News release
From:
Use of controversial weedkiller inadvertently selects for drug-resistant bacteria that can spread to hospitals
Agricultural soils exposed to glyphosate may be unexpected breeding ground for hospital ‘superbugs’
In October 2025, the WHO again sounded the alarm on the emergence of multidrug-resistant bacteria in hospitals around the world. Researchers have now found evidence that the use of weedkillers, in particular glyphosate, can drive the evolution of antimicrobial resistance in soil bacteria as a side-effect of developing resistance to the weedkiller itself. They hypothesized that resistant bacteria can be transmitted between hospitals and impacted soils in both directions through wastewater and other environmental pathways.
Main text: Each year, antimicrobial resistance (AMR) is responsible for an estimated 1.1 to 1.4 million deaths worldwide. Now, scientists have found evidence that the spread of AMR isn’t always driven by bacteria evolving to resist the antibiotics themselves: rather, certain weedkillers can have the same effect.
“Here we show that the most common species of multidrug-resistant bacteria from hospitals are not only resistant to multiple antibiotic classes, but also to high concentrations of the weedkiller glyphosate,” said Dr Daniela Centrón, a researcher at the Institute of Medical Microbiology and Parasitology in Buenos Aires and the senior author of the study in Frontiers in Microbiology.
“These results suggest that weedkillers – which, unlike antibiotics, are widely applied in agricultural environments – may have the unintended side-effect of selecting for AMR among bacterial communities within the soil.”
Resistance is not futile
In 2018 and 2020, Centrón and colleagues had collected 68 bacterial strains from sediments in a nature reserve in the Paraná delta, a wetland of international importance located north of Buenos Aires. Glyphosate is frequently applied to nearby agricultural areas.
The scientists here tested each strain’s degree of resistance to 16 common antibiotics, such as ampicillin combined with sulbactam, meropenem, tetracycline, and vancomycin. They also measured the strains’ resistance to pure glyphosate and glyphosate-based herbicides – chosen because they are among the most frequently used herbicides around the world.
The scientists compared the results with those from 19 strains, including multidrug-resistant species, sampled from local hospitals. Another 15 strains had been isolated from feedlots and herbicide-impacted agricultural soils in the region.
As expected, the hospital strains were each resistant to between 1 and 16 of the antibiotics tested, confirming widespread AMR. Worryingly, 74% were resistant to carbapenems, broad-spectrum antibiotics commonly used as a treatment of last resort. Importantly, all hospital strains also proved highly resistant to glyphosate and glyphosate-based weedkillers.
“This means that if these bacteria enter the environment through untreated wastewater from hospitals, they could go on to thrive in agricultural areas where glyphosate is used,” said first author Dr Camila Knecht from Dr Centrón’s group.
Strains from the Paraná delta spanned 15 genera, including Acinetobacter, Pseudomonas, Exiguobacterium, and Chryseobacterium. Each had at least partial resistance to glyphosate and glyphosate-based weedkillers, even though these have never been used in the reserve itself. Enterobacter strains tolerated the highest concentrations of glyphosate, up to 80 milligram per milliliter. At the other extreme, Bacillus strains, usually found in soils, were particularly susceptible: their growth was already inhibited at a concentration of 2.5 milligram of glyphosate per milliliterAnd high resistance to glyphosate was also found in strains isolated from hospital infections with extreme drug resistance.
All in the family
When the scientists made a ‘family tree’ of all 102 bacterial strains, those most resistant against glyphosate tended to be close relatives, irrespective of their location of origin. For example, the same genera were found to be resistant against glyphosate across hospitals, agricultural areas, and the Paraná delta.
“In the environment, the use of glyphosate leads to the evolution of resistant bacteria in impacted soils, whereas the use of antibiotics favors their evolution in hospitals. Bacteria carrying antibiotic resistance genes can spread and breed between those two niches in both directions and in multiple ways, with the water cycle playing a key role in transmission,” concluded coauthor Dr Jochen A Müller, a group leader at Karlsruhe Institute of Technology.
The use of glyphosate is not without controversy: it is known to harm arthropods (in particular bees), while the International Agency for Research on Cancer has classified it as a probable human carcinogen. For this reason, France, Belgium, and the Netherlands have banned glyphosate for household use, while Germany currently prohibits its use in public spaces.
“Policies for the use of any pesticide, as well as its metabolites, should stipulate the requirement for co-selection testing with antibiotics before marketing. Labels should include a warming that genes for antibiotic resistance can spread from glyphosate-contaminated soils to hospitals through untreated water,” counseled Centrón.
