Leptospirosis is one of the most broadly distributed zoonotic diseases on the planet, yet it remains poorly documented in Papua New Guinea, where published research on the strains actually circulating in the country is scarce. The disease is caused by spirochaete bacteria of the genus Leptospira—exceptionally thin, helical microbes fitted with internal flagella that allow them to screw their way through viscous fluids and anchor themselves in host tissue. More than 300 serovars have been identified over decades of microscopy and serological testing with tools such as the microscopic agglutination test and the enzyme-linked immunosorbent assay, while DNA-based methods have so far distinguished 22 distinct species. Pathogenic and non-pathogenic members of the genus circulate across geographical regions, driving wide variation in how severe infections become. Among cattle, the most consequential strains include Leptospira borgpetersenii serovar Hardjobovis and Leptospira pomona, which are blamed for abortion, neonatal deaths, weak calves and substantial production losses on dairy and beef farms worldwide. Cattle farming itself is no small matter in PNG: it contributes roughly 15 percent of the national livestock subsector and sustains the livelihoods of both commercial operators and smallholder farmers.

Because no molecular investigation of Leptospira in PNG cattle had ever been attempted, the team designed a cross-sectional survey running from January to March 2024 across commercial and smallholder operations in Morobe Province and East New Britain Province. Purposive sampling targeted accessible farms and abattoirs, with animal health officers from the National Agriculture Quarantine and Inspection Authority assisting aseptic collection. In total, the researchers gathered 228 biological samples from nine sites: 150 kidney samples from vaccinated commercial cattle slaughtered at the Ramu Abattoir, 10 kidney samples from non-vaccinated cattle at the PNG University of Natural Resources and Environment abattoir, and material from live animals at the PNG University of Technology farm in Lae, where 10 blood samples were drawn from the tail vein and roughly 60 millilitres of urine was collected from each of 42 cattle. Six non-vaccinated smallholder farms in the Markham Valley, including Agro Venture Limited, Trukai Farms and DAL Warwin, each contributed six urine samples. To probe the wildlife side of the transmission cycle, the team laid cage and glue traps around the university farm over 14 nights and collected 16 rat kidneys, preserved individually in ethanol for the trip to the laboratory.

At the Biotechnology Centre in Lae, the samples underwent a molecular workout. Kidney tissue preserved in 70 percent ethanol was air-dried before processing, and about 25 milligrams of tissue—together with 100-microlitre aliquots of blood and urine—was run through a commercial DNA extraction kit whose wash steps strip out residual ethanol, a notorious inhibitor of the polymerase chain reaction. Spectrophotometry confirmed clean extracts, with absorbance ratios between 1.8 and 2.0. The team then screened every sample by conventional PCR using two primer sets: one targeting the secY gene, a marker for pathogenic Leptospira, and another targeting the rrs gene, which encodes 16S ribosomal RNA and can flag both pathogenic and non-pathogenic relatives. Each reaction mixed primers at 0.4 micromolar with half a unit of Taq polymerase, then cycled 35 times through denaturation at 94 degrees Celsius, annealing at 58 degrees and extension at 72 degrees. Products were resolved on 2 percent agarose gels, where positives were expected as bands of 549 and 525 base pairs. With no positive control available in the laboratory, every candidate band was sent for Sanger sequencing at an accredited facility in Singapore and verified against the NCBI nucleotide database using BLAST.

The results drew a sharp line between the two host groups. All 212 cattle samples—kidneys from both the vaccinated Ramu cattle and the non-vaccinated East New Britain animals, plus every blood and urine specimen—tested negative for Leptospira DNA. The rats told another story. Two of the 16 rodent kidney samples, both collected at the PNG University of Technology farm, amplified successfully, giving the rodents a positivity rate of 12.5 percent and the full 228-sample dataset an overall rate of just 0.88 percent. Curiously, the secY amplicons measured roughly 354 base pairs rather than the 549 the primers were designed to produce. The authors attribute the shortfall to strain-specific genomic variation or to alternative primer binding inside the target gene—phenomena well documented in leptospiral diagnostics that can yield correctly amplified but shorter products. Because PCR alone could not settle the identity of the organism, the positive products were sequenced and searched against the nucleotide collection, and the resulting sequences fed into phylogenetic trees built with 100 bootstrap replicates to test the reliability of every branching pattern.

The genetic evidence left little doubt about what was living in those rat kidneys. Both isolates matched Leptospira interrogans isolate P1D297, and the phylogenetic analysis placed them within a cluster of uncultured Leptospira sequences held in international databases. The PNG rat isolates showed very high similarity—bootstrap support above 95 percent—to isolates labelled NRW30, NRW31, NRW54, NRW73 and NRW72; high similarity, between 85 and 95 percent, to isolates LJR051, NRW24, NRW66 and LJR028; and moderate similarity, between 70 and 85 percent, to Leptospira interrogans strain MORU L1207. The dominance of uncultured isolates among the closest matches is telling in its own right: Leptospira is notoriously difficult to grow in the laboratory, demanding specialised media and long incubation, so many strains known to science exist only as DNA recovered directly from clinical or environmental samples. The close kinship between the PNG strains and globally distributed environmental isolates, the authors suggest, hints that these pathogenic lineages share common ancestral roots with leptospires detected far beyond the Pacific, a signature of potentially wide geographic dispersal.

