Eric Maurice Fèvre

Rift Valley Fever (RVF) transmission is shaped not only by vector-host interactions but also by livestock trade and animal product flows through complex value chains. However, most existing epidemiological models focus on farm-level dynamics and neglect the role of value chains in disease transmission. We developed a compartmental Susceptible-Infected (SI) model integrating mosquito, livestock, human, and meat-mediated transmission across four value chain nodes: producers, traders, butchers, and consumers. The model was parameterized using published data from Kenya and other pastoral settings in Africa and explored under multiple intervention scenarios, including vaccination, trade restrictions, slaughter controls, and vector management strategies. From initial conditions an outbreak of RVF in humans and goats with total infected at the peak of the outbreak reaching 12.5 (9 to 16, 95% CI) and 55.2 (43 to 69, 95% CI) respectively. Compared to baseline a 75% reduction in goat trade, banning producer to trader and trader to butcher goat trade significantly increased the length of outbreaks in goats significantly to 58.94 (43–78.52, 95% CI), 66 (48–89, 95% CI) days and 172.23 (135–219, 95% CI) days. Mosquito contact reduction significantly reduced peak goat and human infections to 4.55 (2–8, 95% CI) and 4.42 (2–8, 95% CI) respectively. The value chain framework reveals that disease control strategies should extend beyond traditional farm-level interventions to target high-risk value chain nodes.

Wesselsbron virus (WSLV) is an understudied mosquito-borne zoonotic pathogen with important veterinary and public health implications in sub-Saharan Africa: it causes abortions and congenital malformations in livestock and can infect humans, resulting in febrile illness. To better characterize its ecological suitability and guide surveillance, we developed a comparative species distribution modelling framework using multiple machine-learning algorithms and ensemble approaches. Using occurrence records and a broad set of climatic, environmental, and anthropogenic predictors, we evaluated the performance of Random Forest, XGBoost, Neural Network, and ensemble models, including Rank Averaging and Stacking. The best-performing models achieved high discriminatory ability, with ensemble approaches providing the most robust and generalizable predictions. Variable importance analyses indicated that WSLV suitability was strongly associated with climatic variability, precipitation-related conditions, altitude, human population density, and livestock density. Spatial predictions identified major areas of suitability across southern, eastern, and western Africa, with uncertainty concentrated in transitional ecological zones where further field data are required. These findings are broadly consistent with previous Maxent-based modelling work, which also highlighted high ecological suitability in equatorial and southern Africa, while suggesting that model choice may influence the spatial extent of predicted risk. Overall, this study demonstrates that WSLV distribution is shaped by interacting climatic and anthropogenic drivers and that ensemble machine-learning frameworks can improve risk estimation for neglected arboviruses with sparse and heterogeneous data. The resulting maps provide a useful basis for One Health surveillance, targeted field validation, and preparedness planning in high-risk livestock-producing regions.

Background: Consumption of food contaminated with antimicrobial-resistant (AMR) Salmonella spp. poses significant public health risks. Methods: We investigated AMR patterns in non-typhoidal Salmonella (NTS) isolated from raw pork, ready-to-eat pork and side salads from butcheries in Busia County, Kenya (April–October 2019). A total of 81 isolates underwent antimicrobial susceptibility testing using disc diffusion according to Clinical and Laboratory Standards Institute 2018 guidelines. Results: Of the isolates, 90% (73/81) were resistant to at least one antibiotic, 22.2% were multidrug resistant and 9.8% exhibited extended drug resistance. High resistance was observed to gentamicin (70.4%), ampicillin (38.3%) and cefuroxime (37%). Conclusions: The high prevalence of AMR Salmonella in pork and salad samples highlights the risk of transmission to humans in Kenya. These findings underscore the urgent need for strengthened food safety practices, antimicrobial stewardship and integrated AMR surveillance across livestock value chains.

Introduction: Soil‐transmitted helminths (STHs) remain a major global public health challenge, with dogs recognised as important reservoirs for several zoonotic species. In Kenya, data on the prevalence and genetic diversity of canine gastrointestinal helminths are scarce. This study investigated the prevalence of zoonotic helminths in dogs and characterised the genetic diversity of canine hookworms in Machakos County, Kenya. Methods: A cross‐sectional household‐based survey was conducted across three sublocations representing low, medium and high human population density. Faecal samples were collected from 104 dog‐owning households and examined using Mini‐FLOTAC and zinc chloride sieving–flotation techniques. Hookworm species were identified using nested polymerase chain reaction targeting the mitochondrial cytochrome c oxidase subunit I gene, followed by DNA sequencing. Phylogenetic diversity of Ancylostoma caninum was assessed using median‐joining haplotype network analyses. Results: Mini‐FLOTAC detected a higher prevalence of hookworm (89.4% vs. 81.6%), Toxocara spp. (45.2% vs. 19.7%) and taeniids (29.8% vs. 5.3%) compared to zinc chloride sieving–flotation. The latter identified additional parasites, including Trichuris spp. (9.2%), Spirocerca lupi (3.9%), Alaria spp. (1.3%) and Clonorchis sinensis (2.6%). A. caninum was the sole hookworm species identified, alongside genetically distinct A. caninum ‐like lineages. High genetic variability was observed, comprising 80 polymorphic sites and 40 haplotypes, of which 31 were singletons, seven occurred in two sequences each and two were represented by five and six sequences. Haplotype network and phylogenetic analyses revealed high genetic diversity and weak population structuring in A. caninum , with Kenyan isolates clustering with global lineages while exhibiting local diversification. We also report the first molecular confirmation of Angiostrongylus vasorum in domestic dogs in Kenya. Conclusions: These findings underscore dogs as key reservoirs of zoonotic helminths and highlight the need for strengthened surveillance and integrated One Health strategies incorporating molecular diagnostics and targeted education within dog health programmes in line with the WHO NTD Roadmap.

