A study of soil microbes showed that drought favors the microorganisms that survive antibiotics. It also found that some of the genes for resistance in soil-dwelling bacteria show up in antibiotic-resistant pathogen samples collected from hospital patients. Because bacteria can easily swap big chunks of genetic information — a process called horizontal gene transfer — any increase in resistance in soil-inhabiting microbes can easily make its way to microbes that infect humans, the study authors said.
“No place is immune,” said Dianne Newman, the study’s senior author and a biologist at Caltech. “If you have a pathogen arise in one part of the world, it very quickly spreads, so this is something of concern regardless of where you live.”
Resistant pathogens
Antibiotic resistance is already a major health problem, with the World Health Organization estimating that antibiotic-resistant pathogens directly caused 1.27 million deaths per year as of 2019 and contributed to another 4.95 million. While antibiotics kill microbes, the drugs people use in the clinic are also derived from microbes (or fungi, such as in the famous case of penicillin). Microbes synthesize antibiotics as part of an evolutionary arms with other microbes, aiming to kill any potential competitors or threats. One of the major battlegrounds for this evolutionary warfare is in soil.
Article continues below
Newman and the new study’s first author, Caltech postdoctoral researcher Xiaoyu Shan, first uncovered a hint that drought could worsen antibiotic resistance in a set of five metagenomics databases that gather soil microbe genetic information from different environments on continents around the world. Some of these databases included samples from the same sites before and after drought.
In every case, the researchers found, antibiotic synthesis genes were more prevalent after a dry period and less prevalent after a drought ended.
“You see this in croplands, in grasslands, in forests, in wetlands, in the U.S., in China, in Switzerland,” Newman told Live Science.
To find out what was going on, Newman, Shan and their colleagues took the question to the lab. They treated sterile soil with the antibiotic phenazine, which is produced by some species of bacteria. Then, they added soil-dwelling bacteria and allowed half of the samples to dry out for three days, while keeping the rest moist.
After this simulated drought, they discovered that the antibiotic in the soil had, unsurprisingly, become more concentrated as the moisture in the soil evaporated. They also found that, in response to this more concentrated antibiotic, bacteria in the soil that were sensitive to antibiotics suffered, while antibiotic-resistant bacteria flourished.
These findings illustrate that antibiotic resistance is driven by evolutionary pressure, Newman said. Only the toughest, most resistant survive when drought concentrates other microbes’ antibiotics to lethal levels.
This is not the time for governments to stop funding scientific research and drug discovery.
Dianne Newman, Caltech biologist
To get a glimpse of this evolutionary battlefield on a genetic level, the researchers returned to the large metagenomics databases. They found that genes for antibiotic resistance became more common in dry periods. This prevalence went hand in hand with the increase in genes for antibiotic synthesis, supporting the idea that drought-stricken microbes boost their antibiotic resistance in response to increased pressure from antibiotic assaults from their neighbors.
Next, the researchers took soil samples from the Caltech campus, added four different antibiotics, and dried out half of the samples. Again, they saw more antibiotic-resistant microbes in the desiccated samples.
A global peril
The next question was whether these patterns could be seen on a global scale. The researchers used existing data on antibiotic-resistant pathogens collected at hospitals around the world, as well as climate and weather data, to quantify the aridity at each hospital. They found that the dryer the region, the more antibiotic-resistant pathogens the hospital reported. The finding held even when the researchers controlled for a country’s socioeconomic status, which might influence pathogen testing.
A final genomic scavenger hunt provided one more piece of bad news: Many of the genes that confer antibiotic resistance in soil microbes were found, replicated exactly, in clinical pathogens known to escape antibiotics. These included the common hospital pathogens Enterococcus faecium, Klebsiella pneumoniae, Acinetobacter baumannii, Pseudomonas aeruginosa and species of Enterobacteria, the researchers reported March 23 in the journal Nature Microbiology. Human pathogens and soil microbes come into contact all the time as people move around the environment, Newman said, and drought-induced resistance can easily transfer from microbes in soil to microbes on the body.
“Continued warming and drying is expected to expand arid conditions,” Timothy Ghaly, a microbial ecologist at Macquarie University in Australia, wrote in an editorial accompanying the study. That means climate change could accelerate the already-serious problem of antibiotic-resistant pathogens, he wrote.
There are ways to wage our own evolutionary battle against the bacteria, Newman said. Beyond limiting climate change, more could be done to get rapid diagnostic tests into clinics so doctors can treat antibiotic-resistant bacteria faster. They can also choose multi-antibiotic treatments that knock out resistant strains. Another crucial step, Newman said, is to fund basic research in drug discovery. Pharmaceutical companies have largely pulled back from seeking out new antibiotics because of lack of profitability, which leaves government and academic scientists at the vanguard of basic research.
“This is not the time for governments to stop funding scientific research and drug discovery,” Newman said.
Shan, X., Cao, K., Jeckel, H., Alcalde, R. E., Trindade, I. B., Kwiecinski, J. V., & Newman, D. K. (2026). Drought drives elevated antibiotic resistance across soils. Nature Microbiology. https://doi.org/10.1038/s41564-026-02274-x