Chemical use, antimicrobial quality and withdrawal time in striped catfish Pangasianodon hypophthalmus aquaculture in the Mekong Delta, Vietnam Tran Minh Phu Section of Food Safety an
INTRODUCTION AND STATE OF THE ART
Project framework
This PhD study was carried out as a part of the EU-funded FP7 project “Sustaining Ethical
Aquaculture Trade” (SEAT project number 222889; www.seatglobal.eu) The SEAT project (2009-
2013) took an interdisciplinary approach and included 12 work packages (WP) and brought together 12 international institutions and organisations in order to address sustainability issues in aquaculture The project aimed to provide better-informed choices to seafood consumers in Europe and other developed countries Investigations of aquaculture value chains, including environmental impacts and impacts of trade on local livelihoods, public health and food safety, were investigated for four main export aquaculture species: Penaeid shrimp [white-leg shrimp (Litopenaeus vannamei) and black tiger shrimp (Penaeus monodon)], fresh water prawn (Macrobrachium rosenbergii), tilapia [Nile tilapia (Oreochromis niloticus) and hybrid red tilapia (O mossambicus ×
O niloticus)] and striped catfish (Pangasianodon hypophthalmus) produced in China, Vietnam, Thailand and Bangladesh The results in this thesis contribute new knowledge about striped catfish aquaculture that is of relevance for food safety and public health (WP6 and WP7) by studying several aspects related to the chemical use, antimicrobial quality and withdrawal times in striped catfish aquaculture in the Mekong Delta, Vietnam.
Striped catfish aquaculture in Vietnam
Striped catfish (Pangasianodon hypophthalmus), also known as cá tra and historically referenced as Pangasius sutchii or Pangasius hypophthalmus, is exported in a range of frozen fillet products This species has air-breathing organs and is an obligate air breather, which helps it tolerate poor water quality—high organic matter, low dissolved oxygen, and high stocking densities Farmed striped catfish production began in Vietnam in the 1940s in small ponds using wild-caught fingerlings, and by the early 1960s, both basa (Pangasius bocourti) and striped catfish were being reared in cages beneath floating houses placed in the Mekong River At that time, the feed was homemade, comprising trash fish, rice bran, and broken rice that was cooked before feeding to the fish In 1981–1982, farmers in Can Tho City advanced to intensive culture of striped catfish in small ponds using wild-caught fingerlings.
(De Silva and Phuong, 2011) Farming practices relied primarily on wild sources of fry caught in the Mekong River Delta region (Quoc et al., 2002; De Silva and Phuong, 2011) In 1994, the
Cambodian authorities banned the capture of striped catfish fry from the wild, and in 2000 it was also banned in the An Giang and Dong Thap provinces in the Mekong Delta of Vietnam (Quoc et al., 2002) Following the success of artificial propagation techniques for striped catfish in the early 2000s, farmers shifted to hatchery-produced fingerlings (Cacot, 1999; Quoc et al., 2002) The dissemination of these techniques to famers stimulated rapid growth and increased yields As a result, striped catfish production increased rapidly and reached approximately 20,000 tonnes by
Figure 1 Striped catfish aquaculture in the Mekong Delta in 2010 - : main hatchery areas and : main nursery areas (adapted from De Silva and Phuong, 2011 and Phan et al., 2011)
From 2000-2004, intensive culture of striped catfish in cages and ponds expanded rapidly due to the hatchery-produced fingerlings that were able to meet demands for stocking In 2005, the rearing of striped catfish in cages and pens collapsed due to disease outbreaks of bacillary necrosis of
Pangasius (BNP), caused by Edwardsiella ictaluri, and farmers shifted to intensive pond culture
(De Silva and Phuong, 2011) At the same time, farmers gradually shifted from homemade feeds to commercial pelleted feeds Striped catfish is now the preferred species cultured in the Mekong Delta compared to basa (P bocourti) because of its higher yield, shorter production cycle (six months rather than eight months for basa) and because it is almost entirely intensively cultured in ponds (De Silva and Phuong, 2011; Halls and Johns, 2013)
Striped catfish farming in Vietnam is concentrated in the Mekong Delta, with Dong Thap, An Giang, and Can Tho as the leading producing provinces The industry has expanded rapidly, exceeding 1.2 million tonnes in 2012, making striped catfish one of the most intensive aquaculture sectors in the world.
2014) Striped catfish is exported to more than 140 countries and was worth US$ 1.8 billion in 2013 (VASEP, 2014) There are plans to expand the production areas of striped catfish in the Mekong Delta, Vietnam to up to 7,600-7,800 ha to reach production of 1.8-1.9 million tonnes in 2020
Figure 2 Production and culture area of striped catfish (De Silva and Phuong, 2011; Directorate of Fisheries, 2014)
Striped catfish aquaculture consists of seed production (hatchery), fry-to-fingerling rearing
(nursery) and grow-out The striped catfish production cycle is presented in Figure 3 Currently, about 200 hatcheries and over 4,000 nursery farms in the Mekong Delta produce approximately two billion striped catfish fingerlings (MARD, 2013) Production of striped catfish starts with the production of larvae/early fry (hatchery) (Fig 3), which are then raised to fingerlings in nursery farms The fingerlings are sold to grow-out farmers who rear the fish to market size
Hatchery size ranges from 0.2 to 15 hectares, with broodstock (male and female) selected mainly from grow-out ponds (about 80% of hatchery farms) and limited use of wild broodstock; broodstock are kept in ponds of about 0.16 ± 0.05 ha for maturation feeding and spawning After fertilisation, eggs are incubated in hatching jars and then transferred to storage tanks; following 16–48 hours of storage, the larvae are sold to nurseries A range of chemicals are used to treat water in the tanks and to disinfect empty tanks, incubators, nets and other equipment Hatcheries occasionally experience disease problems in broodstock caused by parasites and bacterial pathogens such as BNP, spot disease, white gills and head beriberi, prompting farmers to use chemicals and antimicrobials for treatment The fry are then transported to nurseries by truck or motorcycle without the use of any chemicals during transportation (Bui et al., 2010).
The second step in the striped catfish production is the nursery The nursery size is typically from 0.09 to 11 ha and pond depth is 1.8-2.2 m (Bui et al., 2010) Ponds are prepared mainly by drying, sludge removal and the application of lime (calcium carbonate) and sodium chloride Stocking density ranges between 250-2,000 larvae/m 2 The larvae are normally fed live freshwater crustaceans (Moina) and an emulsion of soybean meal and egg yolk for 1-30 days before the diet is changed to commercial pelleted feed, which is provided until fingerlings are sold for grow-out after
60 to 90 days The fry nursery sometimes comprises two stages: from larvae/early fry to a size of small fingerlings (~3 g/fish) and from small fingerlings to fingerlings (20-40 g/fish) The division
Two-stage production is practised because some farmers lack the capacity for nursing the early fry to small fingerlings The present study focuses on nursery efforts to raise early fry to fingerling size (20-40 g/fish) and to convert small fingerlings (~3 g/fish) to fingerling size (20-40 g/fish) Common diseases reported in nursery farms include BNP (25 %), parasite infestation.
