Ornamentals gone wild? Trait clues from urban non-native plants

Le résumé fourni par la source

It is widely known that the ornamental plant trade is the primary introduction pathway for non-native plants to new areas. Cities are key hubs for introduction and spread, where non-native plants—with their charisma and tolerance to urban stressors—are often a dominant component of the urban greenery. Here, non-native plants provide important ecosystem services. However, their widespread planting may be risky when species are not screened for their likelihood to escape cultivation and potential invasiveness. Overlooking this aspect further exacerbates the role of cities in spreading non-native species. Regulations and good practices can prevent the planting of well-known invasive non-native species. However, preventing the establishment and spread of potentially new invaders introduced for urban greenery remains challenging. Determining what drives invasion success remains a major issue in invasion ecology. Invasion success is frequently assumed to be associated with one or more functional traits of non-native species (Gioria et al. 2023). Accordingly, trait-based approaches may provide a useful framework to screen potential future invaders, depending on the context and data availability. Here, we categorized ornamental non-native trees and shrubs cultivated in seven Italian cities (Milan, Turin, Rome, Campobasso, Asti, Pavia, Palermo) into four categories: “cultivated-only”, “casual”, “naturalized” and “invasive”. We analyzed these groups to: characterize functional trait space and main drivers of dissimilarity; assess functional overlap; and detect potential emerging invaders among cultivated taxa functionally similar to invasive and naturalized species. characterize functional trait space and main drivers of dissimilarity; assess functional overlap; and detect potential emerging invaders among cultivated taxa functionally similar to invasive and naturalized species. Trait data were mainly obtained from the ATENA database (Varricchione et al. 2026). Functional traits were combined with ecological variables and reproductive strategies. Twelve of twenty-two variables were retained after distance-based redundancy analysis. We explored functional structure using Non-metric Multidimensional Scaling (NMDS) based on Gower distances. Trait–environment relationships were tested using the “envfit” function (999 permutations). Functional diversity patterns were quantified through functional richness and beta-diversity metrics within the multidimensional trait space. Results showed that species were distributed along gradients mainly driven by structural traits (plant height, leaf economics) and climatic variables (NMDS1), and secondarily by reproductive strategies (NMDS2). Functional differentiation was primarily associated with asexual regeneration, wind dispersal, and key structural traits. Invasive and naturalized species did not form distinct clusters, instead occupying a subset of the broader functional space defined by cultivated-only taxa. Cultivated-only species largely shaped the outer boundaries of this space, although a subset overlapped closely with invasive and naturalized taxa, identifying potential candidates for future invasion risk evaluations. Overall, results indicate a substantial functional overlap among cultivated-only, naturalized, and invasive species, confirming that invasion success is not strictly linked to unique trait syndromes. However, the identification of cultivated species functionally close to invasive and naturalized taxa highlights the usefulness of trait-based screening as an early warning tool. Rather than providing deterministic predictions, this approach effectively flags species deserving closer attention. As such, urban greenery can be a reservoir of functionally pre-adapted species. Integrating trait-based screening with biogeographical and management information may improve early warning strategies aimed at reducing future plant invasions in urban systems.

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.