Systematic Survey of Vibrio spp. and Salmonella spp. in Bivalve Shellfish in Apulia Region (Italy): Prevalence and Antimicrobial Resistance

The emergence of antimicrobial resistance (AMR) is increasingly common across the globe and aquatic ecosystems could be considered a reservoir of antibiotic-resistant bacteria. This study aimed to determine prevalence and antibiotic susceptibility of the potential pathogenic bacteria Salmonella spp. and Vibrio spp. in bivalve molluscs intended for human consumption, collected over a period of 19 months along the northern coast of Apulia region. The AMR profile was also determined in non-pathogenic Vibrio species, common natural inhabitants of seawater and a useful indicator for the surveillance of AMR in the environment. The current study presents data on the AMR of 5 Salmonella and 126 Vibrio isolates by broth microdilution MIC. Multidrug resistance (MDR) was observed in one S. Typhimurium strain towards sulfamethoxazole, trimethoprim, tetracycline, gentamicin, and ampicillin and in 41.3% of the Vibrio strains, mostly towards sulphonamides, penicillin, and cephems. All Vibrio isolates were sensitive to azithromycin, chloramphenicol, tetracycline, amoxicillin/clavulanic acid, gentamicin, streptomycin, amikacin, and levofloxacin. The AMR phenomenon in the investigated area is not highly worrying but not entirely negligible; therefore, in-depth continuous monitoring is suggested. Results concerning the antibiotic agents without available specific clinical breakpoints could be useful to upgrade the MIC distribution for Vibrio spp. but, also, the establishment of interpretative criteria for environmental species is necessary to obtain a more complete view of this issue.


Introduction
The global increasingly spreading antimicrobial resistance (AMR) is such an alarming current issue that many scientific activities have been implemented worldwide to analyse the phenomenon and to conceive effective contrast measures. Several bacterial species are involved in this phenomenon and the zoonotic ones are clearly of greater concern. Resistant bacteria can affect humans directly through consumption of contaminated food or indirectly by transferring mobile genetic elements (e.g., plasmids, transposable elements, super-integron, and integrating conjugative elements genes) for antibiotic resistance to human pathogens [1,2]. Antibiotic-resistant bacteria (ARB) have emerged in the marine environment as a consequence of the excessive use of antibiotics in human, agriculture, and aquaculture systems during the past few decades [3]. Despite the use of antimicrobials in aquaculture being considered a leading cause of development of ARB [1], only five active substances are approved and registered in Italy: amoxicillin, flumequine, oxytetracycline, chlortetracycline, and sulfadiazine-trimethoprim.
Vibrio species could represent an important indicator of the presence of antibiotic resistance in the marine and estuarine ecosystems, as they are natural inhabitants of coastal

Sampling
Overall, 296 shellfish samples were collected from March 2021 to October 2022 along the northern Apulian coast (provinces of Foggia and Barletta-Andria-Trani). Most of them (263/296; 88.85%) derived from the implementation of the official classification programme of bivalve mollusc production and harvesting areas with regard to the Commission Implementing Regulation 2019/627 [21]. The geographical distribution of sampling points was established after a sanitary survey in order to choose the location at highest risk of faecal pollution and ensure that analytical results were representative of the area. At least 12 samples were taken from each sampling point over at least a 6-month period, as recommended in the community guide to the principles of good practice for the microbiological classification and monitoring of bivalve mollusc production and relaying areas with regard to Implementing Regulation 2019/627 [22]. The interval between two sampling occasions was approximatively 2 weeks, depending on the weather conditions and the availability of a commercial-size product. It is important to highlight that search for Salmonella is mandatory according to the European legislation [23]; otherwise, Vibrio detection was carried out only for research purposes.
The remaining samples (33/296; 11.15%) consisted of live bivalve molluscs originated from the same investigated areas and sampled for official control in accordance with the Integrated Regional Control Plan of Apulia region; they were collected in purification and dispatch centres or at retail and also tested for the aim of this study.
