Antimicrobial resistance in bacteria isolated from aquaculture sources in Australia

Authors


Mary D. Barton, School of Pharmacy and Medical Sciences, University of South Australia, GPO Box 2471, Adelaide SA 5001, Australia.
E-mail: mary.barton@unisa.edu.au

Abstract

Aims:  To carry out a preliminary assessment of the occurrence of resistance to antimicrobials in bacteria that has been isolated from a variety of aquaculture species and environments in Australia.

Method and Results:  A total of 100 Gram-negative (Vibrio spp. and Aeromonas spp. predominantly) and four Gram-positive bacteria isolated from farmed fish, crustaceans and water from crab larval rearing tanks were obtained from diagnostic laboratories from different parts of Australia. All the isolates were tested for sensitivity to 19 antibiotics and Minimal Inhibitory Concentrations were determined by the agar dilution method. Plasmid DNA was isolated by the alkali lysis method. Resistance to ampicillin, amoxycillin, cephalexin and erythromycin was widespread; resistance to oxytetracycline, tetracycline, nalidixic acid and sulfonamides was common but resistance to chloramphenicol, florfenicol, ceftiofur, cephalothin, cefoperazone, oxolinic acid, gentamicin, kanamycin and trimethoprim was less common. All strains were susceptible to ciprofloxacin. Multiple resistance was also observed and 74·4% of resistant isolates had between one and ten plasmids with sizes ranging 2–51 kbp.

Conclusions:  No antibiotics are registered for use in aquaculture in Australia but these results suggest that there has been significant off-label use.

Significance and impact of study:  Transfer of antibiotic resistant bacteria to humans via the food chain is a significant health concern. In comparison with studies on terrestrial food producing animals, there are fewer studies on antibiotic resistance in bacteria from aquaculture enterprises and this study provides further support to the view that there is the risk of transfer of resistant bacteria to humans from consumption of aquaculture products. From the Australian perspective, although there are no products registered for use in aquaculture, antimicrobial resistance is present in isolates from aquaculture and aquaculture environments.

Introduction

Antibiotic resistance is a significant human health issue and in recent years there have been many papers reporting a link between antibiotic use in food producing animals, emergence of antibiotic resistance in salmonella, Escherichia coli, enterococci and campylobacter in treated animals, and transfer of these resistant organisms to humans (or their resistance genes to human pathogens) via the food chain (Barton 2000; Stobberingh and van den Bogaard 2000; Angulo et al. 2004). However, there has been less attention paid to the potential for antibiotic use in the aquaculture industries to compromise human health. In addition to transfer of resistant organisms through consumption of contaminated fish and shellfish, there is substantial risk of environmental contamination because of the practice of using medicated feeds to treat whole pens or cages.

The world's aquaculture industry has in recent times grown rapidly to satisfy the demand for seafood, which cannot be met by wild fisheries harvesting as this is currently in a state of decline because of over-fishing, pollution and marine habitat destruction. Aquaculture production is increasing at about 9·25% per year (FAO 1997) and the FAO had previously estimated that half of the world's seafood demand will be met by aquaculture in 2020 (FAO 1995).

A significant challenge to fish farming, however, is disease caused by bacteria such as Aeromonas spp., Vibrio spp., Pseudomonas spp. and Flavobacterium spp. In traditional terrestrial livestock production in most countries, a range of antibiotics is registered for use to control bacterial diseases – although in response to human health concerns some countries no longer permit the registration of products for nontherapeutic purposes such as growth promotant use. Alderman and Hastings (1998) noted that controls on antibiotic use in aquaculture vary widely from country to country. In developed countries such as members of the EU, the USA, Canada and Norway there is a limited number of products, regulatory control is strong and use of antibiotics is declining because of improvements in management and development of effective vaccines (Burka et al. 1997; WHO 2002; Lillehaug et al. 2003). However, 90% of aquaculture production occurs in developing countries (Bondad-Reantaso et al. 2005) where regulatory controls are weak and use of antibiotics appears to be widespread. In Australia and New Zealand with small but rapidly developing industries and with rigorous regulatory control, no antibiotics are registered for use in aquaculture. Unfortunately restrictions on the availability of registered products can lead to pressure to use antibiotics off-label or even totally illegally.

