3.1 Soil nutrient contents
Under four different fertilization treatments, the levels of SOM, AN, and AK in the rhizosphere of alfalfa and awnless brome were significantly higher than those of N0P0K0 (P < 0.05) (Fig. 2). The N2P2K2 treatment resulted in the most notable outcome. In comparison to the N0P0K0 treatment, the alfalfa rhizosphere soil under the N2P2K2 treatment had 29% higher SOM, 40% higher AN, and 23% higher AK contents. Similarly, for awnless brome, the nutrient content analysis in the rhizosphere soil showed that the levels of SOM, AN, and AK increased by 30.3%, 45.1%, and 25.4% respectively under N2P2K2 treatment. Overall, there were significant differences between the various treatments. Furthermore, the rankings for SOM and AN content were N2P2K2 > N2P2K0 > N2P0K2 > N0P2K2. In contrast, the AK content rankings were N2P2K2 > N0P2K2 > N2P0K2 > N2P2K0.
3.2 Soil enzyme activities
The rhizosphere soil enzymatic activities were illustrated in Fig. 3. In the alfalfa rhizosphere soil, the activities of UA, CAT, APA, and SA were 51.00-62.67 mg/g, 0.72–0.95 mg/g, 17.22–27.24 mg/g, 33.40-48.73 mg/g, respectively. The activities in the awnless brome rhizosphere were 51.40–64.2 mg/g, 0.72–0.95 mg/g, 17.96–27.83 mg/g, and 32.55–48.04 mg/g, respectively. Furthermore, the enzyme activity of the four fertilization treatments significantly surpassed that of the N0P0K0 treatment ( P < 0.05 ). Compared to the N0P0K0 treatment, the alfalfa rhizosphere soil treated with N2P2K2 increased by 22.9%, 58.2%, 31.8%, and 45.9% in the activities of UA, CAT, APA, and SA, respectively. Similarly, the activities in the awnless brome rhizosphere soil treated with N2P2K2 increased by 24.9%, 55.0%, 31.8%, and 47.6%, respectively. The activities of CAT, APA, and SA were ranked as N2P2K2 > N2P2K0 > N2P0K2 > N0P2K2, while the ranking for UA activity was N2P2K2 > N0P2K2 > N2P0K2 > N2P2K0.
3.3 Soil biological properties
The results depicted in Fig. 4 showed that the application of N2P2K2 treatment led to the highest levels of SMBC, SMBN, and SMBP in the rhizosphere soil of alfalfa and awnless brome. These levels were significantly higher than those observed under N0P0K0 treatment, with an average increase of 63% and 64%, respectively. In the alfalfa rhizosphere treated with N2P2K2, the contents of SMBC, SMBN, and SMBP were 66.8%, 63.7%, and 59.2% higher than those treated with N0P0K0. Similarly, under N2P2K2 treatment, the SMBC, SMBN, and SMBP contents in the awnless brome rhizosphere were 55.9%, 62.9%, and 66.5% higher, respectively, than those under N0P0K0 treatment. The findings also indicated that the content of SMBC and SMBP ranked as N2P2K2 > N2P2K0 > N0P2K2 > N2P0K2, and the content of SMBN ranked as N2P2K2 > N2P2K0 > N2P0K2 > N0P2K2. Moreover, the contents of SMBC, SMBN, and SMBP had significant differences between different fertilization treatments ( P < 0.05 ).
3.4 Alpha diversity and sample sequencing data information
The bacterial coverage ranged from 93–95%, with the number of sequences per sample varying from 11,687 to 13,627. The sequencing depths of all samples enable a comprehensive interpretation of the microbial community's composition (Table 1, Fig. 5). For the alfalfa bacterial community, the ACE index reached its maximum value under N2P0K2 treatment, with no significant difference among the treatments. The specific ranking was N2P0K2 > N2P2K2 > N0P0K0 > N2P2K0 > N0P2K2; the ranking of Chao1 index was N2P0K2 > N2P2K0 > N2P2K2 > N0P0K0 > N0P2K2; Simpson and Shannon index rank as follows: N2P2K0 > N2P0K2 > N2P2K2 > N0P0K0 > N0P2K2. For the awnless brome community, compared to the N0P0K0 treatment, a downward trend in the ACE, Chao1, Simpson, and Shannon index was observed but did not reach statistical significance. The order of ACE and Simpson index was N0P0K0 > N2P2K0 > N0P2K2 > N2P2K2 > N2P0K2. According to the Chao1 and Shannon index, the order was as follows: N0P0K0 > N0P2K2 > N2P2K0 > N2P2K2 > N2P0K2.