As regards the specimens collected for the official classification purpose, they consisted of oysters (Crassostrea gigas), mussels (Mytilus galloprovinciallis and Modiolus barbatus), clams (Venus gallina/Chamelea gallina), cockles (Acanthocardia tuberculata), and Japanese carpet shells (Ruditapes philippinarum). They were collected from fixed sampling stations, transported to the laboratory on the same day at temperatures between 4 and 10 • C and processed within 24 h of arrival. Environmental parameters, such as water temperature ( • C) and pH, were measured on site during the sampling. The sampling stations are illustrated in Figure 1. subsequent evaluation of the AMR profile of isolates, since antimicrobial resistance surveillance in the marine environment is a crucial aspect to implement effective local antibiotic reduction programs.

Sampling
Overall, 296 shellfish samples were collected from March 2021 to October 2022 along the northern Apulian coast (provinces of Foggia and Barletta-Andria-Trani). Most of them (263/296; 88.85%) derived from the implementation of the official classification programme of bivalve mollusc production and harvesting areas with regard to the Commission Implementing Regulation 2019/627 [21]. The geographical distribution of sampling points was established after a sanitary survey in order to choose the location at highest risk of faecal pollution and ensure that analytical results were representative of the area. At least 12 samples were taken from each sampling point over at least a 6-month period, as recommended in the community guide to the principles of good practice for the microbiological classification and monitoring of bivalve mollusc production and relaying areas with regard to Implementing Regulation 2019/627 [22]. The interval between two sampling occasions was approximatively 2 weeks, depending on the weather conditions and the availability of a commercial-size product. It is important to highlight that search for Salmonella is mandatory according to the European legislation [23]; otherwise, Vibrio detection was carried out only for research purposes.
The remaining samples (33/296; 11.15%) consisted of live bivalve molluscs originated from the same investigated areas and sampled for official control in accordance with the Integrated Regional Control Plan of Apulia region; they were collected in purification and dispatch centres or at retail and also tested for the aim of this study.
As regards the specimens collected for the official classification purpose, they consisted of oysters (Crassostrea gigas), mussels (Mytilus galloprovinciallis and Modiolus barbatus), clams (Venus gallina/Chamelea gallina), cockles (Acanthocardia tuberculata), and Japanese carpet shells (Ruditapes philippinarum). They were collected from fixed sampling stations, transported to the laboratory on the same day at temperatures between 4 and 10 °C and processed within 24 h of arrival. Environmental parameters, such as water temperature (°C) and pH, were measured on site during the sampling. The sampling stations are illustrated in Figure 1.  Detailed information about mollusc species and sampling points are reported in Table 1.

Vibrio Detection and Antimicrobial Susceptibility Testing
Vibrio spp. detection was performed according to the standard ISO 21872-1:2017 [24] by preparing two enrichment broths, which were incubated at 37 ± 1 • C and 41.5 ± 1 • C, respectively, to enhance the recovery of most Vibrio species. All the presumptive Vibrio spp. isolates were identified at species level based on the API ® ID 20E and 20NE systems (BioMérieux, Nürtingen, Germany), the halotolerance test with various concentrations of NaCl (0, 6 and 10%), and the MALDI-TOF MS (Bruker Daltonics, Bremen, Germany) procedure by the direct transfer method, as previously described [25]. Moreover, the isolates recognised as V. parahaemolyticus, V. cholerae, and V. vulnificus were confirmed by the conventional PCR method described in Annex C to ISO 21872-1:2017, which considers, also, the detection of V. parahaemolyticus virulence genes (tdh and trh).
The bacterial inoculum was prepared by dissolving a fresh pure colony in 2.5% NaCl solution using a sterile cotton swab until it achieved the turbidity of the 0.5 McFarland standard. Then, 0.1 mL of this suspension was added to 9.9 mL of cation-adjusted Mueller-Hinton broth (Becton Dickinson, Milan, Italy), the antimicrobial plate wells were inoculated with 50 µL of this suspension, and the plate was incubated aerobically at 36 ± 1 • C for 24 h. Escherichia coli ATCC ® 25922 was used as quality control in each batch. The results were interpreted according to CLSI clinical breakpoints specific for Vibrio spp. No CLSI breakpoints were available for the following agents: ceftriaxone, nalidixic acid, streptomycin, tigecycline, ticarcillin/clavulanic acid, nitrofurantoin, doripenem, minocycline, ertapenem, tobramycin, and aztreonam. Hence, results of broth microdilution assays referring to the above-mentioned molecules are reported in the present study without any interpretative criteria. Only for streptomycin, the breakpoints described by other authors were used [27,28].
