Modelling the distribution of bluetongue vectors.

Recent epizootics of Culicoides-borne disease in the Mediterranean Basin have stimulated the development of climate-driven models for vectors. Predictor variables come from two main sources, weather data and satellites. Generally, models for Culicoides imicola combine temperature and moisture variables. The best weather models explain 75-85% of the variance in observed data for C. imicola, but satellite models perform better (85-95% of variance). Predictions of models for other regions appear mixed, with successes and failures. The failures indicate the need to: explore and incorporate other factors that may affect Culicoides populations, such as soil characteristics, host type and wind speed. Develop more complex models, recognising that different climate variables affect different stages of the life-cycle e.g. biological models. The very rapid spread in the distribution of C. imicola in recent years suggests that global warming may be a less important driver of change than other, currently unknown, factors.