3.1 The effects of selenium on physiochemical properties of crops
In this study, fertilization of selenium affected diverse physiochemical properties of millet plants and those effects differed between two varieties (Table 1). Applying selenium improved the plant height especially during shooting stage, and higher concentration exhibited higher improvement. For Jigu 19, 6% of improvement was observed compared to 5% of 38%. The chlorophyll concentration was obviously elevated during seedling stage, shooting stage, and grain-filling stage. Medium level (Se2) showed the best performance. In maturation stage, no significant effects on chlorophyll concentration were observed. The GSH-Px activity in leaves was significantly improved during seedling stage, grain-filling stage and maturation stage, with Se2 as the optimal fertilization level. Selenium fertilization also facilitated the absorption of macro-elements such as nitrogen and phosphorus (Table 2). During shooting stage, the nitrogen content of aboveground part was increased. Jigu 38 exhibited a higher phosphorous content with Se2 during filling stage and maturation stage. However, too much selenium was adverse to uptake of nitrogen and phosphorus and the contents in aboveground parts decreased accordingly under Se3.
Table 1
Plant height, chlorophyll and GSH-Px activities of millet under different selenium fertilizer levels
Stage | Treatments | plant height(cm) | chlorophyll(SPAD) | GSH-Px (U/L) |
Jigu 38 | Jigu 19 | Jigu 38 | Jigu 19 | Jigu 38 | Jigu 19 |
seedling stage | CK | 31.93 ± 2.08a | 33.13 ± 0.93a | 31.77 ± 1.08b | 30.69 ± 0.41b | 54.39 ± 8.39b | 78.75 ± 3.39b |
Se1 | 32.39 ± 4.40a | 31.48 ± 3.10a | 33.99 ± 1.19ab | 37.40 ± 2.68a | 72.71 ± 8.28ab | 92.98 ± 5.18ab |
Se2 | 32.06 ± 3.62a | 32.05 ± 2.34a | 36.70 ± 1.77a | 35.22 ± 0.64a | 85.19 ± 9.98a | 101.17 ± 8.11a |
Se3 | 29.93 ± 2.88a | 29.57 ± 2.82a | 34.25 ± 1.21ab | 37.02 ± 1.21a | 82.26 ± 9.00a | 88.50 ± 5.43ab |
Shooting stage | CK | 110.21 ± 1.52b | 111.20 ± 1.44b | 44.13 ± 0.74b | 43.39 ± 0.89b | 127.93 ± 6.79a | 143.63 ± 7.94a |
Se1 | 112.46 ± 1.45ab | 114.88 ± 1.08ab | 44.79 ± 1.15b | 45.76 ± 0.63ab | 138.99 ± 11.73a | 149.03 ± 4.40a |
Se2 | 114.54 ± 1.36a | 115.21 ± 1.27ab | 48.67 ± 1.56a | 47.39 ± 0.91a | 142.08 ± 9.01a | 152.12 ± 10.50a |
Se3 | 115.78 ± 1.02a | 118.00 ± 1.68a | 45.45 ± 0.56b | 46.86 ± 0.82a | 117.89 ± 8.32a | 141.83 ± 6.80a |
Filling stage | CK | 114.20 ± 5.08a | 120.97 ± 1.86a | 48.47 ± 0.77b | 47.24 ± 0.84b | 111.54 ± 8.15b | 135.35 ± 8.95b |
Se1 | 120.52 ± 2.85a | 122.03 ± 2.56a | 51.93 ± 1.63ab | 51.37 ± 1.08a | 118.62 ± 3.33b | 138.82 ± 8.00b |
Se2 | 121.69 ± 3.48a | 123.12 ± 3.14a | 52.20 ± 1.27a | 53.44 ± 1.61a | 147.33 ± 8.54a | 165.95 ± 4.19a |
Se3 | 120.25 ± 2.62a | 122.88 ± 3.03a | 51.81 ± 0.61ab | 50.49 ± 1.09ab | 121.93 ± 9.44b | 162.19 ± 2.22a |
Maturation stage | CK | 129.41 ± 1.83a | 130.82 ± 2.39a | 37.84 ± 3.09a | 37.74 ± 2.97a | 166.91 ± 3.83b | 162.62 ± 3.93b |