The epidemiological cut-off value (ECOFF) indicated in the Commission Implementing Decision (EU) 2020/1729 [18] was used as the interpretation threshold of AMR, except for colistin and tigecycline, for which the values defined for Enterobacteriales in the EUCAST clinical breakpoint table [31] were chosen, whereas the MIC breakpoints stated in CLSI document M100 [32] were used for azithromicyn and sulfamethoxazole (MIC values for sulphonamides). Multidrug resistance (MDR) was defined as nonsusceptibility to at least one agent in three or more antimicrobial categories [33].
Finally, the multiple antibiotic resistance (MAR) index was calculated as the ratio between the number of antibiotics to which a strain was resistant to and the total number of antibiotics used [34].

Salmonella Isolates and Their Antimicrobial Resistance Profile
Among the 296 live bivalve molluscs analysed, only 1.7% of them were contaminated with Salmonella spp. Overall, five strains were isolated and tested for susceptibility to antimicrobial agents: three strains from clams (S. enterica subsp. enterica Kasenyi, S. enterica subsp. enterica Typhimurium, and S. bongori 48:z35:-) and two strains from mussels (S. enterica subsp. enterica Fischerhütte and S. enterica subsp. enterica Typhimurium). All isolates were sensitive to almost all antibiotics, except for S. Typhimurium isolated from Venus gallina, which showed resistance to sulfamethoxazole, trimethoprim, tetracycline, gentamicin, and ampicillin. Instead, S. Kasenyi and S. bongori 48:z35:-were resistant only to sulfamethoxazole.

Vibrio Isolates and Their Antimicrobial Resistance Profile
In total, 126 Vibrio strains (mostly V. alginolyticus and V. parahaemolyticus) were detected, with 38.2% (113/296) of shellfish samples contaminated with them. The API 20E system and the MALDI-TOF MS identification results were consistent, whereas the API 20NE displayed disputable and unreliable results. Furthermore, it is well known that the biochemical tests are inadequate for an accurate identification of V. harveyi [12,35] and that profiles for V. harveyi are not included in the bioMérieux database; thus, the API 20E and API 20NE systems do not allow this species to be properly identified via Apiweb™ and the identification of V. harveyi in this study was provided by the MALDI-TOF MS system. Vibrio spp. strains have been detected in shellfish collected from all sampling stations, with the exception of point 13. The temperature range for all isolates was of 9.4-29.9 • C, but isolation of potentially enteropathogenic species (V. cholerae, V. parahaemolyticus, and V. vulnificus) occurred when temperature ranged between 19.5 and 29.9 • C. The correlation between source and Vibrio species is shown in Figure 2. V. alginolyticus was the most prevalent species, found mostly in Mytilus galloprovincialis, followed by Venus gallina/Chamelea gallina. Finally, the multiple antibiotic resistance (MAR) index was calculated as the ratio between the number of antibiotics to which a strain was resistant to and the total number of antibiotics used [34].

Salmonella Isolates and Their Antimicrobial Resistance Profile
Among the 296 live bivalve molluscs analysed, only 1.7% of them were contaminated with Salmonella spp. Overall, five strains were isolated and tested for susceptibility to antimicrobial agents: three strains from clams (S. enterica subsp. enterica Kasenyi, S. enterica subsp. enterica Typhimurium, and S. bongori 48:z35:-) and two strains from mussels (S. enterica subsp. enterica Fischerhütte and S. enterica subsp. enterica Typhimurium). All isolates were sensitive to almost all antibiotics, except for S. Typhimurium isolated from Venus gallina, which showed resistance to sulfamethoxazole, trimethoprim, tetracycline, gentamicin, and ampicillin. Instead, S. Kasenyi and S. bongori 48:z35:-were resistant only to sulfamethoxazole.

