In the context of tropical monsoonal dwarf forest restoration, it is well known that the interaction between soil properties and plant functional traits influences the dynamics and forest composition of plant communities. The main aim of this study was to determine the relationships and variations between the plant functional traits and environmental variables in tropical coastal secondary forests. However, it is undisputed whether trait variation is coordinated and whether there is a relationship between the community-weighted mean (CWM) traits and soil variables. TNNR conducted a ground survey to collect actual ground data on the biophysical characteristics of individual trees and shrubs, along with soil sample data. All soil samples and plant materials were collected at the end of June, in the 2020 growing season (July and August), in each quadrat with a 400-m2 area. To gain insight into these questions, a total of 4 plant functional traits of 167 species and 5 soil nutrients from 128 soil samples were analyzed from 128 plots in a tropical evergreen monsoon forest on Hainan Island. Using SMA (standardized major axis), wood density (WD), and leaf thickness (LT), as well as specific leaf area (SLA) and relative leaf water content (RLWC), were significantly positively correlated. Using Pearson correlation analysis, leaf thickness (LT) was significantly positively correlated with soil organic matter (OM). Using principal component analysis (PCA), we also found that soil TN and OM were the strongest predictors for functional traits, i.e., WD and LT. Moreover, using path analysis, OM and TN have a major impact on plant CWM traits, e.g., SLA and RLWC. Our results show a significant relationship between functional traits and soil pH and soil nutrients in tropical coastal secondary forests. Our results highlight that plant traits can be used to predict specific soil nutrients and ecosystem functioning in tropical secondary forests, but we are concerned about how variation in the physical structure of plant affect ecosystem function in forest communities. This research can help us to better understand the restoration of habitats and green infrastructure design, suggesting that selecting different species across multiple trait axes can help ensure functionality at the maximum level.