Summary
African horse sickness (AHS) is a transboundary and non-contagious
arboviral infectious disease of equids. Infected Culicoidesbiting midges can spread the African horse sickness virus, andCulicoides imicola (C.imicola ) is one of the important
transmission vectors. The disease has spread without any warning from
the sub-Saharan Africa towards the Southeast Asian countries. Therefore,
it is imperative to predict the distribution of the AHS infection risk
along the Sino–Southeast Asian borders. The reported AHS outbreaks were
extracted from the archive of the Food and Agriculture Organization from
December 22, 2005 to September 1, 2020. The occurrence records ofC.imicola were mainly obtained from published literature.
Subsequently, the maximum entropy algorithm was used to model AHS andC.imicola separately and to research the relationship among
bioclimate variables, land cover characterization, horse distribution
density, and the prevalence of AHS infection. Finally, we combined the
AHS risk prediction with the suitability map of C.imicola to
model the risk areas for AHS occurrence in Mainland China. The models
showed the mean area under the curve (AUC) as 0.935 and 0.910 for AHS
and C.imicola , respectively. Using jackknife analysis, we
determined the important factors affecting the AHS outbreak as horse
distribution density, mean temperature of the wettest quarter, and
precipitation of the coldest quarter. The mean temperature of coldest
quarter contributed most to the occurrence of C.imicola , followed
by precipitation of coldest quarter and global land cover
characterization. The overlay of the AHS and C.imicola prediction
map shows that the areas southwest of Hainan and southeast of Fujian are
at high risk of AHS occurrence under current conditions. Furthermore,
the border sectors of Yunnan and Guangxi also presented relatively high
risk.
Keywords: African Horse Sickness; Culicoides imicola ;
MaxEnt; Modelling