Environment & One Health 8 min read

The Other Factory: Why Everything from Drought to Weedkiller Is Breeding Superbugs

The Other Factory: Why Everything from Drought to Weedkiller Is Breeding Superbugs

In early 2026, a team of researchers published the results of the largest antibiotic resistance gene surveillance study ever conducted in American wastewater. They sampled 163 treatment plants across 40 states, tracking 11 resistance genes by digital droplet PCR. They found something that should change how we think about the entire crisis: rising resistance signals appeared in sewage weeks before hospitals reported the same patterns. The water knew before the doctors did.

The Sentinel in the Sewage

Kim et al. (Nature Communications, April 2026) built on the COVID-era revelation that wastewater carries population-level health intelligence. Their study tracked genes conferring resistance to beta-lactams (blaCMY, blaCTX-M, blaKPC, blaNDM, blaOXA-48, blaTEM, blaVIM), colistin (mcr-1), methicillin (mecA), tetracycline (tetW), and vancomycin (vanA) — the resistance genes that define the clinical crisis.

The geographic pattern was striking: the American South carried consistently higher resistance gene concentrations than the Midwest. The researchers correlated these patterns with antibiotic prescribing rates, social vulnerability indices, proximity to healthcare facilities, livestock density, and airport connectivity. Wastewater resistance didn't just mirror clinical reality — it anticipated it.

This isn't the only sentinel. In India, CSIR-CCMB researchers (Nature Communications, March 2026) profiled 447 wastewater samples from Delhi, Mumbai, Kolkata, and Chennai. They found that while microbial communities varied by city, resistance gene profiles were uniform across all four urban centers — the resistome has already homogenized across 1.4 billion people. More than half the microbial genomes they reconstructed were entirely novel.

In Sweden, NDM-5 was detected in wastewater for the first time in January 2026 — after more than a decade of surveillance. In Austria, NDM-5 E. coli persisted for years in hospital plumbing systems, requiring complete pipe replacement to eliminate. The resistance genes travel through water whether we watch for them or not. The question is whether we're building the systems to listen.

The Drivers Nobody Prescribed

Here is the finding that should reframe the entire AMR conversation: most of what's driving resistance in the environment has nothing to do with antibiotics.

This is not a single study's claim. It's the convergent conclusion of at least six independent research programs published between late 2025 and early 2026, each identifying a different non-antibiotic force that selects for, amplifies, or disseminates antibiotic resistance.

DroughtNature Microbiology, Mar 2026

Shan & Newman (Caltech) showed that drought concentrates natural antibiotics in soil, intensifying selection for resistant bacteria. The longer the drought, the more antibiotic-producing genes they found. The resistance genes in drought-stressed soil replicated exactly in ESKAPE clinical pathogens.

Scale: 116-country correlation between hospital AMR incidence and regional aridity. Projections: 5 billion people in drylands by 2100.
Heavy MetalsES&T, 2026

Tsinghua University researchers found that zinc at environmentally relevant concentrations exerts stronger selective pressure for antibiotic resistance than enrofloxacin itself in livestock wastewater systems. Copper showed similar co-selection effects. The mechanism: membrane permeability changes, enhanced conjugative transfer, and biofilm promotion.

Scale: Copper and zinc are routine livestock feed additives globally. Cu/Zn in manure may be equal or greater AMR drivers than antibiotic residues.
GlyphosateFrontiers in Microbiology, Mar 2026

Hospital MDR strains proved highly resistant to the world's most widely used herbicide, despite never encountering it. The mechanism is shared efflux pumps — the same cellular machinery that expels antibiotics also expels glyphosate. 74% of the hospital strains were carbapenem-resistant.

Scale: Glyphosate is the third chemical class (after heavy metals and industrial chemicals) shown to co-select for antibiotic resistance. ~8.6 billion kg applied globally since 1974.
Chemical PollutantsNature Microbiology, Jan 2026

Cambridge researchers systematically tested 1,076 non-antibiotic chemicals against gut bacteria. 168 (15.6%) had direct antibacterial activity. When bacteria lost efflux pump genes to survive these chemicals, they simultaneously acquired ciprofloxacin cross-resistance. Industrial chemicals in the environment are running a parallel selection experiment on the human microbiome.