(20 %) and haemorrhagic disease (9 %) which mostly occur during the wet season Farmers often use chemicals and antimicrobials to treat the diseases, while some farmers mainly increase the water exchange (Bui et al., 2010) As there is limited information available on chemical usage (type, dose etc.), the ability of farmers to diagnose diseases and their fish health management practices in nurseries and hatcheries were studied and reported in Manuscript I
Figure 3 Structure of the hatchery, nursery and grow-out sectors of the striped catfish industry in the Mekong Delta and the movement of stock between each sector (De Silva and Phuong, 2011)
The third step in striped catfish production is grow-out Farm size ranges from 0.2 to 30 ha Earthen ponds (0.08-2.2 ha) are stocked with 18-125 fish/m 2 (mean= 48), equalling 5-31 fish/m 3 (mean) (Phan et al., 2009) The vast majority of farmers use commercial pelleted feed Most striped catfish
24 farms abstract untreated water from, and release untreated water and waste directly into, nearby canals and rivers that make up the water network in the Mekong Delta The exchange of pond water is increased during the production cycle and reaches almost 100 % daily in the fifth or sixth month of rearing until harvest (Phan et al., 2009) Farmers’ practices with regard to the use of chemicals and antimicrobials are described in Section 1.3.2 and information about fish health management at farm level, e.g small-scale vs large-scale farms, is provided in Manuscript I
Most grow-out farms are small-scale and operated by households, whereas large-scale farms are mainly owned by processing and/or feed companies (Belton and Little, 2011; Belton et al., 2011)
Fish health management
1.3.1 Diseases in striped catfish aquaculture
Intensive production in open farming systems, i.e sharing influent and effluent water sources and their use by several adjacent grow-out farms (Directorate of Fisheries, 2012), is associated with frequent disease outbreaks and high mortality rates, mainly as a result of bacterial infection Typical mortality rates are up to 30 % from after stocking to the mid-production cycle and less than 10 % in later months (Phan et al., 2009) Bacillary necrosis of Pangasius (BNP) caused by Edwardsiella ictaluri is the most serious disease, causing acute mass mortalities of up to 60 % (Crumlish et al., 2002; Tu et al., 2004; Oanh and Phuong, 2008; Phan et al., 2009) Motile Aeromonad septicaemia (MAS) or haemorrhagic disease, caused by several opportunistic Aeromonas spp (A hydrophila, A sobria and A caviae), also causes significant losses (Dung et al., 2008a; Ly et al., 2009; Phan et al., 2009; Huong et al., 2011) Furthermore, a number of endo- and ecto-parasite infections reduce growth and increase susceptibility to bacterial infections (Nguyen et al., 2008; Phan et al., 2009) Rapid progress is being made in developing vaccines for bacterial infections of striped catfish in Vietnam (Thinh et al., 2009; Dung, 2011), e.g a commercial BNP vaccine for injection has been approved by the Vietnamese authorities and was launched on the market in early 2013
A study in the Mekong Delta in 2008-2009 showed that E ictaluri isolates were fully resistant to flumequin, combined trimethoprim and sulfamethoxazole and highly resistant to streptomycin (84
In Aeromonas hydrophila isolates, resistance to chloramphenicol and enrofloxacin was observed at about 74% each, while these isolates remained susceptible to ampicillin, amoxicillin, and cephalexin (Huong et al., 2011) Overall, resistance issues appear less severe in A hydrophila, which is susceptible to florfenicol, tetracyclines, and quinolones, but shows resistance to streptomycin (56%) and to the combination of trimethoprim and sulfamethoxazole.
(33 %) (Huong et al., 2011) A more recent study conducted between 2010 and 2012 concludes that
E ictaluri and A hydrophila isolated from striped catfish in the Mekong Delta are resistant to nearly all the approved antimicrobials for use in Vietnamese aquaculture (Phuc et al., 2012) The increase in antimicrobial-resistant bacterial pathogens may be associated with inadequate disease diagnosis and treatment There is limited knowledge about farmers’ fish health management practices and a better understanding of their knowledge of fish disease and treatment practices is required (Manuscript I)
1.3.2 Chemical use in striped catfish aquaculture
A wide range of chemicals are used to disinfect equipment and treat water in tanks and ponds, while various antimicrobials are used for disease prevention and treatment in hatcheries, nurseries and grow-out ponds (Chinh, 2005; Phan et al., 2009; Bui et al., 2010; Phuong, 2010) The term antimicrobial rather than antibiotic is used throughout this thesis as the former includes antibiotics produced by live microorganisms as well as chemically-synthesised compounds, e.g sulfonamides and enrofloxacin Before stocking, 76 % of farms use sodium chloride (78 %) and antimicrobials
Phan et al (2009) reported that 32% of antimicrobial use targeted fingerlings Phuong (2010) shows that the most popular antimicrobials used in striped catfish farming to treat bacterial disease are florfenicol (76.6%), enrofloxacin (67.2%), and amoxicillin (43.8%), with quinolones being the most common class (used by 77% of farmers), followed by the phenicol group (e.g., florfenicol) The Phuong (2010) survey also reveals that farmers use a wider and more diverse range of antimicrobials than in Chinh (2005), where only 50% of farmers reported antimicrobial use However, prior studies largely provide general information about chemical use in striped catfish aquaculture—types of antimicrobials and the share of farmers using them for bacterial disease—without specifying dosage, disease-specific usage, or how the chemicals are prepared and applied; details on these issues are given in Manuscript I.
In 2012, Vietnamese authorities approved 28 antimicrobials for therapeutic use in aquaculture (Tai, 2012) By comparison, mature fish-farming sectors in the USA, Norway, Scotland, Chile and Canada have approved far fewer antimicrobials, with the United States licensing only four antimicrobials for aquaculture—florfenicol, oxytetracycline, and combinations of sulfadimethoxine and ormethoprim (FDA, 2011) The widespread use of antimicrobials in striped catfish culture mirrors the early years of Norwegian salmon aquaculture, where the introduction of effective vaccines caused a dramatic decrease in antimicrobial usage (Burridge et al., 2010) In 2012, a total of 2,913 products were registered for use in Vietnamese aquaculture, including 813 veterinary drugs (Tai, 2012) Clearly, drug-quality testing and the approval of such a high number of products is resource-intensive and costly, and the effectiveness of current approval procedures has been questioned, with concerns raised regarding the quality of the approved antimicrobial products (Manuscript II).
Disinfectants are used widely in striped catfish aquaculture (Chinh, 2005; Phan et al., 2009;
Phuong, 2010) In 2009, a survey showed that grow-out striped catfish farmers use 17 types of chemicals, e.g lime to kill parasites and other fish (95.3 %) during pond preparation, sodium chloride to improve water quality and to prevent and treat bacterial disease (87.5 %), and copper sulfate to prevent and treat ecto-parasites (65.6 %) (Phuong, 2010) The effectiveness, toxicity to fish and environmental impact of disinfectants used in aquaculture have been described, e.g for calcium hypochlorite (Boyd and Massaut, 1999; Emmanuel et al., 2004), copper sulfate (Carvalho and Fernandes, 2006; Marcussen et al., 2014), and benzalkonium chloride (Rico and Van den Brink, 2014) Probiotics are used for a variety of purposes in aquaculture, such as immune stimulation, improved digestion and the competitive exclusion of unfavourable bacterial populations in fish intestines and the pond water/sediment environment (Aguirre-Guzmán et al.,
Current information on differences in chemical usage practices between small-scale and large-scale farms is limited, and there is little understanding of whether stocking densities and farm size correlate with diseases reported and chemical use (Manuscript I).
Occupational health hazards in aquaculture
Occupational health hazards in striped catfish aquaculture have not been described in previous studies Inappropriate handling when using chemicals might lead to occupational health hazards such as respiratory, skin and other health issues associated with direct exposure to chemicals, but there are also risks of infectious diseases, e.g from zoonotic pathogens such as Vibrio spp and Aeromonas spp (Durborow, 1999; Erondu and Anyanwu, 2005; Cole et al., 2009; Moreau and Neis, 2009; Watterson et al., 2012) Inappropriate handling of burnt lime and hydrated lime can cause blindness and severe irritation following eye or skin contact (Boyd and Massaut, 1999;
Erondu and Anyanwu, 2005) Direct contact and inhalation of strong vapour disinfectants or anti- parasite products, e.g chlorine-based products, formaldehyde and pesticides, can cause severe burns or skin irritation and lead to the development of respiratory ailments such as bronchitis, rhinitis and asthma (Karkkainen, 2002; Uronu and Lekei, 2004) Direct contact with antimicrobials may cause antimicrobial allergies (Thong, 2010) The occupational health hazards in striped catfish aquaculture are described in Manuscript I
Quality of antimicrobials used in aquaculture
The antimicrobial quality of drugs used in human medicine has been tested in developing countries
A study of the quality of antimicrobials used to treat sexually transmitted diseases in Myanmar reports that one product did not contain the active drug declared (chlortetracycline) and the product did not show any in vitro activity against the bacteria Moreover, seven of the 21 products did not contain the declared compound concentration: one product contained more than that declared and six products contained less than that declared, and two products contained only 48 % of the declared concentration (Prazuck et al., 2002) Similar findings have been reported for human drugs in Ghana (Bekoe et al., 2014) where some antimicrobials in tablet and capsule form have been found to contain as little as 41 % of the declared concentration (erythromycin), and only four of the
10 tested products contained more than 90 % of the declared concentration Compared to human antimicrobial products (Prazuck et al., 2002; Abdulah, 2012; Bekoe et al., 2014), there seems to be few if any studies on the quality of antimicrobial products used in aquaculture
Poor quality antimicrobial products results in farmers providing medication at sub-therapeutic dosages and subsequent treatment failure, excess fish mortality and economic losses Furthermore, sub-therapeutic doses of antimicrobials are the most important factor selecting for antimicrobial resistance, both among bacterial pathogens associated with specific diseases and the normal bacterial microflora, e.g in aquaculture environments (Depaola, 1995; Cabello, 2006; Huong et al., 2011; Phuc et al., 2012) There is therefore an urgent need to determine the quality of common antimicrobial products used by Vietnamese striped catfish farmers (Manuscript II).
