Fungal biomass in soils represents a dynamic pool of carbon (C) and nutrients, but its incorporation into ecosystem C models has been hindered by uncertainty regarding its decomposition rate and limited understanding of the chemical constituents affecting that rate. To assess the generality of empirical decomposition models in predicting fungal necromass mass loss, we decomposed 28 different fungal residues from 23 species of mycorrhizal and saprotrophic fungi in laboratory microcosms over a 90-day period, comparing a single-pool exponential decomposition model with a two-pool asymptotic model. We also characterized the initial C chemistry of these tissues using fiber analysis methods commonly used for plant tissues along with Fourier transform infrared (FTIR) spectroscopy. Additionally, we quantified melanin and nitrogen (N) concentrations of each substrate, as these have previously been shown to impact decay rates of fungal necromass.
Results/Conclusions
We found that the two-pool asymptotic decomposition model better described decomposition than the single-pool exponential model in all cases. The cell-soluble fraction of the initial necromass was the best predictor of the labile pool (R2=0.56, P<0.001), followed initial N content (R2=0.19, P=0.02). Interestingly, these two predictors were unrelated to each other, suggesting they exert independent control over decomposition rates. The size of the slow decomposing pool was best predicted by the size of the acid non-hydrolyzable carbon fraction of the initial substrate (R2=0.29, P<0.01), which was in turn positively correlated with aliphatic bonds and melanin-associated aromatics. Collectively, our results indicate that the decomposition of fungal tissues in soils can be generally described in having two distinct stages driven by the C chemistry of the substrate.