Expert Reaction
These comments have been collated by the Science Media Centre to provide a variety of expert perspectives on this issue. Feel free to use these quotes in your stories. Views expressed are the personal opinions of the experts named. They do not represent the views of the SMC or any other organisation unless specifically stated.
Dr Mahima Krishnan is a Postdoctoral Research Fellow with the Weed Management Initiative at Adelaide University
"This study reports that environmental glyphosate exposure alters microbial communities in ways that favour species associated with multidrug resistance in human health settings. While the association is plausible, the analysis fails to account for other key contributing factors, particularly the inherent resilience of these taxa to multiple environmental stressors. For example, the authors have not considered how variation in nutrient availability influences the microbial community structure. Likewise, environmental perturbations such as flooding and drying events, known to significantly alter microbial communities, have not been addressed. Notably, the reported community shifts have not been observed in the study region, and the glyphosate application rates used substantially exceed those typically used in the fields.
Several key influencers of microbial community composition remain unaddressed, including the influence of livestock grazing systems and wastewater contamination. While the presented risk map highlights regions of genetically modified soy production, it does not account for recent shifts in herbicide use patterns caused by the increasing incidence of glyphosate-resistant weeds. Additionally, these same regions coincide with intensive animal agriculture, a well-documented reservoir of antibiotic-resistant bacteria, which has not been considered in the analysis.
Multi-drug resistance is a complex, global health issue driven by multiple interacting factors, including extensive antibiotic use in livestock, existing environmental resistomes, and anthropogenic contaminants. The evidence provided here is insufficient to support glyphosate as a primary causal driver."
Professor Ian Rae is an expert on chemicals in the environment from the School of Chemistry at the University of Melbourne. He was also an advisor to the United Nations Environment Programme on chemicals in the environment and is former President of the Royal Australian Chemical Institute
"The impacts on human health of the herbicide glyphosate have been a matter of contention for some years. As a result, some jurisdictions have banned or restricted its use, but many have continued to sanction its use because of the benefits it brings to agriculture and weed control in general. This acceptance of glyphosate rests on the fact that, as a herbicide, it affects growth pathways that exist in plants but not in humans, and can therefore be regarded as safe.
This new study adds substantial detail to a concern that has been growing slowly: what impact does glyphosphate on other species, notably bacteria? Because bacteria multiply so quickly, any impact (chemical or otherwise) on them can rapidly skew the population. Knocking off the weaker members of the microbial population, for example, opens the way to the more robust members of the microbial population to become dominant in the second and subsequent generations. In the case being addressed by this new research, it is pathological organisms that are targeted, and members of the 'more robust' population are multi-drug resistant and therefore of great concern to human health. The 'type case' for this is the precautionary use of low concentrations of an anti-bacterial that doesn't wipe out the whole bacterial population, but skews the population towards more robust and possibly more dangerous members.
There had been suspicions that glyphosate might be acting in this way to drive bacterial populations and make them more dangerous to human health. This new work confirms that possibility and explores the biochemical mechanisms by which glyphosate exerts its affect.
In the environment, glyphosate binds to soil but is only slowly destroyed there, so it stays around for a while and can affect other species than the weeds it was 'aimed' at. The present work concerned microbial populations in wetlands where glyphosate had been used for weed control, and you might wonder if it was really necessary there."
Professor David Ackerley is the Director of Te Matapihipihi – The Centre for Biodiscovery at Victoria University of Wellington
"The authors of this paper have done some comprehensive work characterising over 100 isolated bacterial strains for their resistance to both antibiotics and the herbicide glyphosate. However, their framing of the work is a bit too sensationalist for my liking.
"One Health is an important concept that views the health of humans, animals, plants and ecosystems as being interconnected and seeks sustainable solutions to optimise health across this network. A key focus is on the use of antibiotics in inappropriate settings, but if other human activities such as glyphosate spraying promote antimicrobial resistance (AMR), then we definitely want to know about it.
"Given the title of the work (“Glyphosate resistance as a potential driver for the dissemination of multidrug-resistant clinical strains”), I was surprised that the authors did not seek to compare bacteria from equivalent glyphosate-exposed versus glyphosate-free environments. Instead, they isolated 68 bacterial strains from a herbicide-free wetland, and assessed them alongside 35 previously-isolated strains that included a mix of drug-resistant hospital isolates and bacteria from herbicide-impacted soils.
"A central finding was rather at odds with the work’s title: “Regarding AMR phenotypes (Table S2), no pattern was observed that correlated this trait with glyphosate resistance in all strains”. That is to say, despite suggesting glyphosate resistance may drive antimicrobial resistance, the authors actually saw no correlation between glyphosate resistance and antibiotic resistance.
