At present, uninhibited and uninformed use of AMs is adding to the problem of AMR, which in turn fuels AM use. The challenge is to break this vicious circle. Below we discuss mechanisms that could be used to reduce AM use, based on the underlying factors identified in the previous section.
Biosecurity refers to any step that would prevent entry of pathogens into farms or hatcheries, thereby reducing the risk of disease outbreaks and consequent AM use (Bondad-Reantaso et al. 2012). On pond-based grow-out farms, simple biosecurity measures would include deterrents to keep disease vectors out (such as bird nets and scares, or barriers for crabs), drying of sediments, liming of ponds, and organic waste removal before re-stocking (Yanong 2013). Excessive organic waste build-up serves as a reservoir for bacteria and other microorganisms, while organic loading increases the biological oxygen demand (Yanong 2013). Equipment, such as seine nets, paddle wheels, and vehicles, can also harbor and carry infectious disease between ponds and farms, a risk that can be reduced by adequate cleaning, disinfection, and/or drying between uses (Yanong 2013). More elaborate physical biosecurity measures also involve mechanical aeration and water treatment (e.g. by use of pre- and probiotics) to reduce water exchange rates from local waterways, implying that the risk of transferring vectors and associated pathogens will be minimized. An extreme example of an increasingly used biosecurity system is RAS, where disease vectors can be almost completely excluded through water treatment and re-use. RAS systems, however, still carry the risk of disease exposure through the introduction of new animals and feeds (Martins et al. 2010).
Implementing biosecurity programs at hatcheries can help reduce the incidence of disease. Screening post-larvae coming out of SPF hatcheries for viral pathogens, using rapid testing and screening tools (e.g. PCR), is a preferred option that can ensure absence of pathogens. For example, highly biosecure hatcheries produce SPF Asian tiger shrimp post-larvae that can reduce farm exposure to disease (Bondad-Reantaso et al. 2012).
At a regional level, coordinated planning may further help reduce overall disease prevalence if farms are sited sufficiently far apart to reduce disease transfer (Guerry et al. 2012). Regional management initiatives, including rapid response mechanisms, stronger regulations (e.g. maximum production and biomass loads, or organic waste management protocols), and comprehensive strategies for responsible introduction of live aquatic animals (e.g. ICES Code of Practice on the Introductions and Transfers of Marine Organisms) may also help reduce the risk of disease outbreaks and limit the spread of pathogens.
Aquaculture extension programs
Extension programs are aimed at helping farmers improve farming practices and are generally issued by governments and NGOs (Garrett et al. 1997). Many aquaculture extension programs have been launched to date, but their reach is limited in many lower income regions. In the meantime, much of the overuse of AMs is directly related to uninformed farmer decisions, including incorrect diagnoses, excessive dosage (sometimes due to illiteracy, miscalculations, or having experienced treatment failure at recommended dosages), use of poor quality feeds or drug products, overfeeding, insufficient infrastructure, and poor water management (Garrett et al. 1997). The promotion of probiotics and greenwater techniques, which boost the growth of health-promoting bacteria in cultured animals, has also proven to be an efficient mean of preventing AM use by reducing pathogenic bacteria and animals’ susceptibility to disease (Balcázar et al. 2006; Natrah et al. 2014; Bentzon-Tilia et al. 2016). It has been argued that AMs have negative long-term effects on pond microbial systems in general, by destroying healthy bacterial communities (Lavilla-Pitogo et al. 1998; De Schryver et al. 2014). De Schryver et al. (2014), for example, suggests that AM use (especially prophylactic use) may increase the prevalence of outbreaks of Vibrio spp. in shrimp cultures by destroying mature microbial systems of slow-growing bacteria in the pond water, thereby benefitting faster growing bacteria (incl. Vibrio spp.) that often are pathogenic.
Water temperature is also directly linked to the growth of pathogens, and, as most fish are ectotherms, they use temperature gradients to induce behavioral fever (Cerqueira et al. 2016). This behavioral response is, however, hampered in ponds and cages that restrict the fish’s movement. Thus, providing temperature gradients in the farm medium by helping farmers construct ponds with alternating depths and/or partial shading could help limit the severity of disease outbreaks (MacKenzie, pers com.). Climate change will further influence temperature and consequently the type, spread, and frequency of disease (Burge et al. 2014). This stresses the importance of improving the resilience of the aquaculture industry by maintaining suitable genetic, species, and farming diversity to match variability in the environmental conditions of the production area (Troell et al. 2014; Klinger et al. 2017).