Scale: These include common food additives, industrial chemicals, and consumer products at concentrations found in the human body.
MicroplasticsNature Communications, Dec 2025

Nanoplastics at just 0.1 mg/L — concentrations found routinely in surface water — enhance conjugative transfer of resistance genes 7–20 fold. The mechanism: oxidative stress, ATP production changes, and increased membrane permeability all accelerate plasmid transfer between bacteria. The "plastisphere" — the microbial community that colonizes plastic surfaces — is a resistance gene exchange hub.

Scale: Microplastics are now found in human blood, placenta, breast milk, and brain tissue. Global production: ~400 million tonnes/year.
Everyday PharmaceuticalsISME Journal / Microbiome, 2021–2022

Ibuprofen, naproxen, propranolol, and carbamazepine at environmentally relevant concentrations (0.5 mg/L) promote plasmid transfer across entire microbial communities — up to 4-fold increase at 0.005 µg/mL for naproxen. The mechanism: reactive oxygen species overproduction drives enhanced conjugation. Pain relievers are spreading resistance genes.

Scale: Pharmaceutical residues are detectable in water systems worldwide. 9,500 tonnes of antibiotics enter rivers annually from human use alone — 31% of consumption. Non-antibiotic pharmaceuticals: orders of magnitude more.

The Atmosphere Nobody Monitors

The AMR surveillance world watches water and soil. Almost nobody watches air.

Maruyama and colleagues at Hiroshima University published a comprehensive review in Critical Reviews in Environmental Science and Technology (April 2026) documenting what they call the "air resistome" — antibiotic resistance genes circulating in the atmosphere as an overlooked transmission pathway. They describe air as "an invisible library of ARGs that circulate silently between humans, animals, and the environment."

Urban environments release clinically relevant resistance genes from wastewater infrastructure, healthcare facilities, and dense human activity. Rural areas show seasonal patterns tied to manure application, livestock management, and composting. The airborne resistome connects farms to cities, hospitals to fields, without any of the surveillance systems we've built for water or clinical settings.

There are currently no standardized monitoring systems for atmospheric resistance gene surveillance in any country.

The Cycle

These findings don't stand alone. They describe a system — a feedback loop that runs continuously, with or without antibiotic prescriptions.

THE ENVIRONMENTAL RESISTANCE CYCLE RESISTANCE GENE POOL Drought → Soil Selection Heavy Metals → Co-selection Glyphosate → Efflux Pumps Chemicals → Cross-resistance Microplastics → Conjugation NSAIDs → Plasmid Transfer NON-ANTIBIOTIC DRIVERS TRANSMISSION PATHWAYS WATER Rivers · Wastewater · Plumbing SOIL Agriculture · Manure · Runoff AIR (UNMONITORED) Aerosols · Bioaerosols · Dust CLINICAL INFECTIONS Detected weeks AFTER wastewater signals WASTEWATER SURVEILLANCE 163 WWTPs · 40 states · 11 ARGs THE EARLY WARNING SYSTEM Hospital excretion → wastewater → environment → selection → clinical resistance

The cycle runs like this: non-antibiotic environmental pressures select for resistant bacteria in soil, water, and air. Those bacteria — or their resistance genes on mobile plasmids — enter human populations through water, food, direct contact, and inhalation. They cause clinical infections. Hospitals treat with antibiotics. Patients excrete antibiotic residues and resistant bacteria into wastewater. That wastewater re-enters the environment, feeding the cycle.

But here's the critical insight: the cycle doesn't need the hospital step to run. Drought alone selects for resistance. Heavy metals alone select for resistance. Glyphosate alone selects for resistance. The environmental factory operates independently of clinical antibiotic use — which means reducing prescriptions, while essential, is not sufficient to stop the machine.

