Effects of close-to-nature management of planted forests on soil properties
Among the soil nutrients indices, total carbon, total nitrogen, total phosphorus, available phosphorus, nitrate-nitrogen content and pH all first decreased and then increased with the progress of near-naturalization, while organic carbon, ammoniacal nitrogen, nitrite-nitrogen, dry matter content and C/N ratio presented the opposite trend of first climbing up and then following down. Total carbon, organic carbon, total nitrogen, nitrate-nitrogen and ammoniacal nitrogen were all higher in natural secondary forest soils than in near-naturalized stands, while total phosphorus, available phosphorus, pH and dry matter content were highest in near-natural forest. Among the soil enzyme activities, cellulase and urease activities tended to increase during the near-naturalization process, while acid phosphatase activity tended to decrease and β-glucosidase and dehydrogenase activities first increased and then decreased. The acid phosphatase and urease activities in natural secondary forests were significantly higher than those in the plantation development sequence, while β-glucosidase and dehydrogenase activities were significantly lower than those in the plantation successional sequence. The rate of soil N mineralization ecological processes was significantly faster in near-naturalization sequence stands than in natural secondary forest (Table 2).
Table 2
Comparison of soil properties between near-naturalized stands and natural secondary forest (P: Pinus tabulaeformis forest; M: Mixed forest; NNF: Near-natural forest; NF: Natural secondary forest)
| SOC (mg/kg) | AP (mg/kg) | TP (mg/kg) | NO3−-N (mg/kg) | NH4+-N (mg/kg) | pH |
P | 79.40 ± 14.02b | 38.58 ± 32.35a | 545.80 ± 159.60a | 13.32 ± 2.26b | 8.10 ± 0.72ab | 4.79 ± 0.19b |
M | 95.88 ± 13.18ab | 33.89 ± 43.06a | 472.02 ± 88.10a | 9.97 ± 3.55c | 8.72 ± 7.46ab | 4.64 ± 0.34b |
NNF | 79.94 ± 17.85b | 45.40 ± 10.99a | 563.10 ± 193.30a | 15.42 ± 3.73b | 4.84 ± 3.05b | 5.85 ± 0.38a |
NF | 105.41 ± 21.03a | 40.28 ± 33.67a | 532.09 ± 159.40a | 21.17 ± 8.00a | 10.50 ± 3.98a | 5,74 ± 0.88a |
| Moisture content (%) | Dry matter (%) | TC (%) | TN (%) | C/N | NO2−-N |
P | 38.87 ± 4.93a | 95.71 ± 1.19a | 7.18 ± 3.55a | 0.52 ± 0.26a | 13.87 ± 1.26ab | 1.22 ± 0.63b |
M | 32.64 ± 5.66b | 97.11 ± 0.97a | 5.16 ± 0.37b | 0.36 ± 0.04b | 14.55 ± 0.71a | 2.84 ± 1.39a |
NNF | 32.11 ± 9.31b | 96.73 ± 3.27a | 7.15 ± 1.16a | 0.56 ± 0.12a | 12.97 ± 0.77b | 2.26 ± 2.72ab |
NF | 38.23 ± 5.85a | 95.89 ± 1.90a | 7.55 ± 2.78a | 0.61 ± 0.06a | 12.35 ± 0.86b | 1.70 ± 2.23ab |
| Cellulase (µg/(g·min) | Urease (µg/(g·h)) | β-Glucosidase (µg/(g·h)) | Dehydrogenase (µg/(g·h)) | Acid phosphatase (µg/(g·min) | Nitrification rate (µg/(g·d)) |
P | 0.097 ± 0.086b | 5.23 ± 1.63b | 0.83 ± 0.78ab | 0.39 ± 0.35a | 5.83 ± 1.10ab | 0.26 ± 0.02a |
M | 0.178 ± 0.054ab | 5.24 ± 1.81b | 1.24 ± 0.98a | 0.61 ± 0.54a | 5.29 ± 1.29ab | 0.15 ± 0.14b |
NNF | 0.199 ± 0.093ab | 9.61 ± 1.34a | 1.01 ± 0.51ab | 0.46 ± 0.37a | 4.36 ± 1.52b | 0.27 ± 0.08a |
NF | 0.213 ± 0.062a | 8.98 ± 2.78a | 0.59 ± 0.54b | 0.33 ± 0.38a | 6.85 ± 3.73a | 0.14 ± 0.79b |
| Change in ammonia N (mg/kg) | Change in nitrite N (mg/kg) | Change in nitrate N (mg/kg) | Total net mineralization (mg/kg) | Ammoniation rate (mg/(kg·m)) | |
P | -1.05 ± 2.99ab | 0.53 ± 1.48a | 24.54 ± 6.85b | 24.02 ± 9.06a | 37.04 ± 1.42ab | |
M | -3.34 ± 8.03ab | -1.67 ± 1.31a | 30.30 ± 8.00a | 25.39 ± 15.39a | 54.74 ± 27.24a | |
NNF | 1.33 ± 3.19a | -0.92 ± 2.85a | 26.38 ± 8.91a | 26.79 ± 7.90a | 34.38 ± 12.82ab | |
NF | -5.01 ± 3.49b | -0.52 ± 2.20a | 10.38 ± 15.64c | 14.00 ± 9.59b | 23.42 ± 12.38b | |
Effect of close-to-nature management on the niche width and abundance of various habitat specificity taxa of soil microorganisms
