Discussion
Contrary to the proposed hypothesis of element homeostasis, we observed high flexibility in fungal C:P and even C:N ratios, reaching values far beyond common estimates of microbial stoichiometry (Cleveland & Liptzin 2007; Strickland & Rousk 2010), with maxima of 1488 and 126, respectively. Induced N and P limitations under controlled conditions reduced the relative amount of fungal N and P concentrations on average by 69 and 81%, respectively, causing wide fungal C:nutrient ratios, while increasing C availability in more natural substrate (i.e. SEA) allowed fungi to build up on average eight times more biomass despite strongly widening C:N and C:P ratios. These results show that soil fungi can adjust C:nutrient ratios much more flexibly than expected. Fungal N:P also showed variations, but the consistent parallel shift in N and P concentrations resulted in more homeostatic fungal N:P than C:nutrient ratios. This may be a general pattern in fungi: Gulis et al. (2017) also reported lower 1/H values for N:P than C:P in aquatic hyphomycetes, while Zhang and Elser (2017) realized that fungal N:P was closer to the canonical Redfield ratio than C:N and C:P average values (Redfield 1958). However, potential underlying mechanisms and general stoichiometric patterns revealed here have not been described before.