Metabolic traits

In February 2016, we measured growth rates of each species across gradients in temperature and phosphate concentration. Each species was grown in a factorial experiment at 5 temperatures and 13 phosphate concentrations, with 3 replicates per combination, yielding 1170 cultures (Fig. S2A). We created 13 solutions of COMBO medium with different phosphate concentrations ranging from 0.01 to 50 μmol PO₄ ³⁺ L⁻¹ (Table S2B), a range relevant to phosphate concentrations commonly found in lakes (Downing et al. 2001). Small tissue culture flasks (Nunclon) filled with 40 mL of each solution were inoculated with each species in monoculture at very low density (100 cells mL⁻¹) ensuring that the increase in phosphate concentration due to the inoculum volume was minimal (0.01 μmol L⁻¹). Cells were then grown at 15, 20, 25, 30 and 35 °C, and 90 μmol m⁻² s⁻¹ on a 12 : 12 light‐dark cycle. Samples were shaken and their position rotated within the incubators daily during the month‐long experiment. Every 2 days, 200 μL was taken and 10 μL of 1% sorbitol solution was added as a cryoprotectant. After 1 h of incubation in the dark, samples were frozen at −80 °C until further analysis. Cell density was determined by flow cytometry (BD Accuri C6) on fast flux settings (66 μL min⁻¹), counting 10 μL per sample. During the experiment, some samples failed to grow properly and were therefore removed from the subsequent analyses.