There are now a number of reports of antibiotic resistant bacteria in fin fish in aquaculture settings, (Aoki et al. 1983; Tsoumas et al. 1989; McPhearson et al. 1991; Starliper et al. 1993; Son et al. 1997; Ho et al. 2000; Schmidt et al. 2000; Mirand and Zemelman 2002; Saha and Pal 2002; Michel et al. 2003; Hatha et al. 2005); however, there are few published studies covering other aquaculture industries. Antibiotic resistance has been reported in eels (Alcaide et al. 2005), shrimps (Tjahjadi et al. 1994) and shrimp ponds (Tendencia and de la Pena 2001), shellfish (Ho et al. 2000) and aquaculture environments (McPhearson et al. 1991; Petersen et al. 2000; Rhodes et al. 2000; Tendencia and de la Pena 2001; Chelossi et al. 2003; Kim et al. 2004). Some reports have noted that resistance emerges within few years of treating infections with antibacterial drugs (Sorum 1998, 1999) and this is a factor, which limits their value in the control of bacterial diseases of fish (Smith et al. 1994), apart from any public health concerns. Recognition of the resistance issue has led to calls for intensified surveillance of antibiotic use and antibiotic resistance (Aoki 1992; WHO 2002).

The JETACAR (1999) report to the Australian government highlighted the need for both the medical and agricultural sectors to define the extent of the antibiotic resistance problem in Australia and develop strategies to reduce it. Currently, no antibiotics are registered for use in aquaculture in Australia but there are anecdotal accounts suggesting that they have been and are still being used off-label, under veterinary prescription. At present there is no published information that allows an informed opinion on the extent of antibiotic resistance associated with Australian aquaculture (Barton et al. 2003).

In this study, we carried out a preliminary assessment of the occurrence of resistance to antimicrobials in bacteria that have been isolated from a variety of aquaculture species and environments. We included antimicrobials frequently used in aquaculture in overseas countries such as the USA, Denmark, Norway, UK and Canada, some of which are registered for use in livestock in Australia and others of medical interest.

Materials and methods

Source of bacterial strains

One hundred Gram-negative and four Gram-positive bacterial strains considered to be of clinical or potential clinical significance were received from veterinary diagnostic and aquaculture research laboratories in different states in Australia. The strains were isolated from aquaculture settings and included freshwater and marine water fishes, crustaceans and environmental isolates. Vibrio spp. accounted for approximately 60%, Aeromonas spp. 21%, Photobacterium spp. 4%, Pseudomonas spp. 4%, Citrobacter spp. 2%, Edwardsiella tarda 2%, Hafnia alvei 1%, Flavobacterium spp. 2%, Plesiomonas shigelloides 1% and Gram-positive organisms (Staphylococcus and Micrococcus spp.) 4% of total isolates. A list of isolate source and distribution is shown in Table 1.

Table 1.  Species distribution and source of isolates used in this study
CodeIDSource/tissueLocation (year)
  1. QLD, Queensland; WA, Western Australia; NT, Northern Territory; NSW, New South Wales.

