Structural Features of Root-Associated Fungus-Plant Interaction Networks in the Tropical Montane Rain Forest of Jianfengling, China

Siqi Yang,Qi Zhang,Xiqiang Song,Jian Wang,Yide Li,Han Xu,Shouyu Guo,Qiong Ding
DOI: https://doi.org/10.17520/biods.2018339
2019-01-01
Biodiversity Science
Abstract:: Functionally diverse root-associated fungi may differentially interact with host plants, potentially affecting the assembly processes of belowground plant and fungi communities. Here, we applied the Illumina Miseq sequencing technique to identify root-associated fungi of plants which were co-dominant in a tropical montane rain forest on Hainan Island, China. Structural features of bipartite networks were compared among the whole root-associated, arbuscular mycorrhizal (AM) and ectomycorrhizal (ECM) fungus–plant interactions. A total of 297,831 fungal ITS1 sequences were obtained from eight families including Aceraceae, Annonaceae, Apocynaceae, Aquifoliaceae, Arecaceae, Fagaceae, Lauraceae, and Oleaceae. Fungal sequences were assigned to 1,279 OTUs comprised of Ascomycota (748 OTUs), Basidiomycota (354), Glomeromycota (80), and unidentified fungi (97). At least three functional groups of fungi i.e. putatively ECM (40 OTUs, represented 23.4% of the total fungal reads), AM (40, 13.9%) and saprophyte (83, 19.8%) were prevalent in the core root-associated fungal community (420 OTUs) of the tropical montane rain forest. Network analysis indicated that AM, ECM and root-associated fungus–plant interaction network showed structural features which cannot be predicate by null models assuming species interact randomly. Community level indices behaved differently among different ecotypes of fungus–plant interactions. Specifically, the degree of nestedness (NODF) and connectance were higher, while specialization was lower in the AM interaction network than the expected values from null models. In contrast, the ECM interaction network had a significantly higher degree of specialization and lower nestedness and connectance than the null models. At guild levels, plant niche overlap of AM and ECM interactions are higher and lower than the null model, respectively. Niche breadth of ECM fungi was narrower than that of AM fungi. Co-occurrence patterns of plant and fungus further confirmed competition for resources was intense in ECM interaction network (high C-score of both plants and fungi) and weak in the AM interaction network (low C-score). These findings suggest that at least two modes of interspecific interactions are critical for the assembly and coexistence of root-associated fungal communities, i.e. redundancy (nestedness) of AM interactions, and niche differen-tiation (specialization) of ECM interactions. Here we provide a comprehensive exploration of the interactions among functionally diverse root-associated fungal guild within a forest, which is key to understand the mechanisms maintaining species coexistence in tropical forests.
What problem does this paper attempt to address?