Table 1
Analysis of soil bacteria Alpha diversity index
Treatment | Effective CCS | ACE index | Chao1 index | Simpson index | Shannon index | Coverage/% |
N0P0K0G | 12334 ± 882.9c | 3436.39 ± 441.37a | 2918.58 ± 85.05ab | 0.996 ± 0.0001ab | 9.32 ± 0.12c | 0.93 ± 0.00a |
N0P2K2G | 11687 ± 212.6c | 3284.69 ± 370.75a | 2831.81 ± 96.31b | 0.996 ± 0.0001b | 9.32 ± 0.06c | 0.93 ± 0.00a |
N2P0K2G | 12720 ± 726.1ab | 3654.55 ± 313.32a | 3107.24 ± 36.59a | 0.997 ± 0.0001a | 9.58 ± 0.05ab | 0.93 ± 0.00a |
N2P2K0G | 12581 ± 1330.5a | 3343.32 ± 438.88a | 3029.17 ± 153.14ab | 0.997 ± 0.0001ab | 9.63 ± 0.02a | 0.93 ± 0.00a |
N2P2K2G | 13377 ± 1235.2b | 3487.9 ± 198.36a | 3022.76 ± 121.08ab | 0.997 ± 0.0001ab | 9.48 ± 0.07b | 0.93 ± 0.00a |
N0P0K0Y | 12477 ± 258.9a | 3461.84 ± 390.00a | 3002.82 ± 69.23a | 0.997 ± 0.0001a | 9.43 ± 0.05a | 0.93 ± 0.00a |
N0P2K2Y | 13034 ± 336a | 3092.29 ± 204.44a | 2960.39 ± 220.33a | 0.996 ± 0.0001a | 9.36 ± 0.16a | 0.93 ± 0.00a |
N2P0K2Y | 12790 ± 1066.6a | 2708.42 ± 1137.21a | 2244.32 ± 1266.02a | 0.972 ± 0.043a | 7.81 ± 2.82a | 0.95 ± 0.00a |
N2P2K0Y | 13627 ± 1051.9a | 3294.15 ± 54.45a | 2730.64 ± 86.98a | 0.996 ± 0.0001a | 9.2 ± 0.03a | 0.94 ± 0.00a |
N2P2K2Y | 13063 ± 1102.2a | 2846.47 ± 573.8a | 2541.64 ± 254.27a | 0.994 ± 0.005a | 8.99 ± 0.46a | 0.94 ± 0.00a |
Note: Data represent mean ± SD, different lowercase letters in the same column indicate significant differences between the groups at a significance level of P < 0.05. In the treatment name, "G" represents M. sativa ‘Gannong' No.3, while "Y" represents Bromus inermis ‘Yuanye'. |
3.5 Relative abundances of bacterial communities
In the root zone of the legume-Gramineae mixtures, Fig. 6 illustrated the bacterial community structure at the phylum and family levels (Fig. 6). The analysis revealed that the microbial community structure in the alfalfa and awnless brome rhizosphere soil was similar at both phylum and family levels, with the bacteria showing consistent relative abundance of > 1% and no change in order. The main bacterial phyla in the rhizosphere soil of legume-Gramineae mixtures included Proteobacteria, Acidobacteria, Bacteroidetes, Gemmatimonadetes, Verrucomicrobia, Actinobacteria, Planctomycetes, Chloroflexi, and Nitrosospira; The main bacterial families included Chitinophagaceae, unclassified_Vicinamibacterales, Vicinamibacteraceae, Gemmatimonadaceae, Sphingomonadaceae, unclassified_Bacteria, Pyrinomonadaceae, Nitrosomonadaceae, Blastocatellaceae, Nitrospiraceae.
