Trait-based approaches for predicting mosquito distributions under environmental change

Mosquito-borne diseases cause substantial mortality and morbidity worldwide and are changing in distribution and impact due to environmental change. Invasive mosquitoes have become widely established across Europe this century, with subsequent outbreaks of dengue and chikungunya virus. Species distribution models (SDM) are widely used to inform policy responses to these threats, for example to understand how invasive mosquito vectors might spread following arrival and where vectors might overlap with key hosts to permit transmission. However, insect vector distributions are most commonly modelled species-by-species even though novel statistical community modelling methods are now available. For other insect taxa, a mechanistic understanding of likely environmental change impacts has been gained by modelling and comparing distributional responses simultaneously across multiple species with different ecological traits. The availability of new continental-scale distribution data, of environmental predictors appropriate to mosquito habitats and statistical methods for dealing with species interactions, dispersal limitation and biased recording effort, mean we can now use similar trait-based approaches to understand and predict mosquito responses to environmental change.
The aim of this project is to investigate the role of ecological traits, invasion status, and environmental factors in constraining the distribution and seasonality of insect vectors in temperate environments. The student will develop and apply novel statistical joint distribution models for communities across Europe, combining distribution and seasonality data from VectorNet (a European data-sharing Network, funded by ECDC and EFSA) with a traits database, to analyse whether species responses are linked to broad ecological traits like breeding site and host preferences and whether these responses vary with scale and invasion status.
The student will then generate virtual species distributions with similar traits and invasion status and range of responses to environmental drivers to those of European mosquitoes, sampled with biased and unbiased recording effort. The resulting distributions will be analysed using alternative statistical single-species and community modelling algorithms to understand how distribution modelling for insect vectors might best account for dispersal
limitation and biased recording effort. Finally the student will analyse the impact of mis-specifying mosquito distributions on the utility of such models for policy and in predictions of where mosquito-borne transmission will occur.