A total of 46,873 OUTs attributed to 537 genera were yielded by sequencing and cluster analysis of bacteria, which were divided into 17 generalized genera, 203 neutral genera and 317 specialized genera taxa. 1619 OUTs in sum attributed to 355 genera were yielded through by sequencing and cluster analysis of fungi, with 58 genera of generic, 153 genera of neutral and 144 genera of specialized taxa. The top genera in terms of abundance in each of the fungal groups were Ilyonectria, Archaeorhizomyces(*), Gliocladiospis for the generalist group, Fusarium, Cladosporium, Botryotrichum for the neutral group, Aspergillus, Erysiphe, Paecilium for the specialist group. In the bacterial community they were the generalized taxa: Bauldia, Hyphomicrobium, Chthonomonas; the neutral taxa: Sphingomonas, Halliangium, Rhodoplanes; the specialized taxa: Candidatus-udaeobacter, KB41, Candidatus-solibacter (*), Bradyrhizobium (the addition of * to the genus name indicates significantly difference of abundance between at least two forest stands). Analyzing by genera with the abundance greater than 1%, the abundance and numbers of fungi were significantly higher than those of bacteria, and both showed significantly lower abundance of the generalized taxa than the neutral and specialized taxa (Figs. 1,2).
The niche width of soil fungi is greater than that of bacteria for taxa of the same habitat specificity in each forest stage (Fig. 3 FG-BG, FN-BN, FS-BS), but there is no difference between different taxa of the same forest stage (Fig. 3 FG-FN-FS, BG-BN-BS). The niche width of all taxa increased with near-naturalization and was significantly lower than that of natural secondary forests. The increase showed a pattern of generalized taxa > neutral taxa > specialized taxa, i.e. slowing down with increasing habitat specificity (Fig. 3). In terms of abundance, specialized taxa of bacterial and fungal both decreased with near-naturalization. The abundance of generalized and neutral taxa showed an increase in fungi with near-naturalization and was lower than in natural secondary forests, but a decrease in bacteria and was higher than in natural secondary forests (Fig. 4).
Effect of close-to-nature management on the community structure of soil microbial
The impact of near-naturalization on the β-diversity of soil fungal and bacterial communities showed the trend of generalized taxa>neutral taxa>specialized taxa (Fig. 5 FG-FN-FS, BG-BN-BS). For the generalized taxa of fungi, the coordinate points of P. tabulaeformis forest, mixed forest and near-natural forest could be divided into different areas (Fig. 5 FG). For the neutral taxa, the coordinate points of P. tabulaeformis forest were clearly differentiated from other near-naturalized forest stands. This indicated a significant impact on the structure and distribution of the neutral taxa in the first and middle stages, but the effect was weaker in the later stages (Fig. 5 FN). For the specialized taxa, coordinates of P. tabulaeformis stands could still be separated from the others, but the coordinate of other forest stands were indistinguishable from each other, especially the near-natural forest and natural secondary forest, indicating a very similar community structure between them (Fig. 5 FS). For the soil bacterial community, the structure of the generalized taxa was relatively clearly differentiated with near-naturalization (Fig. 5 BG), while the neutral and specialized taxa only showed a clear distinction between P. tabulaeformis forest and other stands, the heterogeneity of the community structure between mixed, near-natural and natural secondary stands was already minimal (Fig. 5 BN BS).
Influence of soil physicochemical properties and enzyme activity on the community structure of soil microbial taxa
Soil physicochemical properties and enzyme activity explained more of the bacterial community structure than the fungal community, and both decreased with habitat specificity (Figs. 6, 7). Physicochemical content better explained microbial community structure than soil enzyme activity. The structure of fungal generalized and neutral taxa responded most significantly to pH (Fig. 6G N), suggesting that changes in soil pH were the main factor driving variation in the structure of such soil microorganisms during near-naturalization, while the distribution and structure of fungal specialized taxa were mainly influenced by urease activity (Fig. 6 S). The distribution of all soil bacterial taxa was dominated by soil carbon- and nitrogen-related characteristics (Fig. 7).
