On San Cristóbal Island in the Galápagos, a research team from the University of Pennsylvania has spent two years running a microbiology laboratory that fits into backpacks, generating one of the more concrete recent examples of what field-deployable research infrastructure can do for environmental and public-health surveillance in a resource-limited setting.
A lab without a cold chain
The project, led by Daniel P. Beiting (Associate Professor, Department of Pathobiology, Penn’s School of Veterinary Medicine) together with Lisa M. Mattei (Senior Research Investigator, Department of Pathobiology, and Director of Operations at Penn’s Institute for Infectious & Zoonotic Diseases) and Michael Weisberg (Director of Penn’s Global Education and Research Alliance), built a portable microbiology setup designed around the constraints of an island community without reliable cold-chain shipping or constant electricity. According to Penn’s own account of the work, the kit includes a smartphone-controlled qPCR machine described as roughly brick-sized, a nanopore-style sequencer described as about the size of a candy bar, and self-contained reagent chambers modeled on contact-lens blister packs so that field staff don’t need to do precision pipetting on-site. The result is a workflow that can run genomic and metagenomic sequencing in the field rather than shipping samples off-island to a fixed laboratory.
Authors on the resulting study also include collaborators from Penn’s School of Arts & Sciences, the Perelman School of Medicine, and Ecuador’s Galápagos National Park authority (Parque Nacional Galápagos) — a collaboration structure that reflects how this kind of surveillance work depends on local institutional partnership, not just imported equipment.
What two years of sampling found
The team sampled 16 marine sites, two freshwater sites, and two points in the municipal wastewater system around Puerto Baquerizo Moreno, San Cristóbal’s main town, over roughly two years. The surveillance identified several locations where untreated wastewater was bypassing treatment infrastructure and reaching the ocean directly. Bacteria cultured from contaminated sites showed a markedly higher rate of multidrug resistance — resistance to three or more antibiotic classes — than samples from cleaner sites, and the metagenomic data pointed to extensive sharing of antimicrobial-resistance genes among bacterial populations in the marine environment. The findings were published in Nature Communications on August 4, 2026, with an earlier version posted as a bioRxiv preprint.
For a wildlife-dense marine ecosystem like the Galápagos, where sea lions, marine iguanas, and other endemic species share coastal waters with human settlements, wastewater-driven AMR gene flow is not just a local sanitation problem — it’s a pathway by which resistance genes can move between human, animal, and environmental bacterial reservoirs, which is precisely the interface One Health-oriented AMR surveillance is designed to track.
Why the delivery model is the story, not just the finding
What distinguishes this project from a typical AMR survey is less the result than the infrastructure choice behind it. Sequencing-in-the-field, rather than sequencing-after-shipping, removes a dependency that has historically limited environmental genomic surveillance to sites with reliable logistics: cold-chain sample transport, laboratory access, and turnaround times measured in weeks. A backpack-portable kit that a small team can carry to a wastewater outfall or a beach sampling point compresses that timeline and, notably, makes it feasible to keep local partners — in this case, Galápagos National Park staff — involved in ongoing monitoring rather than as one-time sample collectors for an external lab. Beiting has described the broader goal in terms of narrowing the distance between what researchers do and what the public actually sees of that work, a framing that lines up with a wider push toward community-engaged and locally embedded environmental monitoring rather than parachute-style sampling trips.
For research-infrastructure planning more broadly, the project is a useful illustration of a design principle: portable, low-power, field-operable instrumentation changes not just where data can be collected but who can be involved in collecting and interpreting it on an ongoing basis. That has implications for how funders and institutions think about equipping field sites in low-resource or logistically isolated settings, whether the target is AMR, water quality, or other forms of environmental genomic surveillance.
Funding
Penn’s account of the project credits support from the U.S. National Science Foundation, Penn’s School of Arts & Sciences and Global Research Institute, and several Penn undergraduate-research and mentoring awards that funded student participants over the project’s multi-year run — a funding mix that itself reflects how this kind of infrastructure-building field research tends to get resourced: a mix of federal science funding and internal university programs supporting sustained, multi-year student and staff involvement rather than a single large grant.