3-46976-2TCBSGVibrio splendidusCrustacean/haemolymphQLD (2003)
0-55526-3TCBSGVibrio mimicusFish/intestineQLD (2000)
3-43413-3BTCBSYVibrio tubiashiiFish/finQLD (2000)
Mp1906Vibrio tubiashiiMudcrab larval rearing tank/waterNT (2002–2003)
Mp1492Vibrio tubiashiiMudcrab larval rearing tank/waterNT (2002–2003)
04-625#3Vibrio tubiashiiPink snapper/skinWA (2004)
04-944#3Vibrio tubiashiiSeahorse/tank wallWA (2004)
0-53535-HVibrio parahaemolyticusCrustacean/haemolymphQLD (2000)
Mp2249Vibrio proteolyticusMudcrab larval rearing tank/waterNT (2002–2003)
Mp1830Vibrio proteolyticusMudcrab larval rearing tank/waterNT (2002–2003)
03-2945#9Vibrio proteolyticusTank/waterWA (2003)
03-2945#11Vibrio proteolyticusTank/waterWA (2003)
04-0034#2Vibrio choleraeLung fish/flankWA (2004)
04-0173#5Vibrio choleraeBarramundiWA (2004)
04-790#1Vibrio cholerae non-01Cherax rotundis/liverWA (2004)
0-53535-5Vibrio alginolyticusCrustacean/haemolymphQLD (2000)
Mp2583Vibrio alginolyticusMudcrab larval rearing tank/waterNT (2002–2003)
Mp2278Vibrio alginolyticusMudcrab larval rearing tank/waterNT (2002–2003)
04-617#1Vibrio alginolyticusClownfish/gill, liverWA (2004)
04-711#1Vibrio alginolyticusSeahorse/skinWA (2004)
04-944#1Vibrio alginolyticusSeahorse/tank wallWA (2004)
03-3416#2Vibrio alginolyticusBarramundi/skinWA (2003)
BR-S3Vibrio alginolyticusPrawnNSW
GR-S2Vibrio alginolyticusPrawnNSW
2-PK-S2Vibrio alginolyticusPrawnNSW
BK-S3Vibrio alginolyticusPrawnNSW
3-47001-1Vibrio harveyiCrustacean/haemolymphQLD (2003)
3-44722-2TCBSYVibrio harveyiFish/kidneyQLD (2003)
0-53535-CVibrio harveyiCrustacean/haemolymphQLD (2003)
J/R3Vibrio harveyiLarval rearing tank/waterNSW (2003)
L/R2Vibrio harveyiLarval rearing tank/waterNSW (2003)
N/R9Vibrio harveyiLarval rearing tank/waterNSW (2003)
Mp2533Vibrio harveyiMudcrab larval rearing tank/waterNT (2002–2003)
Mp1238Vibrio harveyiMudcrab larval rearing tank/waterNT (2002–2003)
Mp1009Vibrio harveyiMudcrab larval rearing tank/waterNT (2002–2003)
Mp0589Vibrio harveyiMudcrab larval rearing tank/waterNT (2002–2003)
03-2779#1Vibrio harveyiPink snapper/gall bladderWA (2003)
04-711#4Vibrio harveyiSeahorse/skinWA (2004)
04-756#1Vibrio harveyiYellowtail/intestine-diseasedWA (2004)
04-756#4Vibrio harveyiYellowtail/intestine-diseasedWA (2004)
04-944#2Vibrio harveyiSeahorse/tank wallWA (2004)
03-2668#1Vibrio harveyiBlack bream/bloodWA (2003)
03-3123#1Vibrio harveyiSnapper/skin lesionWA (2003)
03-3123#5Vibrio harveyiSnapper/skin lesionWA (2003)
04-625#4Vibrio harveyiPink snapper/skinWA (2004)
3-46984-3TCBSYVibrio sp.Crustacean/haemolymphQLD (2003)
A1/R3Vibrio sp.Larval rearing tank/waterNSW (2003)
A2/R3Vibrio sp.Larval rearing tank/waterNSW (2003)
D/R14Vibrio sp.Larval rearing tank/waterNSW (2003)
F/R15Vibrio sp.Larval rearing tank/waterNSW (2003)
I/R7Vibrio sp.Larval rearing tank/waterNSW (2003)
P/R5Vibrio sp.Larval rearing tank/waterNSW (2003)
O/R5Vibrio sp.Larval rearing tank/waterNSW (2003)
03-2945#3Vibrio sp.Tank/waterWA (2003)
03-2668#3Vibrio sp.Snapper/gut contentWA (2003)
04-625#6Vibrio sp.Pink snapper/gutWA (2004)
04-711#2Vibrio sp.Seahorse/skinWA (2004)
04-756#6Vibrio sp.Yellow tail/intestine-healthyWA (2004)
04-712#1Vibrio sp.Aquarium fish/kidneyWA (2004)
04-944#7Vibrio sp.Seahorse/tank wallWA (2004)
GR-S2Vibrio sp.PrawnNSW
04-977#1Vibrio sp.Kingfish/skin lesionWA (2004)
3-50146 1AAeromonas hydrophilaFish/AbdomenQLD (2003)
03-3416#1Aeromonas hydrophilaBarramundi/skinWA (2003)
04-0200#1Aeromonas hydrophilaBarramundi/brainWA (2004)
04-900#1A. hydrophila subsp. hydrophilaKoi/swabWA (2004)
04-594#1A. hydrophila subsp. hydrophilaTrout/skinWA (2004)
04-594#2A. hydrophila subsp. hydrophilaTroutWA (2004)
04-0018#2A. hydrophila subsp. hydrophilaBarramundi/spleenWA (2004)
04-0034#1A. hydrophila subsp. dhakensisLung fish/flankWA (2004)
04-841#1A. hydrophila subsp. dhakensisRummynose Tetra/gillWA (2004)
04-842#1A. hydrophila subsp. dhakensisGolden Tiger Barb/gillWA (2004)
04-976#1A. hydrophila subsp. dhakensisBarramundi/brainWA (2004)
04-900#2Aeromonas veronii sobriaKoi/swabWA (2004)
04-943#1Aeromonas veronii sobriaGoldfish/internalWA (2004)
04-944#6Aeromonas veronii veroniiSeahorse/tank wallWA (2004)
3-47910-2CAeromonas sobriaFish/peritoneal fluidQLD (2003)
3-5146-1BAeromonas sobriaFish/abdomenQLD (2003)
04-486#1Aeromonas sp.Barramundi/brainWA (2004)
04-486#2Aeromonas sp.Barramundi/heartWA (2004)
04-486#4Aeromonas sp.Barramundi/spleenWA (2004)
04-841#2Aeromonas sp.Rummynose Tetra/liverWA (2004)
04-843#1Aeromonas sp.Barramundi/bloodWA (2004)
04-0173#9Aeromonas sp.BarramundiWA (2004)
0-54612-1Edwardsiella tardaFish/Gill, heartQLD (2000)
04-842#3Edwardsiella tardaGolden Tiger Barb/gillWA (2004)
04-843#3Pseudomonas fluorescensBarramundi/bloodWA (2004)
04-310#2Pseudomonas fluorescensGoldfishWA (2004)
89-22073-13Pseudomonas sp.Crustacean/waterQLD (1989)
04-625#7Pseudomonas sp.Pink snapper/liverWA (2004)
04-233#2Hafnia alveiMarron/hepatopancreasWA (2004)
04-233#3Citrobacter braakiiMarron/hepatopancreasWA (2004)
04-0034#3Citrobacter sp.Lung fish/flankWA (2004)
2-42882-BPhotobacterium damselaeFish/finQLD (2002)
03-3123#2Photobacterium damselaeSnapper/skin lesionWA (2003)
04-625#1P. damselae subsp. damselaePink snapper/skinWA (2004)
04-977#3P. damselae subsp. damselaeKingfish/liverWA (2004)
1-PK-S2Plesiomonas shigelloidesPrawnNSW
04-0018#6Flavobacterium sp.Barramundi/gillWA (2004)
03-3326#5Flavobacterium sp.Koi carp/skinWA (2003)
04-310#4Micrococcus sp.GoldfishWA (2004)
04-0173#14Staphylococcus sp.BarramundiWA (2004)
WH-S8Staphylococcus sp.PrawnNSW
B5Staphylococcus spPrawnNSW

