2.1 Experimental design
In summer 2017, we set up a plant-plant interaction (PPI) experiment and a plant-soil interaction (PSI) experiment with three common central European grassland forb species (Geranium pratense L.,Leucanthemum vulgare (Vaill.) Lam., and Ranunculus acrisL.). We chose these species based on their representation in the Trait-Based Experiment of the Jena Experiment (Ebeling et al., 2014),i.e. , monocultures of each species, all two-species mixtures, and the three-species mixture were established (see below). Prior to each experiment, we germinated seedlings of each species from non-sterilized seeds (Rieger-Hofmann GmbH, Blaufelden-Raboldshausen, Germany). To assure that we would use similarly developed seedlings and to account for species-specific differences in germination, we treated the seeds as follows: all seeds of Geranium pratense were gently scarified with sandpaper, placed in a petri dish, and treated with 3 mL 1 g/L gibberellic acid for 24 h at 7°C. The same procedure was followed forRanunculus acris seeds, but they were treated with 0.66 g/L gibberellic acid. No treatment was necessary for Leucanthemum vulgare seeds. Following the treatment, all seeds were transferred to plastic boxes half-filled with glass beads (50 seeds per box, only one species per box). Each box was covered with a transparent lid, and seeds were watered daily with tap water. All boxes were transferred to growth chambers (CLF Plant Climatics, Percival E-36L, Wertingen, Germany) with a photoperiod of 16 h light at 20°C and 8 h darkness at 12°C, and 50% relative humidity. Seeds of Geranium pratense andLeucanthemum vulgare were left in the growth chamber for 14 days, and those of Ranunculus acris for 28 days, until the seedlings reached similar sizes.