It is now accepted that increased antimicrobial resistance (AMR) in bacteria affecting humans and animals in recent decades is primarily influenced by an increase in usage of antimicrobials for a variety of purposes, including therapeutic and non-therapeutic uses in animal production. Antimicrobial resistance is an ancient and naturally occurring phenomenon in bacteria. But the use of antimicrobial drugs – in health care, agriculture or industrial settings – exerts a selection pressure which can favour the survival of resistant strains (or genes) over susceptible ones, leading to a relative increase in resistant bacteria within microbial communities. It has been observed that, in countries where use of particular substances (e.g. fluoroquinolones) is banned in animal production, there are low levels of resistance to these antimicrobials in livestock populations. The rate of AMR emergence in ecosystems such as the human or animal gut is likely to be highly dependent on the quantity of antimicrobials used, along with the duration and frequency of exposure. In animal production, the prolonged use of antimicrobial growth promoters (AGPs) at subtherapeutic levels in large groups of livestock is known to encourage resistance emergence, and is still common practice in many countries today. Due to the interdependence and interconnectedness of epidemiological pathways between humans, animals and the environment, determining the relative importance of factors influencing AMR emergence and spread in animal production is a significant challenge, and is likely to remain one for some time. In intensive livestock production systems, resistant bacteria can spread easily between animals and this can be exacerbated if biosecurity is inadequate. While some studies have shown reduced levels of AMR on organic farms, a high prevalence of multidrug-resistant (MDR) Campylobacter strains has been detected in organic pig farms in the United States even in the absence of antimicrobial usage (AMU). In aquaculture, AMR can develop in aquatic and fish gut bacteria as a result of antimicrobial therapy or contamination of the aquatic environment with human or animal waste. The extent and persistence of antimicrobial residues in aquatic systems is unknown and current evidence is conflicting. Furthermore, no international guidelines currently exist for maximum antimicrobial residue limits in water. Water is an important vehicle for the spread of both antimicrobial residues and resistance determinants, since contaminated water can be consumed directly by humans and livestock and used to irrigate crops. Food is likely to be quantitatively the most important potential transmission pathway from livestock to humans, although direct evidence linking AMR emergence in humans to food consumption is lacking. There is a theoretical risk of widespread dissemination of AMR due to the increasingly global nature of food trade and human travel. This would mean that strains of resistant bacteria could now very quickly reach parts of the world where they had previously not been present. Agricultural systems in emerging economies such as China and India have changed radically in recent years, becoming increasingly intensive in order to meet growing domestic and global demands for animal protein. This is likely to heighten the occurrence and spread of infectious diseases in these systems, thereby leading to increased AMU and therefore resistance. If the selection pressure resulting from AMU in animals and humans were to be removed, this would still not completely halt the emergence and global spread of AMR due to the ability of AMR genes to move between bacteria, hosts and environments, and the occurrence of spontaneous mutations.
However, the release of large quantities of antimicrobials or resistant bacteria into the environment is still thought to be an important point for control, and therefore measures which encourage the prudent use of antimicrobials are likely to be extremely useful in reducing the emergence and spread of AMR. Future development of quickly biodegradable antimicrobials could help to reduce environmental contamination, and pharmacodynamic studies in livestock can be used to inform the optimization of AMU. Improved hygiene and biosecurity should be a major focus for all types of animal production systems so that the risks of introducing pathogens and resistance genes – and the spread of these within animal populations – can be reduced. Detailed, specific recommendations for countries to move towards more prudent AMU in different agricultural settings are, however, beyond the scope of this paper. An improved understanding of the epidemiology of AMR emergence and spread in animal production will provide an essential foundation for successful mitigation strategies. There are still considerable gaps in our understanding of the complex mechanisms that lead to the emergence of AMR in bacteria, and the interactions that take place within microbial ecosystems enabling the transfer of resistance between bacteria. There are insufficient data at present to determine quantitatively how important the selection pressure of AMU is for the emergence of AMR in bacteria. Evidence regarding AMR transmission pathways between food animals and humans is lacking, especially from low- and middle-income countries (LMICs). Such pathways are likely to be highly complex and multi-directional, especially in LMICs, but are still largely unknown. There remains little doubt, however, that the most significant factor in AMR emergence in humans is AMU for human treatment and prevention. It is clear that both human and animal AMU can contribute to environmental contamination, although collection of meaningful data is challenging. The relationships between different types of farming systems and both AMU and the emergence and spread of AMR are discussed in this paper, including extensive and organic systems, but there is still a notable lack of knowledge on the role that sustainable agriculture systems can play in combatting AMR. Most importantly, future research needs to involve an interdisciplinary (e.g. One Health) approach, integrating agricultural, medical, environmental and social sciences, and especially recognizing the importance of human behaviour. A set of specific recommendations to fill current knowledge gaps is presented in the final section of this technical paper.