The relative abundance of Proteobacteria increased by 2.33% in the rhizosphere soil of alfalfa after the N2P2K2 treatment, as well as a 0.67% increase in Gemmatimonadetes and a 0.71% increase in Verrucomicrobia, compared to the N0P0K0 treatment. Inversely, Acidobacteria, Actinobacteria, and Chloroflexi experienced a decrease in relative abundance, with declines of 4.23%, 0.46%, and 0.05%, respectively. Similarly, the application of N2P2K2 treatment increased the relative abundance of Gemmatimonadaceae, Sphingomonadaceae, Nitrosomonadaceae, and Nitrospiraceae. In the rhizosphere soil of awnless brome, compared to the N0P0K0 treatment, the N2P2K2 treatment resulted in a higher relative abundance of Proteobacteria, Acidobacteria, Bacteroidetes, Planctomycetes, and Nitrosospira by 1.45%, 2.18%, 0.49%, 0.07%, and 0.13%, respectively. Oppositely, the relative abundance of Gemmatimonadetes, Verrucomicrobia, Actinobacteria, and Chloroflexi decreased by 0.63%, 0.46%, 1.91%, and 0.41%, respectively. Notably, Gemmatimonadaceae and Nitrospiraceae showed a significant increase in the N2P2K2 treatment compared to N0P0K0, while the trend for Chitinophagaceae was the opposite.
3.6 Identification of microbial indicators using LEfSe analysis
In alfalfa, the primary bacterial group biomarkers in N0P0K0 were identified as Lysobacter and Xanthomonadacceae, while Blastocatellales was determined to be the most significant bacterial group biomarker in N0P2K2. Sphingomonas was recognized as the most significant bacterial group biomarker in N2P2K0, while Nitrosomonadaceae was found to be the most important bacterial group biomarker in N2P2K2 (Fig. 7). In awnless brome, the main bacterial group biomarkers in N0P2K2 were Acidobacteria bacterium WWH8, RB41, and Pyrinomonadaceae. Flavobacteriaceae was identified as the most important bacterial group biomarker in N2P0K2, while Gemmatimonadaceae was determined to be the most significant bacterial group biomarker in N2P2K0 (Fig. 7).
3.7 Correlation between soil nutrients, enzyme activities, and dominant microbes
Redundancy analysis (RDA) was conducted to investigate the correlation between the bacterial community structure and soil environmental factors in alfalfa and awnless brome at the phylum level. The redundancy analysis for alfalfa revealed that the first ordination axis (45.84%) and the second ordination axis (27.75%) collectively accounted for 73.59% of the variation in the impact of soil environmental factors on soil bacterial community composition (Fig. 8a). The results revealed that soil environmental factors were positively correlated with phyla Verrucomicrobiota, Planctomycetota, Bacteroidota, Chloroflexi, Gemmatimonadota, and Proteobacteria in rhizosphere, respectively. Particularly, CAT activities (P = 0.042) and SOM content (P = 0.046) were two major factors that influenced the bacterial community composition in the rhizosphere of alfalfa, explaining 20.7% and 17.8% of the variation, respectively (Table 2).
