Urban green spaces are important hotspots for urban biodiversity and primary recreational areas for residents. Understanding the spatiotemporal characteristics of soundscapes and their influencing factors is of significant importance for biodiversity conservation and soundscape management. The earlier soundscape researchs primarily focused on the soundscape characteristics and their differential analysis on forests, natural reserves, and other natural vegetation areas. However, the human interferences present within urban green spaces may lead to spatiotemporal characteristics of soundscapes that differ from those of natural vegetation areas. In particular, there is insufficient understanding regarding the soundscape characteristics among different vegetation types under highly interfered urban environment, as well as the intrinsic relationships among these soundscapes, the surrounding environment, and human interferences. In this study, three different vegetation types, namely secondary forest, plantation forest, and lawn with woodland within the Baiyun Mountain Scenic Area in Guangzhou were selected for a 28 days soundscape monitoring. Spatiotemporal characteristics of soundscapes in green space were revealed based on six soundscape indices and soundscape power. Additionally, the impacts of environmental factors and human interferences on the soundscape patterns were investigated using a random forest model and redundancy analysis. Results showed that the variation pattern of the Acoustic Complexity Index (ACI) significantly reflected the dawn (6:00-8:00) chorus phenomenon of birds. The acoustic signal intensity decreased from plantation forest and secondary forest and then to lawn with woodland, indicating that appropriate human interferences enhanced the vocalization intensity of birds, while excessive human interferences restricted their vocalizations. The Normalized Difference Soundscape Index (NDSI) showed that the daytime soundscapes of the secondary forest and plantation forest were dominated by biophony, while lawn with woodland was dominated by anthrophony. The dominant frequency bands and their influencing factors varied significantly among different green spaces. The secondary forest was mainly dominated by mid-frequency biophony (4-8kHz), primarily influenced by altitude (ALT), shrub richness (SR), and tree height diversity (THD). Plantation forest mainly comprised low-frequency biophony (2-4kHz) and high-frequency biophony (8-11kHz), with soundscape power positively correlated with slope (SLO) and negatively correlated with distance to hardened path (DHP). Lawn with woodland featured more anthrophony (1-2kHz), with the distance to hardened main road (DHMR), distance to hardened path (DHP), and distance to non-hardened path (DNHP) all having a positive impact on it. This study reveals the spatiotemporal characteristics and main influencing factors of soundscapes in different vegetation types, providing scientific references for the design of healthy soundscapes and biodiversity conservation in urban green spaces.