.1398. [DOI] [Google Scholar]
  17. Forster  B, Pinedo  CA. Bacteriological Examination of Waters: Membrane Filtration Protocol. Washington, DC: American Society for Microbiology, 2015. [Google Scholar]
  18. Frère  L, Maignien  L, Chalopin  M  et al.  Microplastic bacterial communities in the Bay of Brest: influence of polymer type and size. Environ Pollut. 2018;242:614–25. 10.1016/j.envpol.2018.07.023. [DOI] [PubMed] [Google Scholar]
  19. Ge  Z, Ma  Z, Zou  J  et al.  Purification of aquaculture wastewater by macrophytes and biofilm systems: efficient removal of trace antibiotics and enrichment of antibiotic resistance genes. Sci Total Environ. 2023;901:165943. 10.1016/j.scitotenv.2023.165943. [DOI] [PubMed] [Google Scholar]
  20. Haldar  C. Vibrio related diseases in aquaculture and development of rapid and accurate identification methods. J Mar Sci Res Dev. 2012;1:002. 10.4172/2155-9910.S1-002. [DOI] [Google Scholar]
  21. Hudzicki  J.  Kirby-Bauer Disk Diffusion Susceptibility Test Protocol. Washington, DC: American Society for Microbiology, 2009. [Google Scholar]
  22. Husna  A, Masudur Rahman  M, Badruzzaman  ATM  et al.  Extended-spectrum β-lactamases (ESBL): challenges and opportunities. Biomedicines. 2023;11:2937. 10.3390/biomedicines11112937. [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Ina-Salwany  MY, Al-saari  N, Mohamad  A  et al.  Vibriosis in fish: a review on disease development and prevention. J Aquat Anim Health. 2019;31:3–22. 10.1002/aah.10045. [DOI] [PubMed] [Google Scholar]
  24. Jousset  A, Bienhold  C, Chatzinotas  A  et al.  Where less may be more: how the rare biosphere pulls ecosystems strings. ISME J. 2017;11:853–62. 10.1038/ismej.2016.174. [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Junaid  M, Ali Siddiqui  J, Sadaf  M  et al.  Enrichment and dissemination of bacterial pathogens by microplastics in the aquatic environment. Sci Total Environ. 2022;830:154720. 10.1016/j.scitotenv.2022.154720. [DOI] [PubMed] [Google Scholar]
  26. Kassambara  A. Rstatix: pipe-friendly framework for basic statistical tests. CRAN, 2023. https://cran.r-project.org/web/packages/rstatix/index.html. [Google Scholar]
  27. Kelly  MR, Whitworth  P, Jamieson  A  et al.  Bacterial colonisation of plastic in the Rockall trough, North-East Atlantic: an improved understanding of the deep-sea plastisphere. Environ Pollut. 2022;305:119314. 10.1016/j.envpol.2022.119314. [DOI] [PubMed] [Google Scholar]
  28. Krumperman  PH. Multiple antibiotic resistance indexing of Escherichia coli to identify high-risk sources of fecal contamination of foods. Appl Environ Microbiol. 1983;46:165–70. 10.1128/aem.46.1.165-170.1983. [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Lafferty  KD, Harvell  CD, Conrad  JM  et al.  Infectious diseases affect marine fisheries and aquaculture economics. Ann Rev Mar Sci. 2015;7:471–96. 10.1146/annurev-marine-010814-015646. [DOI] [PubMed] [Google Scholar]
  30. Latva  M, Dedman  CJ, Wright  RJ  et al.  Microbial pioneers of plastic colonisation in coastal seawaters. Mar Pollut Bull. 2022;179:113701. 10.1016/j.marpolbul.2022.113701. [DOI] [PubMed] [Google Scholar]
  31. Lear  L, Padfield  D, Dowsett  T  et al.  Bacterial colonisation dynamics of household plastics in a coastal environment. Sci Total Environ. 2022;838:156199. 10.1016/j.scitotenv.2022.156199. [DOI] [PubMed] [Google Scholar]
  32. Lee  J-W, Nam  J-H, Kim  Y-H  et al.  Bacterial communities in the initial stage of marine biofilm formation on artificial surfaces. J Microbiol. 2008;46:174–82. 10.1007/s12275-008-0032-3. [DOI] [PubMed] [Google Scholar]
