The elevational patterns of biological communities have long been a central focus in ecology. Compared with biodiversity distribution of plants, animals, and belowground microorganisms, few studies concentrated on the interplay between above- and below-ground biotic communities. In forests, multispecies association networks resulting from the woody plants and mycorrhizal fungi interactions contribute to plant growth and productivity. Inconsistent elevational patterns between plants and fungi contribute to the complexity of mycorrhizal associations, and remains poorly understood. In this study, an oak-dominated forest (1020—1770 m, 10 forest transects) in Dongling Mountain, Beijing was used as a model system.
Along altitude, soil fungi alpha diversity showed a linearly increasing pattern in an oak dominated forest. Diversity of the dominant soil fungal groups was mainly affected by soil bulk density, pH and available phosphorus content, while fungal species richness and diversity of rare taxa were affected by soil sand content and herb richness. Soil fungi showed a significant Mantel correlation with elevation and herb composition.Partial Mantel test suggested that herb composition was a significant factor influencing soil fungal community. Beta diversity of herbs and soil fungi did not show significantly elevational patterns.
Unlike soil fungi, alpha diversity of Quercus wutaishansea root fungi did not show a significantly elevational pattern. Yet, significantly lower fungal beta diversity in Quercus wutaishansea roots was observed than that in soil pools. Two-way ANOVA suggested significant interactions between beta diversity difference and elevation,indicating that the selective intensity of Quercus wutaishansea in choosing partners from soil fungal pools could be circumstance dependent.
The overall mycorrhizal association network, and transect-based overall mycorrhizal association network exhibited significant nested web structures, with NODFAll values being 68.715 and 44.508 respectively. In the two mycorrhizal association networks, cumulative distribution of plants and mycorrhizal fungi all conformed to the truncated power law. In the mid-domain T5 transect, local mycorrhizal association network is composed of 16 plants and 271 mycorrhizal fungi, with a connectance of 0.3700. NODFfungi, NODFplant, and NODFAll values of the local network were 57.271, 68.539 and 57.308 respectively. Interactive intensity of species within a trophic level was quite different. Cumulative distribution of host plants followed an exponential curve, while mycorrhizal fungi were conformed to the truncated power law distribution.
Along attitude, significantly specialized patterns in all 10 plant-symbiotic fungi networks, with higher network specialization (H2’) emerging at mid-domain and treeline adjacent transects. Host plant richness and crown density, rather than soil properties that significantly correlated with fungal diversity, were the significantly driving forces in shaping the network specialization. Species specificity (d’) of symbiotic fungi was positively correlated with competition index of plants. Additionally,relative d’ of 123 co-observed OTUs showed significant partial Mantel correlations with the diameter at breast height (DBH) and plant abundance. Therefore, high canopy competitions and carbon source diversification could promote specialization of the symbiotic network.
Our study proposes a mechanism that likely shapes plant-symbiotic fungi networks in addition to the traditional matrix property theory and contributes to the understanding of the driving forces or limiting factors of community evolution.