BACKGROUND: Yeast-like fungi inhabit soils throughout all climatic zones in a great abundance. While recent estimations
predicted a plethora of prokaryotic taxa in one gram of soil, similar data are lacking for fungi, especially yeasts.
METHODOLOGY/PRINCIPAL FINDINGS: We assessed the diversity of soil yeasts in different forests of central Germany using
cultivation-based techniques with subsequent identification based on rDNA sequence data. Based on experiments using
various pre-cultivation sample treatment and different cultivation media we obtained the highest number of yeasts by
analysing mixed soil samples with a single nutrient-rich medium. Additionally, several species richness estimators were
applied to incidence-based data of 165 samples. All of them predicted a similar range of yeast diversity, namely 14 to 16
species. Randomized species richness curves reached saturation in all applied estimators, thus indicating that the majority of
species is detected after approximately 30 to 50 samples analysed.
CONCLUSIONS/SIGNIFICANCE: In this study we demonstrate that robust species identification as well as mathematical
approaches are essential to reliably estimate the sampling effort needed to describe soil yeast communities. This approach
has great potential for optimisation of cultivation techniques and allows high throughput analysis in the future.