The means (±SE) for each variable followed by different uppercase letters indicate signficant differences between grazing and exclusion sites (P < 0.05). The lowercase letters indicate significant difference among different grassland types (P< 0.05). SWC, soil water content; BD, bulk density; SOC, soil organic carbon; TN, total nitrogen; TP, total phosphorus; C:N, carbon:nitrogen; C:P, carbon:phosphorus; N:P, nitrogen:phosphorus.
3.2. Soil fungi community OTU numbers and composition as affected by grazing exclusion and grassland type
The Wayne analysis (Fig. 1) showed that in the 0-5 cm soil layer (Fig. 2A), the number of fungal OTUs in mountain meadows (619) accounted for 38.98% of the total OTUs; the number of fungal OTUs in temperate steppe (512) accounted for 32.24%; and the number of fungal OTUs in temperate deserts (457) accounted for 28.78%. In addition, mountain meadows and temperate steppes shared 12.15% of fungal OTUs (193); mountain meadows and temperate deserts shared 0.56% of OTUs (9); temperate steppes and temperate deserts shared 1.32% of fungal OTUs (21) and the three grasslands shared 1.95% of fungal OTUs (31). After exclusion (Fig. 2B), the number of fungal OTUs in mountain meadows and temperate steppes decreased by 18.1% and 18.36%, respectively, compared with those before exclusion, but the number of fungal OTUs in temperate deserts increased by 38.51% compared with those before exclusion.
In the 5-10 cm soil layer (Fig. 2C), the number of fungal OTUs in mountain meadows (512) accounted for 32.04% of the total OTUs; the number of fungal OTUs in temperate steppes (596) accounted for 37.30%; and the number of fungal OTUs in temperate deserts (490) accounted for 30.66%. After exclusion (Fig. 2D), the number of fungal OTUs in temperate steppes decreased by 20.47% compared with that before exclusion, while the number of fungal OTUs in mountain meadows and temperate deserts increased by 2.7% and 15.71% respectively, compared with that before exclusion.
At the phylum level (Fig. 3A), Ascomycota andBasidiomycota were the dominant fungi in the three grassland types in the 0-5 cm soil layer before exclusion, accounting for 94.03% and 5.67% of the total fungi in temperate deserts, 80.76% and 18.51% of the total fungi in temperate steppes, and 78.43% and 18.52% of the total fungi in mountain meadows respectively. After exclusion, theAscomycota decreased by 14.72% and 9.74% in temperate deserts and mountain meadows respectively, and increased by 4.82% in temperate steppes. In contrast to Basidiomycota , increased by 238.8% and 47.68% in temperate deserts and mountain meadows respectively, but it decreased by 23.12% in temperate steppes. In addition, significant differences were found between the three grassland types for theGlomeromycota and Mucoromycota (P <0.05; Supplementary Table 4). The interaction between grazing exclusion and grassland type had no significant effect on the other fungi. In the 5-10 cm soil layer (Fig. 3B), the main fungi of the three grassland types were the same as in the 0-5 cm soil layer, and the analysis revealed that the Mortierellomycota was not affected by either grazing exclusion or grassland type, but the interaction between the two had a significant effect on the Mortierellomycota (P < 0.05; Supplementary Table 5). In addition, grassland type significantly influenced the Blastocladiomycota (P < 0.05), whereas grazing exclusion had no significant effect on any of the fungi measured (P > 0.05).
At the phyla level in the 0-5 cm soil layer (Fig. 3C), the dominant fungi in the three grassland types before and after exclusion includedDothideomycetes , Archaeorhizomycetes ,Sordariomycetes and Agaricomycetes , while in the temperate desert Archaeorhizomycetes almost disappear in temperate deserts. Grassland type had a significant effect on Glomeromycetes ,Orbiliomycetes and Tremellomycetes (P < 0.05; Supplementary Fig. 6). In the 5-10 cm soil layer (Fig. 3D),Sordariomycetes increased significantly by 279.32% in the temperate desert before exclusion compared to the 0-5 cm soil layer, and the interaction of grazing exclusion, grassland type and both significantly altered Sordariomycetes and Pezizomycetes(P < 0.05; Supplementary Fig. 7). Grassland type had a highly significant effect on Ustilaginomycetes (P< 0.001) and a significant effect on Eurotiomycetes andOrbiliomycetes (P < 0.05). In addition, grassland type had a highly significant effect on Dothideomycetesand Archaeorhizomycetes in both the 0-5 cm and 5-10 cm soil layers (P < 0.001; Supplementary Tables 6 and 7).