Effect of leaf morphology of perennial plants and planting pattern on PM2.5 reduction for urban forest

Urban forests play a crucial role in mitigating PM 2.5 pollution; however, the effectiveness of phytoremediation depends on the plant species and planting patterns. This study evaluated the PM 2.5 reduction efficiency of 22 perennial plants and investigated the influence of leaf morphology and planting configurations. Plants were classified into five groups based on their 24-h PM 2.5 reduction efficiency. Group high efficiency, including Cassia bakeriana , Lonicera frutescens , Wrightia religiosa , and Dalbergia cochinchinensis , exhibited the highest reduction efficiency, while Ficus annulata was the least effective. Diurnal variations were observed, with species like L. frutescens accumulating more PM 2.5 at night (24.082 ± 26.776 μg/m 2 ) and C, grandis showing exceptional daytime accumulation (113.405 ± 52.631 μg/m 2 ). PCA analysis highlights the positive influence of leaf count, stomata density, vein number, and trichome density and PM reduction, while stomatal length and width show negative correlations. Nighttime PM accumulation is associated with primary vein and abaxial trichome density, whereas daytime PM accumulation correlates more strongly with secondary vein and their ratio to primary veins. Multi-layered configurations, such as combining W. religiosa and C. bakeriana , achieved the highest reduction (40.5 ± 5.0%), whereas single-layer designs were less effective (6.2 ± 1.4%). Dense arrangements with taller plants in the second layer outperformed sparse layouts. These findings underscore the importance of selecting high-efficiency species with favorable leaf traits and optimizing planting designs to enhance urban forest phytoremediation. Integrating species such as C. bakeriana and W. religiosa in multi-tiered systems can maximize PM 2.5 mitigation, providing practical solutions for improving urban air quality.