  33. Lemonnier  C, Chalopin  M, Huvet  A  et al.  Time-series incubations in a coastal environment illuminates the importance of early colonizers and the complexity of bacterial biofilm dynamics on marine plastics. Environ Pollut. 2022;312:119994. 10.1016/j.envpol.2022.119994. [DOI] [PubMed] [Google Scholar]
  34. Letchumanan  V, Loo  K-Y, Law  JW-F  et al.  Vibrio parahaemolyticus: the protagonist of foodborne diseases. Progr Microbes Mol Biol. 2019;2. 10.36877/pmmb.a0000029. [DOI] [Google Scholar]
  35. Liang  H, Haan  WPd, Cerdà-Domènech  M  et al.  Detection of faecal bacteria and antibiotic resistance genes in biofilms attached to plastics from human-impacted coastal areas. Environ Pollut. 2023;319:120983. 10.1016/j.envpol.2022.120983. [DOI] [PubMed] [Google Scholar]
  36. Lu  J, Zhang  Y, Wu  J  et al.  Effects of microplastics on distribution of antibiotic resistance genes in recirculating aquaculture system. Ecotoxicol Environ Saf. 2019;184:109631. 10.1016/j.ecoenv.2019.109631. [DOI] [PubMed] [Google Scholar]
  37. Mah  T-F. Biofilm-specific antibiotic resistance. Fut Microbiol. 2012;7:1061–72. 10.2217/fmb.12.76. [DOI] [PubMed] [Google Scholar]
  38. Martínez-Campos  S, Pissaridou  P, Drakou  K  et al.  Evolution of prokaryotic colonisation of greenhouse plastics discarded into the environment. Ecotoxicol Environ Saf. 2022;232:113213. 10.1016/j.ecoenv.2022.113213. [DOI] [PubMed] [Google Scholar]
  39. McMurdie  PJ, Holmes  S. Phyloseq: an R package for reproducible interactive analysis and graphics of microbiome census data. PLoS One. 2013;8:e61217. 10.1371/journal.pone.0061217. [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. Metcalf  R, Oliver  DM, Moresco  V  et al.  Quantifying the importance of plastic pollution for the dissemination of human pathogens: the challenges of choosing an appropriate “control” material. Sci Total Environ. 2022;810:152292. 10.1016/j.scitotenv.2021.152292. [DOI] [PubMed] [Google Scholar]
  41. Mustapha  S, Moulay Mustapha  E, Nozha  C. Vibrio alginolyticus: an emerging pathogen of foodborne diseases. Maejo Int J Sci Technol. 2013;2:9. [Google Scholar]
  42. Naudet  J, Roque d’Orbcastel  E, Bouvier  T  et al.  Identifying macroplastic pathobiomes and antibiotic resistance in a subtropical fish farm. Mar Pollut Bull. 2023;194:115267. 10.1016/j.marpolbul.2023.115267. [DOI] [PubMed] [Google Scholar]
  43. Ngajilo  D, Jeebhay  MF. Occupational injuries and diseases in aquaculture—a review of literature. Aquaculture. 2019;507:40–55. 10.1016/j.aquaculture.2019.03.053. [DOI] [Google Scholar]
  44. Nwafia  IN, Ike Orabueze  AC IN, Nwafia  WC. Carbapenemase producing Enterobacteriaceae: environmental reservoirs as primary targets for control and prevention strategies. Niger Postgrad Med J. 2022;29:183–91. 10.4103/npmj.npmj_95_22. [DOI] [PubMed] [Google Scholar]
  45. Oberbeckmann  S, Osborn  AM, Duhaime  MB. Microbes on a bottle: substrate, season and geography influence community composition of Microbes colonizing marine plastic debris. PLoS One. 2016;11:e0159289. 10.1371/journal.pone.0159289. [DOI] [PMC free article] [PubMed] [Google Scholar]
  46. Oksanen  J, Simpson  GL, Guillaume Blanchet  F  et al.  Vegan: community ecology package. CRAN, 2022. https://CRAN.R-project.org/package=vegan. [Google Scholar]
  47. Olaru  ID, Walther  B, Schaumburg  F. Zoonotic sources and the spread of antimicrobial resistance from the perspective of low and middle-income countries. Infect Dis Poverty. 2023;12:59. 10.1186/s40249-023-01113-z. [DOI] [PMC free article] [PubMed] [Google Scholar]
