Highly pathogenic avian influenza (HPAI) is a severe, multispecies disease that endangers wild birds, poultry, and mammals, with devastating ecological and economic consequences. Recently, H5 HPAI 2.3.4.4 viruses have circulated globally, with subclade 2.3.4.4b undergoing reassortment, spreading beyond traditional waterfowls to seabirds and mammals, while 2.3.4.4h remained mainly confined to domestic poultry. Vaccination targeting 2.3.4.4b and 2.3.4.4h were implemented in China in 2022 and 2019, respectively. Understanding the factors shaping HPAI spread and evolution is essential for surveillance and control. Anseriformes (e.g., ducks, geese) differ from Galliformes (e.g., chickens, turkeys) in their immune responses and often act as asymptomatic carriers, facilitating undetected virus spread across regions and hosts. To determine the potential for unobserved HPAI transmission in China, we simulated duck trade networks and reservoir-to-human networks based on duck production data from extensive (free-range) and intensive systems, GPS tracking of duck transport, and infected wild bird distributions. Then using discrete trait phylogeographic analyses, we inferred the transmission dynamics of H5 HPAI subclades 2.3.4.4h and 2.3.4.4b in Anseriformes. We found the duck trade from extensive farms to intensive farms is significantly correlated with 2.3.4.4h transmission, while vaccination effectively reduced its spread. Contrastingly, the 2.3.4.4b clade transmission network is not significantly correlated with the simulated networks, suggesting more complex transmission drivers involving wild birds. Therefore, we apply phylogenetic generalized linear models (GLM) under an epoch model, integrating factors such as duck trade flows, genetic distance, wild bird abundance, road density, and land use to explore and quantify predictors influencing virus transmission.