Identification of bio-climatic determinants and potential risk areas for Kyasanur forest disease in Southern India using MaxEnt modelling approach

Abstract

Abstract Background Kyasanur forest disease (KFD), known as monkey fever, was for the first time reported in 1957 from the Shivamogga district of Karnataka. But since 2011, it has been spreading to the neighbouring state of Kerala, Goa, Maharashtra, and Tamil Nadu. The disease is transmitted to humans, monkeys and by the infected bite of ticks Haemaphysalis spinigera. It is known that deforestation and ecological changes are the main reasons for KFD emergence, but the bio-climatic understanding and emerging pathways remain unknown. Methods The present study aims to understand the bio-climatic determinants of distribution of tick vector of KFD in southern India using the Maximum Entropy (MaxEnt) model. The analysis was done using 34 locations of Haemaphysalis spinigera occurrence and nineteen bio-climatic variables from WorldClim. Climatic variables contribution was assessed using the Jackknife test and mean AUC 0.859, indicating the model performs with very high accuracy. Results Most influential variables affecting the spatial distribution of Haemaphysalis spinigera were the average temperature of the warmest quarter (bio10, contributed 32.5%), average diurnal temperature range (bio2, contributed 21%), precipitation of wettest period (bio13, contributed 17.6%), and annual precipitation (bio12, contributed 11.1%). The highest probability of Haemaphysalis spinigera presence was found when the mean warmest quarter temperature ranged between 25.4 and 30 °C. The risk of availability of the tick increased noticeably when the mean diurnal temperature ranged between 8 and 10 °C. The tick also preferred habitat having an annual mean temperature (bio1) between 23 and 26.2 °C, mean temperature of the driest quarter (bio9) between 20 and 28 °C, and mean temperature of the wettest quarter (bio8) between 22.5 and 25 °C. Conclusions The results have established the relationship between bioclimatic variables and KFD tick distribution and mapped the potential areas for KFD in adjacent areas wherein surveillance for the disease is warranted for early preparedness before the occurrence of outbreaks etc. The modelling approach helps link bio-climatic variables with the present and predicted distribution of Haemaphysalis spinigera tick.