Identifying inter-epizootic transmission routes of Rift Valley fever virus in Tanzania to inform targeted control strategies for outbreak response

Globally, land transformation is proceeding rapidly from natural to agricultural and along a rural-urban continuum. The expansion of agricultural areas and increasing density of human settlements have consequences for disease risk. Land use change alters conditions for livestock, wild animals, and insect populations in addition to the pathogens they transmit, creating new opportunities for exposure of people to animal pathogens.

One exemplar disease is Rift valley fever (RVF) in Africa. The causative virus (RVFV) is transmitted primarily between cattle, sheep, and goats by mosquitoes. Humans can become infected either by handling an infected animal or from the bite of an infectious mosquito.

Large epidemics of RVF occur when heavy rains cause floods leading to the creation of extensive mosquito larval habitats. Between epidemics, RVFV was thought to be maintained only by female mosquitoes passing virus to their offspring. However, there is emerging evidence that mosquito density is continually sufficient for transmission to livestock between epidemics, importantly in peri-urban, as well as rural areas. Transmission to livestock between epidemics has consequences for the frequency of localised outbreaks, the sources of pathogen spread at the start of an epidemic, and consequently risk to humans in space and time. In peri-urban settings, human activities create a mosaic of human settlements, livestock and vector habitats that are juxtaposed in ways that do not occur in rural or urban locations.

Our hypothesis is that peri-urbanisation and crop cultivation is creating suitable conditions for new patterns of RVFV transmission, with implications for outbreak risk and control. To address this hypothesis, we will quantify how the numbers, diversity and feeding behaviour of mosquito vectors varies between peri-urban and rural areas, and between grassland and crop habitats, through field studies in northern Tanzania. The data produced will be used within models of RVFV transmission, alongside data on: i) livestock density, movement and turnover; and ii) variation in vectors species' ability to transmit RVFV, to determine how peri-urban settings and crop cultivation are affecting transmission routes for RVFV.

We will validate model predictions of where inter-epidemic transmission risk is highest through testing livestock for antibodies against RVFV. The resulting model, developed and validated with field data, will be used to simulate epidemics and determine how spread, outbreak size and risk to humans is influenced by the identified transmission routes in peri-urban and rural areas prior to the start of the outbreak. Through simulations of outbreaks we will then identify optimal interventions and investigate the effect of targeting these interventions to high risk areas. Simulated interventions will include livestock vaccination and movement bans in addition to vector control using indoor residual spraying or insecticide-treated livestock.

Through stakeholder engagement, we will also outline a list of ecological interventions, or 'barriers' that may be more sustainable in the long-term. We will use the models to establish the effect that these barriers would need to have on either: reducing vector numbers, contacts between vectors and livestock, or contacts between vectors and humans to reduce pathogen transmission and disease risk to humans. These barriers may include introduction of mosquito larval predators in irrigation systems, the use of nets for livestock while housed, and human housing adjustments (window screens, eave tubes).

Our project will generate knowledge to enable policy makers in Tanzania to develop rational strategies to monitor and control RVFV and identify areas at risk of RVFV. Our whole-system approach will provide a framework for other mosquito-borne pathogens of humans and livestock.