Recent European heatwaves have significantly impacted forest ecosystems, leading to increased plant water stress. Advances in land surface models aim to improve the representation of vegetation drought responses by incorporating plant hydraulics into the plant functional type (PFT) classification system. However, reliance on PFTs may inadequately capture the diverse plant hydraulic traits (PHTs), potentially biasing transpiration and vegetation water stress representations. The detection of vegetation drought stress is further complicated by the mixing of different tree species and forest patches. This study uses the Community Land Model version 5.0 to simulate an experimental mixed-forest catchment with configurations representing standalone, patched mixed, and fully mixed forests. Biome-generic, PFT-specific, or species-specific PHTs are employed. Results emphasize the crucial role of accurately representing mixed forests in reproducing observed vegetation water stress and transpiration fluxes for both broadleaf and needleleaf tree species. The dominant vegetation fraction is a key determinant, influencing aggregated vegetation response patterns. Segregation level in PHT parameterizations shapes differences between observed and simulated transpiration fluxes. Simulated root water potential emerges as a potential metric for detecting vegetation stress periods. However, the model’s plant hydraulic system has limitations in reproducing the long-term effects of extreme weather events on needleleaf tree species. These findings highlight the complexity of modeling mixed forests and underscore the need for improved representation of plant diversity in land surface models to enhance the understanding of vegetation water stress under changing climate conditions.