"The authors did find that most of their hospital-isolate bacteria were strongly resistant to glyphosate, and quite reasonably suggest that the same mechanisms that promote drug resistance can also cause glyphosate resistance. A more accurate title of their paper might therefore have been that “Antibiotic resistance is a potential driver of bacterial glyphosate resistance”, rather than pitching things the other way around. However, that would be less ‘One Health’ relevant and likely not garner quite the same headlines."
Dr Sohinee Sarkar is a Senior Researcher (Infection Immunity & Global Health) at the Murdoch Children's Research Institute and an Honorary Senior Fellow at The University of Melbourne
"Antibiotic resistance is rapidly becoming one of the most significant global health threats. Although healthcare settings receive most of the attention, resistance emerges through multiple pathways, including the extensive use of antibiotics in agriculture and food production. Previous research shows that when glyphosate and antibiotics are present together, they can increase the activity of bacterial efflux pumps that expel these compounds from the cell. This not only promotes tolerance to glyphosate but also raises the likelihood of mutations that confer antibiotic resistance. Given glyphosate’s widespread agricultural use, its contribution to this problem is especially concerning.
Growing evidence indicates that antibiotic resistance genes found in clinical settings are linked to those present in environmental bacteria, including soil organisms. The current study builds on this by identifying shared resistance signatures between clinical and environmental bacteria in a region of Argentina. Notably, highly resistant clinical isolates also displayed strong herbicide resistance, and were closely linked to environmental bacteria in protected natural areas where herbicides are not applied. These findings highlight the interconnected movement of resistance factors across natural ecosystems and human populations. While the exact pathways remain unclear, the study reinforces the need for a holistic One Health approach that coordinates human, animal, and environmental strategies to address rising antimicrobial resistance."
Prof Mark Blaskovich is Director of Translation at the Institute for Molecular Bioscience at The University of Queensland, Director of the ARC Training Centre for Environmental and Agricultural Solutions to Antimicrobial Resistance, and co-founder of the Community for Open Antimicrobial Drug Discovery
"The study adds to the increasing body of evidence that bacteria exposed to a range of chemicals in different environments can develop increased resistance against being killed by antibiotics.
Traditionally, it was thought that antibiotic resistance was just driven by overexposure to antibiotics, both within humans and in the environment. However, in recent years there have been reports the things like the antiseptic triclosan (used in handwash) and other drugs used for non-infection therapy can lead to bacteria surviving concentrations of a range of antibiotics.
In the current study, the authors have found that the common weed killer glyphosate, extensively used in the environment, might have a similar effect, though in a subtly different manner. They tested bacteria isolated from soil samples in Argentina, and compared them to bacteria isolated from Argentinian patients. They found that the types of environmental bacteria that survived higher concentrations of glyphosate were similar to the pathogenic antibiotic-resistant bacteria that infect humans. Therefore, glyphosate treatment of crops may lead to the soil microbiome (the collection of different species of bacteria normally found in soil) being enriched in the types of bacteria that are more dangerous to humans.
The study did not establish a direct connection of transmission between the environmental bacteria and the resistant human bacteria, nor did it show that glyphosate causes bacteria to develop resistance, which are areas for future research."
Professor Naresh Singhal is Director of the Water Research Centre, University of Auckland
“Antibiotic-resistant bacteria pose one of the biggest threats to humanity. Despite over a decade of research, many basic questions remain unanswered. For example, could the herbicides we use for weed control also accelerate the spread of antibiotic-resistant bacteria?
"A new study from Argentina uses a combination of over 100 bacterial strains and genomic analysis to conclude that the answer is “yes”. The approach enables direct comparisons between clinical and environmental settings, yielding several troubling insights.
"Firstly, the study finds that clinical multidrug-resistant bacteria exhibit high tolerance to glyphosate, which is unusual given that glyphosate is not applied to humans.
"Secondly, the study reports that some environmental strains, e.g., Enterobacter species, and hospital pathogens show similar antibiotic resistance profiles. This has significant implications, as it indicates that antibiotic resistance in agricultural environments and clinical settings could form a continuum, something our resistance control policies have historically ignored.
"Lastly, glyphosate resistance did not primarily result from changes to the herbicide's target enzyme. Instead, it arose from efflux pumps and phosphonate-degrading pathways, the same mechanisms that drive antibiotic resistance. This sharing of mechanisms between non-antibiotics and antibiotic triggers suggests that the development of antibiotic resistance in soils, water, and humans could be deeply interconnected. Also, this implies that resistant environmental bacteria could inhabit the human body, leading to unforeseen infections.
"If we are to succeed in dealing with this calamitous threat, our policies for managing antibiotic resistance need to be urgently updated to acknowledge this continuum of effects."