Table 2
Simple-term effects of environmental variables based on Monte Carlo permutation tests from the redundancy analysis(Alfalfa)
Name | Explains % | Contribution % | pseudo-F | P |
CAT | 20.7 | 26.6 | 3.4 | 0.042 |
SOM | 17.8 | 22.8 | 3.5 | 0.046 |
SMBP | 8 | 10.3 | 1.7 | 0.208 |
SMBC | 8.7 | 11.2 | 2 | 0.142 |
AN | 9.2 | 11.8 | 2.3 | 0.108 |
AK | 4.5 | 5.7 | 1.1 | 0.318 |
SA | 8.2 | 10.6 | 2.5 | 0.108 |
UA | 4.8 | 6.1 | 1.6 | 0.246 |
SMBN | 4.2 | 5.4 | 1.5 | 0.136 |
APA | 1.4 | 1.8 | 0.5 | 0.544 |
For awnless brome, based on the redundancy analysis results, RDA1 and RDA2 explained a total variation of 33.97% and 25.35%, respectively (Fig. 8b). AK and UA showed a positive correlation with phyla Bacteroidota and Proteobacteria in the rhizosphere; other soil environmental factors (except AK and UA), exhibited a positive correlation with phyla Acidobacteriota, Actinobacteriota, Gemmatimonadota, Chloroflexi, and Planctomycetota in the rhizosphere. In Particular, AK (P = 0.002) and CAT (P = 0.018) were identified as two major factors that influenced the bacterial community composition in the rhizosphere of awnless brome, explaining 22.6% and 8.5% of the variation, respectively. Additionally, SA (P = 0.01) and SMBN (P = 0.01) were identified as other factors significantly correlated with the differences in the composition of rhizosphere bacterial communities, accounting for a total of 40.0% of the variation (Table 3).
Table 3
Simple-term effects of environmental variables based on Monte Carlo permutation tests from the redundancy analysis(Awnless brome)
Name | Explains % | Contribution % | pseudo-F | P |
SMBN | 19.3 | 24.2 | 3.1 | 0.01 |
SA | 20.7 | 25.9 | 4.1 | 0.01 |
AK | 22.6 | 28.3 | 6.6 | 0.002 |
CAT | 8.5 | 10.7 | 3 | 0.018 |
APA | 4.3 | 5.4 | 1.6 | 0.186 |
AN | 3.7 | 4.7 | 1.4 | 0.216 |
UA | 4.2 | 5.3 | 1.8 | 0.162 |
SMBP | 3.4 | 4.2 | 1.5 | 0.154 |
SOM | 2.4 | 3 | 1.1 | 0.340 |
SMBC | 0.8 | 1 | 0.3 | 0.806 |
3.8 Estimating the function of the microbial community
FAPROTAX analysis revealed that the alfalfa sample had functional groups that were related to chemical nutrition (including oxidative nutrition), chitin decomposition, animal parasitism, nitrate reduction, and nitrogen fixation, accounting for 50.6%, 3.9%, 3.2%, 3.0%, and 2.6% of the total bacterial community, respectively (Fig. 9a). Notably, the proportion of bacteria consorted with chitinolysis and nitrogen fixation was highest in the no-fertilized treatment, while the treatment with N0P2K2 showed the highest proportion of bacteria involved in nitrogen respiration. In the N2P2K0 treatment, a higher proportion of bacteria that are linked to chemoheterotrophy (including aerobic chemoheterotrophy), nitrate reduction, and ureolysis was observed compared to other fertilization treatments. Additionally, the N2P2K2 treatment exhibited the highest proportion of bacteria associated with animal parasites or symbionts and human pathogens (including pneumonia).
In the case of awnless brome, based on FAPROTAX analysis, the functional groups with the highest relative abundance were related to combined nutrition (including oxidative nutrition), animal parasitism, fermentation, nitrate reduction, and nitrogen fixation, accounting for 53.2%, 5.0%, 2.8%, 3.6%, and 2.7% of the total bacterial community, respectively (Fig. 9b). It is worth noting that the proportion of bacteria related to human pathogens (including pneumonia) was highest in the no-fertilization treatment, while the N0P2K2 treatment showed a higher proportion of bacteria linked to predatory or ectoparasitic compared to other fertilization treatments. Additionally, the N2P2K0 fertilization treatment had a higher proportion of bacteria associated with animal parasites or symbiotic relationships and fermentation than other fertilization regimens. The N2P2K0 treatment resulted in a higher abundance of bacteria linked to chemoheterotrophy, nitrate reduction, and nitrogen fixation compared to other fertilization regimens. Meanwhile, the proportion of bacteria associated with chitinolysis peaked in the N2P2K2 treatment.