Are they there, how many, and how big? Investigating potential trap biases in the surveillance of La Crosse virus vectors.

Abstract

Several methods of mosquito collection are used for the surveillance of the primary La Crosse virus (LACV) vectors, Aedes triseriatus (Say, 1823), Ae. albopictus (Skuse, 1895), and Ae. japonicus (Theobald, 1901). However, little is known about how the choice of collection method may confound inferences made from LACV vector surveillance data. Therefore, the objective of this study was to investigate potential biases in the surveillance of LACV vectors using the Biogents BG-Sentinel 2 (BGS), CDC-Light Trap (CDC-LT), Biogents Gravid Aedes Trap (BG-GAT), and standard oviposition cup (ovicup). The traps were deployed simultaneously at 10 sites in Knovxille, Tennessee, USA for 20 consecutive weeks. Surveillance results differed widely among the traps, demonstrating a strong potential for trap biases in LACV vector surveillance. The BGS and CDC-LT were effective for collecting Ae. albopictus but were not sensitive to the presence of Ae. triseriatus or Ae. japonicus. The ovicup was the best trap for detecting Ae. triseriatus, while the BG-GAT was the only trap that regularly collected Ae. japonicus. Surveillance conducted with the CDC-LT or BGS indicated that Ae. albopictus was dominant at all sites, but the ovicup and BG-GAT suggested a much larger relative abundance of Ae. triseriatus and Ae. japonicus, respectively. Aedes albopictus and Ae. triseriatus collected in the BG-GAT were significantly larger than those collected from the BGS and CDC-LT, indicating that the traps sampled different sub-populations. A multi-method surveillance approach is recommended to reduce potential biases when conducting surveillance of LACV vectors.