Elimination and withdrawal time of antimicrobials
Enrofloxacin (ENR), a fluorinated quinolone carboxylic acid derivative, is used by most striped catfish farmers to treat BNP and MAS due to its broad spectrum of activity (Brown, 1996; Chinh,
Enrofloxacin (ENR) is oxidatively de-ethylated to ciprofloxacin (CIP) by cytochrome P450 enzymes in many animal species, including fish and crustaceans, with the structures of ENR and CIP shown in Figure 5 An experimental study on the elimination of ENR and its metabolite CIP in striped catfish fed medicated diets under tank conditions reported residual ENR levels of about 98–188 µg/kg and CIP levels of about 13–17 µg/kg.
CIP/kg in muscle tissue (including skin) seven days after treatment (Danyi et al., 2010) However, the withdrawal time has not been set for the use of enrofloxacin in striped catfish under grow-out conditions in the field The pharmacokinetics of ENR and CIP have been reported for Atlantic salmon (Salmo salar) (Stoffregen et al., 1997), European seabass (Dicentrarchus labrax) (Intorre et al., 2000), gilthead seabream (Sparus aurata L.) (della Rocca et al., 2004), rainbow trout
(Oncorhynchus mykiss) (Luccheti et al., 2004), Nile tilapia (Oreochromis niloticus) (Xu et al.,
2006), Korean catfish (Silurus asotus) (Kim et al., 2006), and black tiger shrimp (Penaeus monodon) (Tu et al., 2006), but not for striped catfish A withdrawal period of 62 days is required for rainbow trout treated with ENR at a dose of 10 mg/kg of body weight per day for five consecutive days in field conditions (12-13 °C) to meet 100 àg/kg, which is the maximum residue limit (MRL) accepted by the European Commission (Luccheti et al., 2004) Drug elimination rates appear to be positively correlated with water temperature (Luccheti et al., 2004; Xu et al., 2006) Stoffregen et al (1997) conclude that ENR shows a slower elimination in the skin of Atlantic salmon as compared to muscle tissue The elimination of ENR in striped catfish was studied in grow-out field conditions and the results are presented in Manuscript III
Other than ENR, mixtures of sulfamethoxazole (SMX) and trimethoprim (TMP) (Fig 5) are also used in striped catfish and tilapia aquaculture (Rico et al., 2013; Phu et al., 2013) in order to obtain an antibacterial synergistic effect (Wormser et al., 1982) Pharmacokinetic studies on sulfonamides were undertaken for Romet 30 , a commercial product containing sulfadimethoxine and ormetoprim, in Atlantic salmon (Samuelsen et al., 1997), Nile tilapia, summer flounder (Paralichthys dentatus), walleyes (Sander vitreus) (Kosoff et al., 2007) and channel catfish (Ictalurus punctatus) (Rawles et al., 1998) The elimination of sulfonamides and trimethoprim depends on fish species and rearing temperature (Kosoff et al., 2007) To meet US-FDA safety levels (50 àg/kg for ormethoprim and
Withdrawal times for sulfadimethoxine are species-dependent: Atlantic salmon require 11 and 38 days after five days of oral Romet 30 treatment, while residue testing shows that after five days of medicated feed, sulfadimethoxine and ormetoprim were not detected in the flesh of Nile tilapia after six days, in summer flounder after 21 days, and in walleyes after 10 days (Kosoff et al., 2007) In 2013, tilapia imported from Vietnam to Europe showed residues of trimethoprim (76–323 µg/kg) and sulfadiazine (199 µg/kg), as reported by RASFF (2014) This pattern highlights the urgent need to assess drug elimination kinetics and establish clear withdrawal guidelines to protect seafood safety.
33 withdrawal time of trimethoprim and sulfonamides in both tilapia and striped catfish (Manuscript IV)
Figure 5 Chemical structure, molecular formula and weight of studied compounds (PubChem, 2013)
In order to establish the withdrawal time for antimicrobials or their pharmacokinetics in animals, the methods for determining residues in specific sample matrices should be validated Methods of analysis of sulfonamides and trimethoprim in various animal sample types have been developed, mainly following the introduction of liquid chromatography coupled to mass spectrometry (LC- MS/MS) (González et al., 2007; Stubbings and Bigwood, 2009; Cháfer-Pericás et al., 2010;
Fernandez-Torres et al., 2011; Lopes et al., 2012b) To extract sulfonamides and trimethoprim from
34 edible food, different extraction methods are applied, such as liquid-solid phase extraction
(González et al., 2007) or enzymatic-microwave assisted extraction (Fernandez-Torres et al., 2011)
QuEChERS (quick, easy, cheap, effective, rugged and safe) extraction has been applied and validated for pesticide and antimicrobial residue analysis in fish and environmental samples, offering a fast, inexpensive workflow with fewer steps that shortens extraction and cleanup times This approach is supported by published validations (Stubbings and Bigwood, 2009; Lopes et al., 2012a, 2012b) In this context, the validation of a UPLC-MS/MS method coupled with the Agilent Bond Elut QuEChERS extraction kit for monitoring sulfamethoxazole (SMX) and trimethoprim (TMP) in striped catfish muscle is described in Manuscript IV.
Legislation in importing countries
Regulations about veterinary drug residues in food used for human consumption differ between seafood-importing countries The maximum residue limit (MRL) of 100 à g/kg (sum of ENR and CIP) set by the European Commission (EC) and that of 5 àg/kg by US Food and Drug
Administration (US-FDA) of fish for human consumption (EC, 2010; Love et al., 2011) (Section 1.6) Furthermore, EU-MRL for SMX and TMP in fish is 100 and 50 àg/kg respectively (EC,
2010) The US-FDA has not listed SMX and TMP for residue analysis testing in fish for human consumption, while 100 àg/kg of sulfadimethoxine and 50 àg/kg of ormethoprim have been set as safety levels (Love et al., 2011) The US-FDA also has a zero tolerance level for sulfamerazine The EU-MRL for the sum of ENR and CIP, SMX and TMP residue in fish as stated in Regulation (EU) N°37/2010 of the EC is for composite fish muscle and skin samples.
Regulations and advisory services on chemical use in aquaculture in Vietnam
Vietnam’s Ministry of Agriculture and Rural Development (MARD) released Circular No
Vietnam updated its list of drugs, chemicals, and antimicrobials banned or restricted for use in aquaculture in 2009 with Decision 15/2009/TTBNN, replacing the 2005 version (MARD, 2009) A high rate of contamination in aquaculture products—particularly enrofloxacin (ENR) and trifluralin in species such as striped catfish and shrimp—was reported by the United States, Japan, and Canada (Love et al., 2011), prompting regulatory action to protect public health: trifluralin was banned in 2010, and ENR, cypermethrin, and deltamethrin bans followed in 2012 (MARD, 2010a; 2012a) Moreover, in 2012, Decision 1471/QĐ-BNN-QLCL was issued by the Ministry of Agriculture and Rural Development (MARD) to reinforce these controls.
MARD lists the drugs, chemicals and antimicrobials that must be checked for in exported products (i.e striped catfish and shrimp) following the food safety requirements of importing countries (MARD, 2012b)
To help meet the food safety requirements of export markets such as the European Union, USA and Japan, the Vietnamese National Agro-Forestry-Fisheries Quality Assurance Department
NAFIQAD operates a national aquaculture residue monitoring programme (NRMP) to track antimicrobial residues in farmed fish Data from the NRMP show a decrease in antimicrobial residue findings in striped catfish from 2011 to 2013 (NAFIQAD, 2014) However, ENR residues detected in striped catfish in 2014 indicate that some farmers continued using ENR despite a ban by authorities in 2012.
