Aim: Alien species successfully colonize new ranges if they encounter favourable environmental conditions there and possess traits that match new challenges. Climate matching approaches comparing native and exotic ranges mostly consider temperature and precipitation niches of alien species, but have largely ignored UV-B radiation. UV-B fundamentally differs between hemispheres, with much higher levels at southern than at northern latitudes. Consequently, UV-B might act at the global scale and present a so far neglected filter that species need to overcome when invading high-UV-B environments.

Location: We performed two multi-species common-garden experiments, conducted in the native European range (Germany) and the high-UV-B exotic range (New Zealand) to test for preadaptation to UV B.

Methods: We used 25 herbaceous species from open habitats, which we exposed in each range to three UV radiation treatments: (i) natural sunlight, (ii) exclusion of UV-B while allowing natural UV-A, and (iii) exclusion of UV-B and UV-A. We additionally used plant traits (TRY), global distribution data (GBIF, GloNAF) and global UV-B satellite data (glUV) to determine species-specific characteristics as fostering agents of UV-B tolerance. The joint analysis of experimental and macroecological data allowed quantification of species plasticity and identification of beneficial species traits in high-UV-B environments.

Results: Our results showed an overall limiting effect of UV-B in both common gardens but the UV-stress response tended to be more pronounced in the invaded range. Across all species, we found little evidence for preadaptation by functional plant traits. In contrast, preadaptation to climatic conditions related to the species’ native UV-B niche was of greater importance for plant performance in the presence of UV-B radiation.

Main conclusions: For predicting alien species’ ability to expand into high-UV-B environments, macroclimatic niche characteristics of the species’ native range might be better predictors than functional traits and should be more considered in future projection models.