Soil fungi play vital roles in natural ecosystems, however, their community distribution patterns along different environmental gradients and ecological assembly processes remain unclear. In this study, Illumina MiSeq sequencing was used to investigate the soil fungal community structures of five different forest types along an elevational gradient, and a framework based on a null model was adopted to quantify the relative contribution of deterministic and stochastic ecological assembly processes. The results showed that the majority of soil fungal OTUs were derived from Zygomycota, Basidiomycota, and Ascomycota. Soil fungal community structure differed significantly among the five sites (P < 0.01), and the fungal α-diversity decreased as elevation increased (P < 0.01). The null model showed that the relative contribution of stochastic processes (37.78–73.33%) was higher than that of deterministic processes (26.67–62.22%) within the same forest type, while that of deterministic processes (35.00–93.00%) was higher than stochastic processes (7.00–65.00%) between forest types. These results suggest that forest soil fungal diversity decreased significantly with increasing elevation, and that deterministic processes may be key factors influencing soil fungal community assemblies among forest types. The results of this study provide new insight into soil fungal distribution patterns and community assembly processes in natural forest ecosystems.

Land-cover change has long been recognized as having marked effect on the amount of soil organic carbon (SOC). However, the microbially mediated processes and mechanisms on SOC are still unclear. In this study, the soil samples in a degenerative succession from alpine meadow to alpine steppe meadow in the Qinghai-Tibetan Plateau were analysed using high-throughput technologies, including Illumina sequencing and geochip functional gene arrays. The soil microbial community structure and diversity were significantly (p < .05) different between alpine meadow and alpine steppe meadow; the microbial ɑ-diversity in alpine steppe meadow was significantly (p < .01) higher than in alpine meadow. Molecular ecological network analysis indicated that the microbial community structure in alpine steppe meadow was more complex and tighter than in the alpine meadow. The relative abundance of soil microbial labile carbon degradation genes (e.g., pectin and hemicellulose) was significantly higher in alpine steppe meadow than in alpine meadow, but the relative abundance of soil recalcitrant carbon degradation genes (e.g., chitin and lignin) showed the opposite tendency. The Biolog Ecoplate experiment showed that microbially mediated soil carbon utilization was more active in alpine steppe meadow than in alpine meadow. Consequently, more soil labile carbon might be decomposed in alpine steppe meadow than in alpine meadow. Therefore, the degenerative succession of alpine meadow because of climate change or anthropogenic activities would most likely decrease SOC and nutrients medicated by changing soil microbial community structure and their functional potentials for carbon decomposition.

Evergreen broad‐leaved (EBF) and deciduous broad‐leaved (DBF) forests are two important vegetation types in terrestrial ecosystems that play key roles in sustainable biodiversity and global carbon (C) cycling. However, little is known about their associated soil fungal community and the potential metabolic activities involved in biogeochemical processes. In this study, soil samples were collected from EBF and DBF in Shennongjia Mountain, China, and soil fungal community structure and functional gene diversity analyzed based on combined Illumina MiSeq sequencing with GeoChip technologies. The results showed that soil fungal species richness (p = 0.079) and fungal functional gene diversity (p < 0.01) were higher in DBF than EBF. Zygomycota was the most dominant phylum in both broad‐leaved forests, and the most dominant genera found in each forest varied (Umbelopsis dominated in DBF, whereas Mortierella dominated in EBF). A total of 4, 439 soil fungi associated functional gene probes involved in C and nitrogen (N) cycling were detected. Interestingly, the relative abundance of functional genes related to labile C degradation (e.g., starch, pectin, hemicellulose, and cellulose) was significantly higher (p < 0.05) in DBF than EBF, and the functional gene relative abundance involved in C cycling was significantly negatively correlated with soil labile organic C (r = −0.720, p = 0.002). In conclusion, the soil fungal community structure and potential metabolic activity showed marked divergence in different broad‐leaved forest types, and the higher relative abundance of functional genes involved in C cycling in DBF may be caused by release of loss of organic C in the soil.