The role of wild birds in transmitting highly pathogenic avian influenza in Denmark: An exploration using a spatiotemporal model

We developed a stochastic spatiotemporal simulation model, DanHPAIwild, to explore the transmission dynamics of highly pathogenic avian influenza (HPAI) in wild bird populations and generate a time-dependent risk map in Denmark. Denmark and its near-coast waters were rasterised into 10 x 10 km cells. Weekly changes in abundance of five waterbird species over an epidemiological year (October to September) were estimated by adjusting missing reports and then comparing the estimates with raw population numbers. The model consisted of two parts: bird abundance and environmental transmission. Dynamics of bird abundance were modelled using voluntarily reported weekly bird counts, published literature, and expert opinion on bird ecology. Environmental transmission simulated HPAI virus (HPAIV) exposure via consumption of contaminated water, together with viral shedding in the rasterised cells. We calibrated and initialised the model with passive surveillance data from the 2020/21 season, refining inputs to minimise deviations between simulated and observed outcomes. Sensitivity analyses revealed that variations in the median infection dose and peak time of contacts notably influencing the simulated bird mortality. The model also examined removing infectious dead birds as a potential control strategy, showing that removing 50 % reduced the annual HPAIV mortality by 38 % compared to implementing no interventions. The model reflects well the real-life spatiotemporal patterns, highlighting highrisk areas close to coastline and water areas. These findings offer valuable insights into HPAIV dynamics in wild birds in Denmark and can aid resource allocation for interventions and surveillance.