Can we predict where arboviruses will emerge?

Over the past decades, viruses transmitted by ticks and mosquitoes have expanded their distribution and caused an increasing number of human outbreaks. Spatial modelling of ecological risk factors for virus circulation can help identify areas of potential emergence.

The growing threat of arboviruses

Viruses transmitted by arthropod vectors, such as ticks and mosquitoes, are a growing public health concern in Europe. Some of these so-called arboviruses are endemic, i.e. they naturally occur in certain parts of Europe, such as West Nile virus (WNV), tick-borne encephalitis virus (TBEV), Crimean-Congo haemorrhagic fever virus (CCHFV), and louping ill virus (LIV).

The geographic range of these viruses has expanded over the past decades, and they increasingly cause disease in humans. For example, CCHFV did not occur in Turkey before 2002, but the country is now basically the epicenter of the disease. In 2018, a massive outbreak of WNV occurred in Southern and Central Europe. That same year, Germany experienced record-high numbers of TBE-cases.

Meanwhile, there is concern that exotic arboviruses, such as Japanese encephalitis virus (JEV) and Rift Valley fever virus (RVFV), might become established in Europe.

The Netherlands: a country at risk

Hazard map for the establishment of WNV in the Netherlands. Black circles indicate locations where antibodies against WNV in birds were detected. Source:

The Netherlands is considered particularly vulnerable to arbovirus emergence, owing to its high densities of livestock, great global connectivity in trade and travel, and local vector and wildlife species that can support and transmit arboviruses. In fact, the six arboviruses mentioned above have all been marked as top-priority for the Netherlands based on their epidemiology and societal and economic impact.

Surveillance of ticks, mosquitoes, and sentinel hosts is therefore part of ongoing efforts to detect these viruses as early as possible. Because such surveillance programs are time-consuming, expensive, and inherently test only a small proportion of vectors and hosts, it is important that they target areas where these arboviruses are most likely to emerge. However, identifying areas of potential emergence is a major challenge.

Using GIS to identify hotspots for surveillance

Our study aimed to determine which areas in the Netherlands have suitable ecological conditions for endemic circulation of the six aforementioned arboviruses (WNV, TBEV, LIV, CCHFV, JEV, RVFV). Arbovirus circulation depends on the presence of competent vector and host species as well as specific climatic conditions and suitable habitat.

Each of these is considered an ecological risk factor. Based on our earlier systematic review, we first identified which risk factors were relevant for the Netherlands (surprise: elevation was not). We then gathered nationwide continuous data on each risk factor, such as temperature, precipitation, vegetation cover, abundance of mosquitoes, livestock, and migratory birds, and used geographic information system (GIS-)based tools to characterize their spatial variation. Each risk factor represented a separate GIS-layer, which were then combined to create hazard maps for the introduction and/or establishment of each virus.

It’s important to stress that these maps portray relative hazard rather than actual hazard, i.e. arbovirus circulation is more likely in certain locations than in others.

Although there were considerable differences between the maps of each arbovirus, overlaying all of the establishment maps showed that overall, the environmental suitability for arbovirus circulation was highest in the southern part of the country; a region characterized by a warmer climate, which positively affects the abundance of vectors and their capacity to transmit viruses. This provides opportunities for targeted sampling of vectors and/or sentinel hosts in areas where multiple arboviruses may emergence, thereby increasing the efficient use of limited resources for surveillance.

Emergence of TBEV and WNV

It is almost ironic that while our study was ongoing, TBEV actually emerged in the Netherlands. Antibodies against the virus were initially detected in roe deer serum samples, after which TBEV-infected Ixodes ricinus ticks were collected from the same area. Human cases followed shortly after, and are now reported each year. Moreover, within a week after our study was published in Parasites & Vectors, WNV was detected in a common whitethroat (Curruca communis) and in Culex pipiens mosquitoes. Four weeks later, the first autochthonous human case was reported from the same area.

The emergence of TBEV and WNV in the Netherlands underlines the importance of surveillance programs for early detection and risk management, as both viruses were detected in wildlife sentinel hosts and arthropod vectors before human cases occurred. It also illustrates how difficult it is to predict where arboviruses will emerge.

Past models had predicted that TBEV would not become established in Northwestern Europe, as the region lacked the specific climatic conditions required for virus transmission. Yet there is now sustained circulation of TBEV in the Netherlands, and the virus has recently spread to the UK.

When we consider our maps, many of the TBEV-infected ticks, wildlife and humans have been reported from areas that were classified as relatively suitable for TBEV establishment. But there are also large, suitable areas that as of yet have no evidence of TBEV circulation. Only time will tell whether these are misclassifications or whether TBEV has simply not yet arrived.

Likewise, birds with antibodies against WNV were previously found in a region that is classified on our map as the most suitable for WNV establishment. However, the recent (and definitive) emergence of WNV in the Netherlands occurred just outside this region. Clearly, predicting where arboviruses will emerge remains a tricky business, but one that we can continue to improve on while we keep learning about arbovirus ecology.

View the latest posts on the BugBitten homepage