Deep DNA sequencing of fungal communities in the Andean Yungas reveals high diversity and strong structuring among altitudinal forest types

JOZSEF GEML; NICOLÁS PASTOR; LISANDRO FERNANDEZ; SILVIA PACHECO; TATIANA SEMENOVA; EDUARDO R. NOUHRA

Freising

Congreso; Annual Conference of the Society for Tropical Ecology (gtö) - Tropical Ecosystems - Between Protection and Production; 2014

The Yungas, a system of subtropical montane forests on the eastern slopes of the Andes, are extremely diverse and severely threatened by anthropogenic pressure and climate change. Previous mycological works in the Argentine Yungas focused on macrofungi (e.g., agarics, polypores) and mycorrhizae in Alnus acuminata forests, while fungal diversity in most of the Yungas has remained mostly unexplored. We carried out Ion Torrent sequencing of ITS2 rDNA from soil samples taken at 24 sites along the entire latitudinal extent of the Yungas in Argentina (appr. S 22.2?27.4°). The sampled sites represent the three altitudinal forest types: the piedmont (400?700 m asl), montane (700?1500 m asl), and montane cloud (1500?3000 m asl) forests. Based on our deep sequence data (i.e. analyzing 51 264 reads per sample following quality filtering and subsampling to equal sequencing depth), we delimited 14 039 non-singleton 97% sequence similarity fungal operational taxonomic units (OTUs). NMDS analyses suggested that fungal community composition correlated most strongly with altitudinal forest type (P=0.00002), explaining ca. 70% of the observed variation, with many OTUs showing strong preference for a certain altitudinal zone. For example, ectomycorrhizal and root endophytic fungi were far more diverse in the montane cloud forests, particularly at sites dominated by Alnus acuminata, while the diversity values of various saprobic fungal groups were highest at lower elevations. Despite the strong altitudinal structuring, fungal diversity were comparable across the different zonal forest types. In addition, fungal community structure correlated with latitude in a somewhat weaker, but still significant (P=0.0007) manner that explained ca. 17% of the variation. Besides elevation, soil pH, N, P, and organic matter contents strongly correlated with fungal community structure as well, although most of these variables were co-correlated with elevation and forest type. Our data offer an unprecedented insight into the high diversity and spatial distribution of fungi in the Yungas forests.