Leveraging non-traditional vector surveillance data for reliable species distribution modeling: Insights from Common house mosquito Culex pipiens in Europe

This is the authors' abstract. We don't add key points for this paper.

Gaps in health surveillance systems expose the shortage of real-time data required for public health decision-making. Non-traditional data (NTD) from citizen science can help bridge these gaps, but such data are often affected by heterogeneous sampling effort and spatiotemporal biases. This study investigates how NTD from the Mosquito Alert citizen science platform can be used to accurately model the distribution of the common house mosquito, Culex pipiens , a competent vector for West Nile virus across Europe. We address the inherent sampling biases of opportunistic data by applying thermal niche filters and using Mosquito Alert sampling effort as a proxy to generate reliable inferred-absences . Our findings demonstrate that citizen observations outdoors yield the most ecologically consistent predictions when validated against independent trap data, whereas indoor observations reflect the host-seeking and shelter-seeking behavior of mosquitoes combined with more opportunistic sampling by citizens. By accounting for spatial and temporal heterogeneity, we provide a methodological framework for integrating citizen-contributed images into public health surveillance. Ultimately, the research proves that carefully filtered citizen science data can enhance the resolution of species distribution models for critical disease vectors.