Interpreting the all-clear on the cattle demanded as much care as the positive rat result. Serological surveys conducted years earlier in PNG had detected antibodies against Leptospira in ruminants, but those tests register immune memory—to past infections, or even to vaccines—whereas PCR registers the bacterium itself. Vaccination may partly explain the pattern: the Ramu cattle were routinely vaccinated against leptospirosis, a practice known to reduce bacterial circulation and urinary shedding and thus narrow the window in which DNA can be caught. Yet the non-vaccinated farms were also negative, pointing to other forces. Leptospiral bacteraemia is brief and confined to the acute phase of infection; urinary shedding is intermittent even in chronically infected animals; and conventional PCR for leptospirosis carries a reported sensitivity of only about 56 to 62 percent, meaning low bacterial loads or degraded DNA can slip past detection. Wet-season sampling may even work against the assay, diluting leptospires in urine rather than concentrating them. The strongest argument, the authors note, is that kidney—the bacterium’s preferred sanctuary, where it colonises the renal tubules—still came back negative in abattoir cattle, supporting the conclusion that the herds were not actively infected at the time of sampling.

The exclusive detection of the pathogen in rats reinforces their established role as maintenance hosts of leptospirosis. Rodents tolerate chronic infection and shed bacteria continuously in their urine, seeding soil and standing water with organisms that persist for weeks in warm, wet conditions. On a cattle farm, that contamination creates ready transmission pathways—through drinking water, feed stores and shared paddocks—to incidental hosts such as livestock and people. The implications reach well beyond the herd. Farm workers, veterinarians and livestock handlers at facilities harbouring infected rodents risk exposure through contact with contaminated environments, and domestic animals such as dogs and pigs sharing the same ground can act as intermediate or amplifying hosts, widening the zoonotic loop further. This is leptospirosis operating as the textbook example of a One Health disease, a zoonosis playing out at the junction of animal, human and environmental health. The authors argue that only surveillance which monitors multiple host species and the shared environment simultaneously can realistically map and manage transmission across PNG’s agricultural landscapes.

The study also marks a milestone in the country’s sparse leptospirosis record. It is the second molecular confirmation of Leptospira diversity in PNG rodents, following a 2022 survey of bats and rats, and the first such documentation in Morobe Province specifically. The sequences generated in Lae now feed into international databases, sharpening the global picture of Leptospira diversity while giving national authorities a molecular reference point for future surveillance. The findings also translate into a concrete checklist for the university farm and similar facilities: secure feed storage against rodent intrusion, clear away debris piles and unused equipment that provide harborage, run regular rodent population monitoring, and equip workers who handle animals or move through potentially contaminated ground with gloves and boots. Integrated pest management, the authors stress, is not mere housekeeping but a frontline public health measure, because reducing rodent numbers and severing contact between livestock and rodent habitats directly lowers the risk of spillover to animals and humans alike.

Even with zero positives in the cattle, the researchers treat the survey as a foundation rather than a failure. The standardized protocols, the negative screening across nine sites and the two characterized isolates now constitute the first molecular baseline for leptospirosis in PNG cattle systems, and they set the stage for the multi-site, multi-season investigations—with statistically determined sample sizes—that the authors say are needed to track how rainfall, farm management and rodent ecology shape the bacterium’s movement across the country’s agricultural landscapes. For now, the two positive rat kidneys carry the study’s central message: in the fields of Morobe Province, the threat is not sweeping through the herd. It is scurrying quietly among the rodents, waiting for a chance to cross.

The H5N1 influenza virus is primarily recognized for its impact on avian species, but sporadic reports of infection in mammalian hosts, including poultry-adjacent livestock, have raised alarm about its potential adaptation to cattle. Given the immense scale of dairy farming in the United States, understanding the virus’s transmission networks within these environments is paramount. The team pioneered a compartmentalized approach to model the spread, subdividing the cattle population into distinct health states—susceptible, exposed, infectious, and recovered—while incorporating herd demographics and behavioral parameters, including spatial interactions and management practices.

What sets this model apart is its nuanced incorporation of multiple transmission pathways. Beyond direct contact between animals, the model accounts for indirect transmission via contaminated fomites, aerosolized droplets under various environmental conditions, and seasonal variations that influence viral persistence. The researchers meticulously parameterized these components using a blend of field data collected from US dairy farms and viral shedding profiles obtained through experimental virology studies. This multi-disciplinary synthesis ensures the model mirrors real-world complexity, enhancing its predictive power and relevance.

