Ecological niche modeling to estimate the distribution of Japanese encephalitis virus in Asia

Abstract

Background: Culex tritaeniorhynchus is the primary vector of Japanese encephalitis virus (JEV), a leading cause of encephalitis in Asia. JEV is transmitted in an enzootic cycle involving large wading birds as the reservoirs and swine as amplifying hosts. The development of a JEV vaccine reduced the number of JE cases in regions with comprehensive childhood vaccination programs, such as in Japan and the Republic of Korea. However, the lack of vaccine programs or insufficient coverage of populations in other endemic countries leaves many people susceptible to JEV. The aim of this study was to predict the distribution of Culex tritaeniorhynchus using ecological niche modeling.

Methods/principal findings: An ecological niche model was constructed using the Maxent program to map the areas with suitable environmental conditions for the Cx. tritaeniorhynchus vector. Program input consisted of environmental data (temperature, elevation, rainfall) and known locations of vector presence resulting from an extensive literature search and records from MosquitoMap. The statistically significant Maxent model of the estimated probability of Cx. tritaeniorhynchus presence showed that the mean temperatures of the wettest quarter had the greatest impact on the model. Further, the majority of human Japanese encephalitis (JE) cases were located in regions with higher estimated probability of Cx. tritaeniorhynchus presence.

Conclusions/significance: Our ecological niche model of the estimated probability of Cx. tritaeniorhynchus presence provides a framework for better allocation of vector control resources, particularly in locations where JEV vaccinations are unavailable. Furthermore, this model provides estimates of vector probability that could improve vector surveillance programs and JE control efforts.

Figure 1. Japanese encephalitis virus endemic area.

Map adapted from CDC.

Figure 2. Distribution of known Cx. tritaeniorhynchus locations and documented human cases of JE within endemic region.

Figure 3. Maxent model estimation of the probability of Cx. tritaeniorhynchus distribution in the JE endemic region.

Darker areas indicate areas that are likely to have suitable habitat for this vector species while lighter areas indicate areas of that are less suitable for the vector.

Figure 4. Human JE cases categorized by color based on the estimated probability of Cx. tritaeniorhynchus presence.

Figure 5. Percent of 30 meter pixels classified as rice land cover within 1 one square kilometer derived from the GeoCover Land Cover product.

Gray areas indicate no data.