  48. Parada  AE, Needham  DM, Fuhrman  JA. Every base matters: assessing small subunit rRNA primers for marine microbiomes with mock communities, time series and global field samples. Environ Microbiol. 2016;18:1403–14. 10.1111/1462-2920.13023. [DOI] [PubMed] [Google Scholar]
  49. Petrocelli  A, Portacci  G, De Gasperis  E  et al.  Preliminary results of plastic substitution in mussel farming with natural vegetal fibers. In: Proceedings of the 2021 International Workshop on Metrology for the Sea; Learning to Measure Sea Health Parameters (MetroSea). Piscataway, NJ: IEEE, 2021, 366–70. 10.1109/MetroSea52177.2021.9611562. [DOI] [Google Scholar]
  50. Pinto  M, Langer  TM, Hüffer  T  et al.  The composition of bacterial communities associated with plastic biofilms differs between different polymers and stages of biofilm succession. PLoS One. 2019;14:e0217165. 10.1371/journal.pone.0217165. [DOI] [PMC free article] [PubMed] [Google Scholar]
  51. Pollet  T, Berdjeb  L, Garnier  C  et al.  Prokaryotic community successions and interactions in marine biofilms: the key role of flavobacteriia. FEMS Microbiol Ecol. 2018;94:fiy083. 10.1093/femsec/fiy083. [DOI] [PubMed] [Google Scholar]
  52. Reverter  M, Sarter  S, Caruso  D  et al.  Aquaculture at the crossroads of global warming and antimicrobial resistance. Nat Commun. 2020;11:1870. 10.1038/s41467-020-15735-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
  53. Rivas  A, Lemos  M, Osorio  C. Photobacterium damselae subsp. damselae, a bacterium pathogenic for marine animals and humans. Front Microbiol. 2013;4. 10.3389/fmicb.2013.00283. [DOI] [PMC free article] [PubMed] [Google Scholar]
  54. Sathicq  MB, Sabatino  R, Corno  G  et al.  Are microplastic particles a hotspot for the spread and the persistence of antibiotic resistance in aquatic systems?. Environ Pollut. 2021;279:116896. 10.1016/j.envpol.2021.116896. [DOI] [PubMed] [Google Scholar]
  55. Schröttner  P, Tille  E, Lück  C  et al.  Wound infection caused by Photobacterium damselae in a 32-year-old woman: case report and review of the literature. GMS Infect Dis. 2020;8:Doc23. 10.3205/id000067. [DOI] [PMC free article] [PubMed] [Google Scholar]
  56. Sivaraman  GK, Rajan  V, Vijayan  A  et al.  Antibiotic resistance profiles and molecular characteristics of extended-spectrum beta-lactamase (ESBL)-producing Escherichia coli and Klebsiella pneumoniae isolated from shrimp aquaculture farms in Kerala, India. Front Microbiol. 2021;12:622891. 10.3389/fmicb.2021.622891. [DOI] [PMC free article] [PubMed] [Google Scholar]
  57. Song  J, Beule  L, Jongmans-Hochschulz  E  et al.  The travelling particles: community dynamics of biofilms on microplastics transferred along a salinity gradient. ISME Commun. 2022;2:1–12. 10.1038/s43705-022-00117-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
  58. Stenger  KS, Wikmark  OG, Bezuidenhout  CC  et al.  Microplastics pollution in the ocean: potential carrier of resistant bacteria and resistance genes. Environ Pollut. 2021;291:118130. 10.1016/j.envpol.2021.118130. [DOI] [PubMed] [Google Scholar]
  59. Stentiford  GD, Sritunyalucksana  K, Flegel  TW  et al.  New paradigms to help solve the global aquaculture disease crisis. PLoS Pathog. 2017;13:e1006160. 10.1371/journal.ppat.1006160. [DOI] [PMC free article] [PubMed] [Google Scholar]
  60. Tamburini  E, Turolla  E, Fano  EA  et al.  Sustainability of mussel (Mytilus galloprovincialis) farming in the Po River Delta, Northern Italy, based on a life cycle assessment approach. Sustainability. 2020;12:3814. 10.3390/su12093814. [DOI] [Google Scholar]