The provincial aquaculture departments in the Mekong Delta are responsible for the management of aquaculture activities in the provinces They provide training on disease management and other technical aspects of aquaculture, e.g regulation updates, disease outbreaks and the VietGAP aquaculture standards Chemical companies and feed companies, sometimes in cooperation with academic institutes, provide training on chemical usage and disease treatment with the main purpose of promoting their products On-farm or laboratory-based disease diagnosis is occasionally provided by so-called chemical shops, drug and feed companies, provincial aquaculture departments or academic institutions (unpublished data) According to Phan et al (2009), the majority of farms have acquired striped catfish farming experience through family tradition (39 %), training (40 %) and advice from other farmers (40 %) Information about striped catfish diseases, how farmers diagnose and treat disease, and the role of advisory services is described in details in Manuscript I
SPECIFIC OBJECTIVES AND HYPOTHESES OF THE THESIS
This thesis consists of studies that collectively contribute to increasing knowledge about chemical usage in striped catfish aquaculture, with an emphasis on fish health management practices, the quality of the antimicrobials used, and their withdrawal periods The specific objectives were to evaluate current patterns of chemical usage, assess antimicrobial quality, and determine appropriate withdrawal periods.
− to evaluate fish health management practices and occupational health hazards associated with striped catfish aquaculture in Vietnam (Manuscript I)
− to determine the quality of antimicrobial products commonly used in striped catfish aquaculture (Manuscript II)
− to establish the withdrawal period for enrofloxacin following treatment of striped catfish through on-farm trials (Manuscript III)
− to validate a UPLC-MS/MS method for the detection of sulfamethoxazole and trimethoprim and to establish the withdrawal periods for the two antimicrobials following treatment of striped catfish and hybrid red tilapia to meet EU food safety regulations (Manuscript IV)
− Fish health management practices differ between small and large-scale grow-out farms, including during the various production stages
− The quality of antimicrobial products used in striped catfish aquaculture is poor
− Current withdrawal times are appropriate for enrofloxacin used in striped catfish and sulfamethoxazole and trimethoprim used in striped catfish and hybrid red tilapia aquaculture
METHODOLOGY
Fish health management practices and occupational health hazards in striped catfish (Manuscript I)
In order to evaluate the fish health management practices in striped catfish (Pangasianodon hypophthalmus) farming, semi-structured interviews were conducted with 83 aquaculture enterprises (15 hatcheries, 32 nursery farms and 36 grow-out farms) located in five provinces culturing striped catfish (Fig 1) Grow-out farms were aggregated into large-scale (10) and small- scale (26) (the selection criteria are described in Section 1.2.3) which were randomly selected from a sample frame of 212 striped catfish farms identified in a previous SEAT survey (SEAT; Murray et al., 2013) Although not a stratification criteria, this resulted in the selection of farms located in the provinces of An Giang (22 farms), Can Tho (six farms), Dong Thap (five farms), Tra Vinh (one farm) and Ben Tre (two farms) (Fig 1) The hatcheries and nurseries were randomly selected from a list provided by the provincial aquaculture departments
3.1.2 Survey design and data collection
Three semi-structured questionnaires were developed for face-to-face interviews with the owners or managers of hatcheries, nursery farms and grow-out farms (Annex 2) Information was collected on the educational background of the respondents (farm owner/manager) and farm infrastructure (e.g pond area, volume and stocking density) as well as details of antimicrobial, disinfectant and probiotic usage during the last crop of striped catfish and whether these compounds were used for disease prevention or treatment Interviews also incorporated less structured in-depth discussions on the types and frequency of diseases and the respondents’ understanding of the clinical symptoms of these diseases Types, dosages and rationales for decision-making on chemical usage were also explored Farmers were asked about perceived health hazards associated with their work, e.g protection equipment used when handling chemicals, knowledge of toxicological and exposure risks associated with chemical use, and types of accidents that had occurred due to chemical use The interviews were conducted between July and December 2011
Multivariate analyses were used to evaluate correlations between farm characteristics (independent variables) and reported diseases (dependent variable) in the grow-out and nursery farms, and between reported diseases (independent variables) and the chemical treatments used (dependent variable) The significance of any correlation between the independent variables and the variance in the dependent variable dataset was tested by redundancy analysis (RDA; performing 499 Monte Carlo permutations) using the CANOCO 5 software package (Ter Braak and Smilauer, 2012).
Quality of antimicrobial products used in striped catfish aquaculture (Manuscript II)
Twenty-one antimicrobial products marketed by ten Vietnamese companies were obtained during visits to nine so-called chemical shops located in Dong Thap, An Giang and Can Tho provinces in the Mekong Delta, Vietnam The products were collected with the aim of determining the actual concentration of antimicrobials in the products compared to the declared concentrations on the product labels Two different batches of 11 products were obtained for comparison analysis The selected products represented the antimicrobials most commonly used by the striped catfish farmers and included 11 products with a single antimicrobial (amoxicillin, doxycycline or florfenicol) and
Ten products contained a mixture of two antimicrobials (Table 3, Section 4.2) Sixteen antimicrobial products were in powder form, with the rest in liquid form All products carried expiry dates of up to two years, and the antimicrobial content analysis was performed at least one year before expiry Samples were stored at ambient temperature, as in chemical shops, until laboratory analysis.
Ultra high-performance liquid chromatography mass spectrometry (UPLC-MS/MS) was used to analyse the concentrations of sulfonamides (sulfamethoxazole, SMX; sulfadiazine, SDZ; sulfamethazine, SMZ) and trimethoprim (TMP) (Zhao and Stevens, 2012); amoxicillin (AMX) and cefalexin (LEX) (USDA, 2011); and ciprofloxacin (CIP) (Pearce et al., 2009), while concentrations of florfenicol (FFC) (USDA, 2010) and doxycycline (DOX) (Kamakura et al., 1994) were analysed by high-performance liquid chromatography with UV detection (HPLC-UV) Coded samples
39 without product labels were analysed at the NAFIQAD laboratory in Can Tho, Vietnam which is ISO/IEC 17025 accredited (ISO, 2010) Analytical results were presented as the percentage (%) of the active antimicrobial compound concentration as compared with the concentration declared on product label Details on the antimicrobial quality analysis are given in Manuscript II
Product label information was evaluated for its coverage of prophylactic use, the diseases treated, the therapeutic dosage, the method for calculating the dose, instructions for preparing medicated feed, and the required withdrawal time.
Elimination of enrofloxacin in striped catfish following on-farm treatment (Manuscript III)
This experiment was set up one month before the ban of the use of fluoroquinolones in animal production in Vietnam (MARD, 2012a) Three striped catfish farms located in Thot Not district, Can Tho City, Vietnam were asked to participate voluntarily in the trials, each allocating one pond for the experiment Enrofloxacin (ENR) (20 % w/v, Vimenro®, Can Tho City, Vietnam) was applied once a day for five consecutive days at a dose of 10 mg/kg body weight in three ponds (Fig
6) The preparation of medicated feed and feeding were carried out by the farmers according to their normal practices
At each sampling time, ten fish (30-60 g) were collected from each pond using a small cast net Sampling was always undertaken in the morning on the day before (D0) the medicated feed was applied and again each day after the first (D1), third (D3) and fifth (D5) application of the medicated feed Sampling continued 7 (D7), 15 (D15), 30 (D30) and 45 days (D45) after the last treatment (Fig 6) Temperature, pH and dissolved oxygen concentrations in the water were measured in the pond at each sampling time The concentration of ENR and CIP in muscle/skin samples was determined for a period of 45 days On the day of harvest, skin samples were collected for ENR and CIP residue analysis Pooled samples consisting of muscle/skin from ten fish (100 g) and medicated feed (100 g) were placed in clean plastic bags and kept on ice in an insulated box and transported to the laboratory, where they were homogenised using a meat blender (Moulinex,
France) and stored at -20 °C until analysis At the time of fish harvest, only fish skin (100 g) was collected
Figure 6 Treatment with enrofloxacin and sampling times (day)
ENR and its metabolite CIP were analysed according to the validated method of Danyi et al
(2010) For the extraction of ENR and CIP residues, 1.0 ± 0.05 g of homogenised fish muscle/skin, feed or skin was introduced into a 50 mL Falcon tube and spiked with 100 à L of 3 à g/mL lomefloxacin solutions as the internal standard Samples were then extracted by acetonitrile, followed by a cleaning step in solid phase extraction cartridges SDB-RPS (3M Bioanalytical technologies, St Paul, MN, USA) Detection and quantification were performed by LC/MS-MS on a
Instrumentation comprised a Waters Alliance Separations Module that integrates an auto-sampler, solvent delivery system, and column heater (Waters, Milford, USA) linked to a Quattro Ultima Platinum triple quadrupole mass spectrometer (Waters Micromass, Manchester, UK) Details on sample extraction and analytical procedures are provided in Manuscript III.