The organisms were cultured on Blood agar or Tryptone soy agar (supplemented with 1·5% NaCl for Vibrio species) and maintained at −80°C in snap freeze medium and when required, stored organisms were streaked on the same media.

Identification

Isolates that were already identified before receipt were checked to confirm identity to Genus level: Aeromonas spp.: Gram-negative rods, motile, positive for oxidase, catalase, ortho-nitrophenyl-b-d-galactopyranoside (ONPG), arginine decarboxylase, fermentation of glucose, negative for urease and resistant to the vibriostatic agent 0/129; Vibrio spp.: green or yellow colonies on thiosulfate citrate bile sucrose (TCBS) agar, Gram-negative rods, positive for oxidase, catalase, indole, and fermentation of glucose, negative for arginine decarboxylase and sensitive to the vibriostatic agent 0/129; Photobacterium spp.: green colonies on TCBS agar, Gram-negative rods, positive for oxidase, fermentation of glucose, arginine decarboxylase, negative for indole and sensitive to the vibriostatic agent 0/129; Pseudomonas spp.: Gram-negative rods, fluorescent colonies on Pseudomonas agar, positive for oxidase, catalase, citrate, arginine decarboxylase and negative for the fermentation of glucose (oxidative).

Unidentified Gram-negative isolates were identified using Microbact 24E (MEDVET Diagnostic, Thebarton, South Australia, Australia); and the Gram-positive bacteria were identified using ID 32 STAPH (Biomerieux).

Antibiotics

The 19 antimicrobial agents selected for susceptibility testing belong to eight different classes of antibiotics, which have been reported as used in aquaculture (in countries other than Australia) or in livestock production in Australia (and therefore, available for off-label use in aquaculture) or of interest in human medicine. They were amoxycillin, ampicillin, cephalothin, cephalexin, cefoperazone, ceftiofur, nalidixic acid, oxolinic acid, ciprofloxacin, chloramphenicol, florfenicol, gentamicin, kanamycin, erythromycin, tetracycline, oxytetracycline, sulfamethoxazole, trimethoprim and potentiated sulfonamide (trimethoprim-sulfamethoxazole).