Vibrio Isolates and Their Antimicrobial Resistance Profile
In total, 126 Vibrio strains (mostly V. alginolyticus and V. parahaemolyticus) were detected, with 38.2% (113/296) of shellfish samples contaminated with them. The API 20E system and the MALDI-TOF MS identification results were consistent, whereas the API 20NE displayed disputable and unreliable results. Furthermore, it is well known that the biochemical tests are inadequate for an accurate identification of V. harveyi [12,35] and that profiles for V. harveyi are not included in the bioMérieux database; thus, the API 20E and API 20NE systems do not allow this species to be properly identified via Apiweb™ and the identification of V. harveyi in this study was provided by the MALDI-TOF MS system. Vibrio spp. strains have been detected in shellfish collected from all sampling stations, with the exception of point 13. The temperature range for all isolates was of 9.4-29.9 °C, but isolation of potentially enteropathogenic species (V. cholerae, V. parahaemolyticus, and V. vulnificus) occurred when temperature ranged between 19.5 and 29.9 °C. The correlation between source and Vibrio species is shown in Figure 2. V. alginolyticus was the most prevalent species, found mostly in Mytilus galloprovincialis, followed by Venus gallina/Chamelea gallina.   Table 2 by specifying Vibrio species, resistance pattern, and source.  Table 3.  Overall, 41.3% (52/126) of strains displayed MDR, mostly towards sulphonamides, penicillin, and cephems. The most prevalent MDR profile was FIS-AMP-PIP-FAZ, followed by FIS-AMP-FAZ, both of them found primarily in V. alginolyticus and secondarily in V. parahaemolyticus.
The MIC values are shown in Supplementary Table S2, including those relating to the antibiotic agents without available specific CLSI breakpoints.

Discussion
Overall, few shellfish samples (5/296) have been found to be contaminated with Salmonella spp. in the current study and only one S. Typhimurium strain showed multidrug resistance. Despite the low prevalence of Salmonella, the detection of pathogenic serovars raises concerns about consumer health, especially given the recent finding of S. Typhimurium strains of human origin, collected from the same region and resistant mainly to ampicillin, tetracycline, azithromycin, and sulfamethoxazole [17], which are the same antimicrobials depicted in our multidrug-resistant strain, except for azithromycin replaced with gentamicin. Giacometti et al. [36] carried out a similar investigation on 102 Salmonella isolates from bivalve molluscs and water samples collected during the official monitoring programme in the area of the province of Ferrara (northwestern area of the Adriatic Sea) between 2001 and 2017. The most common resistances observed by the authors were to streptomycin (58.8%), ampicillin (52%), and tetracycline (45.1%), whereas 44.12% of isolates were MDR. Thus, it seems to suggest that resistance to these antimicrobial classes (penicillin, aminoglycoside, and tetracycline) is widely spread in the Adriatic Sea, in accordance with the AMR profile exhibited by our S. Typhimurium strain. However, any comparison with our results is difficult due to the long period of sampling (7 years), the type of samples (molluscs and water), and the method used to evaluate the AMR (agar disk diffusion). It is easy to understand that the AMR patterns vary according to the geographical origin of the samples; indeed, results concerning 27 Salmonella strains isolated by Lozano-Leon et al. [37] from mussel samples harvested in Galicia are completely different: all isolates showed MDR and were resistant to cefuroxime and cefuroxime/axetil; the majority of them expressed resistance to cefoxitin and gentamicin and some resistance was observed towards ampicillin, amikacin, cephalothin, and tobramycin, although the antimicrobial susceptibility was performed only on some strains.
As regards Vibrio spp., three different identification techniques were performed simultaneously in the present work to obtain a more reliable result, given that differentiation of closely related Vibrio species can be difficult. Although MALDI TOF MS is considered a valid tool, its discriminative power depends on the fullness of the reference library, which mainly contains clinically relevant species; thus, the reference library could be increased by inserting environmental bacterial isolates, especially marine bacteria, to make the MALDI TOF MS result more consistent. The API 20E and MALDI TOF MS systems gave the same results, except for the V. harveyi species because of the lack in the bioMérieux database. Therefore, the MALDI TOF identification of this Vibrio species was considered reliable and might perhaps be supported by biomolecular analysis, such as whole genome sequencing, which offers a great identification accuracy. Vibrio detection occurred during the entire investigation period but it is worth noting that potentially enteropathogenic species were isolated when water temperature was above 19.5 • C. It is well known that occurrence and densities of V. parahaemolyticus in molluscs are positively correlated to water temperatures [38] and outbreaks occur mainly during the warmer months in temperate zones [39]. However, the increase in surface temperature in coastal European seas in recent years has been linked to outbreaks caused by Vibrio cholerae nonO1-nonO139, V. parahaemolyticus, and V. vulnificus in several European countries [40]. Among potentially enteropathogenic species, V. parahaemolyticus was the most frequent one, isolated mainly from Mytilus galloprovincialis, but the only strain harbouring trh gene, hence proving to be harmful to humans, was isolated from Venus gallina o Chamelea gallina. V. cholerae isolation from Acanthocardia tuberculata cockles is a rare finding and certainly noteworthy given the scarcity of studies concerning this shellfish, mostly carried out in Morocco. For example, Boutaib et al. [41] found Salmonella spp. and non-pathogenic Vibrio parahaemolyticus in seven and five Acanthocardia tuberculata samples, respectively, but no results concerned V. cholerae, since this species was not included in the study.
