The most common symbiosis of plants in terrestrial ecosystems is the arbuscular mycorrhiza (AM), which influences plant nutrition, resistance, diversity, and also productivity in the vast majority of ecosystems.

Only about 200 AM fungal species are described, usually mainly based on the characters of their asexual spores, but recently also by means of molecular phylogenetic data. Species recognition is complicated by the fact that many AM fungi (AMF) form spores with different morphotypes, and that spore characters useful for identification are very limited for many taxa. Obviously, the low species number is in contrast to the ubiquity and the age of this fungal group. Molecular ecological studies indicate that at least ten-times more species exist, but these cannot yet be identified, and the species level could usually not be resolved. Therefore, an important challenge is to establish reliable molecular identification tools and DNA barcodes for these ecologically (and economically) important fungi.

The aim of this study was to test if the ITS rDNA region can be reliably used for the discrimination of AMF species. The background is, that up to now it is unclear whether the high intraspecific variability (also present within one multinucleated spore) allows the use of the ITS region as a DNA barcode for this fungal group.

Results/Conclusions

We show that the ITS region is suited for AMF DNA barcoding and can also be used for ecological studies. We followed a cloning and sequencing (after RFLP) approach, to cover and define the intraspecific variability from AMF from the Diversisporacea, Ambisporaceae, and Glomeraceae, also comprising frequently used "model species". Moreover, we also sequenced the LSU and SSU rRNA genes. Based on the ITS-region it was possible to separate all investigated AMF species. No overlap of the sequences between different species was observed, and analyses of environmental data show that several (unknown) species are hidden behind the phylotype definitions often used in molecular ecological studies. Our results demonstrate that, after a partial characterisation of intraspecific variability, the ITS-region is well suited to discriminate AMF species, also from field studies.