Antimicrobial susceptibility testing

The antibiotic susceptibility testing of all isolates was carried out using the agar dilution method according to the NCCLS (2003) guidelines using a multipoint inoculator (manufactured in the School of Pharmacy and Medical Sciences workshop). Isolates were streaked on nonselective medium [Blood agar and Tryptone soy agar (supplemented with 1·5% NaCl for Vibrio spp.)] and incubated at 30°C for 24–48 h. Single colonies were emulsified in 0·85% NaCl to obtain a bacterial suspension that corresponded in turbidity to a 0·5 McFarland standard and used as inoculum for the agar dilution on Mueller–Hinton agar. This was incubated overnight at 30°C. E. coli ATCC 25922, Staphylococcus aureus ATCC 25923, Staph. aureus ATCC 29213 and Enterococcus faecalis ATCC 29212 were used as controls.

Isolates were defined as susceptible, intermediate or resistant in accord with the NCCLS Enterobacteriaceae breakpoints.

Plasmid isolation

Plasmid DNA was isolated from organisms that were resistant to at least one antibiotic. Isolates were grown overnight on Tryptone soy agar or Tryptone soy agar with added 1·5% NaCl for Vibrio spp. For each isolate, 2–3 colonies were picked off and extracted by alkaline lysis of cells (Sambrook et al. 1989), and examined by agarose gel electrophoresis.

Results

Antimicrobial resistance

One hundred Gram-negative and four Gram-positive bacteria were tested for their susceptibility to 19 antibiotics; the frequency of resistance to the various antimicrobials for all bacteria is presented in Fig. 1. Generally, resistance was widespread for penicillins and some of the cephalosporins as well as erythromycin. More than half (54·8%) of all the isolates were found to be resistant to ampicillin and amoxycillin, 41·4% were resistant to cephalexin, 23·1% to cephalothin, 5·8% to ceftiofur and 1·9% to cefoperazone, while 47·1% were resistant to erythromycin. Tetracycline resistance was found in 16·4% of the isolates, and oxytetracycline resistance in 19·2%. Resistance to gentamicin and kanamycin was detected in 2·9% and 5·8% of the isolates respectively; 13·5% of the isolates were found to be resistant to nalidixic acid and nine (8·7%) to oxolinic acid. Resistance to chloramphenicol and florfenicol was found in 6·7% and 8·7% of the isolates respectively. Resistance to trimethoprim was seen in 4·8% of the isolates and 16 (15·4%) of the isolates were resistant to sulfamethoxazole while resistance to trimethoprim-sulfamethoxazole was found in four (3·9%) isolates. All the isolates were susceptible to ciprofloxacin. Of the 104 isolates tested, 18 were sensitive to all 19 antimicrobial agents tested.

Figure 1.

Frequency of resistance to antibacterials for 104 isolates obtained from aquaculture settings in Australia. Antibacterial abbreviations: Amp, ampicillin; Amx, amoxicillin; Cfx, cephalexin; Cft, cephalothin, Cf, ceftiofur, Cfp, cefoperazone; Chl, chloramphenicol; Flo, florfenicol; Nal, nalidixic acid; Oxl, Oxolinic acid; Cip, ciprofloxacin; Gt, gentamicin; Kn, kanamycin; Ery, erythromycin; Tet, tetracycline; Oxy, oxytetracycline; Trim, trimethoprim; Sul, sulfamethoxazole; T-S, trimethoprim-sulfamethoxazole.

Vibrio spp.

Forty-six out of 62 Vibrio spp. (74%) were found to be resistant to at least one antibiotic. Twenty five (40%) were found to be resistant to ampicillin and amoxycillin resistance was detected in 28 (45%) of the isolates. Resistance to cephalexin was found in 21 (34%) of the isolates and cephalothin resistance in nine (14%). Gentamicin and kanamycin resistance was found in one isolate each. Erythromycin resistance was found in 21 (34%) of the isolates, tetracycline resistance in three (5%) and resistance to oxytetracycline in four (6·5%) of the isolates. No resistance to ceftiofur, cefoperazone, chloramphenicol, florfenicol, oxolinic acid, ciprofloxacin, trimethoprim, sulfamethoxazole or trimethoprim-sulfamethoxazole was found.

Aeromonas spp.

