3.2 Effect of amendments on soil MBC, MBN and MBC/MBN
Amended treatments and incubation time had significant effects on MBC, MBN (P <0.001) and MBC/MBN (P <0.01), and had no significant interactions between the two factors. During the entire incubation period, MBC and MBN in all amended treatments were significantly higher than those in CK treatment, while MBC/MBN only in C+M+N treatments were significantly lower than that of CK treatment (Table 2). Specifically, MBC and MBN contents increased significantly after addition of M (in C-M-N and C-M treatments), in which the M2 and N2 level supported the higher contents. MBC/MBN decreased significantly in N2 levels (C+M2+N2 and C+M1+N2)(Fig.3). With the increasing incubation time (Fig.3), MBC, MBN and MBC/MBN in CK treatment remained almost unaltered, and with the average value of 42 mg/kg, 5 mg/kg and 9.52, respectively. MBC and MBN (P <0.01) increased significantly up to day 45, and then decreasing during the remaining incubation time(day 45 to 90). MBC/MBN in all amended treatments exhibited significant changes with incubation time but no obvious regularity.
3.3Diversity and composition of soil bacterial community
A total of 1,532,205 bacterial sequences were obtained from the complete data set, of which 13,769 bacterial OTUs belonged to 33 phyla, 257 classes, 257 orders, 475 families and 1,000 genera. The rarefaction curves of bacteria showed clear asymptotes, which indicated a near complete and true sampling of the community. The dominant phyla (relative abundance >1%) were Proteobacteria (48.26-82.32%), Actinobacteria (16.59-4.09%), Bacteroidetes (4.49-15.47%), Firmicutes (0.55-9.20%), Gemmatimonadetes (1.76-8.18%), Chloroflexi (0.89-5.21%), Patescibacteria (0.87-5.17%), and Acidobacteria (0.09-2.72%), together accounting for > 98% of bacterial sequences across all samples (Fig. S1a). Notably, the relative abundance of Bacteroidetes increased significantly in C-M-N treatments (P <0.001), in which N2 level increased the higher value (Fig. S1a). Firmicutes also increased significantly in C-M-N treatments (P <0.01), in which N1 level increased the higher value. However, the relative abundance of Acidobacteria decreased significantly (P <0.001) with addition of N fertilizer treatments, in which N1 level (ordered by C+N1, C+M1+N1, C+M2+N1) decreased obviously (Table 3).
Alpha-diversity estimated by Chao1 estimator, and Shannon and Simpson indices showed significant differences in species richness and diversity of soil bacterial community between different treatments (P <0.05). Chao1 estimator was significantly higher in C (C-N0), C+M (C-M1-N0, C-M2-N0) and C+M+N2 (C-N2, C-M1-N2, C-M2-N2) treatments than in CK treatment (P <0.05), in which N2 level treatment supported higher value (highest in C-M2-N2 treatment). Simpson and Shannon indices were significantly higher in all amended treatments especially at the N2 level (P <0.05), but no significant difference observed between amended treatments (Table 4).
PCoA analysis based on Bray-Curtis distances accounted for 47.3% of total variance among bacterial communities, with axes 1 and 2 explaining 25.4 and 21.9% of the variance, respectively (Fig. 4a). PCoA analysis showed that bacterial communities were divided into three major groups. Treatments with N1 and N2 level (C-N1, C-N2, C-M1-N1, C-M1-N2, C-M2-N1 and C-M2-N2) tended to group together, N3 level (C-N3, C-M1-N3 and C-M2-N3) clustered into another group, and treatments without N addition (C-N0, C-M1-N0, C-M2-N0) grouped together with CK (R2 = 0.555 > 0.5, P < 0.001; PERMANOVA, Test statistic= 5.0189, P = 0.001). Overall, three groups exhibited significant differences in bacterial community composition and were separated mainly by N level.