For the generalized taxa of fungi, the RDA axis 1 explained 20.81% of the variation and the axis 2 explained 12.16%, with soil pH being the most significant influence on the RDA axis 1 and the most influenced by dry matter content on the RDA axis 2 (Fig. 6G). The structure of neutral taxa was most influenced by pH on the RDA axis 1 and by dehydrogenase activity on axis 2 (Fig. 6N) In combination with the characteristics presented on the distribution of each community coordinate, the increase in pH may be the main factor driving the difference in the structure of soil fungal generalized and neutral taxa between pre-near-naturalized and later stands. The structure of specialized taxa was mainly influenced by urease activity on the RDA axis 1 and total nitrogen on the RDA axis 2 (Fig. 6S).
The RDA axis 1 explained 32.36% of the structural variation in bacterial generalized taxa and axis 2 explained 16.51%, with C/N ratio being the most strongly correlated with the axis 1 and water content with the axis 2 (Fig. 7G). For neutral taxa, total phosphorus and C/N ratio were the most explained indicators on axis 1 and total carbon and total nitrogen on axis 2 (Fig. 7N). For specialized taxa, C/N ratio and total nitrogen interpreted the most highly on axis 1 and total phosphorus and acid phosphatase on axis 2. Combined with the characteristics of the community's coordinate distribution, the decrease in C/N ratio drove the difference in bacterial structure between early to mid and late period of near-naturalization to the greatest extent, while the decrease in total phosphorus content also drove the difference between generalized and neutral taxa. (Fig. 7S).
Explanation of soil ecological processes by dominant genera abundance with different degrees of habitat specificity
The community structure of the top 1% of genera in terms of abundance of each taxon well explained the soil nitrogen mineralization. The community structure of each taxon manifested significant difference between near-naturalized sequence and natural secondary forests, which can be clearly shown in the RDA diagram (Fig. 8,9). The values of natural secondary forest sample sites were significantly lower than those of near-naturalized stands on axis 1 of RDA plot, consistent with the conclusion that nitrogen mineralization rates were lower in natural secondary forests. This revealed that heterogeneity in community structure drives the differences in nitrogen mineralization processes between near-naturalized and natural forest stands.
The community structure of fungal specialized taxa explained the most of the nitrogen mineralization process with 58.6% of the variation explained by the RDA axis 1 and 22.71% by the RDA axis 2. This was followed by the generalized taxa with 50.4% explained by the RDA axis 1 and 20.14% by the RDA axis 2. The structure of neutral taxa explained 49.78% of the variance by the RDA axis 1 and 17.73% by the RDA axis 2. The amount of net nitrogen mineralization was the most highly interpreted indicator of generalized and neutral taxa on the RDA axis 1. Among the generalized taxa, Staphylotrichum was the genus with the highest strength and positive correlation with soil nitrogen mineralization, Gliocladiopsis also showed a strong positive correlation with the amount of ammonia nitrogen change and nitrification rate, but the high abundant Archaeorhizomyces and Ilyonectria showed a negative correlation with soil nitrogen mineralization (Fig. 8G). Of the neutral taxa, Purpureocillium was the most strongly positively correlated with net mineralization and Condenascus with nitrification rates as well as Botryotrichum with ammonification rates, but the high abundant Cladosporium also showed a negative correlation with nitrogen mineralization (Fig. 8N). Soil nitrification rate was the most highly interpreted indicator of the structure of specialized taxa, and the strongest correlation was with Paecilomyces. Erysiphe showed the strongest positive correlation with the amount of ammoniacal nitrogen change while Russula showed a clear negative relationship with the nitrogen mineralization. However, the highest abundance of Aspergillus was not strongly correlated with nitrogen mineralization (Fig. 8S).
The structure of the generalized taxa of bacteria mostly explained soil nitrogen mineralization at 62.95% on the RDA axis 1 and 16.22% on the axis 2. Net nitrogen mineralization was the highest explained by the generalized and specialized taxa on the RDA axis 1. Among the generalized taxa, Nocardia was highly negatively and strongly correlated with the nitrogen mineralization, the highest correlation with net nitrogen mineralization was Hyphomicrobium, and Bauldia positively but not strongly correlated with nitrification rates (Fig. 9G). Among the specialized taxa, Bradyrhizobium was most strongly negatively correlated with each process of mineralization, while Candidatus-solibacter and Nitrospira were significantly positively correlated with nitrification rate (Fig. 9S). Soil nitrification rate was the most highly interpreted indicator for neutral taxa on the RDA axis 1 and strongest positive correlations were with Hirschia and Chtoniobater. Rhodoplanes was strongly negatively correlated with the mineralization process (Fig. 9N).