Withdrawal time for sulfamethoxazole and trimethoprim following treatment of striped catfish and hybrid
Striped catfish (40.3±7.2 g) and red tilapia (10.7±4.5 g) fingerlings were provided by commercial striped catfish and red tilapia nursery farms located in Can Tho City, Vietnam The absence of sulfamethoxazole (SMX) and trimethoprim (TMP) residues in the fingerlings was confirmed before the experiment After two weeks’ acclimation, the fingerlings were transferred into six experimental
500 L tanks at a stocking density of 50 fish/m 3 for striped catfish and 100 fish/m 3 for red tilapia, resembling normal stocking densities on the farms After another two weeks’ acclimation, striped
41 catfish and red tilapia were treated once a day with medicated feed containing SMX (1000 mg/kg feed) and TMP (200 mg/kg feed) for five consecutive days The medicated feed was prepared according to the instruction labels of the commercial product Trimesul® (SMX 20 % and TMP
Medicated feed in this study was supplied by Vemedim Co Ltd, Vietnam, and followed farmers’ common dosage practices and methods for incorporating antimicrobials into feed pellets The fish in the tanks were fed twice daily, and all prepared medicated feed was used the same day.
To determine the elimination of sulfamethoxazole (SMX) and trimethoprim (TMP) in on-farm trials, two red tilapia cages were installed at a single farm on the Mekong River in Thot Not District, Can Tho City, Vietnam Each cage has a volume of 128 m^3 and a width of 4 m, enabling controlled sampling of antibiotic residues in a real-world cage aquaculture setting.
An on-farm trial was conducted in ponds sized 8 m in length and 4 m in depth, with striped catfish ponds not available for the trial; red tilapia were stocked at 100 fish per cubic meter and reared for 1.5 months without any prior exposure to SMX and TMP before the experiment, with fish weighing 117±15 g at the start and a biomass in each cage of about one tonne; Trimesul® was applied once daily for five consecutive days, as farmers diluted 250 g of Trimesul® powder in 5 L of water and mixed it by hand with 50 kg of pelleted feed, the medicated feed being placed on a large plastic sheet for 15 minutes to allow the antimicrobials to be absorbed before being fed to the fish in the morning, all medicated feed being used on the day of preparation and non-medicated feed fed in the afternoon, with farmers ensuring by observation that all the medicated feed was consumed.
At each sampling time, five striped catfish and ten red tilapia were collected from each tank with a small hand net, and the fish were killed by a blow to the head with a metal stick Separated muscle and skin samples from striped catfish and muscle/skin samples from red tilapia were collected one day before the first feeding with medicated feed, as well as on days 1 and 5 during medication and on days 3, 7 and 15 after medication termination At each sampling, muscle and skin from five striped catfish and muscle/skin from ten tilapia were pooled into individual composite samples, minced and homogenised using a Moulinex meat blender (France) A 20 g subsample of the medicated feed was collected and minced in the blender Twenty grams of each composite sample were stored at -20 °C until analysis for SMX and TMP residues by UPLC-MS/MS Temperature, pH and dissolved oxygen concentrations in tank water were monitored during the experiment.
42 the on-farm trial with the red tilapia cages, the fish were collected, processed and stored as described for fish used in the tank experiment
3.4.3 Sulfamethoxazole and trimethoprim residue analysis
An analytical method was adapted from a LC-MS/MS method using Agilent Bond Elut QuEChERS to determine sulfonamides in bovine liver (Zhao and Stevens, 2012) For the extraction of SMX and TMP residues, 2.0 ± 0.05 g of homogenised striped catfish skin, striped catfish muscle, tilapia muscle/skin and feed were introduced into a 15 mL Falcon tube and spiked with 50 à L of 1 àg/mL sulfapyridine solution as the internal standard Samples were then extracted using 10 mL acetonitrile: CH3COOH (100:1, v:v) and cleaned up by Agilent Bond Elut QuEChERS EN Kits (Agilent, CA, USA) Detection and quantification were performed by a Waters ACQUITY ultra high-performance LC system coupled to Xevo™ TQ triple quadrupole mass spectrometry (Waters, Milford, MA, USA) Details on sample extraction and analysis are given in Manuscript IV
Because no certified reference material was available, validation was performed in accordance with Commission Decision 2002/657/EC by spiking a blank matrix with known amounts of analytical compounds and assessing recovery and precision Validation parameters, including recovery and precision (repeatability and within-laboratory reproducibility), were determined using 18 reference matrix samples of striped catfish muscle fortified at three levels corresponding to MRL/2, MRL, and MRL×1.5 Repeatability was calculated from the average concentration (µg/kg) and the coefficient of variation (CV, %) of results obtained from six replicate analyses conducted on the same day, while within-laboratory reproducibility was calculated from results across six independent days, with mean and CV computed accordingly The recovery for each sample was calculated by comparing the measured concentrations to the fortified concentrations, and the mean recovery (%) was reported to indicate method accuracy.
CV (%) were calculated from all the results of six days at each fortification level The decision limit CCα (α = 5 % for substances with a MRL) (EC, 2002) was calculated as the average concentration of the reference matrix fortified at the MRL concentration of SMX (100 àg/kg) and TMP (50 àg/kg) in within-laboratory reproducibility conditions, plus 1.64 times the corresponding standard deviation The limit of detection (LOD) was determined as the concentration of the compound for which the signal-to-noise ratio (S/N) was higher than 3 The limit of quantification (LOQ) was
43 considered to be the lowest concentration of the matrix calibration curves (for which the S/N was much higher than 10)
OVERVIEW OF RESULTS AND DISCUSSION
Fish health management practices (Manuscript I)
The apparent absence of reported disease in hatcheries and brood stock ponds is possibly a result of a very short production period from hatching eggs to selling the striped catfish fry (1-3 days) and keeping broodstock in well-managed ponds at low stocking densities In 2009, Bui et al (2010) observed that 27 % of 45 hatcheries surveyed experienced diseases due to bacterial and parasite pathogens It is not clear whether these differences in reported diseases were due to improved management or to underreporting
BNP and MAS remain the most prevalent and severe diseases in striped catfish aquaculture, affecting both nursery and grow-out farms, with farmers reporting one to five outbreaks per crop and experiencing high mortalities and economic losses; antimicrobial treatments are widely used, yet analyses show no significant link between the specific bacterial disease and the antimicrobials selected, and GlobalGAP implementation did not significantly reduce disease occurrence Vaccination has been limited, with about 20 million BNP vaccine doses used in 2013, representing only around 1% of fingerlings stocked, largely due to the need for manual injection and the per-fish cost (2.35–3.29 US cents); this underscores the need for innovative control strategies, such as immersion vaccines similar to those used in rainbow trout and the development of immune-stimulant approaches for striped catfish.
E ictaluri infection has revealed that levamisole and lipopolysaccharide enhance the immunological response in striped catfish (Hang et al., 2014) The future use of antimicrobials in
Striped catfish production should be grounded in knowledge of antimicrobial pharmacokinetics and the susceptibility of relevant bacterial pathogens, ensuring treatments are effective while minimizing resistance Strengthening fish health management by applying aquaculture standards such as GlobalGAP and VietGAP (Section 4.1.2) and advancing zonal aquaculture at a regional level can efficiently reduce disease spread and the impact of disease outbreaks, as highlighted in SFP (2014).