Background: Taenia solium is responsible for a substantial global disease burden in pig-raising and pork-consuming regions of the developing world. The study assessed the reported spatial pattern of Taenia solium taeniosis and cysticercosis (TSTC) in Kenya, with the focus on risk mapping of TSTC in Busia County. Methodology: The study employed a mixed approach, incorporating routinely collected data and open-source resources. A literature review and collection of pig population data were used to map the TSTC and pig population distribution at the national level. In Busia County, the retrospective study retrieved reports on porcine cysticercosis and epilepsy (as a proxy indicator of neurocysticercosis) from hospitals and meat inspection records. The cross-sectional study assessed risk behaviors associated with TSTC in Busia County. The TSTC reports and proportions of porcine cysticercosis and epilepsy (as a proxy indicator of neurocysticercosis) were aggregated at the county and sub-county levels and visualized using QGIS. In addition, the World Health Organization’s (WHO’s) T. solium mapping tool was employed for risk mapping T. solium in Busia County. Results: The pig population data showed a linear growth trend, with a higher population reported in Central and Western Kenya. A systematic search of the literature yielded a total of fourteen research reports, with the reported cases ranging from 1.8% to 49.9% for porcine cysticercosis, 1.6% to 31.1% for human cysticercosis, and 0.18% to 19.9% for T. solium taeniosis, respectively. The retrospective data showed cases of porcine cysticercosis and epilepsy (as a proxy indicator of neurocysticercosis) in all sub-counties of Busia. The WHO risk mapping tool categorized Bunyala, Teso South, Nambale, and Butula sub-counties as high-risk areas. The questionnaire survey highlighted semi-confinement as the predominant pig husbandry practice (61.1%), with 32% of pigs having access to sewage, and there was poor community awareness about TSTC in Busia County. Conclusion: The study indicated the presence of TSTC transmission in Western Kenya and identified Busia County as a high-risk area based on multiple layers of evidence. This mixed approach utilized readily available data to generate new evidence that could support sustainable One Health strategies for interrupting the T. solium transmission cycle.

Background: Ecological variables that vary across time and space shape mosquito populations, creating microenvironments that can become disease transmission hotspots. Rift Valley fever virus (RVFV), a priority zoonotic arbovirus, thrives in diverse conditions and has been detected in over 50 mosquito species. This diversity complicates efforts to identify the key vectors involved in transmission and highlights the need to understand how environmental conditions shape mosquito abundance in high-risk landscapes. Methods: This study investigated spatio-temporal variation in mosquito abundance across the semi-pastoral landscape of Loitokitok sub-county, Kajiado County, Kenya. Over a full year, inclusive of the 2023–2024 El Niño rains, repeated mosquito trapping events were conducted at households enrolled in a human clinical cohort study, with weather station data linked to each trapping event. Results: A total of 441 mosquitoes were captured across 39 trapping events, with an average of 11.3 mosquitoes per event. The highest rainfall occurred in November 2023, while mosquito abundance peaked in April 2024. Traps placed at households in cropland areas hosted significantly more mosquitoes overall and were associated with more Anopheles spp., predominantly Anopheles gambiae (Kruskal–Wallis χ ² = 6.9, df = 2, P = 0.03), while those in shrubland areas had more Aedes aegypti (Kruskal–Wallis χ ² = 11.9, df = 2, P = 0.002). Multivariable models showed that land use/land cover (LULC) consistently improved model fit, though temporal weather factors were stronger predictors. Weather conditions from the prior month better predicted mosquito abundance than weather conditions over shorter time frames, with temperature consistently included in top models and humidity outperforming rainfall as an additional covariate in the best-fit model that included LULC, temperature, and humidity. Conclusions: These findings highlight the role of weather patterns and LULC in shaping mosquito dynamics, with irrigated cropland likely creating persistent breeding sites and shrubland providing niches for Ae. aegypti . This emphasizes the need for targeted, community-driven vector control strategies to mitigate RVFV transmission risk and highlights the importance of altered agricultural landscapes in driving vector dynamics. Graphical

Diverse salmonellae have the potential to cause disease and may be carried asymptomatically within the intestine of many vertebrate species. The relative contribution of human, animal, and environmental hosts to the transmission of Salmonella is unknown within and between households in low-income settings, especially where humans and animals may live in close contact and sanitary infrastructure is often inadequate. Between November 2018 and December 2019, we isolated Salmonella spp. from thirty households in urban and rural locations in Malawi, sampling at three time points from the stool of humans, animals, and their household environment. Using whole genome sequencing and fine-resolution bioinformatic and phylogenetic analyses we found evidence of sharing of Salmonella species and strains between humans, animals and the environment, both within and between households. The intricate web of interconnected salmonellae within this ecosystem underscores the importance of adopting a multi-faceted ‘One Health’ strategy when considering control of Salmonella in low-intensity agricultural systems.

