Using contact networks and next-generation sequencing for wildlife epidemiology

Abstract

Horizontal transmission of infectious diseases is strongly determined by the contact network between hosts [1] . Environmental and ecological variables that define the probability of contacts between individuals, such as food resource quality and quantity, have been largely neglected in models studying the dynamics and epidemiology of wildlife diseases. In this work, we used pollinators and their pathogens to study the role of contact networks in a natural multi-host pathogen community. In agricultural landscapes, wild flowers are the most important food resources of insect pollinators such as bees and flies. To boost yields of agricultural crops and ensure pollinator conservation, the UK Environmental Stewardship scheme prompted farmers to grow pollinator-friendly wild flower margins along their fields ( Fig. 1 ). Wild flower margins ( Fig. 1 ) have proven to be efficient to increase both density and diversity of bees in agricultural areas, and to increase pollination success of surrounding crops. As the rate of disease transmission should increase with host density, we hypothesize that the success of wild flower margins may generate hubs for pathogen exchange within the bee community. Several recent studies have illustrated the frequent transmission of infectious diseases between managed and wild bees, potentially via the shared use of flowers [2] . However, the environmental and ecological drivers of disease dynamics between pollinators remain uncharacterised [3] . To understand the role of flower density and diversity for bee disease transmission, we reconstructed high-resolution plant–insect visitor networks from flower visitation data collected in ten farms in Southern England (five farms participating in the scheme vs. five control farms), as well as a record of bee pollen collection to describe the resource bees were exploiting (pollen vs. nectar). When analysing the plant–pollinator networks, we found that flower density and diversity strongly define pollinator density and foraging behaviour, and influence the structure of indirect ‘contact networks’ among pollinators (via the use of flower species). Particularly, we found flower diversity to be positively correlated with the reduction of niche overlap between insect species. This response of insect pollinators suggest that wild flower margins with high plant species richness may reduce insect competition for resources, and potentially reduce the risk of inter-specific disease transmission by supporting diverse diet for insects exploiting different flowers. To test the effect of plant diversity on pathogen dynamics in bees, we sampled pollinators on these farms and characterized their virome by deep transcriptome sequencing. We are now combining these environmental data to virus discovery in order to reveal the impact of the agri-environmental scheme on viral dynamics. Ultimately, we aim to identify environmental (flower density, agricultural practices) and ecological factors (plant taxa, insect community assemblage) that significantly enhance the transmission of plant and pollinator viral diseases within our model to eventually improve agricultural practises and wildlife management.