In the current survey, the majority of Vibrio strains displayed resistance to ampicillin, sulfisoxazole, and cefazolin. Resistance to ampicillin was prevalent, as in previous studies [1,2,42,43].
If we compare our results with similar studies previously conducted in Italy, data seem to be quite varied. For instance, in the study carried out by Ottaviani et al. in 2001 [44] on several Vibrio species isolated from seafood, all strains exhibited susceptibility to imipenem, meropenem, chloramphenicol, and tetracycline, except 58% of V. alginolyticus, which were resistant to tetracycline; more than 90% of isolates showed susceptibility to oxolinic acid, cefotaxime, flumequine, doxycycline, and trimethoprim-sulphamethoxazole and more than 80% to nalidixic acid and ciprofloxacin. Moreover, resistance to streptomycin and lincomycin was found in more than 90% of isolates and many strains of V. alginolyticus, V. harveyi, V. vulnificus, and V. parahaemolyticus were resistant to penicillin, carbenicillin, ampicillin, cephalothin, and kanamycin, while V. alginolyticus and V. parahaemolyticus strains were resistant to rifampicin. A later investigation of antimicrobial resistance in V. parahaemolyticus from indigenous bivalves collected from harvesting areas along Italian coasts of the south Adriatic Sea, the central Tyrrhenian Sea, and the central Adriatic Sea reported that all isolates were resistant to ampicillin and amoxicillin but no resistances were observed to chloramphenicol, tetracycline, oxytetracycline, doxycycline, and trimethoprim/sulfamethoxazole [45]. Some strains exhibited resistance to cefotaxime (24.1%), cefalothin (43.7%), cefalexin (67.8%), colistin sulphate (13.8%), erythromycin (20.7%), and streptomycin (32.2%), whereas very low resistance percentages were found towards polymyxin B, nalidixic acid, oxolinic acid, nitrofurantoin, ciprofloxacin, kanamycin, and neomycin. Lopatek et al. [28] investigated the antimicrobial susceptibility of V. parahaemolyticus strains isolated from different species of raw shellfish and marine fish originated from various countries. They found that the majority of the strains were resistant to ampicillin and streptomycin and were recovered mainly from Italian samples; resistance to gentamicin was found in 12.5% of the strains, isolated from Italian, Dutch, and Norwegian samples, whereas only one strain was resistant to ciprofloxacin, isolated from Italian clams.
It should be noted that the majority of the studies on AMR in Vibrio spp. performed the agar disk diffusion method; thus, few works can be properly compared with our results.
In the current study, the V. vulnificus strain exhibited a poor resistance pattern, showing resistance only to cefazolin and an intermediate resistance to cefoxitin. However, Baker-Austin et al. [46] found numerous coastal and, also, septicaemia isolates in the USA resistant to antibiotics routinely prescribed for V. vulnificus infections, such as doxycycline, tetracycline, aminoglycosides, and cephalosporins, thus suggesting the importance of continued monitoring. In contrast, Bier et al. [27] reported that most antimicrobial agents recommended for treatment of V. vulnificus and V. cholerae non-O1/non-O139 infections were effective in vitro; likewise, in our study, both V. vulnificus and V. cholerae isolates did not display any worrisome resistance.
Banerjee and Farber [42] characterised 1021 Vibrio strains isolated from molluscs harvested in Canada between 2006 and 2012 and found that only 4.9% of them were sensitive to all tested drugs, while the antibiotics contributing the most to AMR were ampicillin, cephalothin, erythromycin, kanamycin, and streptomycin, although a declining trend in the frequency of MDR/AMR Vibrio spp. was registered until 2012.
