Context-dependent dominance: Pollution sources vs. tree species in shaping leaf-deposited PM characteristics

Urban particulate matter (PM) pollution poses a significant threat to human health and urban ecosystems. While urban trees are increasingly promoted as a nature-based solution through leaf surface deposition. However, a fundamental uncertainty remains as to when leaf-deposited PM characteristics are governed primarily by pollution source heterogeneity and when they are governed by tree species traits. This knowledge gap limits the context-specific design of green infrastructure for targeted PM removal. Here, we combined a PM resuspension chamber (PMRC) with a single-particle aerosol mass spectrometer (SPAMS) and applied an ART-2a neural-network-based source apportionment to characterize the size distribution and chemical composition of leaf-deposited PM on two common urban tree species across three contrasting pollution environments. To move beyond a qualitative "who dominates" description, we further introduced a quantitative, distance-based context dominance index (CDI) that compares inter-environment and inter-species dissimilarities in characteristics of leaf-deposited PM using hierarchical clustering. The results show that under highly heterogeneous pollution sources, traffic-dust-dominated site imprint their ambient PM signatures onto leaf surfaces irrespective of species identity, resulting in convergence of leaf-deposited PM_0.5-1.0 and enrichment of EC/OC and levoglucosan markers. In contrast, when pollution sources are relatively homogeneous, species-specific traits become the primary regulators: Sabina chinensis selectively accumulates coarse and metal-rich particles, whereas Cinnamomum camphora preferentially retains ultrafine and levoglucosan-rich particles. The CDI objectively captures this switching from source-dominant to species-dominant regimes. These findings illustrate that the dominance of pollution sources versus tree species in shaping leaf-deposited PM is context dependent and can be quantified. The results underscore the necessity of considering both pollution regimes and species-specific traits for strategic selection to maximize the air quality benefits of urban ecosystem.