In-person participation for the workshop (free of charge) is limited to 40 places and will be allocated on a first-come, first-served basis. To register, send an email to uli.kluemper@tu-dresden.de indicating your interest in attending.

3/3

As part of our new EU Project “One Bridge”, led by the University of Patras, we are organizing an 'AMR & #OneHealth Workshop in Dresden, Germany on the 6th of October 2025. The workshop will take place at BUSINESSPARK DRESDEN, Bertolt-Brecht-Allee 22-24, 01309 Dresden from 8:30 – 17:00.

2/3

Our findings urge a #OneHealth rethink of antibiotic use in aquaculture.

Check the full study led by Jin Huang (Central South University) & Bing Li (Tsinghua Uni Shenzhen Graduate School) here:

www.nature.com/articles/s44...

🚨Key message🚨
Regulations based solely on chemical residues overlook microbial risks. Withdrawal times must also consider ARG abundance, mobility, and pathogen enrichment to protect food safety & the environment.

7/8

Nearly half of the enriched potential pathogens (e.g., Aeromonas, Vibrio, Citrobacter) carried ARGs identical to those on plasmids, showing strong plasmid–chromosome links for resistance spread.

6/8

Resistance genes weren’t just more abundant, they were also more mobile! Integrons, transposons (carrying floR), and diverse plasmids persisted after withdrawal, keeping ARGs ready for transfer.

5/8

ARG abundance spiked up to 17× during treatment. It dropped after stopping antibiotics but stayed significantly above control levels even long past the 15-day withdrawal time required for residue compliance.

4/8

We simulated standard (5 & 10 days) and prolonged (30 & 90 days) florfenicol treatments in common carp and tracked gut microbiota, ARGs, and mobile genetic elements (MGEs) before, during, and after treatment.

3/8

We reveal that antibiotic resistance #AMR risks in fish farming don’t stop after the legal withdrawal time. Our study shows that oral florfenicol in carp leaves a lasting “resistome” footprint even after fish are deemed market-ready.

2/8

Big thanks to co-authors Peiju Fang, Bing Li, Xia Yu, Dominic Frigon, Kerry Hamilton, Hunter Quon, Thomas Berendonk & Magali de la Cruz Barrón. Also to the Explore-AMR project that allowed me to spend a month at @tsinghuauniversity.bsky.social where this paper's concept and outline were created.
3/3

We argue that, in addition to quantification, assessing which and what proportion of ARGs can move via MGEs is key for predicting environmental #AMR risks. We outline tools, from ddPCR linkage to hybrid metagenomics, to make this part of routine monitoring and integrate it into QMRA.

2/3

Data & R code are open:
🔗 github.com/enviresist/P...
Interactive lookup: enviresist.github.io/PNEC_res

This could help set antibiotic discharge limits that truly protect against AMR selection in the environment. 🧬💊
🌍🛡

9/9

In conclusion:

Our framework is:
💊grounded in ecology & evolution
💊transparent & easy to compute
💊adjustable to different risk tolerances
💊ready for use in environmental AMR regulation
💊easily updateable when new cost or MIC data emerges

8/9

Next, we created a probabilistic model of resistance cost distributions. Instead of a fixed “/10” factor, we now use a probabilistic cost tied to a chosen protection level.

Example: To protect against selection for 95% of resistances, you’d use ~1/250 of MIClowest (25x lower than before).

7/9

Next, we used published data to map the global distribution of plasmid-borne resistance costs. Plasmid borne resistance is the most worrying type, because plasmids spread ARGs fast across species and surprisingly cheaper on average than chromosomal resistance costs.

6/9

We tested it:
26 strain–antibiotic pairs
13 antibiotics, 3 model bacteria, plasmid & chromosomal resistance
➡ 66% of predicted MSCs were within a factor of 2 of experimental values!

5/9

Our approach:
We link the minimum selective concentration (MSC) directly to the fitness cost of resistance.

For most high-level resistances (f=MICres/MICsus > 20),
MSC ≈ MIC × cost of resistance
✅ Simple
✅ Based on real evolutionary biology

4/9

Regulators need PNECres (predicted no-effect concentrations for AMR selection) to set discharge and pollution limits. Until now, the most used method (doi.org/10.1016/j.envint.2015.10.015) took the lowest MIC in databases & divided by 10

Problem: That “factor of 10” isnt biologically justified

3/9