Revisiting soil fungal biomarkers and conversion factors: Interspecific variability in phospholipid fatty acids, ergosterol and rDNA copy numbers

Refined conversion factors for soil fungal biomarkers are proposed. High interspecific variability is present in all fungal biomarkers. A modeling approach supports the validity of biomarker estimates in diverse soils. ITS1 copies vary strongly, but are fungal-specific with least phylogenetic bias. A combination of fungal biomarkers will reveal soil fungal physiology and activity. The abundances of fungi and bacteria in soil are used as simple predictors for carbon dynamics, and represent widely available microbial traits. Soil biomarkers serve as quantitative estimates of these microbial groups, though not quantifying microbial biomass per se. The accurate conversion to microbial carbon pools, and an understanding of its comparability among soils is therefore needed. We refined conversion factors for classical fungal biomarkers, and evaluated the application of quantitative PCR (qPCR, rDNA copies) as a biomarker for soil fungi. Based on biomarker contents in pure fungal cultures of 30 isolates tested here, combined with comparable published datasets, we propose average conversion factors of 95.3 g fungal C g −1 ergosterol, 32.0 mg fungal C µmol −1 PLFA 18:2ω6,9 and 0.264 pg fungal C ITS1 DNA copy −1 . As expected, interspecific variability was most pronounced in rDNA copies, though qPCR results showed the least phylogenetic bias. A modeling approach based on exemplary agricultural soils further supported the hypothesis that high diversity in soil buffers against biomarker variability, whereas also phylogenetic biases impact the accuracy of comparisons in biomarker estimates. Our analyses suggest that qPCR results cover the fungal community in soil best, though with a variability only partly offset in highly diverse soils. PLFA 18:2ω6,9 and ergosterol represent accurate biomarkers to quantify Ascomycota and Basidiomycota . To conclude, the ecological interpretation and coverage of biomarker data prior to their application in global models is important, where the combination of different biomarkers may be most insightful.