Diversity and species composition of fungal endophytes in Avenella flexuosa under different sheep grazing regimes

Abstract

Fungal endophytes have been shown to be extremely widespread and abundant in terrestrial plants. It is also known that fungal endophytes may affect the grazing intensity of the host plant. In this study the occurrence of fungal endophytes in the perennial grass Avenella flexuosa was investigated under different grazing regimes in a field site in Norway. The main aim was to reveal to what degree the grazing influenced the fungal diversity and species composition. The fungal endophytes were analyzed using both culturing coupled to Sanger sequencing as well as 454 sequencing. The ITS2 rDNA region was used as a fungal DNA barcode. Based on the 454 data, a higher number of fungi were detected in stems of grazed compared to non-grazed plants. However, no significant difference was observed in the inflorescences and leaves. A distinct difference in fungal community composition was observed between leaves, stems and inflorescences. Apparently, the fungal community composition varied more in non-grazed compared to grazed plants. Saturation curve analyses indicated that most fungi remained undetected. Noteworthy, a high discrepancy was observed between the two methods, both in OTU number and overlap. From the 37 and 896 operational taxonomic units obtained from cultivation and 454 sequencing respectively. Only three OTUs were recovered by both methods. No known toxin producing endophytes were detected in A. flexuosa. Instead, Phaeosphaeria/Stagonospora spp. dominated both culture and 454 data, constituting potentially biological important endophytes of this grass. Intriguingly, various basidio-yeasts with high prevalence in the 454 dataset were not recovered by cultivation. This study indicates that grazing may have some influence on the diversity and composition of fungal endophytes. However, the different approaches underlines that the results must be interpreted with great care since there are many potential biases inherent both with the culturing and the 454 sequencing approach.