Improved farm management can be achieved through extension programs that improve information flows. Extension agencies can help communicate the risks of misuse and overuse and demonstrate the efficacy of judicious use, best use practices, alternative treatment options, and disease avoidance techniques (Hernández Serrano 2005), all of which can limit AM use. While extension agencies are common throughout most higher income countries, similar programs are often absent or underfunded in lower income countries where most aquaculture currently takes place (Aker 2011). Top-down extension efforts can also be unsuccessful at engaging farmers and other supply chain actors if they do not allow for bidirectional communication and grassroots participation, further complicating extension in lower income countries (Umesh et al. 2010).
Improved farm support for diagnostics and treatment
The capacity to accurately diagnose disease is essential to effective treatment. Challenges associated with new species, established species in new environments, and new diseases may lead to delayed diagnosis and treatment or treatment without a proper diagnosis. Most high and many upper middle income countries have veterinarians, technicians, diagnostic labs, and research facilities available to provide fish health advice, but in lower middle income countries the lack of disease diagnostic capacity, including fish health experts, hinders rapid and proper diagnosis, and often leads to inappropriate use of AMs. For new or emerging diseases, however, diagnostics will remain obstructed until confirmatory diagnostic methods become available. These confirmatory diagnostics should preferably enable farmers to test their animals and determine a diagnosis, as has been achieved by, e.g. IE WSSV and YHV strip test (Wangman et al. 2016).
Increased frequencies of AMR may decrease the effectiveness of AM treatments, and, if farmers respond to this decrease in efficacy by using greater amounts of AMs, the resistance itself becomes a proximate driver of increased usage. The lack of AM alternatives thus easily results in AM abuse. Rotating the active ingredient has been proposed for reducing the likelihood of such situations, by avoiding selection and co-selection of AMR genes (Niederman 1997; Miranda et al. 2013). Clinical studies have, however, shown limited success for this strategy, as AMR genes can persist for long periods after the removal of the relevant selection pressure, which in this case would be any specific active ingredient (Taylor et al. 2011; Lee et al. 2013). Ecological theory instead suggests that mixing compounds actually yields better results than cycling (Bergstrom et al. 2004; Levin and Bonten 2004). Although AM mixing has shown great potential, random mixing without proper disease diagnosis and previous toxicological tests might result in inappropriate usage of AMs, thus suggesting that more clinical studies are needed (Lee et al. 2013). Nevertheless, excessive use or abuse of AMs will lead to an alteration of the resistome in target and non-target bacteria in environments surrounding farms and potentially spread to downstream farms. The problem of AM use and AMR, therefore, needs to be tackled internationally by adopting one common approach, to avoid the spread of AMR genes, including a coordinated ‘One Health’ approach with the livestock sector and human medicine practitioners (AVMA 2008; onehealthinitiative.com, accessed 29-Aug-2017).
Limiting AM access
Changes in access to AMs can be due to regulations that either permit or do not explicitly disapprove the use of some substances. AMs may also be inaccessible to farmers due to costs or lack of access to pharmaceutical markets or distribution networks. The most efficient ways of regulating access to AMs largely varies among countries.
Most upper middle and high income countries have lists of explicitly approved AMs. For example, in Canada there are four AM products registered for aquaculture, containing: oxytetracycline, florfenicol, trimethoprim/sulfadiazine, and ormetoprim/sulfadimethoxine (DFO 2017). These products are administered through medicated feeds and require veterinary prescriptions. Other AMs may be obtained through an ‘Emergency Drug Release’ provided by veterinarians in special cases (Health Canada 2017), such as erythromycin use in broodstock. The importance of oxolinic acid and flumequine (quinolone AMs) in human medicine has led to a prohibition of their use for treating salmon in Canada and Scotland (Burridge et al. 2010). In the US, the US Food and Drug Administration regulates AM use in aquaculture with specific applications for specific species and only three approved ABs (florfenicol,oxytetracycline, and sulfamethoxine/ormethoprim) (USFDA 2017).
