Effects of environmental drivers in field populations
Trait variation for in situ field populations was associated with both environmental drivers and biogeographic range (native vs. non-native) in the best models, although their effects did not interact in most cases (see blue lines in Fig. 5; Table S4; Fig. S7-S9). Biomass was positively correlated with Vegetation Cover and Mowing, and was higher in non-native populations (Fig. 5a). SLA did not differ between ranges, although we found an interaction between Temperature and biogeographic range, whereby SLA decreased with increasing Temperature in the native but not the non-native range (Fig. 5b). Biomass-corrected probability of flowering was affected negatively by Vegetation Cover and positively by Mowing, and was lower in non-native populations (Fig. 5c). Biomass-corrected fecundity was positively affected by Aridity and Temperature, and the effect of Aridity was stronger in the native than the non-native range (Fig. 5d). When biomass was excluded as a covariate from the analyses, the best model of probability of flowering lost the effects of Vegetation Cover and Range, and the best model of fecundity incorporated the effects of Mowing and the interaction between Range and several source drivers (Table S8). Non-native populations showed significantly higher temperature and seasonality of moisture index than native populations, and lower values in moisture index (Table S9).