4.1.2 Disease diagnosis between small-scale and large-scale farms
Many of the large-scale farms (7/10) had basic diagnostic equipment, particularly microscopes, and managers had received formal fish health training to make effective use of them (Manuscript I) Most large-scale grow-out farms (9/10) had a fish health management plan that included descriptions of disease symptoms, means of diagnosis and methods of preventing and treating diseases In contrast, none of the small-scale grow-out and nursery farms surveyed had such plans in place Statistical analysis revealed a positive correlation between keeping records of chemical use and reporting specific diseases because the maintenance of such records is required by certification standards The results from this study indicated that disease diagnostic capacities are far superior in large-scale grow-out farms compared to small farms, for example there is the delegation of health management to staff with formal training, the development of fish health management plans and the keeping of farm records on production performance and chemical usage These are all requirements for GlobalGAP certification (GlobalGAP, 2014) It should be noted that having such certification was a selection criteria for large-scale farms included in the study Many international aquaculture standards aim to promote good management practices, of which fish health management is an essential and integrated part (Belton et al., 2011) Hatchery, nursery and small-scale farmers were not certified by any scheme and there was very limited capability on their farms to diagnose diseases Although these farmers stated that they could easily recognise the symptoms of BNP and MAS, symptoms of white spot disease were misdiagnosed as those of BNP even though the former is caused by infection by endo-parasites Some farmers ascribed the two groups of MAS symptoms to different pathogens However, Dung et al (2008a) suggest that these could be different stages of MAS or cases of dual infections with bacterial and parasite pathogens
Results with regard to aquaculture knowledge and training participation showed that fewer than
25 % of nursery owners had a college or bachelor degree in aquaculture Few small-scale farmers (9/26) had participated in any aquaculture training, while all (10/10) managers of the large-scale
46 farms had participated in one or more training courses and/or workshops, e.g on certification standards Most nursery farmers (25/32) participated in workshops (16/32) and training courses (10/32) to learn about good aquaculture management practices in events organised by state extension service units (12/32), universities (11/32), and feed companies and chemical suppliers (11/32) Limited aquaculture knowledge and participation in aquaculture training resulted in limitations with regard to disease diagnosis by small-scale grow-out and nursery farmers
Moreover, large-scale farmers were also more likely to follow the instructions of veterinarians or fish health technicians when applying chemicals, whereas other farmers primarily applied chemicals based on their own previous experience (22/32 nursery farmers, 12/15 hatchery farmers, 11/26 small-scale farmers) Aquaculture standards require that veterinarians or fish health technicians be hired or employed to offer instructions on disease diagnosis and treatment, which would enhance fish health management practices on small-scale grow-out and nursery farms
The limited capability to diagnose diseases represents a serious problem since (i) diagnoses based on gross visual symptoms alone are particularly challenging in juveniles and requires the use of laboratory facilities (e.g a microscope) and (ii) BNP and other bacterial diseases cause the greatest losses of young fish during the first half of the grow-out cycle (Dung et al., 2008a; Phan et al.,
2009), pointing to a need for improved early diagnosis and treatment Although international certification schemes, e.g GlobalGAP, have established procedures for improved fish health management, chemical use and food safety, they also have significant limitations Certified farms are mainly large-scale farms owned by processing plants or feed companies, while a limited number of small-scale farms have the resources and knowledge needed to fulfil certification requirements This led to the development of the VietGAP (Vietnamese Good Aquaculture Practice) standards in
Since 2013, the promotion of aquaculture quality assurance has targeted small-scale farmers, with VietGAP offering a simpler and less costly approval pathway than GlobalGAP and enabling training through local government extension services on fish health management that requires consultation with a trained veterinarian; VietGAP standards and Better Management Practice (BMP) schemes are designed to strengthen smallholders’ ability to use chemicals prudently and to reduce contamination risks in processed and exported products, while also highlighting the need to expand government-delivered diagnostic services to support improved disease and health monitoring in the sector.
47 extension services and the private sector, e.g the role of chemical shops in disseminating information on the prudent use of antimicrobials and other chemicals
4.1.3 Chemical use in striped catfish
All nursery and grow-out farmers stated that they did not use antimicrobials prophylactically due to their high costs A majority of nursery farms (22/32) and grow-out (23/26 small-scale and 9/10 large-scale) farms used antimicrobials to treat bacterial diseases during the nursing of fingerlings and the grow-out period In 2012, a total of 28 antimicrobials were authorised for therapeutic use in Vietnamese aquaculture (Tai, 2012), of which 24 were reported as being in use by farmers in this PhD study (Table 2) This is far higher than the range of antimicrobials approved for use in more mature, consolidated fish farming sectors in the USA, Norway, Scotland, Chile and Canada In the USA, for example, there are only four licensed antimicrobials for use in aquaculture: florfenicol, oxytetracycline and combinations of sulfadimethoxine and ormethoprim (FDA, 2011) The relevant Vietnamese authorities should consider limiting the number of approved antimicrobials to those that can effectively treat the main disease causing bacterial pathogens in different cultured aquatic species, and avoid the use of antimicrobials that are essential in human medicine (WHO, 2011) Moreover, authorities should perform more stringent controls to ensure the quality of antimicrobial products used in striped catfish farming (Section 4.2 and Manuscript II)
Enrofloxacin and florfenicol were the most common antimicrobials used in the treatment of BNP and MAS, which is consistent with other studies (Chinh, 2005; Phuong, 2010) (Table 2) No farmers reported using any internationally banned antimicrobials such as chloramphenicol or those of the nitrofuran class This was also confirmed by personal communication with shop owners selling chemicals for aquaculture about the unavailability of banned antimicrobials This study’s findings did not reveal that specific antimicrobials were being used by farmers to treat BNP and MAS, most likely due to a combination of misdiagnosis and/or resistance of the bacterial pathogens involved Antimicrobial susceptibility testing in grow-out and nursery farms has not been undertaken regularly when farms experience disease outbreaks caused by bacterial pathogens
(unpublished data) There is therefore a requirement for increased capability to provide antimicrobial susceptibility testing services A database containing information about the antimicrobial susceptibility of the main bacterial pathogens should be established, together with an
Effective dissemination of information regarding which antimicrobial is recommended for the treatment of specific diseases in striped catfish is essential for guiding appropriate therapy The improved use of antimicrobials should go hand in hand with the development and introduction of vaccines, integrating treatment with preventive strategies to enhance overall fish health and production.
Table 2 Antimicrobials reported as being used by nursery and grow-out striped catfish farms in 2011
Oxytetracycline and thiamphenicol - - 1 a number of farmers reported to use the antimicrobial.
Excessive antimicrobial use in striped catfish aquaculture can negatively affect the aquatic environment by altering ecological balance and driving the buildup of antimicrobial resistance As resistant fish pathogens emerge, disease management becomes harder within the culture system, threatening production and ecosystem health In addition, humans may face increased exposure to antimicrobial resistance through consumption of cultured aquatic organisms, raising health risks and highlighting the need for responsible antimicrobial use and resistance surveillance in striped catfish farming.
(Capello et al., 2013) Estimates are that 106 tonnes of antimicrobials were used to produce 1.14 million tonnes of striped catfish in 2011 (FAO, 2012; Rico et al., 2013) Of this amount, approximately 29 tonnes of active compounds were discharged into the rivers and canals of the Mekong Delta, representing risks to the aquatic environment (Rico and Van den Brink, 2014) In 2008-2012, studies showed an increase in the antimicrobial resistance of E ictaluri, the cause of BNP, and Aeromonas spp., the cause of MAS, isolated from striped catfish in the Mekong Delta, and demonstrated that they were resistant to nearly all the antimicrobials approved for use in
Quality of antimicrobial product and medicated feed
Analytical results of antimicrobial products used for striped catfish collected at chemical shops revealed that only 4/11 products declared as containing a single antimicrobial and 2/10 products with a mixture of antimicrobials contained active substances within ±10 % of the concentrations declared on the product labels, which is the level of variation accepted by the Vietnamese authorities (Manuscript II) (Table 3; MARD, 2003) Two products with antimicrobial mixtures did not contain any of the declared antimicrobials In the comparison of the two batches, an analysis of
11 products revealed that only one product contained a concentration of active compound that varied by less than 10 % in both batches Several product labels (5/21) stated that the antimicrobials could be used for prophylactic treatment, which will further add to the selection of antimicrobial resistance (Manuscript II)
Manuscript I shows that farmers prepare medicated feed by dissolving antimicrobial powders or solutions in pond water, scooping up the resulting mixture, spraying it onto feed pellets, and then mixing the medicated feed with their bare hands; this practice may hinder the uniform distribution of the antimicrobial within pellets It also points to a lack of more water-stable, commercially manufactured medicated feeds, unlike those available to American channel catfish farmers (FDA, ).