Other countries, from our experience, take an alternative approach and explicitly ban AM substance groups that are known or suspected to cause carcinogenic or mutagenic effects in consumers (i.e., nitrofurans, nitroimidazoles, malachite green, and chloramphenicol and its derivatives), while use of non-banned AMs is tacitly allowed. Liu et al. (2017), for example, describe a Chinese ban on erythromycin in 2002, but also report continued use in 2012 based on literature sources. In the same review, Liu et al. (2017) also document 20 different ABs being used in Chinese aquaculture, while only 13 ABs were authorized. Thailand, on the other hand, only approves the use of five AM substances (enrofloxacin, oxytetracycline, sulfamethoxine/ormethoprim, and amoxicillin) (Baoprasertkul and Somsiri 2012), while Vietnam approves 27 different active ingredients, including substances used in human medicine (VMARD 2012). Rico et al. (2013) also reported 17 different AB compounds being applied in pangasius aquaculture, belonging to ten different AB classes, some which also are of critical relevance for human medicine (e.g. Kanamycin) (WHO 2011). Three years later, Ali et al. (2016) identified seven different ABs being used within Bangladeshi aquaculture (oxytetracycline, chlortetracycline, amoxicillin trihydrate, sulfadiazine, sulfamethoxazole, trimethoprim, and doxycycline).
Permitting a limited number of AMs for use in aquaculture seems to be the better approach, as it is easier to regulate and track the use of a few compounds. However, such restrictions are easier to implement in countries that only produce a handful of species and with a strong consolidated industry. Further, this type of regulation does not necessarily restrict overuse of those AMs that remain allowed. Historically, the approved and banned AM lists available in low- and lower middle income countries have been based on food safety hazards and national or international export quality standards, while most high-income nations also consider potential risks to the environment and their efficacy to kill target fish pathogens as key criteria for their acceptance and registration of AMs.
The inconsistent and poor quality of the AM products available to the aquaculture industry in certain regions (e.g. Phu et al. 2015) presents a serious concern. In these regions, AMR could occur because the product label does not accurately describe the product. The global extent of this problem is unknown, but lack of quality control could have serious implication on animal and human health.
Vaccines can efficiently prevent bacterial disease outbreaks in finfish, but they do not work in the same way for crustaceans or mollusks, as they do not have an adaptive immune system (Du Pasquier 2001). A prime example of the success of vaccines is salmon farming in Norway (Fig. 2), which over the past four decades managed to develop effective vaccines for most important bacterial diseases. While infectious diseases still cause mortality in the region, these are primarily caused by viruses [e.g. Pancreas Disease (PD), Heart and Skeletal Muscle Inflammation (HSMI), and Cardiomyopathy Syndrome (CMS)], against which AMs are not effective (Bondad-Reantaso et al. 2012). Vaccines have also been successfully implemented to treat Grass Carp Hemorrhagic Virus (GCHV) in China, where AM use is widespread throughout the Southern parts of the country (Mi et al. 2013). In other aquaculture sectors, vaccines have been less successful, as the cost of development and administration remains high (Secombes 2008). For example, a number of vaccines have been designed and commercialized against Piscirickettsia salmonis in Chile with low to moderate efficacy (Marshall and Tobar 2014), and Vietnamese farmers have shown overall skepticism to pangasius vaccines due to high costs, extensive labor efforts to inject individual fish, and limited survival improvements (Phu et al. 2016).
In conclusion, the success of vaccines in Norway was due to predominantly bacterial pathogens, high vaccine efficacy, and sufficient resources. For GCHV in China, vaccines have also been a success, with a cost–benefit ratio of 1:7 (Mi et al. 2013). While GCHV is a viral pathogen, reduced symptoms of disease and mortality rates will surely reduce the number of misdiagnoses and consequent AM use. The situations in Chile and Vietnam remain more difficult, suggesting that vaccines will not be a ‘silver bullet’ for all pathogens. The approaches towards designing new vaccines are, however, constantly developing, with the potential for considerably cheaper vaccines with higher efficacy in the future (Secombes 2008).
Regulations are typically applied to either direct use of AMs or the level of AMs in products. Direct use regulations include mandates on how specific AMs can be applied and under which circumstances. Regulation of AM use in northern European salmon farms has subsequently contributed to lower usage relative to other countries with less stringent regulations (Burridge et al. 2010).
In our opinion, product-based food safety regulations have been among the most effective drivers for reducing AM use in aquaculture to date. Product-based regulations are generally applied in the form of maximum residue limits, where samples of seafood products are screened for detectable levels of banned compounds or high concentrations of regulated compounds (Costello et al. 2001). In addition to regulations governing access to AMs (as discussed in Sect. “Limiting AM access”), national level regulations can also limit the amount of allowable AM residues in products meant for human consumption, thereby reducing overuse and misuse (Bondad-Reantaso et al. 2012). Screening consignments for AM residues is, however, resource intensive, meaning only a small sample of all products are often tested.
