Assessing Near-Naturalness of PlantCommunities in Urban Green Spaces UsingComprehensive Index and Random ForestMethods: A Case Study in Hefei, China

As urbanization rapidly progresses, the ecological resilience and health of urban parks and green spaces face significant threats, underscoring the crucial need for effective ecological quality assessments.This study developed a comprehensive evaluation system, uniquely designed with sixteen indicators across four dimensions -community composition, structure, succession, and disturbance -to systematically assess the near-naturalness of 87 urban park plant communities in central Hefei and to explore the influencing factors.Employing comprehensive index and random forest algorithms, we find the near-naturalness index ranging from 0.135 to 0.752, categorizing only 18.39% of samples as 'nearnatural' or 'semi-natural,' while a majority, 81.61%, fell into 'far-natural' or 'artificial' categories.Ideal near-natural conditions emerged in settings with 5 to 10 dead plants, litter cover of 10 to 30%, 10 to 15 species, and mild natural disturbances.Our findings indicate that the overall near-naturalness in Hefei's urban green spaces is moderate to low, suggesting a need for focused enhancements in community succession and structural adjustments.This research establishes a novel framework for advancing urban green space ecological planning and sustainable development, providing fresh insights and a robust scientific basis for enhancing the ecological quality of urban parks and green spaces.