Quantifying and modelling the acquisition and retention of lumpy skin disease virus by haematophagus insects reveals clinically but not subclinically-affected cattle are promoters of viral transmission and key targets for control of disease outbreaks.

Lumpy skin disease virus (LSDV), a poxvirus that causes severe disease in cattle, has in the last few years rapidly extended its distribution from Africa and the Middle East into Europe, Russia, and across Asia. LSDV is believed to be primarily spread mechanically by blood-feeding arthropods, however the exact mode of arthropod transmission, the relative ability of different arthropod species to acquire and retain the virus, as well as their comparative importance for LSDV transmission, remain poorly characterised. Since the vector-borne nature of LSDV transmission is believed to have enabled the rapid geographic expansion of this virus, the lack of quantitative evidence on LSDV transmission has impeded effective control of the disease during the current epidemic. Obtaining high quality data on virus transmission by arthropods is challenging, and practical limitations often result in inadequate arthropod numbers or model hosts, limiting the transferability of experimental findings to the natural transmission scenario. We have addressed these limitations in this study. Using a highly representative bovine experimental model of lumpy skin disease we allowed four representative vector species (Aedes aegypti, Culex quinquefasciatus, Stomoxys calcitrans and Culicoides nubeculosus) to blood-feed on LSDV-inoculated cattle in order to examine the acquisition and retention of LSDV by these species in unprecedented detail. We found the probability of LSDV transmission from clinical cattle to vector correlated with disease severity. Subclinical disease was more common than clinical disease in the inoculated cattle, however the probability of vectors acquiring LSDV from subclinical animals was very low. All four potential vector species studied had a similar rate of acquisition of LSDV after feeding on the host, but Aedes aegypti and Stomoxys calcitrans retained the virus for a longer time, up to 8 days. There was no evidence of virus replication in the vector, consistent with mechanical rather than biological transmission. The parameters obtained in the in-vivo transmission experiments subsequently enabled enhanced modelling approaches to determine the basic reproduction number of LSDV in cattle mediated by each of the insect species. This was highest for Stomoxys calcitrans (19.1), C. nubeculosus (7.4), and Ae. aegypti (2.4), surprisingly indicating these three species are all potentially efficient transmitters of LSDV. These results reveal that currently applied LSDV control measures such as stamping out of all cattle on affected premises or insect control measures targeting single species need to be urgently reconsidered. Overall our studies have highlighted that the combination of highly relevant in-vivo experiments and mathematical modelling can be directly applied to devise evidence-based proportionate and targeted control programmes.