The combination of abundance and infection rates of Culicoides sonorensis estimates risk of subsequent bluetongue virus infection of sentinel cattle on California dairy farms

Abstract

Bluetongue (BT) is an important viral disease of ruminants that is transmitted by hematophagous Culicoides midges. We examined the seasonal patterns of abundance and infection of Culicoides sonorensis at four dairy farms in the northern Central Valley of California to develop estimates of risk for bluetongue virus (BTV) transmission to cattle at each farm. These four farms were selected because of their similar meteorological conditions but varying levels of vector abundance and BTV infection of cattle. C. sonorensis midges were collected weekly at each farm during the seasonal transmission period, using three different trapping methods: traps baited with either carbon dioxide (CO2) alone or traps with CO2 and UV light, and by direct aspiration of midges from sentinel cattle. Analysis of BTV-infected midges using group and serotype-specific quantitative reverse-transcriptase polymerase chain reaction (RT-qPCR) assays confirmed that BTV serotypes 10, 11, 13 and 17 are all present in the region, but that midge infection rates and the number of BTV serotypes circulating differed markedly among the individual farms. Furthermore, more serotypes of BTV were present in midges than in sentinel cattle at individual farms where BTV circulated, and the virus was detected at each farm in midges prior to detection in cattle. BTV infection rates were remarkably lower among female C. sonorensis midges collected by CO2 traps with UV light than among midges collected by either animal-baited aspirations or in CO2 traps without light. A subsample of female midges examined from each collection method showed no overall differences in the proportion of female midges that had previously fed on a host. Findings from this study confirm the importance of using sensitive surveillance methods for both midge collection and virus detection in epidemiological studies of BTV infection, which is especially critical if the data are to be used for development of mathematical models to predict the occurrence of BTV infection of livestock.