Introduction
Ecological stoichiometry represents an important field originally developed in aquatic ecosystems, with its principles being now also applied to terrestrial soil systems (Sterner & Elser 2002). Models use stoichiometric theory for predicting soil nutrient cycles and carbon storage (Sardans et al. 2012). Based on the general observation that C:N:P (carbon:nitrogen:phosphorus) ratios of soil microbial communities are more narrow than resource C:N:P, critical ratios of C:N and C:P are defined to predict nutrient demands of soil microbes, mineralization versus immobilization patterns and C sequestration versus respiration (Manzoni et al. 2012; Zechmeister-Boltensternet al. 2015). Thus, ecological stoichiometry facilitates the incorporation of extremely complex processes into global models and predictions, especially in the context of global change (Hall et al. 2011a).
These models make crucial assumptions that are critical for their validity (Mooshammer et al. 2014; Spohn 2016), one of them that heterotrophic soil microbes are homeostatic, i.e. maintain stable C:N:P ratios independently of the soil nutrient status (Persson et al.2010). Even though some models allow microbial C:N:P to vary slightly (McGill et al. 1981; Nicolardot et al. 2001), homeostatic flexibility is currently only attributed to microbial community shifts, not an actual stoichiometric flexibility in individuals (Buchkowskiet al. 2019). Fixed microbial stoichiometric ratios are interpreted as an indicator of nutrient demands: If microbes need to maintain their narrow C:N ratios, N will be in limiting supply in substrates characterized by wider C:N ratios (Manzoni et al.2010). These assumptions of soil microbial homeostasis are supported by analyses of entire soil communities, which indeed show relatively little variation in microbial C:N:P ratios compared to soil resource variability (Cleveland & Liptzin 2007; Hartman & Richardson 2013). By contrast, we know astonishingly little about the stoichiometry of individual soil microbial groups.
Indeed, flexibility of C:N:P ratios in individual microbial species has rarely been analyzed, and the few studies available mainly use aquatic isolates and partly result in contradictory results depending on the methods applied (Danger et al. 2016). Surprisingly, some of these studies suggest that microbial C:N:P ratios may be less homeostatic than assumed for heterotrophic organisms. For aquatic bacteria Scott et al. (2012) and Godwin and Cotner (2018) demonstrated high variation in C:P ratios for some but not all isolates (see also Makino et al.2003; Danger et al. 2008). In case of saprobic fungi, aquatic hyphomycetes responded to varying element supply with non-homeostatic adjustments in C:P ratios, while C:N remained stable (Danger & Chauvet 2013; Gulis et al. 2017). In soils there are also indications that fungal C:N:P may exceed common textbook assumptions of C:N ~10-20 and C:P ~100-300 (Jennings 1995; Strickland & Rousk 2010). A recent meta-analysis demonstrated wide C:N and C:P ratios in few fungal samples (Zhang & Elser 2017), as also reported for C:N in wood decomposing fungi under low N conditions (Levi & Cowling 1969), while other studies indicate again only little flexibility (Heck 1928; Egli & Quayle 1986; Mouginot et al.2014).
Physiological mechanisms causing non-homeostasis in microbes are still unknown. Most authors assume P storage, i.e. element uptake in excess as the exclusive mechanism (Scott et al. 2012; Danger & Chauvet 2013; Mooshammer et al. 2014), though this only explains stoichiometric shifts in response to high element supply. The significance of such stoichiometric shifts for fungal growth and activity also remain unresolved. Available data show that N or P uptake can be independent of fungal growth, potentially related to storage mechanisms or primary C limitations (Levi & Cowling 1969; Guliset al. 2017).
In order to understand stoichiometric adjustments in soil saprotrophic fungi in detail, we analyzed fungal mycelial element concentrations in response to varying N, P and C supply in different growth media specifically developed for this question (varying from highly controlled to natural substrates). Derived from common assumptions in soil ecological stoichiometry, we tested the hypothesis that saprobic fungi are homeostatic, especially regarding their C:N ratios. Our results not only further question the general assumption of microbial homeostasis, but also provide insights into the physiological mechanisms of element allocation in mycelia, and the significance of stoichiometric shifts for fungal growth and activity under varying conditions.