  61. Tobias-Hünefeldt  SP, Wenley  J, Baltar  F  et al.  Ecological drivers switch from bottom–up to top–down during model microbial community successions. ISME J. 2021;15:1085. 10.1038/s41396-020-00833-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
  62. Tolentino  ARC, Opeda  ARP, Naniong  LL  et al.  Prevalence of blaCTX-M, blaTEM, and blaSHV genes of extended-spectrum beta lactamase-producing Escherichia coli in aquaculture in Asia: a systematic review. Asian J Biol Life Sci. 2022;11. 10.5530/ajbls.2022.11.31. [DOI] [Google Scholar]
  63. Trachoo  N. Biofilm removal technique using sands as a research tool for accessing microbial attachment on surface. J Sci Technol. 2004;26. [Google Scholar]
  64. Uruén  C, Chopo-Escuin  G, Tommassen  J  et al.  Biofilms as promoters of bacterial antibiotic resistance and tolerance. Antibiotics. 2021;10:3. 10.3390/antibiotics10010003. [DOI] [PMC free article] [PubMed] [Google Scholar]
  65. Vaksmaa  A, Knittel  K, Abdala Asbun  A  et al.  Microbial communities on plastic polymers in the Mediterranean Sea. Front Microbiol. 2021;12. 10.3389/fmicb.2021.673553. [DOI] [PMC free article] [PubMed] [Google Scholar]
  66. Viršek  MK, Lovšin  MN, Koren  Šp  et al.  Microplastics as a vector for the transport of the bacterial fish pathogen species Aeromonas salmonicida. Mar Pollut Bull. 2017;125:301–9. 10.1016/j.marpolbul.2017.08.024. [DOI] [PubMed] [Google Scholar]
  67. Wardeh  M, Risley  C, McIntyre  MK  et al.  Database of host-pathogen and related species interactions, and their global distribution. Sci Data. 2015;2:150049. 10.1038/sdata.2015.49. [DOI] [PMC free article] [PubMed] [Google Scholar]
  68. Wright  RJ, Langille  MGI, Walker  TR. Food or just a free ride? A meta-analysis reveals the global diversity of the plastisphere. ISME J. 2021;15:789–806. 10.1038/s41396-020-00814-9. [DOI] [PMC free article] [PubMed] [Google Scholar]
  69. Yang  Y, Liu  G, Song  W  et al.  Plastics in the marine environment are reservoirs for antibiotic and metal resistance genes. Environ Int. 2019;123:79–86. 10.1016/j.envint.2018.11.061. [DOI] [PubMed] [Google Scholar]
  70. Young  KM, Isada  MJ, Reist  M  et al.  A scoping review of the distribution and frequency of extended-spectrum β-lactamase (ESBL)-producing Enterobacteriaceae in shrimp and salmon. Epidemiol Infect. 2023;151:e1. 10.1017/S0950268822001819. [DOI] [PMC free article] [PubMed] [Google Scholar]
  71. Yu  X, Zhang  Y, Tan  L  et al.  Microplastisphere may induce the enrichment of antibiotic resistance genes on microplastics in aquatic environments: a review. Environ Pollut. 2022;310:119891. 10.1016/j.envpol.2022.119891. [DOI] [PubMed] [Google Scholar]
  72. Zettler  ER, Mincer  TJ, Amaral-Zettler  LA. Life in the “plastisphere”: microbial communities on plastic marine debris. Environ Sci Technol. 2013;47:7137–46. 10.1021/es401288x. [DOI] [PubMed] [Google Scholar]
  73. Zhang  Y, Lu  J, Wu  J  et al.  Potential risks of microplastics combined with superbugs: enrichment of antibiotic resistant bacteria on the surface of microplastics in mariculture system. Ecotoxicol Environ Saf. 2020;187:109852. 10.1016/j.ecoenv.2019.109852. [DOI] [PubMed] [Google Scholar]
  74. Zhou  A, Zhang  Y, Xie  S  et al.  Microplastics and their potential effects on the aquaculture systems: a critical review. Rev Aquacult. 2021;13:719–33. 10.1111/raq.12496. [DOI] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

fiaf076_Supplemental_Files
fiaf076_supplemental_files.zip (887.4KB, zip)

Articles from FEMS Microbiology Ecology are provided here courtesy of Oxford University Press

RESOURCES