Recently, Chahouri et al. [47] conducted a similar microbiological investigation in the Agadir Bay (Morocco), performing the search and AMR characterisation of Vibrio and Salmonella strains from mussels, sediment, and water samples. In accordance with our results, they isolated Vibrio strains at a high frequency, while a low percentage was noted for Salmonella. The eight Salmonella isolates showed resistance to ampicillin (100%), chloramphenicol (87.5%), and amoxicillin/clavulanic acid (62.5%) but were sensitive to all the other antibiotics used. Vibrio strains were mainly resistant to ampicillin (57.7%), cephalothin (62%), amikacin (60.6%), and, to a lesser extent, to ciprofloxacin (26.8%). Moreover, the authors perfectly agree with us when they highlight the importance of environmental survey to properly assess the microbiological quality in aquatic ecosystems.
In our study, almost all Vibrio strains exhibited MAR index values ranging from 0.064 to 0.193, below the arbitrary value of 0.2, indicating low-risk contamination sites [34], with the only exception of one V. alginolyticus isolate, which was resistant to seven antibiotics (MAR = 0.225). Since indices between 0.20 and 0.25 are in a range of ambiguity [34] and no Vibrio strains displayed MAR indices above 0.25, it could be stated that isolates originated from a low-risk environment where antibiotics are not regularly used perhaps. Moreover, resistance was observed mostly towards unusual antimicrobial compounds, such as ampicillin, piperacillin, sulfisoxazole, and the first-generation cephalosporin cefazolin, while few strains exhibited resistance to trimethoprim/sulfamethoxazole, piperacillin/tazobactam, and ampicillin/sulbactam (five, two, and one V. alginolyticus, respectively), proving that the antimicrobial combination therapy is still quite effective. In conclusion, the level of AMR in the tested Salmonella and above all Vibrio strains, which represent a consistent bacterial population in the marine environment, seems to indicate a poor diffusion of this phenomenon in the investigated area, suggesting that the actual condition is not highly worrying but perhaps promising, especially in view of the strategies currently implemented for the prudent use of antimicrobials in both human medicine and the zootechnical sector. However, our findings are not negligible, as some V. alginolyticus isolates exhibited resistance towards the critically important antimicrobials for human medicine. The epidemiological value of our study is of great relevance given the methodical approach used for monitoring: the survey was conducted over a period of 19 months and in a rather large area particularly devoted to shellfish farming and shellfish harvesting. Thus, the large amount of systematic data collected give an overview of the current scenario in one of the most representative regions for the Italian shellfish production sector [48].
The present study, like other previous ones, suggests the opportunity to implement a national/European programme to monitor the prevalence and distribution of antimicrobial resistance pattern in several not only pathogenic, but also environmental bacteria. Special consideration should be given to bacteria isolated from seafood which are generally eaten raw or undercooked and could represent a great threat to human health by transferring mobile genetic elements for antibiotic resistance to human pathogens. With regard to the latter aspect, it is worth mentioning, for example, the isolation of a V. parahaemolyticus strain carrying the bla NDM-1 gene from seafood and displaying in vitro carbapenemase activity but not phenotypical resistance [49]; this is a certainly highly worrisome finding, which could lead to therapeutic failure; thus, further research should be encouraged. Investigation of AMR in microorganisms within a specific area is necessary to formulate effective antibiotic reduction programmes. Indeed, although antibiotic resistance is an ancient and natural phenomenon, it is generally recognised that it occurs mainly in bacteria inhabiting the gastrointestinal tract of subjects receiving antibiotics and that distribution of antibiotic resistance genes (ARGs) in the aquatic ecosystem reflects mostly the faecal contamination by ARB. Furthermore, antibiotic pollution contributes to promoting the emergence and maintenance of ARGs and ARB in a delimited area [50]. Hence, insightful information on trends of AMR distribution in site-specific microbial populations is fundamental to better understand the local use and abuse of antibiotics and realise appropriate corrective measures. In this perspective, it would be appropriate to fix MIC breakpoints, also, for environmental Vibrio species and our results could be useful to upgrade the MIC distribution for Vibrio spp. relating to the antibiotic agents without available specific CLSI breakpoints.
Supplementary Materials: The following supporting information can be downloaded at: https: //www.mdpi.com/article/10.3390/microorganisms11020450/s1, Table S1. Information on Vibrio isolates: origin, resistance profile and MAR index. Table S2: Results for the broth microdilution assays of Vibrio isolates.