One of the pivotal revelations of the study is the identification of key risk factors that exacerbate the spread of H5N1 within herds. Population density emerged as a critical determinant, with tightly packed housing increasing contact rates and facilitating rapid viral dissemination. Moreover, the model spotlights the role of calf housing areas as potential “hotspots” due to younger animals’ heightened susceptibility and immune naivety. Importantly, the model predicts that without timely intervention, outbreaks could quickly escalate, leading to substantial morbidity and jeopardizing milk production.

The researchers employed a rigorous sensitivity analysis to dissect which parameters hold the greatest sway over transmission dynamics. Contact rate coefficients, environmental viral decay constants, and latency periods were among the most influential, revealing critical leverage points for disease control. For instance, accelerating the removal of infectious animals from the population and optimizing cleaning protocols for shared equipment could significantly curb virus spread — insights that are both actionable and economically feasible for farmers and veterinarians alike.

To simulate intervention efficacy, the team integrated vaccination strategies into their model, exploring scenarios ranging from partial to full herd immunization. The outcomes suggest that even moderate vaccination coverage could drastically reduce outbreak size, delay peak infection times, and enhance herd immunity thresholds. These simulations underscore the potential benefits of adopting preemptive vaccination programs tailored to specific farm structures and seasonal risk windows, providing a datapoint for policymakers contemplating regulatory measures.

Beyond internal herd dynamics, the model extends to appraise inter-farm transmission risks, factoring in cattle movement patterns, such as transport to markets and shared grazing lands. This broader network perspective reveals that controlling disease at the individual farm level is insufficient if regional transmission corridors remain open. As such, the study advocates for coordinated surveillance and movement restrictions during outbreak periods, drawing parallels with successful containment protocols used in other livestock diseases.

Technically, the mathematical framework hinges on a system of coupled ordinary differential equations (ODEs) that describe the temporal evolution of each compartment. The researchers supplemented these with stochastic elements to capture random fluctuations, which are especially pertinent during early outbreak phases when case numbers are low. This hybrid deterministic-stochastic paradigm affords robustness against uncertainties inherent in biological systems, which often defy purely deterministic forecasting.

A notable strength of this model lies in its extensibility. The modular architecture enables rapid incorporation of new viral strains, variable host susceptibilities, or alternative management practices, making it a valuable platform for ongoing surveillance in a landscape where influenza viruses continually mutate. The authors envision adapting the framework to other susceptible livestock species, potentially creating an integrated tool for multi-host influenza ecology.

The implications of this research reverberate beyond the realm of agricultural biosecurity. Considering the zoonotic potential of H5N1, insights from dairy cattle transmission models could inform human health risk assessments, particularly for farm workers and communities situated near intensive livestock operations. The modeling approach also contributes to the global understanding of influenza virus ecology, feeding into One Health initiatives that strive to bridge veterinary and human medical sciences.

This study’s methodological rigor was balanced by transparency regarding limitations. The authors acknowledge the paucity of longitudinal data on H5N1 prevalence in US cattle, which necessitated certain assumptions and parameter estimations. Future studies will benefit from targeted surveillance to validate and refine model parameters, facilitating dynamic updating as new data emerge. Furthermore, the model currently excludes viral evolution dynamics, an aspect critical in influenza research, earmarked for next-generation iterations.

One of the most captivating facets of the work is its emphasis on real-world applicability. By partnering with dairy industry stakeholders during model development, the researchers ensured that their findings have immediate translational potential. Recommendations such as modifying pen designs to reduce animal density or adjusting ventilation systems to mitigate airborne spread could be implemented swiftly at the farm level with demonstrable impacts on disease control.

The article also stimulates discussion around the economic trade-offs inherent in disease mitigation. While vaccination and enhanced biosecurity measures incur upfront costs, the model’s projections of outbreak severity and duration enable quantitative cost-benefit analyses, enabling producers to make informed decisions. This aligns with the increasing trend toward data-driven farm management where epidemiological models serve as decision support tools.

Looking ahead, the interdisciplinary outlook of this research heralds a new chapter in infectious disease modeling. By marrying mathematical sophistication with biological realism and practical farming insights, the study exemplifies how computational epidemiology can transcend theoretical abstraction to become an indispensable asset in safeguarding food production systems. The prospect of expanding such models to incorporate climate change effects or socio-economic variables further enriches their potential.

In sum, the work by Rawson and colleagues represents a landmark contribution to our understanding of H5N1 influenza in dairy cattle, elevating the discourse on livestock disease transmission through state-of-the-art mathematical modeling. Its comprehensive approach, spanning molecular biology to farm management, offers a beacon of guidance for researchers, agriculturalists, and policymakers confronting the multifaceted challenges posed by zoonotic pathogens in a globally interconnected world.

Subject of Research: Mathematical modeling of H5N1 influenza transmission in US dairy cattle

Article Title: A mathematical model of H5N1 influenza transmission in US dairy cattle

Article References:
Rawson, T., Morgenstern, C., Knock, E.S. et al. A mathematical model of H5N1 influenza transmission in US dairy cattle. Nat Commun 16, 4308 (2025). https://doi.org/10.1038/s41467-025-59554-z

Image Credits: AI Generated