Results
The common garden experiment showed that the soil microbiome modulated fitness-related variables, particularly aboveground biomass of T. officinale plants (Suppl. Table 2). Nevertheless, this effect was more pronounced in T. officinale plants growing in non-native South American soils, where individuals from both origins exhibited a significant reduction in biomass when the soil microbiome was removed (M+ vs. M). Specifically, plants from the native range (Europe) that grew in South American M soil accumulated less biomass and produced fewer flower heads than their counterparts growing in the untreated soil (M+). Plants from the introduced range (South America) that grew in M South American soil not only accumulated less biomass and produced fewer flowers but also exhibited reduced survival (Fig. 1). Conversely, when T. officinale plants were grown on native soils, only the biomass of the South American plants appeared to be affected by the soil microbial removal; in all other cases, the loss of the soil microbiome did not result in significant performance responses, either for European or South American plants (Fig. 1). In all cases in which soil microbiome removal affected the performance of T. officinale , the reinoculation treatment successfully restored the previously lost performance in the plants (Mr vs. M+; Fig. 1). There was also a subtle but significant genotype effect (i.e. , associated with origin) on the aboveground biomass of T. officinale . Irrespective of the soil treatment, the South American plants presented on average, greater biomass than their European counterparts (Fig. 1).
The soil microbiome influenced the performance of T. officinaleplants along the latitudinal gradient in their invasive range (Fig. 2). While plants grown in control (M+) and reinoculated (Mr) soils showed similar responses in terms of performance variables along the latitudinal gradient, the performance of plants in sterilized soil (M) deviated from the referred pattern (Fig. 2a). Across the entire latitudinal gradient, survival of plants was high in M+ and Mr treatments but low in the Mtreatment. Additionally, only under this latter treatment, survival rates were lower at the extremes of the gradient compared to the center (Suppl. Table 3). Overall, there was a positive association between plant aboveground biomass and latitude, increasing from south to north (Fig. 2); nonetheless, the biomass of plants in the northern extreme was lower when grown M soil (Fig. 2b). The number of flowers per plant was higher in the middle of the latitudinal gradient and lower at both extreme south and north (Fig. 2c). However, despite this overall pattern, the removal of the soil microbiome was associated with a significant reduction in the number of flowers per plant across the entire south-north distributional range (Fig. 2c). The same pattern of responses to variation in latitude and soil microbiome treatment was observed in integrated index of plant performance (Fig. 2d).