Across the barns, cattle sheds, and racetracks of the Republic of Korea, a painstaking five-year trapping campaign has produced one of the most detailed pictures yet of the country’s Culicoides biting midges, the tiny blood-feeding insects that serve as the primary biological vectors of some of the most feared livestock viruses in the world. Between 2021 and 2025, researchers from South Korea’s Animal and Plant Quarantine Agency systematically collected more than 41,000 midges from nine livestock-associated sites, identified them to species, mapped how their communities shifted across space and time, and screened them for three groups of arboviruses with devastating potential for agriculture. The results, published in the open-access journal Parasites & Vectors, offer both a reassuring virological snapshot and a methodological template for how long-term entomological surveillance should be conducted in an era of changing climates and expanding vector ranges.
The stakes behind this seemingly modest insect are considerable. Culicoides biting midges, some of which measure barely a millimeter in length, are the principal transmitters of bluetongue virus, African horse sickness virus, and a suite of Simbu serogroup viruses including Akabane, Aino, and Peaton viruses. Bluetongue has wrought hundreds of millions of dollars in losses across European ruminant industries, while African horse sickness carries mortality rates in susceptible equine populations that can approach ninety percent. Simbu serogroup viruses, for their part, are notorious for causing congenital deformities in cattle, sheep, and goats when pregnant animals are infected. Because there are no curative treatments for these diseases and vaccination strategies vary in coverage and effectiveness, surveillance of the vectors themselves remains a cornerstone of veterinary public health. Knowing which midge species dominate at which sites, in which months, and in which years is the foundation upon which risk models, vaccination timing, and movement restrictions are built.
The research team, led by Jong-Uk Jeong and Jihun Ryu of the Foreign Animal Disease Division at the Animal and Plant Quarantine Agency in Gimcheon, with corresponding authors In-Soon Roh and SeEun Choe, structured their surveillance with unusual temporal density. Culicoides were collected biweekly from May through October—the window of peak midge activity on the Korean peninsula—during five consecutive field seasons. Trapping took place at seven cattle sheds and two horse racetracks, a deliberate mix of facility types designed to capture midge communities associated with both ruminants and equines, the two host groups most threatened by the viruses under investigation. Ultraviolet blacklight traps, the standard tool of Culicoides surveillance, were deployed at each site, yielding a total of 476 collection events over the study period. Of these, 277 produced Culicoides specimens, while 199 traps returned nothing, a pattern that itself carries ecological information about the patchiness of midge activity.
In total, 41,259 individual midges were collected, and each was identified morphologically using the taxonomic characters of wing patterns, palpal morphology, and antennal ratios that distinguish Culicoides species. Where morphology alone was insufficient to resolve an identification, the team turned to molecular barcoding, sequencing the mitochondrial cytochrome c oxidase subunit I gene—the COI marker that has become the workhorse of insect DNA taxonomy—to confirm species assignments. This combination of classical taxonomy and molecular confirmation is critical because misidentified vectors can silently corrupt surveillance data, leading risk assessments astray. The rigor paid off in a strikingly clear result: the Korean livestock-associated midge fauna proved overwhelmingly dominated by just three species, Culicoides arakawae, Culicoides punctatus, and Culicoides oxystoma, which together accounted for 95.1 percent of all individuals collected. The remaining species, though present, were numerically marginal.
One of the study’s most important contributions lies in its demonstration of how dramatically midge populations fluctuate from year to year. Mean abundance per collection event peaked during 2021 and 2022 and then fell substantially during the 2023, 2024, and 2025 seasons. Such swings are not merely academic; they have direct implications for how surveillance programs are evaluated. A single-season or single-site study, the authors note, can easily be misled by the idiosyncrasies of one year’s weather, one site’s microhabitat, or one farm’s management practices. Only by integrating data across multiple years and multiple locations can the underlying structure of a vector community be distinguished from its noisy year-to-year expression. This is precisely why multi-year datasets of this kind remain rare and why the Korean program’s persistence is scientifically valuable.
The temporal dimension extended to the composition of the community as well as its size. The relative proportions of the dominant species changed among years, meaning that even the identity of the primary vector candidates at a given site cannot be assumed to remain constant over time. When the researchers analyzed dominance at the level of individual sites, a consistent pattern emerged: most locations were dominated by either C. arakawae or C. punctatus, with C. oxystoma and other species playing secondary roles. This site-level dominance structure suggests that local environmental conditions—perhaps host availability, breeding substrate, moisture, or farm-level management—exert a stronger organizing force on midge communities than the regional forces that drive annual variation.
To test this formally, the team employed two complementary multivariate techniques that have become standard in community ecology. Permutational multivariate analysis of variance, or PERMANOVA, applied to Bray–Curtis dissimilarity matrices—a measure of how different two communities are in species composition and abundance—revealed that site effects on community composition exceeded year effects. In other words, where you trap matters more than when you trap, at least at the scale of this dataset. Non-metric multidimensional scaling, or NMDS, was used as a supplementary ordination tool to visualize these patterns, with the event-level and site-year-level ordinations presented as diagnostic figures rather than independent inferential tests. The NMDS analysis, based on the eight species whose total abundance exceeded one hundred individuals, showed clear clustering of collection events by dominant taxon, reinforcing the picture of a community structured primarily by local conditions.
The seasonal timing of each dominant species adds another layer of practical value. Mean seasonal abundance peaked in June for C. punctatus, in August for C. arakawae, and in September for C. oxystoma—a staggered phenology that effectively extends the window of vector risk across the entire summer and early autumn. For veterinarians and animal health officials, this staggered peak has concrete consequences. Bluetongue vaccination campaigns, movement restrictions, and vector control measures such as insecticide-treated shielding of livestock housing would ideally be timed to the local dominant species’ activity curve. A farm dominated by C. punctatus faces its highest-risk period in early summer, while one dominated by C. oxystoma remains at risk well into September. A one-size-fits-all national calendar, the data suggest, may obscure important regional and site-level differences in transmission risk.
