Seed collection, growth conditions and experimental setups
We collected seeds from Taraxacum officinale populations in places where the species is native (Europe) or introduced (South America). Populations in the native range were París, Lyon and Marsella (France) while those of the introduced range were Antofagasta (23° 37´ S), Caldera (27° 06´ S), La Serena (29° 54´ S), Viña del Mar (33° 01´ S), Talca (35° 27´ S), Concepción (36° 57´ S), Valdivia (39° 80´ S), Coyhaique (45° 61´ S) and Punta Arenas (53° 10´ S). Four to five seeds were collected per individual (N = 60 maternal plants) and pooled by population. For each population, 200 seeds were germinated on wet paper in Petri dishes at room conditions (24 ± 4 ºC). Those plants were grown in individual pots (3 L) filled with commercial soil in a greenhouse and watered on demand. This first generation (F1) of plants from each population were grown until they produced seeds, which were used to obtain the experimental F2 plants.
Besides the collection of seeds from each population, we also collected rhizospheric soil at each site. We dig the soil around the roots of 50T. officinale plants and carefully transferred it to a common bag for each site. For the experiments, T. officinale plants were randomly assigned to three treatments: “native soil microbiome (M+)”, “removed soil microbiome (M-)”, and “re-inoculated soil microbiome (Mr)”. The M+ treatment consisted in groups of 500 cc plastic pots (one seedling per pot) filled with a mix of native rhizospheric soil and sand in a 2:1 proportion. For the M treatment, pots were filled with the same native rhizospheric soil but previously sterilized by autoclaving (121 °C, 90 min.). Lastly, pots in the Mr treatment were initially the same as those in M treatment that were later irrigated with 100 mL of a solution containing the native soil microorganisms (i.e., 200 gr of native soil diluted in 1,0 L of tap-water). Autoclaving eliminated soil microorganisms, allowing us to study the effect of local microbiomes on the fitness-related traits inT. officinale plants. Since autoclaving can also alter the availability of soil nutrients (de Deyn et al. 2004), we evaluated the content of nitrogen (N) and phosphorus (P) in autoclaved and non-autoclaved soils (n = 12 each). No differences were found in the availability of the two nutrients (t = 0.71; p = 0.26 andt = 0.97; p = 0.41).
One week after appearance of the first true leaf, the seedlings in the plastic pots were transferred to a glasshouse under semi-controlled conditions (air temperature: 26 ± 3ºC; relative humidity: 75 ± 3%; and solar radiation: 1160 ± 80 µmol m-2s-1). Each pot was irrigated with 75 mL of tap-water (or tap-water with diluted rhizospheric soil in Mr) every two days. In each experiment, the pot positions were weekly randomized within the glasshouse. Inter-pot distance was sufficient to prevent mutual shading influence. The two experiments lasted 100 days and in both, plant survival as well as flower production were recorded every 5 days. At the end of the experiment, we recorded total dry biomass (including fallen leaves) of plants. The plant material was oven-dried at 70 °C for 72 h.
In experiment 1, we studied the effect of local and introduced soil microbiome on the performance of T. officinale plants from populations of both the native (Europe) and the introduced (South America) range. This experiment was carried out simultaneously in Europe (52º 12´N; 00º 04´E, Cambridge, UK) and in South America (35º 24´S; 71º 37´W, Talca, Chile). Greenhouse conditions in the UK were similar to those in Talca. We used F2 seeds from Paris, Lyon, and Marseille to represent the native range, while those from La Serena, Talca, and Punta Arenas were selected to represent the introduced range. In each location, a total of 225 seedlings per population were established where 75 was the number of plants randomly assigned to any of the soil microbiome treatments [control (M+), sterilized (M) or sterilized and reinoculated (Mr)]. To compare the effect of the microbiome on the fitness-related traits of T. officinale between native and introduced range,
In experiment 2, we studied the effect of the soil microbiome on the performance of T. officinale plants from populations collected along a latitudinal gradient in the introduced range (Antofagasta, Caldera, La Serena, Viña del Mar, Talca, Concepción, Valdivia, Coyhaique and Punta Arenas) (Suppl. Table 1; Suppl. Figure 1). This experiment was carried out in Talca, Chile. A total of 90 seedlings per population were established where 30 were randomly assigned to any of the soil microbiome treatments [control (M+), sterilized (M) or sterilized and reinoculated (Mr)]. Apart from the common procedures indicated above, in this experiment the individual performance variables were integrated through the calculation of a “Performance Index”. For each plant, we summed the scaled values (i.e. , between 0 and 1) of their three recorded response variables (survival, biomass and flower number), after being weighted by their proportional contribution to the fitness. The proposed weighting values correspond to 0.5 for survival, 0.3 for flower production, and 0.2 for biomass. For example, a performance index of 1 corresponds to a plant with the highest performance in all variables, while lower index values indicate lower performance. It is worth mentioning that in the case of survival, each individual value corresponds to the proportion of survivors within their respective experimental group at the end of the experiment.