Germination and growing conditions
We first performed whole plant greenhouse experiments. We usedMedicago truncatula as a host (genotype Jemalong A17, courtesy of dr. Bettina Hause, Leibniz Institute of Plant Biochemistry, Halle, Germany), as previously (Kiers et al. 2011; Whiteside et al. 2019). We scarified M. truncatula seeds by submerging them in 95% sulfuric acid for 5-10 minutes, after which we rinsed the seeds with excess ddH2O. We stored the seeds in petri dishes on moist filters, first two days in the dark at 5 ˚C, then one day at 20 ˚C in the dark, followed by two days in the light at 20 ˚C. We planted the germinated seeds in autoclaved germination soil. After 11 days, we selected seedlings of 3-4 cm with at least three leaves to transplant to three-compartment boxes with a 6 L capacity (Garcia et al.2006). The boxes were divided into three equal compartments with a 50 µm pore size nylon mesh (Cell Micro Sieves, Gentaur) glued onto a PVC window. This limited the plant roots to the outermost compartments, but allowed the fungal hyphae to grow in the central fungus-only compartment. We filled each compartment with autoclaved quartz sand and supplemented the central fungus-only compartment with 1 g hydroxyapatite per kilogram quartz sand as a phosphorus source (Pel et al.2018). We planted oneM. truncatula seedling in each of the two outer compartments of the three-compartment box.
As fungal inoculum, we homogenized in-vitro Ri T-DNA Daucus carota L. transformed root organ cultures containing each fungus and added 16 mL of the mixture to the roots (~700 spores). We randomly assigned one plant as the “focal plant”. This focal plant was consistently inoculated with the strain ‘A5’. The other root was designated as the partner plant, and inoculated with either A5 (n =8), B12 (n =9) or Agg (n =9) (Fig. 1a). After inoculation, we added 10 mL water to the roots and we fertilized the plants once with an adjusted Hoagland solution with 25% phosphorus (5.5 mM KNO3; 4.0 mM CaCl2.2H2O; 7.25 mM NH4NO3; 0.5 mM KH2PO4; 1.0 mM 20 mM MgSO4.6H2O; Fe(Na)EDTA; 1.0 mL/L micronutrients). We placed all experimental units in a randomized grid, and we rotated them every two weeks to avoid local effects. After a week, we covered the sand with a one cm layer of sterile low-density polyethylene beads (Fardem Packaging, Edam, the Netherlands) to limit evaporation. Temperature in the greenhouse fluctuated between 20 ˚C and 30 ˚C. We watered the plants twice a week with dH2O keeping the water content between 10 and 12.5 % and fertilized the plants every fourth day with 35 mL of an adjusted Hoagland solution, containing no phosphorus but extra nitrogen (5.5 mM KNO3; 4.0 mM CaCl2.2H2O; 7.25 mM NH4NO3; 0.5 mM KCl; 1.0 mM MgSO4.6H2O; 20 mM Fe(Na)EDTA; 1.0 mL/L micronutrients). To confirm that our three fungal strains did not differ significantly in their mutualistic quality, we also grew single M. truncatula plants in standard pots of 880 mL, filled with sterile quartz sand. We inoculated host plants with either A5, B12, or Agg, and grew and fertilized plants as above. We found no statistically significant difference in either root or shoot biomass of plants grown with our three strains, confirming that they did not significantly differ in their nutrient provisioning strategy (Fig. S3).