Objective criteria to assess representativity of soil fungal community profiles

Soil fungal community structures are often highly heterogeneous even among samples taken from small field plots. Sample pooling is widely used in order to overcome this heterogeneity, however, no objective criteria have yet been defined on how to determine the number of samples to be pooled for representatively profiling a field plot. In the present study PCR/RFLP and T-RFLP analysis of fungal 18S rDNA in ten soil samples obtained from a grassland plot of 400 m(2) also revealed this known heterogeneity in fungal community structures. Based on these data a three-step approach to assess representativity of fungal community profiles was established. First, soil DNA quantities needed for robust community profiling were determined. Second, profiles of single or multiple samples were theoretically averaged to test for statistically significant clustering in order to determine the minimal number of samples to be pooled to achieve representativity. Third, DNA extracts of single or multiple samples were pooled prior to profiling in order to test for statistically significant clustering. Analyses revealed robust profiles for 50 ng soil DNA but not for 5 ng. Averaged T-RFLP profiles from five or more soil samples and experimental T-RFLP profiles from pools of seven or more samples formed one significant branch. Theoretical averaging and experimental pooling revealed that five to seven samples have to be pooled for robustly representing the field plot. Our results demonstrate that representativity of soil fungal community profiles can objectively be determined for a field plot with only little deviation between theoretical and experimental approaches. This three-step approach will be of assistance for designing sampling and pooling strategies for comparative analyses of soil fungal communities in ecological studies.