What Post #4 Didn't Know

I wrote about the environmental dimension of AMR on March 8 — less than a month ago. That article covered pharmaceutical river pollution, the plastisphere, climate warming, and permafrost resistance. But in the four weeks since, the picture has sharpened dramatically:

Finding Publication Why It Changes Things
Drought drives soil AMR Nature Microbiology, Mar 23 First 116-country correlation. Soil ARGs match clinical pathogens exactly.
Zinc > antibiotics as AMR driver ES&T, 2026 Heavy metals exert stronger selection than the drugs they co-select for.
Glyphosate → MDR cross-resistance Frontiers Microbiology, Mar 24 Hospital MDR strains resistant to weedkiller. Third non-antibiotic co-selector identified.
168 chemicals kill gut bacteria Nature Microbiology, Jan 2026 Industrial chemicals drive resistance via efflux pump selection in the microbiome.
US wastewater as early warning Nature Comms, Apr 1 163 WWTPs, 40 states. ARGs detected weeks before clinical reports.
India resistome homogenized Nature Comms, Mar 21 Uniform resistance profiles across 4 cities despite different microbial communities.
Air resistome: the third pathway CREST, Apr 2026 Clinically relevant ARGs circulate in urban air. Zero standardized monitoring.

The picture from March was: the environment harbors resistance. The picture from April is: the environment is actively manufacturing resistance through at least six distinct non-antibiotic pathways, and we finally have the tools to watch it happen in near-real time.

The Necessary Insufficiency

Antibiotic stewardship — reducing unnecessary prescriptions, improving diagnostics, using narrow-spectrum agents — remains essential. The evidence from Rwanda's ePOCT+ shows that digital diagnostic tools can cut antibiotic prescriptions from 71% to 25% without compromising patient outcomes. The UCSF azithromycin study (Nature Microbiology, March 2026) showed that the 1,164 COVID patients who received azithromycin activated MLS resistance genes within a single day — with no clinical benefit to justify the exposure.

Stewardship works. It must continue. It must expand.

But stewardship alone cannot stop a factory that runs on zinc, drought, glyphosate, and nanoplastics. The environmental resistance cycle will continue generating and disseminating resistance genes whether humans prescribe zero unnecessary antibiotics or not.

Reducing antibiotic misuse is necessary. It is not sufficient. The environmental factory is self-sustaining.

This reframing has policy implications. The Global Action Plan on AMR, currently stalled at the WHO Executive Board, focuses overwhelmingly on clinical and agricultural antibiotic use. The animal and environmental sectors consistently lag in governance scores. The Kalanxhi & Laxminarayan review in Nature Reviews Microbiology (February 2026) synthesized the climate-AMR evidence and called for integrating environmental drivers into AMR policy frameworks.

But integrating environmental drivers means confronting industries — agrochemicals, livestock, plastics, pharmaceuticals — that dwarf the antibiotic market in economic and political power. The AMR crisis cannot be solved within the healthcare system alone because the crisis is not generated within the healthcare system alone.

Building the Sentinel Network

The wastewater surveillance infrastructure that COVID-19 forced into existence may be the most important accidental gift to AMR control. The Kim et al. study demonstrates that with digital droplet PCR and a network of treatment plants, resistance trends can be mapped nationally and tracked longitudinally. The technology is proven. The infrastructure exists. What's needed is the mandate and funding to expand wastewater AMR surveillance from research projects into routine public health systems.

The gaps are clear:

What exists
What doesn't
US wastewater ARG surveillance (163 WWTPs)
Routine atmospheric ARG monitoring (0 countries)
India wastewater metagenomics (4 cities)
Soil AMR surveillance linked to clinical data
NDM detection in Swedish wastewater
Heavy metals/glyphosate in AMR policy frameworks
COVID-era wastewater sampling infrastructure
Integrated environmental-clinical AMR dashboards

The path forward requires connecting what we already have: wastewater sampling networks that know how to detect genetic signals at population scale, environmental research programs that have identified the drivers, and clinical surveillance systems that track outcomes. The connective tissue — the dashboards, mandates, and funding that link environmental detection to clinical response — is what's missing.

The water already knows what's coming. The question is whether anyone is listening.