New insights into the spread of West Nile virus in the Netherlands

West Nile virus (WNV) circulates between mosquitoes and birds and can also infect humans and horses. Understanding where and when the virus can spread is challenging. Transmission depends on interactions between mosquitoes, different bird species, pathogens and environmental conditions – and not all of these processes can be observed directly.

In her PhD research at Wageningen University & Research, Mariken de Wit investigated how data-driven mathematical models can help uncover these hidden drivers of transmission. She studied both WNV and the closely related Usutu virus (USUV), which share important aspects of their ecology, transmission cycles and surveillance.

The research started with a systematic review of mathematical models used to study WNV transmission. These models have become increasingly sophisticated and now incorporate factors such as multiple host species, host movement, and environmental conditions.

At the same time, an important challenge remains: models are only as useful as the biological and epidemiological data on which they are based. The research therefore highlights the importance of closer collaboration between field and laboratory researchers and mathematical modellers. “Combining their expertise can improve both the realism of transmission models and their value for surveillance and disease control,” according to De Wit.

One of the challenges in understanding WNV is determining which bird species contribute to virus circulation. Research into the emergence of USUV in the Netherlands demonstrates why this matters. When USUV spread across the Netherlands, it caused substantial mortality among Eurasian blackbirds. Yet transmission models calibrated with multiple surveillance datasets showed that blackbirds alone could not explain the observed patterns. “Our models suggest that other bird populations may play an important role in spreading Usutu virus, even though they are much less visible in surveillance data. Compared with blackbirds, this unidentified population appears to experience less mortality following infection, lives longer and potentially disperses over greater distances.”

The finding illustrates a broader challenge for surveillance of mosquito-borne viruses: species that attract relatively little attention may nevertheless influence how a virus spreads. Understanding the role of different host species can help improve the design of early-warning and surveillance systems for viruses such as USUV and WNV.

What could happen as the Dutch climate and landscape change? To explore this, transmission models for USUV and WNV were combined with future Dutch One Health scenarios incorporating climate and land-use change.

Across all scenarios, outbreaks were projected to become larger, last longer and grow more rapidly. Rising temperatures were the main driver behind the increased transmission potential. Changes in the local distribution of hosts and mosquitoes further influenced regional patterns.

“These findings show why preparedness for mosquito-borne diseases requires looking beyond the virus itself. Climate, mosquitoes, wildlife and the environment are interconnected and together determine where opportunities for transmission arise,” concludes the PhD student.

The simultaneous circulation of WNV and USUV also creates challenges for surveillance, explains De Wit. “Because the viruses are closely related, antibodies produced after infection can cross-react in laboratory tests. This can make it difficult to determine which virus caused a previous infection.”

Using neutralisation test data from blackbirds, the research developed a Bayesian model to estimate virus circulation and antibody responses simultaneously. The analysis revealed substantial differences in antibody responses to USUV and WNV and indicated that seroprevalence for both viruses was probably higher than previously estimated.

“Based on these findings, we developed a model-informed classification method that can improve the interpretation of serological surveillance data when the two viruses co-circulate.”

The research demonstrates how mathematical modelling can bring together information from wildlife surveillance, laboratory testing and environmental data to reveal epidemiological processes that cannot be observed directly.

For WNV, these insights are particularly relevant as the virus continues to expand geographically in Europe. “Better understanding which hosts are involved, how environmental change affects transmission and how surveillance results should be interpreted can contribute to earlier detection and more robust risk assessments,” says De Wit.

The approach developed in this PhD research therefore goes beyond WNV and USUV. “My research shows how data-driven modelling can help prepare for emerging mosquito-borne diseases in a changing environment – connecting animal, human and environmental health through a One Health approach,” concludes De Wit.

Mariken de Wit defended her PhD thesis Hiding in plain sight – Revealing transmission dynamics through data-driven modelling of Usutu and West Nile virus at Wageningen University & Research on 15 October, 2026. A recording of the defence is available online (Hiding in plain sight – Revealing transmission dynamics through data-driven modelling of Usutu and West Nile virus | WUR).

Read the thesis online