Effective surveillance of mosquito populations is critical to monitoring and mitigating the spread of mosquito-borne diseases (MBDs). This study evaluated the relative trapping efficiency of two widely used adult mosquito traps—the Biogents BG-Sentinel 2 (BGS2) and the Mosquito Magnet® Executive (MM)—on British mosquitoes across four wetlands in south-east England over a 12-week period. A third trap, a Box-Gravid trap fitted with an FTA card, was deployed to detect arboviruses such as West Nile virus (WNV) via saliva collection. A total of 11,584 adult female mosquitoes representing 15 species were collected. The MM trap captured a significantly higher total number of mosquitoes, while the BGS2 trap demonstrated greater species evenness and was significantly more effective at catching Culex (Culex) pipiens L., 1758. Spatial variation strongly influenced catch rates, with significant differences between wetlands. No evidence of WNV was detected in any mosquito pools or FTA cards. While both trap types yielded similar species richness, the MM trap may be optimal for collecting large sample sizes of mammalophagic species, whereas the BGS2 is better suited for capturing enzootic vectors such as Culex pipiens s.l., and broader spectrum of species. These findings provide evidence-based recommendations for future UK wetland surveillance and enhance preparedness for emerging vector-borne disease risks.

Mosquito traps were operated in four wetlands in Kent in south-east England, targeting wetlands in the region where Culex (Barraudius) modestus Ficalbi, 1890, a vector of WNV, is known to occur (Supp. Info figure 1; Cliffe [latitude: 51.47364, longitude: 0.47749], Northward Hill [51.46209, 0.54479], Stodmarsh [51.30388, 1.18728], and Sandwich [51.27818, 1.36086]). Broadly, Cliffe, and Northward Hill are dominated by grazing marsh including flooded grassland and vegetated permanent ditches bounding grassland grazed by livestock. Stodmarsh is a large nature reserve including wet woodland, grazing marsh, vegetated permanent ditches, wet grassland, and reedbed. The Sandwich wetland is dominated by grazed arable fields, a large permanent ditch, wet grassland, and arable fields, noticeably less wetland habitat than the other wetlands studied here. This study did not quantify the scale of available habitat at each wetland. The traps were continuously operational across 11 weeks, (calendar week 27 (5th July 2022) to week 38 (20th September 2022)). At each wetland, a Mosquito Magnet® (MM) Executive trap (Woodstream Corporation, St. Joseph, MO, USA), using a propane cylinder (13kg propane supplied by Calor Ltd, Warwick, UK) and baited with an octenol (1- Octen-3-ol) lure (Woodstream Corporation) and a Biogents BG-Sentinel (BGS) baited with BG-Lure and CO2 (CO2 VB cylinder, 6.35kg, supplied by BOC Ltd, Woking, UK). Traps were located at three locations (A, B, C) with over 50 metres between them, apart from at Sandwich where traps were 20 metres apart due to access restrictions. Each week, the MM and the BGS traps were rotated between locations, so that the trap previously at location A was moved to location B, and vice-versa. The Box Gravid trap remained at location C for the duration of the study. The BGS2 traps were powered using a photovoltaic panel system, comprising a 12V 15Ah lead acid battery, 50W photovoltaic panel, and 10A 12V/24V charge controller (Photonic Universe PU1024BW). Box-Gravid traps (figure 1), based on the design of the modified Reiter-Cummings gravid trap (Cummings, 1992; Fynmore et al., 2021; Reiter, 1983). These traps consisted of two main components: a lower tray with hay infusion (5 L of liquid attractant, consisting of a hay infusion made by mixing 0.9 kg of hay with 114 L of water), and the main trap body which held the components and BugDorm (model TM -4M1515 Insect Rearing Cage [175mm x175mmx175mm]). The lower tray was a stackable black plastic tray (470 x 350 x170mm) with drainage hole, while the main trap body was a black toolbox (Stanley Tool Box 19" 2 Pieces, 250 mm x 485 mm). The main body stood on two construction bricks placed within the tray. Two 60 mm diameter holes were drilled centrally into the bottom and left-hand side of the toolbox using a hole saw. The lower hole was used to insert a 3Dprinted intake pipe (figure 1) and net holder with embedded magnets to affix it in place. The fan was bolted in place over the left-hand side hole. The intake pipe was sealed at one end, with an exit hole on the side of the tube, to prevent mosquitoes from falling directly out of the trap and into the water. Inside the toolbox, a 120mm 12V DC axial fan, powered by the battery stowed within the toolbox, was bolted to the interior covering the left-hand drilled hole. The fan was powered by a 12V 15aH leadacid battery connected to a charge controller and 50W photovoltaic panel, allowing for unlimited run time. The BugDorm was placed over entrance hole, oriented with the entrance port facing downwards in line with the intake pipe, and the entrance port sleeve was tucked into the sleeve holder on the 3D printed inlet, and the main trap body was then positioned over the lower tray. Whatman™ FTA™ Classic cards (GE Healthcare Life Sciences, Buckinghamshire, UK) soaked in Manuka honey solution (10% honey to distilled water) were prepared using a procedure similar to previously published methods (Flies et al., 2015; Wipf et al., 2019). The cards were cut into equal quarters, soaked in honey solution overnight, and placed into plastic resealable bags, with a 1.5 cm square hole cut in the front to provide mosquitoes access to the FTA™ cards. A cotton pad, soaked in the honey solution, was placed behind the FTA™ card to keep it moist during deployment. The FTA™ card was then affixed to the inside of the BugDorm TM on the opaque plastic side. All traps were checked weekly on a Tuesday, and FTA™ cards, BugDorms, and catch bags removed and labelled, and placed on dry ice. Upon return to the laboratory, mosquitoes were removed and stored at -80oC. FTA™ cards were stored at room temperature. Mosquito identification Female mosquitoes were identified morphologically using taxonomic keys (Becker et al., 2010; Cranston et al., 1987; Snow, 1990). Where morphological identification was not possible, due to requiring DNA analysis methods, species were grouped as follows: Culex pipiens sensu lato (s.l.) = Cx. (Culex) pipiens Linnaeus, 1758 and Cx. (Culex.) torrentium Martini, 1925; Aedes cantans/annulipes = Ae. (Ochlerotatus) cantans (Meigen, 1818) and Ae. (Ochlerotatus) annulipes (Meigen, 1830).