Over the past two decades, more stringent AM regulations can be associated with reduced use in European and North American aquaculture (given that Atlantic salmon makes up about half of production in these regions; FAO 2016a; Fig. 2) and possibly also reduced use in aquaculture production providing imports to these regions (Rico et al. 2013; Henriksson et al. 2015). However, import regulations do not address AM use throughout the production cycle, as farmers can limit AM concentrations in products by shifting the time of application or active substance. Instead, it would be more comprehensive to require farmers to register the quantities of AMs applied throughout the grow-out period, as is the case in Norway, Scotland, Chile, and some Canadian provinces (Burridge et al. 2010). This, however, requires a certain level of regulatory capacity to keep track of AM sales and use, and enforce accurate record keeping.
Product regulations have been shown to be effective at reducing at least late stage use of AMs in internationally traded products. For example, in response to regulation and monitoring of export-oriented products, contamination of Thai shrimp samples aimed for export dropped from 24% to 5% over a four-year period (Holmström et al. 2003). Product-based regulations can, however, incentivize different AM use practices among farmers dedicated to domestic or less regulated international markets. Similarly, differences in food safety standards between import countries (e.g. US vs. EU) can influence AM use practices in producing countries. Consequences of AM residues in seafood products also differ between importing countries, from consignments being held and thereby often lost, to bans on all animal imports from the country of origin (McCracken et al. 2013). In some cases, repeatedly stopped consignments have forced some countries to modify national regulations and AM use practices. For example, enrofloxacin has been banned in Vietnam since 2012 due to the repeated consignment rejections of pangasius catfish in the US (VMARD 2012), which has a zero tolerance limit.
Avoidance of food safety alerts have also resulted in better cooperation among aquaculture farmers, as processing plants and/or government officials now often screen consignments before shipping (caa.gov.in, accessed 02-May-2017; and pers. comm. processing plant managers in China and Indonesia). Thus, it is in all country-level producer’s interest to comply with regulations to enable product exports, especially to the EU and the US. However, these screenings and occasional consignments stopped by customs can become costly, rendering Europe or America less attractive as trading partners until regulations are enforced (Love et al. 2011; CBI 2013). For example, market prices for shrimp have been increasing in China with a doubling in consumption between 2005 and 2015,Footnote 1 incentivizing many processing plants to shift their target market from the EU to China (CBI 2013). This could potentially result in less restrictive use of AMs, a shift that might be further influenced by the US withdrawal from the Trans Pacific Partnership.
Rapid unregulated expansions of aquaculture have also paved the way for disease outbreaks, with subsequent AM abuse as a consequence. The Chilean salmon industry, for example, grew rapidly for nearly 30 years before being hit by an outbreak of Infectious Salmon Anemia (ISA) (Bustos-Gallardo 2013). This outbreak was the consequence of a pursuit of increases in production, with the government failing to take into account scientific advice on disease risks (Bustos-Gallardo 2013).
To comply with most types of certification, farmers need to limit and report on AM use. Existing organic certification standards have similar restrictions and all forbid prophylactic usage of AMs, but with some individual differences related to accepted practices and types of AMs (Table 1). The support provided on disease diagnostic and aquatic health management by certifying schemes can also influence the number and dosages used by farmers, as well as a number of issues related to record keeping, withdrawal periods, and food safety aspects.
Market driven aquaculture certification is becoming an increasingly powerful tool to enforce compliance of the industry to international standards (Jonell et al. 2013). Market access and pricing strategies are slowly attracting more and more aquaculture farmers to adopt and implement Better Management Practices (BMPs) and Aquaculture Improvement Projects (AIPs) in order to comply with national and international standards, and transition to third party certification (e.g. ASC and BAP) (Jonell et al. 2013). BMPs also often promote farm level biosecurity and sustainable intensification practices, reducing the overall risks for disease (e.g. bapcertification.org).
News media can also help change AM use and improve production practices by influencing public perception, consumer awareness, and aquaculture firms’ ‘social license to operate’ (Leith et al. 2014). For example, several articles in popular press outlets have highlighted excessive use of AM drugs in Chilean salmon farming and Vietnamese pangasius farming.Footnote 2 The coverage resulted in reduced purchases of Chilean salmon by consumers and decreases in the price of pangasius (Little et al. 2012). In this sense, the popular press can act as an enforcement mechanism for judicious use of AMs and general sustainability (Hosono et al. 2016), although it may also contribute to severe miscommunications (Murk et al. 2016).
Improved general public awareness would not only empower certification labels, but it would also help reduce unnecessary AM use among humans. Since human run-off is intermixed with aquaculture irrigation in many parts of the world, enlightening the public about responsible AM use would benefit people, animals, and the environment (Robinson et al. 2016).