Abstract: Forest is the main component of terrestrial ecosystems that harbors about 40% of the existing species on the earth. As a vital component of biodiversity, phyllosphere microbes in the canopy play a critical and unique role in maintaining plant health, improving host resistance, and influencing global biogeochemical cycle. However, the studies on the community structure of phyllosphere fungi in natural forests are scarce as compared to that on rhizosphere microbes. Consequently, we know litter about how phyllosphere fungi associates with leaf traits. In this study, we analyzed fungal community composition of canopy leaves of six dominant tree species (i.e., Pinus koraiensis, Tilia amurensis, Quercus mongolica, Acer mono, Fraxinus mandshurica, and Ulmus japonica), in a broad-leaved Korean pine forest of Changbai Mountain Nature Reserve in Jilin Province, using high-throughput sequencing. We compared the differences of phyllosphere fungal community structure and functional groups of different dominant tree species. Moreover, 14 key leaf functional traits of their host trees were measured to investigate the relationships between fungal community composition and leaf functional traits. We found that the dominant phyla and class of phyllosphere fungi were Ascomycota and Basidiomycota, and Dothideomycetes and Taphrinomycetes, respectively. Results of LEfSe analysis indicated that all the tree species except Ulmus japonica had significant biomarkers, such as the Eurotiomycetes of Pinus koraiensis and the Ascomycetes of Quercus mongolica. The main functional groups of phyllosphere fungi were pathotroph. The results of redundancy and envfit analysis showed that functional traits related to plant nutrient acquisition as well as resistance to diseases and pests were the main factors influencing the community structure of phyllosphere fungi.

Key words: broad-leaved Korean pine forest, ITS high-throughput sequencing, community structure, functional prediction, leaf functional traits