All 22 Aeromonas spp. were found to be resistant to at least one antibiotic. Nineteen (86%) out of 22 were found to be resistant to ampicillin and amoxycillin resistance was also detected in 19 (86%) of the isolates. Resistance to cephalexin was found in (41%) and cephalothin resistance in ten (46%) of the isolates. Nalidixic acid resistance was found in five (23%) while resistance to oxolinic acid was found in only two (9·1%) isolates; erythromycin resistance was detected in 16 (73%). Tetracycline resistance was found in nine (41%) while oxytetracycline resistance was found in ten (45·5%) of the isolates and one isolate was resistant to trimethoprim. Single isolates were resistant to ceftiofur and florfenicol and two isolates were resistant to chloramphenicol. Resistance to sulfamethoxazole was found in 13 (59%) of the isolates and one isolate was also resistant to trimethoprim-sulfamethoxazole. However, no resistance was found to cefoperazone, ciprofloxacin gentamicin or kanamycin.

Pseudomonas spp.

One of the four Pseudomonas spp. tested was susceptible to all antibiotics. Two out of the three resistant isolates were resistant to ampicillin and amoxycillin, all three were resistant to cephalexin and two were resistant to cephalothin. Resistance to ceftiofur was found in three of the isolates and two of the isolates were resistant to cefoperazone, chloramphenicol, florfenicol, trimethoprim, sulfamethoxazole and trimethoprim-sulfamethoxazole. Nalidixic acid and erythromycin resistance was found in three of the isolates; however, only one was resistant to oxolinic acid. One isolate was resistant to both tetracycline and oxytetracycline. All isolates were susceptible to ciprofloxacin, gentamicin and kanamycin.

Photobacterium spp.

Three out of four Photobacterium spp. were resistant to at least one antibiotic. Resistance to ampicillin was found in one isolate, amoxycillin resistance in two isolates and erythromycin resistance in one isolate. All isolates were susceptible to cephalexin, cephalothin, ceftiofur, cefoperazone, chloramphenicol, florfenicol, nalidixic acid, oxolinic acid, ciprofloxacin, gentamicin, kanamycin, tetracycline, oxytetracycline trimethoprim, sulfamethoxazole and trimethoprim-sulfamethoxazole.

Flavobacterium spp.

The two Flavobacterium spp. tested were susceptible to both ciprofloxacin and cefoperazone. Both isolates were found to be resistant to ampicillin, one to amoxycillin, two to cephalexin, cephalothin, chloramphenicol, florfenicol, nalidixic acid, gentamicin, kanamycin and erythromycin and oxytetracycline. However, only one isolate was resistant to tetracycline, trimethoprim, sulfamethoxazole and trimethoprim-sulfamethoxazole.

Citrobacter spp.

The two Citrobacter spp. tested were found to be resistant to ampicillin, amoxycillin, cephalexin, cephalothin, florfenicol and erythromycin. They were susceptible to ciprofloxacin, cefoperazone, chloramphenicol, nalidixic acid, oxolinic acid, gentamicin, kanamycin, tetracycline, oxytetracycline, trimethoprim, sulfamethoxazole and trimethoprim-sulfamethoxazole.

Edwardsiella tarda

The two Edw. tarda tested were found to be resistant to ampicillin, amoxycillin, erythromycin, chloramphenicol and tetracycline and oxytetracycline. They were susceptible to cephalexin, cephalothin, cefoperazone, ciprofloxacin, florfenicol, nalidixic acid, oxolinic acid, gentamicin, kanamycin, tetracycline, trimethoprim, sulfamethoxazole and trimethoprim-sulfamethoxazole.

Hafnia alvei

Only one H. alvei was tested and it was found to be resistant to ampicillin, amoxycillin, cephalexin, cephalothin, florfenicol, erythromycin and sulfamethoxazole. It was susceptible to all other antibiotics.

Plesiomonas shigelloides

One isolate was tested and it was found to be resistant to kanamycin and susceptible to all other antibiotics.

Staphylococcus and Micrococcus spp.

All four organisms were resistant to oxolinic acid while only one out of the three Staphylococcus spp. was resistant to nalidixic acid. Another isolate was resistant to tetracycline, oxytetracycline nalidixic acid, erythromycin and trimethoprim while the third was resistant to tetracycline, oxytetracycline, nalidixic acid, erythromycin, kanamycin, florfenicol, ampicillin and cephalexin. The Micrococcus sp. was found to be resistant only to nalidixic acid. The Gram-positive isolates were sensitive to the other antibiotics tested.

The frequency distributions of the MIC values for all antibiotics tested are shown in Table 2.