Research indicates that variations in how antimicrobials are applied on farms inevitably produce large disparities in the actual dose delivered Analyses of enrofloxacin (ENR)-medicated feed prepared by farmers reveal substantial differences in ENR concentration (Manuscript III) In striped catfish farming, several antimicrobials are used, but inconsistent preparation and dosing across farms highlight the need for standardized medicated-feed protocols and quality control to ensure consistent therapeutic exposure.
55 farmers are only slightly soluble in water, e.g trimethoprim and sulfamethoxazole are soluble at levels of 0.4 and 0.61 mg/mL respectively (PubChem, 2013) Furthermore, the mixing of medicated feed by small-scale grow-out and nursery farmers using their bare hands poses a high risk to the farmers’ health (discussed in Section 4.3)
Table 3 Active antimicrobial compound concentration compared with the declared concentration on the product label (%) in products used for striped catfish
Antimicrobial type and concentration declared on product label a
SMX SDZ TMP DOX FFC AMX LEX CIP
IV SDZ 33 % + TMP 6.7 % - ND-ND c 47.7-32.4 - - - - -
VI FFC 20 % + AMX 5 % - - - - ND ND e - -
VII AMX 12 % + CIP 10 % - - - - - ND-ND - ND-ND
Eight veterinary antibiotics were analyzed: SMX (sulfamethoxazole), SDZ (sulfadiazine), TMP (trimethoprim), DOX (doxycycline), FFC (florfenicol), AMX (amoxicillin), LEX (cephalexin), and CIP (ciprofloxacin) The two percentage figures shown correspond to contents found in two different batches In one case, the declared SDZ was not detected (ND); instead, the product contained sulfamazine at 11.0% and 10.5% in the two batches In another case, 10.1% AMX was found instead of LEX, and in a separate instance, 11.3% LEX was found instead of AMX Some compounds were not analysed.
Farmers reported the calculation of the therapeutic dosage to be based on fish biomass (5/21 products), the amount of feed used (3/21 products) or both means of calculation (13/21 products) (Manuscript II) The correct dosage should be calculated based on the antimicrobial concentration in the prepared medicated feed, together with the amount of feed consumed by a particular fish biomass Recommended doses on labels varied greatly between products, e.g for doxycycline the dose was 2.50 to 40.0 mg/kg fish when calculated based on fish biomass, and 75.0 to 667 mg/kg feed when calculated based on amount of feed used (Manuscript II) The variation in antimicrobial concentration declared on different product labels made it difficult for farmers to prepare medicated feed with the correct antimicrobial concentration Only the dose for florfenicol (5 and 10 mg/kg body weight) to treat E ictaluri infections was in agreement with the recommended dosage for the treatment of channel catfish (Gaunt et al., 2004; FDA, 2011) Labels on 18/21 products provided information about withdrawal time (15 to 30 days), despite the limited availability of information about the pharmacokinetics of different antimicrobials used in striped catfish culture Consequently the correct therapeutic dosage and withdrawal time have not been established for most approved antimicrobials, e.g florfenicol, sulfonamides and tetracyclines The withdrawal times of enrofloxacin, sulfamethoxazole and trimethoprim in striped catfish are given in Section 4.4
The Vietnamese authorities (MARD, 2010b) have registered and approved all the antimicrobial products analysed The total number of registered products for use in Vietnamese aquaculture was
In 2012, a total of 2,913 products were approved, including 813 so-called veterinary drugs (Tai, 2012) The sheer volume of approvals demands substantial resources and entails significant cost It remains unclear whether the observed inferior product quality results from inadequate testing at the approval stage, or more plausibly from post-approval changes to antimicrobial composition and concentration.
Altogether, these findings show that it is most likely that diseased striped catfish in the Mekong Delta receive a sub-therapeutic antimicrobial dosage, with the subsequent high risk of treatment failure and antimicrobial resistance development Increasing levels of antimicrobial resistance among E ictaluri and A hydrophila isolated from striped catfish in the Mekong Delta have been identified in studies conducted between 2008-2012, with both pathogens being resistant to nearly all the antimicrobials approved for use in Vietnamese aquaculture (Dung et al., 2008b; Huong et al., 2011; Phuc et al., 2012) One feed mill did begin manufacturing medicated feed (Viet Thang Feed Company, personal communication), however production was terminated since farmers apparently
57 preferred to prepare their own medicated feed There is an urgent need to investigate the manufacturing of good quality commercial medicated feed and promote its use by farmers
Introducing the use of commercial medicated feed in aquaculture standards could be a way of encouraging the use of medicated feed.
Farmer health risks associated with chemical use
The farmers in this study had frequent and direct exposure to various chemicals, including antimicrobials, disinfectants and anti-parasitic compounds There was a particularly high risk of exposure to hands and arms due to the common practice of farmers preparing and mixing antimicrobial solutions with pelleted feeds using their bare hands and during the preparation and application of disinfectant solutions to ponds (see Section 4.2) A few farmers used burnt lime and hydrated lime, which can cause blindness and severe irritation following eye or skin contact (Boyd and Massaut, 1999) Instead most farmers used agricultural limestone, which is safer to handle and apply As farmers did not use masks, there is also a real risk of inhaling antimicrobials and disinfectants when handling these compounds Documented health risks associated with exposure to antimicrobials and disinfectants include skin allergies, organ-specific reactions (e.g blood dyscrasias, liver and kidney problems), and systemic reactions (e.g anaphylaxis, drug induced hypersensitivity syndrome) or combinations of these Some of the antimicrobials commonly used by farmers, such as ampicillin, cotrimoxazole and quinolones, are also among the most common causes of antimicrobial allergies (Thong, 2010) Despite such documented risks, relatively few farmers, mainly large-scale farmers, perceived these exposures to represent a risk to their personal health Only half of the small-scale farmers used protective measures, including gloves and dust masks, whereas such measures were available on all large-scale farms
GlobalGAP standards require farmers to keep written records of all chemicals used, store chemicals safely, wear protective gear when handling them, and receive training on safety measures As certification becomes more widespread and standards tighten, these measures are likely to reduce exposure to toxic chemicals and improve health outcomes for farmers and their families However, because small farms still predominate in the sector, these findings point to an occupational health issue that merits further investigation, for example through expanded field studies and monitoring.
58 involvement of health specialists There is an urgent need to focus training on safe chemical handling procedures for such farmers
Table 4 Reported use of chemicals and perceptions of occupational health hazards by farmers
Grow-out Nursery Hatchery Small-scale Large-scale
Chemicals administered according to: safety instructions on product packaging 13 9 10 3 instructions by veterinarian/technician 1 7 7 2 instruction by extension staff - - - - farmer experience 11 7 22 12 information from friends/others 1 - 2 1
Record keeping of chemical use 6 10 9 1
Direct contact between skin and chemicals 21 0 18 5
Direct contact between skin and water containing chemicals
Use of protection during handling of chemicals 11 10 17 4
Farmers/workers are instructed on how to handle chemicals safely
Farmers/managers are informed about the health and environmental risks associated with chemical use
Farmers/managers are informed about banned chemicals
Common clinical manifestations following use of chemicals (skin lesions, coughing)
16 a 0 8 b 1 a : 13/16 farmers reported skin lesions and 3/16 farmers reported coughing problems b : 7/8 farmers reported skin lesions and 1/8 farmers reported coughing problems
Antimicrobial withdrawal time
4.4.1 Elimination of enrofloxacin in striped catfish following on-farm treatment
On-farm experiments were conducted with striped catfish on three different farms to determine the accumulation and elimination of enrofloxacin residue in striped catfish (Fig 8) (Manuscript III) The results on enrofloxacin (ENR) residue in feed pellets (n=2) used on the first day showed variations in ENR concentration, i.e 106±10.1 mg/kg in pond 1, 49.9±13.3 mg/kg in pond 2 and 39.8±8.6 mg/kg in pond 3 (Manuscript III) This variation in ENR concentration was due to the farmers mixing the feed with their hands, making it difficult to obtain a uniform distribution of the antimicrobial (see Section 4.2) Although the ENR concentration in feed used during the remaining
59 treatment period was not analysed, it is likely that differences in ENR concentrations in the feed accounts for some of the variations found in ENR concentrations in muscle/skin samples (Fig 8) Moreover, the amount of feed used by striped catfish farmers for preparing medicated feed seemed to vary between farmers, which also would lead to a variation of ENR concentrations in medicated feed It should also be noted that farmers paid little attention to the concentration of the active antimicrobial compound in the prepared medicated feed since they calculated the antimicrobial dosage based on fish biomass and not per volume of feed applied Again, if commercial medicated feed were available, this would reduce the variation in antimicrobial concentrations in medicated feed
Figure 8 Accumulation and elimination of enrofloxacin (ENR) and ciprofloxacin (CIP) in muscle/skin samples of striped catfish during five days of treatment of three ponds (P1, P2 and P3) with feed containing ENR ENR and CIP residue levels (àg/kg) are shown on the day before feeding with medicated feed (D0) and during treatment (D1, D3 and D5) (a) ENR and CIP concentrations after termination of medication (D7, D15, D30 and D45) are shown together with the EU-MRL value (100 àg/kg) and US-FDA action level (5 àg/kg) (b)
ENR is oxidatively de-ethylated into CIP by cytochrome P450 enzymes in many animal species, including fish and crustaceans (Vaccaro et al., 2003; Tang et al., 2006) Before starting the experiment (D0), muscle/skin samples of the stocked juvenile fish from ponds 1, 2 and 3 contained
Residue data show ENR and CIP levels in nurseries supplying striped catfish juveniles, with ENR residues at 171, 11.5 and 102 ag/kg and CIP residues at 5.4, 8.9 and 18.7 ag/kg (Fig 8a) This suggests ENR was used in the nurseries that produced the juveniles In striped catfish nurseries, bacterial diseases commonly occur and the use of antimicrobials seems to be inevitable (Bui et al., 2010) Survey data on chemical use indicate ENR was the most widely used antimicrobial among nursery farmers (Table 2, Manuscript I) However, in this study it was not possible to trace back the specific antimicrobials used in the nurseries that delivered the juveniles, or their exact concentrations.