In recent decades there has been a huge increase in the export of cut flowers from countries in Africa and elsewhere to European flower markets, with the vast majority first entering the Netherlands for local use or for export. Coincidentally, three significant livestock disease outbreaks caused by viruses associated with Africa or other tropical regions were first detected in the Netherlands (bluetongue virus serotype 8 ( BTV ‐8), 2006, and BTV ‐3, 2023) and in western Germany about 200 km from the Netherlands border (Schmallenberg virus, SBV , 2011). This study aimed to determine whether Culicoides biting midges (Diptera: Ceratopogonidae), the vectors of BTV and SBV , are present within flower‐packaging plants in East Africa, and therefore whether Culicoides could be unknowingly exported during the shipping of cut flowers. Field sampling was undertaken at a flower‐packaging facility in Kenya, East Africa. The facility undertook all stages of cut flower production from maintaining rootstock through to packaging and shipping to an airport for international export. Trapping was undertaken at each stage of production (rootstock, propagation, inside growing greenhouses, in the packing‐house, inside cold‐storage rooms, during transportation) using Centers for Disease Control and Prevention ( CDC ) Light Emitting Diode ( LED ) light traps. Hand‐held aspirators were used to obtain individual insects directly from flowers and around composting sites, while emergence traps studied insect emergence from compost, leaf litter and flowers discarded at quality control checkpoints. A maximum nightly catch of 269 Culicoides was identified on a half‐acre smallholding, containing 15 ruminants and 40 birds, located 20 m from the nearest greenhouse. The greatest numbers of Culicoides were trapped at a pond ( n = 23) and leaf‐litter compost site ( n = 19) within the curtilage of the flower‐packaging plant. Of the seven greenhouses sampled, three had Culicoides trapped overnight (mean = 4, range: 1–9), and no Culicoides were trapped in the propagation units. No Culicoides were trapped in the pack house, cold‐store, or during transportation of the flowers to the airport for shipment. No Culicoides emerged from emergence traps or were trapped when aspirating directly from flowers. This is the first study to investigate whether Culicoides are present within flower packaging plants in Africa. The results highlight that although present in small numbers both outside and within greenhouses, the presence of Culicoides declined with each stage of production. Therefore, the risk of exporting Culicoides with packaged cut flowers is non‐zero but likely very small.

Enterotoxigenic Escherichia coli (ETEC) is a significant cause of diarrhoea in livestock and humans. The epidemiology of ETEC in animals remains understudied, prompting an investigation into the virulence factors and associated adhesins of ETEC in livestock from Western Kenya. Also, there is limited evidence supporting the role of livestock as possible zoonotic reservoirs for ETEC. ETEC strains harbour colonization factors/adhesins and enterotoxins, with animal ETECs exhibiting various adhesins (F4, F5, F6, F17, F18 and F41). Enterotoxins include heat-labile (LT) and heat-stable (ST) toxins and are further divided into LT-I and LT-II and STa and STb, respectively. Additional toxin combinations occur, with ETEC and Shiga toxin-producing E. coli (STEC) hybrids garnering public health significance. Here, we analysed faecal and mesenteric lymph node samples from diverse livestock across three Western Kenyan counties (Busia, Bungoma and Kakamega), using whole-genome sequencing. In silico screening determined the presence of AB₅ and A₂B₅-like toxin genes, including cytolethal distending toxin (cdtABC) along with associated adhesins. To broaden the screening panel, adhesin genes identified were further characterized to identify both known and novel alleles, particularly focusing on human-ETEC colonization factors. Two estA alleles (estA-4-06, estA-6-02) and six eltAB-II toxin alleles (eltAB-II-a2-01, eltAB-II-a3-01, eltAB-II-c1-02, eltAB-II-c6-03, eltAB-II-c6-04 and eltAB-II-c7-02) were identified in livestock. Hybrid ETECs identified were ETEC/STEC present in 6.7% (4/60) of ETEC strains and ETEC with cdtABC type I. An A₂B₅-like tripartite toxin, potentially resembling the typhoid toxin, was detected in 8.7% (4/46) of the eltAB-II-positive strains. It may have unique effects on enterocytes distinct from known toxins. These findings expand our understanding of ETEC pathogenicity and genetic diversity in animal reservoirs, while also highlighting potential zoonotic risks. They broaden the toxin repertoire, offer adhesin-based vaccine candidates for livestock and provide valuable insights for future vaccine development and public health strategies in the Lake Victoria Crescent ecosystem and beyond.