Environment, vector, or host? Using machine learning to untangle the mechanisms driving arbovirus outbreaks

Climatic, landscape, and host features are critical components in shaping out-breaks of vector-borne diseases. However, the relationship between the outbreaks of vector-borne pathogens and their environmental drivers is typically complicated, nonlinear, and mayvary by taxonomic units below the species level (e.g., strain or serotype). Here, we aim tountangle how these complex forces shape the risk of outbreaks of Bluetongue virus (BTV); avector-borne pathogen that is continuously emerging and re-emerging across Europe, with sev-ere economic implications. We tested if the ecological predictors of BTV outbreak risk wereserotype-specific by examining the most prevalent serotypes recorded in Europe (1, 4, and 8).We used a robust machine learning (ML) pipeline and 23 relevant environmental features to fitpredictive models to 24,245 outbreaks reported in 25 European countries between 2000 and2019. Our ML models demonstrated high predictive performance for all BTV serotypes (accu-racies>0.87) and revealed strong nonlinear relationships between BTV outbreak risk andenvironmental and host features. Serotype-specific analysis suggests, however, that each of themajor serotypes (1, 4, and 8) had a unique outbreak risk profile. For example, temperature andmidge abundance were as the most important characteristics shaping serotype 1, whereas forserotype 4 goat density and temperature were more important. We were also able to identifystrong interactive effects between environmental and host characteristics that were also sero-type specific. Our ML pipeline was able to reveal more in-depth insights into the complex epi-demiology of BTVs and can guide policymakers in intervention strategies to help reduce theeconomic implications and social cost of this important pathogen.