60 initial efforts to obtain juveniles that had not been treated with antimicrobials were unsuccessful However, ENR was not applied in the period between stockings in the grow-out ponds until the beginning of the experiment
ENR and ciprofloxacin (CIP) accumulation and elimination in striped catfish mixed muscle/skin samples are shown in Figure 8 (Manuscript III) After one day of feeding with ENR-medicated pellets, fish muscle/skin samples from ponds 1, 2 and 3 contained 11042, 1817 and 1063 àg/kg of ENR and 642, 156 and 73 àg/kg of CIP respectively The variation in ENR and CIP levels in fish flesh was mainly due to the differences of ENR quantity in the medicated feed used in the three ponds The CIP levels in the flesh of fish from ponds 1, 2 and 3 accounted for 5.4±0.02 %,
13.1±0.14 % and 7.7±0.05 % respectively of the sum of ENR and CIP residues (Fig 8) These findings were consistent with the results of Danyi et al (2010) that found that 3-4 % of ENR was metabolised into CIP after three days of medication in aquarium experiments at 29-32 °C, which is similar to the temperature measured in the ponds in this study (29-30 °C)
The levels of ENR and CIP residues found in muscle/skin samples varied greatly in fish from the three ponds on days 15 and 30 This is probably due to the different concentrations of ENR in the medicated feed provided After 15 days (D15, Fig 8b), the sum of ENR and CIP residue levels in muscle/skin samples from pond 1 and pond 2 was found to be below the maximum residue limit (MRL) (100 àg/kg) assigned by the European Commission (EC, 2010) for the sum of ENR and CIP residues, while after 30 days (D30, Fig 8b), only the fish from pond 2 contained residue levels below the EU-MRL In pond 1, the ENR concentration on days 15 and 30 was close to 100 àg/kg (EU-MRL) It seemed higher on day 30 than on day 15, but this difference could be caused by both experimental error and the different amount of medicated feed consumed by the individual fish This suggests that the elimination of ENR in the fish is not linear, e.g initially there is a drastic decrease between day 7 and day 15 after the medication was terminated, while between day 15 and day 45 after medication, the ENR residue levels decreased much more slowly, particularly in ponds
In tank conditions, striped catfish fed enrofloxacin (ENR) at 1 g/kg feed showed total ENR and ciprofloxacin (CIP) residues of 100–200 µg/kg seven days after the medication was terminated (Danyi et al., 2010) However, contrary to what Danyi et al (2010) found, the present study detected ENR and CIP residues beyond that time frame, up to the end of the observation period.
45 days after the termination of medication Based on a pond water temperature of 30 °C, a withdrawal time for ENR and CIP of 45 days would be required to meet the EU-MRL For
61 comparison, a 62-day withdrawal period is needed for rainbow trout treated with ENR at a dose of
10 mg/kg of body weight per day for five consecutive days in field conditions (12-13 °C) (Luccheti et al., 2004) The fact that the suggested withdrawal time of 45 days for striped catfish treated with ENR is lower than the one for treated rainbow trout (62 days) seems to be due to the higher temperature (29-31 °C), since in other aquatic species the drug elimination rates appear to be faster when water temperatures are higher (Luccheti et al., 2004; Xu et al., 2006) It should be noted that the suggested withdrawal time of 45 days for striped catfish treated with ENR to meet the EU-MRL is longer than the 30 days’ withdrawal time suggested by ENR producers In black tiger shrimp, the elimination of ENR was faster, with 10.2 à g ENR+CIP/kg in shrimp muscle 28 days after feeding with ENR-medicated feed was terminated (4 g ENR 98 %/kg feed) (Tu et al., 2006)
Forty-five days after medicated feed was terminated, the total ENR and CIP concentrations in all mixed muscle/skin samples from the three ponds were below the EU-MRL of 100 µg ENR+CIP/kg product, but above the action levels defined by the US FDA (5 µg/kg) and the Japanese authorities (10 µg/kg), as reported in Manuscript III (Love et al.).
Regulatory frameworks for ENR residues differ across regions: the European Commission sets maximum residue limits (MRLs) based on acceptable daily intake (ADI), while the US FDA and Japanese authorities enforce zero tolerance for ENR residues in aquaculture products As the European Union, Japan, and the United States are major importers of striped catfish and shrimp, these regulatory disparities created challenges for Vietnam’s export-oriented aquaculture sector, leading Vietnamese authorities to ban ENR use in aquaculture in 2012 (MARD, 2012a) Nonetheless, follow-up monitoring showed ENR residues in striped catfish farms during 2013 and 2014, indicating that some farmers apparently continued to use ENR in aquaculture (NAFIQUAD, 2014).
The total ENR and CIP concentration in skin samples at the time of harvest was 16.4±9.9 àg/kg, which may represent a food safety issue in Vietnam if skin and products including skin are used for human consumption In Vietnam, consumers will normally consume the skin part when whole fish are served, which is in contrast to consumers who are served fillets, e.g in EU countries Stoffregen et al (1997) conclude that ENR is eliminated more slowly in the skin of Atlantic salmon as compared to muscle tissue, but no explanation for this is given The present study’s findings of ENR residues in striped catfish skin more than 200 days after medication warrant further study of the pharmacokinetics of ENR in striped catfish skin
4.4.2 Withdrawal time for sulfamethoxazole and trimethoprim following treatment of striped catfish and hybrid red tilapia
4.4.2.1 Method validation for sulfamethoxazole and trimethoprim analysis
The analytical method developed to determine sulfonamide in bovine liver (Zhao and Stevens,
2012) was validated in this study for striped catfish muscle (reference matrix) (Manuscript IV) Slope ratios between matrix-matched and solvent calibration ranged from 0.8 to 1.2 (Lopes et al., 2012b), corresponding to a low matrix effect for trimethoprim (TMP) (ratio =1.52) and no matrix effect for sulfamethoxazole (SMX) (ratio =1.18) in this study, indicating that the QuEChERS clean- up procedure offers a compromise in terms of matrix effect (Stubbings and Bigwood, 2009; Lopes et al., 2012b; Zhao and Stevens, 2012) The high coefficient of determination (R 2 above 0.99), calculated from both matrix and solvent TMP and SMX calibration curves, indicates a linear relationship between the responses and the targeted spiking levels The very low coefficient of variation for repeatability and reproducibility (not exceeding 5 %) indicates a good performance of the analytical method for both TMP and SMX when compared to the Commission Decision