Nature-based Solutions (NbS) are recognized as a strategic approach to improve environmental conditions and human well-being by the provision of multiple ecosystem services. Specifically, urban vegetation can improve air quality by removing atmospheric particulate matter (PM) through adsorption processes. However, not all species are equally effective, as their capacity to capture and retain PM mainly depends on the morpho-functional characteristics of their leaves. Despite the importance of this ecological function in urban areas, the understanding of the role of leaf traits remains challenging.

To fill this gap and address species selection for future management of urban areas, this study investigates: i) the performance of different tree functional types (i.e. conifers, evergreen and deciduous broadleaves) in adsorbing PM, and ii) the role of leaf traits involved in the process. To this end, leaf sampling from 10 different tree species were collected in Ferrara (Northern Italy) over two campaigns (winter and summer seasons). The vacuum filtration method was used to estimate the amount of PM₁₀ and PM_2.5 in the samples. The following functional traits were measured for broadleaf species: Length/Width (L/W) ratio, petiole length, wettability, wax content, stomata density and trichome density. Multiple regression with background selection was performed to select statistically significant leaf traits in removing PM.

Conifers were found to be more efficient at removing atmospheric PM (mean PM₁₀ = 38.5 μg cm⁻², PM_2.5 = 24.6 μg cm⁻²) than broadleaf trees (PM₁₀ = 7.1 μg cm⁻², PM_2.5 = 4 μg cm⁻²), because of high L/W ratio and wax concentrations. Among leaf traits, L/W ratio, petiole length and stomata density were found to significantly influence PM₁₀ adsorption.

These results can inform urban green management strategies and reforestation efforts by selecting tree species with morphological characteristics favourable to PM removal, thereby contributing to improve air quality in highly-polluted urban areas.