Evaluation of a novel West Nile virus transmission control strategy that targets Culex tarsalis with endectocide-containing blood meals

Abstract

Control of arbovirus transmission remains focused on vector control through application of insecticides directly to the environment. However, these insecticide applications are often reactive interventions that can be poorly-targeted, inadequate for localized control during outbreaks, and opposed due to environmental and toxicity concerns. In this study, we developed endectocide-treated feed as a systemic endectocide for birds to target blood feeding Culex tarsalis, the primary West Nile virus (WNV) bridge vector in the western United States, and conducted preliminary tests on the effects of deploying this feed in the field. In lab tests, ivermectin (IVM) was the most effective endectocide tested against Cx. tarsalis and WNV-infection did not influence mosquito mortality from IVM. Chickens and wild Eurasian collared doves exhibited no signs of toxicity when fed solely on bird feed treated with concentrations up to 200 mg IVM/kg of diet, and significantly more Cx. tarsalis that blood fed on these birds died (greater than 80% mortality) compared to controls (less than 25% mortality). Mosquito mortality following blood feeding correlated with IVM serum concentrations at the time of blood feeding, which dropped rapidly after the withdrawal of treated feed. Preliminary field testing over one WNV season in Fort Collins, Colorado demonstrated that nearly all birds captured around treated bird feeders had detectable levels of IVM in their blood. However, entomological data showed that WNV transmission was non-significantly reduced around treated bird feeders. With further development, deployment of ivermectin-treated bird feed might be an effective, localized WNV transmission control tool.

Fig 1. Cx. tarsalis survival in bioassays with IVM and WNV.

(A) Cx. tarsalis survival following a membrane blood meal containing IVM at 73.66 ng/mL (LC₇₅) and WNV at a titer of (A) 5x10⁵ PFU/mL or (B) 10⁷ PFU/mL. (C) Cx. tarsalis survival following a second membrane blood meal of IVM at 73.66 ng/mL given 10 days after a first blood meal of 10⁷ PFU/mL WNV. Error bars indicate standard error.

Fig 2. Blood feeding on Ivomec-treated chickens increased Cx. tarsalis mortality.

Cx. tarsalis survivorship following direct blood feeding on chickens given Ivomec-formulation diet at a concentration of 200 mg IVM/kg of diet for 3 (left panels: A, C, E) and 7 (right panels: B, D, F) days. (Top panels: A, B) Blood feeding occurred on the last day treated diet was given to the IVM groups. (Middle panels: C, D) Blood feeding occurred one day after treated diet was withdrawn from the IVM groups. (Bottom panels: E, F) Blood feeding occurred on the second day after treated diet was withdrawn from the IVM groups. Error bars indicate standard error.

Fig 3. Blood feeding on powder IVM-treated doves increased Cx. tarsalis mortality.

Cx. tarsalis survival following direct blood feeding on captured wild Eurasian collared doves fed powder IVM-diet at a concentration of 200 mg IVM/kg diet for 7 to 10 days. Error bars indicate standard error.

Fig 4. Characterization of chicken IVM sera concentrations.

(A) Chicken sera IVM concentrations measured in IVM-treated groups and taken on the last day treated diet was given, or one or two days after treated diet was withdrawn. Lines indicate median values, boxes indicate 25–75 percentiles, whiskers indicate minimum and maximum values. (B) Individual chicken IVM serum concentrations versus corresponding cumulative Cx. tarsalis mortality on day 5 post blood feeding for both direct and serum-replacement blood feeds on chickens given powder IVM-formulation feed at a concentration of 200 mg IVM/kg of feed.

Fig 5. Blood feeding on wild-caught grackle serum resulted in increased Cx. tarsalis mortality.

Cx. tarsalis survival following a serum-replacement blood feed using serum from a wild-caught grackle in which IVM was detected compared to control calf blood.

Fig 6. Cx. tarsalis abundance and infection for 2017 field sites.

Dotted lines indicate the start of IVM-treated feed at IVM sites. (A) Historical data from the same sites (2006–2016) were fit to non-linear regression (smoothed solid lines) with 95% CIs (smoothed dashed lines), and 2017 data were plotted as connected points. (B) Modeled Cx. tarsalis abundance from each IVM (red) and control (black) site over the season. Red lines indicate IVM treatment. (C) Average infection rate (MLE) from IVM and control sites. (D) Number of WNV-positive and WNV-negative pools at IVM and control sites is shown.