Table 2.  Frequency of distribution of minimum inhibitory concentration (MIC) of isolates against 19 antimicrobial agents
AntibioticNo. of isolatesNumber of isolates with a MIC (mg/l)Resistant
≤0·10·250·51248163264≥128No.%
Ampicillin104 10589852710   405754·8
Amoxicillin104 1363127241511   315754·8
Cephalexin104 3245161813219   134341·4
Cephalothin104 5924237753    212423·1
Ceftiofur10418311620762211 6 5·8
Cefoperazone10422297131610412  2 1·9
Chloramphenicol104 195513331322    37 6·7
Florfenicol104 2252127234      29 8·7
Nalidixic acid104 1084621321 3   101615·4
Oxolinic acid104 92 312  6  9 8·7
Ciprofloxacin10470151053      00
Gentamicin104921137324612  3 2·9
Kanamicin104 359124518513    36 5·8
Erythromycin104 71191315241253    54947·1
Tetracycline104 2937126 3841    41716·4
Oxytetracycline104 2623302 3394    42019·2
Trimethoprim104151223221862  1    45 4·8
Sulfamethoxazole104    252116227   311615·4
    0·25/4·75 0·5/9·5 1/19 2/38 4/76 8/15216/30432/60864/1216>128/2432  
Trimethoprim-sulfamethoxazole104  981 111       24 3·9

Multiple drug resistance

Twenty-three isolates (26·7%) were resistant to just one class of antibiotic, 27 (31·4%) were resistant to two classes, 15 (17·4%) were resistant to three classes, 12 isolates (13·9%) were resistant to four classes, two isolates (2·3%) were resistant to five classes, one isolate (1·2%) was resistant to six classes, four isolates (4·7%) were resistant to seven classes while two isolates (2·3%) was found to be resistant to eight out of the ten classes of antibiotics tested (Fig. 2) and a total of 44 different resistance patterns were identified (data not shown).

Figure 2.

Antibacterial multiresistance of isolates obtained from aquaculture settings in Australia.

Plasmid profile

To check the involvement of plasmids in the resistance to antibiotics observed in some of our isolates, plasmid DNA was extracted from all 86 isolates that were resistant to at least one antibiotic. Plasmids of varying numbers and sizes were found in 74·4% of resistant isolates, while 25·6% did not posses any plasmids. Resistant isolates carried from one to ten plasmids with sizes ranging from about 2–51 kbp (Fig. 3). The various profiles are listed in Table 3.

Figure 3.

Plasmid profile of the different genus of isolates. M (2·5 kbp marker), Lane1–7 (Aeromonas spp.) lane 8–9 (Edwardsiella spp.), lane 10–12 (Pseudomonas spp.), lane 13–14 (Citrobacter spp.), lane 15 (Hafnia alvei), lane 16–18 (Photobacterium spp.), lane 19–20 (Flavobacterium spp.), lane 21–28 (Vibrio spp.), M (1 kbp marker).

Table 3.  Number and size of plasmids
 Number of plasmidsSize of plasmid (kbp)
Aeromonas spp.1–10 2–51
Vibrio spp.1–6 2–51
Photobacterium spp.1–5 6–41
Pseudomonas spp.1–3 3–34
Edwardsiella spp.4 and 7 3–36
Citrobacter spp.1 and 234–41
Hafnia alvei134
Flavobacterium spp.1–5 5–36

Discussion

A variety of resistance patterns was observed in our study but we are unable to correlate them with antibiotic use as none are actually registered to use in aquaculture in Australia. However, the results are not too dissimilar from those seen in countries where antibiotics are known to be used in aquaculture. Resistance to ampicillin, amoxycillin, first and second generation cephalosporins and erythromycin was widespread in this study. The resistance of Gram-negative bacteria to erythromycin is to be expected because of intrinsic resistance of many such organisms to macrolide antibiotics. In addition, the resistance found to ampicillin and amoxycillin is not surprising as it corresponds to what has been found in many studies in a number of countries (Grant and Laidler 1993; Ho et al. 2000; Schmidt et al. 2000; Mirand and Zemelman 2002; Saha and Pal 2002; Chelossi et al. 2003; Hatha et al. 2005). Beta-lactamases have been reported commonly in Aeromonas spp. from aquaculture settings (Alderman and Hastings 1998).