The virological screening, meanwhile, delivered the study’s most consequential negative result. All 766 Culicoides pools—groups of midges pooled by site, date, and species for testing—were analyzed by reverse transcription polymerase chain reaction, or RT-PCR, for the RNA of bluetongue virus, African horse sickness virus, and Simbu serogroup viruses. Not a single pool tested positive. Over five years and more than 41,000 midges, none of the screened arboviruses was detected in the vector population at these livestock-associated sites. The authors are appropriately measured in their interpretation: a negative screening result does not prove the absence of viral circulation in the country or in wild vector populations beyond the sampled sites, and the study’s design—a surveillance dataset rather than a spatially balanced nationwide survey—means the finding cannot be generalized nationally. Still, the absence of viral RNA over such a sustained effort provides a meaningful baseline against which future detections can be judged, and it demonstrates that the surveillance pipeline, from trap to molecular assay, functions reliably.
The study is candid about its limitations, which strengthens rather than weakens its value. The authors explicitly state that the dataset does not constitute a spatially balanced nationwide survey; the nine sites were operational surveillance locations, not a statistically representative sample of Korean livestock environments. Nevertheless, they argue that integrating multi-year livestock-associated collection data provides a descriptive baseline for the spatio-temporal occurrence of dominant Culicoides species that can directly support future vector surveillance and risk assessment. Funding for the work came from the Animal and Plant Quarantine Agency of the Republic of Korea, and the authors acknowledge the regional Animal Disease Control centers, the Korea Racing Authority, and the many livestock producers who granted access to their facilities.
For the international community of vector biologists and veterinary epidemiologists, the Korean dataset speaks to a broader question: how should countries with limited surveillance resources allocate them? The finding that site effects dominate over year effects suggests that establishing a network of well-characterized sentinel sites, each monitored consistently over many years, may be more informative than rotating trapping efforts across many locations for short bursts. The staggering of seasonal peaks among the three dominant species suggests that sentinel trapping should extend across the full May-to-October window rather than concentrating on a single assumed peak month. And the clean virological results underscore the importance of screening not just for one virus but for the full panel of pathogens a vector community is capable of transmitting, since the same midges that carry bluetongue are candidates for Simbu serogroup circulation.
As climate change alters temperature and precipitation patterns across East Asia, the phenology, abundance, and geographic distribution of Culicoides species are all expected to shift, potentially lengthening transmission seasons and allowing vector species to colonize new territory. In that context, the five-year baseline established by the Korean team is not merely a record of what was, but a reference point against which future change will be measured. Whether the low abundances of 2023 through 2025 represent a temporary dip or the beginning of a longer trend, and whether the current virological silence in Korean midge populations persists, will be answerable only through continued surveillance of exactly the kind this study exemplifies—patient, methodical, and sustained across the years.
Subject of Research: Spatio-temporal occurrence and arbovirus screening of Culicoides biting midges at livestock-associated sites in the Republic of Korea, 2021–2025
Subject of Research: Biology
Article Title: Five-year spatio-temporal occurrence of Culicoides biting midges and arbovirus surveillance at livestock-associated sites in the Republic of Korea, 2021–2025
Article References: Jeong, J.-U., Ryu, J., Kim, H.-J., Han, Y., Jeoung, H.-Y., Kang, H.-E., Roh, I.-S., & Choe, S. (2026). Five-year spatio-temporal occurrence of Culicoides biting midges and arbovirus surveillance at livestock-associated sites in the Republic of Korea, 2021–2025. Parasites & Vectors. https://doi.org/10.1186/s13071-026-07674-w
Image Credits: AI Generated
DOI: 10.1186/s13071-026-07674-w
Keywords: Culicoides, Biting midges, Dominant species, Spatio-temporal occurrence, Seasonal occurrence, Vector surveillance, Arboviruses, Bluetongue virus, African horse sickness virus, Simbu serogroup viruses, Republic of Korea, Livestock
Cite Scienmag News
APA MLA Chicago
William Thompson. (September 6, 2026). Five-year surveillance of Culicoides biting midges and arboviruses in South Korea. Scienmag. https://scienmag.com/five-year-surveillance-of-culicoides-biting-midges-and-arboviruses-in-south-korea/
William Thompson. “Five-year surveillance of Culicoides biting midges and arboviruses in South Korea.” Scienmag, 6 September 2026, https://scienmag.com/five-year-surveillance-of-culicoides-biting-midges-and-arboviruses-in-south-korea/. Accessed 6 September 2026.
William Thompson. “Five-year surveillance of Culicoides biting midges and arboviruses in South Korea.” Scienmag. September 6, 2026. https://scienmag.com/five-year-surveillance-of-culicoides-biting-midges-and-arboviruses-in-south-korea/
Copy citation Download RIS
Tags: African horse sickness virusarbovirus screening in insectsarbovirus screening in midgesarbovirus surveillancearbovirus surveillance in South Koreabluetongue virus transmissionclimate change and vector range expansionclimate change impact on vector rangesCulicoides biting midgeslivestock disease outbreak preventionlivestock disease vectorslivestock virus epidemiologylong-term entomological monitoringparasitology and vector-borne diseaseSouth Korea animal health researchSouth Korea vector researchvector ecology and distributionvector species identificationvector-borne livestock viruseszoonotic disease risk assessment
