Several studies have revealed the positive effects of green space and certain activities on counteracting the physical and mental pressure felt by human beings. However, how self-reported stressed people perceive restorative green spaces, their preferences for specific characteristics and activities, and what characteristics of urban green space can induce various types of activities for stress recovery has not been fully examined in the high-density cities of China. Using an on-site questionnaire survey conducted in the People’s Park (PP) in Baoji, China, this study is the first to consider the relationship among eight sensory dimensions, activity types and stress recovery in Chinese green space. Results showed that the highest-stressed respondents were more likely to spend their time in multi-layered woodlands adjacent to water, with more experiences of serene but less about prospect. They preferred serene dimension more, while the culture and social dimensions were least preferred. Sports and leisure activities and quiet activities were the most popular among highest-stressed respondents, which were positively related to the serene and nature dimensions, respectively. Results suggested that the most restorative environment for stress recovery can be regarded as multi-layered woodlands adjacent to water with more serene and nature, less prospect and few or no culture and social dimensions.

Evidence has been accumulating of the psychological and physiological benefits and well-being gained by individuals from recreational activities in urban green spaces due to their ability to sustain biodiversity, but maximizing both biodiversity and recreational values of green spaces has become increasingly difficult in practice. In order to better maximize the biodiversity and recreational value of urban green space, this study was conducted through the utilization of an onsite questionnaire to understand people’s perceptions of and preferences for biodiversity and recreational values of urban green spaces in Baoji City, China. The results showed that respondents were able to correctly perceive biodiversity and preferred to engage in recreational activities in the high biodiversity environment. However, the respondents’ perceptions of the eight perceived sensory dimensions (PSDs) in the different habitats were divergent, and an environment which is high in biodiversity does not necessarily have higher PSDs. Moreover, ‘living environment’ and ‘frequency to the park’ had significant impacts on perception of and preference for urban biodiversity. In addition, ‘education level’, ‘living environment’, ‘age’, and other indicators were more likely to influence the perception of the eight PSDs. Therefore, the presented findings can be applied by urban landscape planners to assess the qualities of urban green spaces in order to maintain urban biodiversity and meet the satisfaction of human recreation in the future.