Chloramphenicol resistance has also been reported in Chile (Mirand and Zemelman 2002), France (Michel et al. 2003) and in a Western Mediterranean study (Chelossi et al. 2003). In contrast, only 6·7% of our isolates were resistant to chloramphenicol, reflecting the fact that it has not been registered for use in livestock since 1982. The extent of florfenicol resistance in this study is of concern as this antibiotic was only registered for use in livestock in Australia in early 2003 – suggesting that there has been a rapid extension of (off-label) use in aquaculture. Florfenicol is used in some countries in aquaculture with Michel et al. (2003) noting little emergence of resistance in France, whereas Ho et al. (2000) in Taiwan and Mirand and Zemelman (2002) in Chile reported significant resistance levels.

Resistance to tetracyclines and/or oxytetracycline has been commonly reported (for example, Park et al. 1994; Son et al. 1997; Ho et al. 2000; Rhodes et al. 2000; Schmidt et al. 2000, 2001; Kim et al. 2004; Hatha et al. 2005). The modest extent of resistance in our study is surprising as tetracyclines are widely used for treatment of livestock in Australia. Resistance to sulfadiazine-trimethoprim is also less frequent in our study in comparison with overseas reports (Schmidt et al. 2000; Mirand and Zemelman 2002; Chelossi et al. 2003).

Quinolone resistance has been reported in environmental isolates at a relatively low frequency (McKoen et al. 1995; Guardabassi et al. 1998); however, Chelossi et al. (2003) reported resistance to nalidixic acid in 70% of their isolates. Oxolinic acid resistance was found by Schmidt et al. (2001) in 20% of Aeromonas isolates. This is a higher rate of resistance than detected in our study. However, neither nalidixic acid nor oxolinic acid is registered for use in food producing animals (although in the past occasional permits allowing oxolinic acid use have been issued) so these results are of some concern. In contrast, no resistance to ciprofloxacin were observed, in keeping with the fact that fluoroquinolones are not registered for use in livestock in Australia.

Our results showed resistance to more than one class of antibiotic. Multiple drug resistance has been reported in a number of studies of fish pathogens and aquaculture environments (McPhearson et al. 1991; Schmidt et al. 2000; Hatha et al. 2005). Vibrio spp. isolated from shrimp hatcheries in Indonesia have demonstrated multiple antibiotic resistance to antimicrobials such as ampicillin, tetracycline, amoxycillin and streptomycin (Tjahjadi et al. 1994) and Mirand and Zemelman (2002) reported that bacteria resistant to six to ten antibacterials were common. Studies on the antibiotic resistance in bacteria from shrimp ponds (Tendencia and de la Pena 2001) demonstrated a correlation between multiple bacterial antibiotic resistance levels and use of particular drugs. Our results indicate a possible use of one or more antibiotics in Australian aquaculture, providing selection pressure for the emergence of multiresistant strains.

Nearly three quarters of the isolates we tested were found to carry different-sized plasmids, including some with high molecular weight. The presence of high molecular weight plasmids in fish farms and environments is well documented (Rhodes et al. 2000; Furushita et al. 2003) and transferable R-plasmids conferring resistance to various antimicrobials including tetracycline, chloramphenicol and streptomycin (and associated with aquaculture) were reported many years ago (Aoki et al. 1983; Aoki and Takahashi 1987). Some of these plasmids are capable of being transferred to other organisms and so transferring resistance genes such as tetracycline resistance determinants (Aoki 1988). Furushita et al. (2003) noted that tetracycline resistance genes from fish farm isolates were similar to those detected in clinical isolates.

In conclusion, the results obtained from this study indicate that antibiotic resistance is common in bacteria isolated from aquaculture species and environments in Australia, supporting the view that antibiotics have been used off-label. The extent of the resistance found and in particular the significant levels of multiple resistance are of concern. Follow-up studies are required to investigate the extent of antibiotic use in Australian aquaculture farms and environments and to determine the molecular basis of antimicrobial resistance to the different antibiotics, the potential for transfer of resistance genes from aquaculture isolates to human pathogens, some assessment of the risk of transfer of resistant organisms (or genes) to humans via the food chain and the threats imposed by environmental contamination with antibiotic resistant bacteria.

Acknowledgements

We wish to thank Dr Nicky Buller of the Department of Agriculture, Government of Western Australia; Dr Richard Callinan of the Aquatic Animal Health Unit, Regional Veterinary Laboratory, Wollongbar, New South Wales; Mr Morris Pizzuto of Berrimah Veterinary Laboratory, Department of B.I.R.D, Darwin, Northern Territory; Dr Paul Smith of the University of Western Sydney, New South Wales and Dr Annette Thomas of the Department of Primary Industries and Fisheries, Queensland for providing the isolates for this study.

Ancillary