Se1 | 129.37 ± 1.07a | 127.82 ± 3.09a | 40.59 ± 4.72a | 39.60 ± 2.81a | 171.66 ± 0.55ab | 168.03 ± 1.39ab |
Se2 | 131.10 ± 0.87a | 134.46 ± 1.37a | 43.43 ± 2.19a | 45.59 ± 3.65a | 178.71 ± 0.41a | 170.52 ± 1.18a |
Se3 | 131.02 ± 1.56a | 133.33 ± 2.41a | 41.19 ± 3.42a | 41.56 ± 4.69a | 175.10 ± 2.44a | 170.19 ± 1.22a |
Note: different letters in the table represent statistical difference among different selenium application treatments (P ≤ 0.05). |
Table 2
Nitrogen and phosphorus contents in aboveground parts millet under different selenium fertilizer levels
Stage | Treatments | N content(g/kg) | P content (g/kg) |
Jigu 38 | Jigu 19 | Jigu 38 | Jigu 19 |
Seedling stage | CK | 37.64 ± 0.68a | 37.37 ± 2.18a | 3.45 ± 0.19a | 3.46 ± 0.21a |
Se1 | 39.81 ± 1.15a | 38.05 ± 2.03a | 3.56 ± 0.24a | 3.64 ± 0.49a |
Se2 | 38.46 ± 1.52a | 39.23 ± 1.60a | 3.87 ± 0.36a | 3.81 ± 0.43a |
Se3 | 38.26 ± 0.71a | 42.18 ± 1.07a | 3.98 ± 0.42a | 3.83 ± 0.36a |
Shooting stage | CK | 25.17 ± 0.42b | 22.27 ± 1.26b | 2.07 ± 0.18a | 2.03 ± 0.21a |
Se1 | 26.59 ± 0.48a | 25.57 ± 1.63a | 1.81 ± 0.19a | 2.12 ± 0.23a |
Se2 | 26.28 ± 0.37ab | 24.27 ± 1.64ab | 2.09 ± 0.13a | 2.26 ± 0.34a |
Se3 | 26.47 ± 0.17a | 24.25 ± 0.82ab | 2.44 ± 0.41a | 2.14 ± 0.16a |
Filling stage | CK | 18.53 ± 0.67a | 17.97 ± 0.33ab | 1.87 ± 0.10b | 2.28 ± 0.03a |
Se1 | 20.12 ± 0.85a | 18.74 ± 0.22a | 2.08 ± 0.01a | 2.12 ± 0.18a |
Se2 | 19.07 ± 0.49a | 18.22 ± 0.42ab | 2.10 ± 0.01a | 2.32 ± 0.13a |
Se3 | 19.20 ± 1.57a | 17.51 ± 0.34b | 1.86 ± 0.06b | 1.62 ± 0.11b |
Maturation stage | CK | 11.89 ± 1.23a | 11.71 ± 0.66a | 0.86 ± 0.09ab | 0.80 ± 0.09a |
Se1 | 13.48 ± 0.52a | 12.00 ± 0.99a | 0.63 ± 0.09b | 1.10 ± 0.11a |
Se2 | 13.81 ± 0.50a | 13.87 ± 1.13a | 0.94 ± 0.09a | 1.03 ± 0.20a |
Se3 | 12.29 ± 0.50a | 13.03 ± 0.54a | 0.69 ± 0.11ab | 0.74 ± 0.12a |
Note: different letters in the table represent statistical difference among different selenium application treatments (P ≤ 0.05). |
For millet grain, the yield of Jigu 38 was improved by Se1 while the yield of Jigu 19 was decreased by Se3 (Table 3). Thousand-grain weight was barely affected and panicle length of Jigu 19 was improved by Se1 and Se2. In maturation stage, selenium fertilization remarkably enhanced the selenium contents of different aboveground parts (Table 4). Roots showed the highest selenium content followed by stem/leaves and panicles. Without selenium fertilization, the selenium content was ordered by steam/leaves, roots and panicles. The selenium content of grain was also effectively improved (Fig. 1). Se2 showed the best performance and under the same fertilizing level, Jigu 38 absorbed more selenium than Jigu 19. At the same time, selenium fertilization also increased the content of vitamin b1 in grain, while the dietary fiber content was limited affected.
Table 3
Panicle length, thousand-grain weight (TKW) and yield of foxtail millet under different selenium fertilizer levels
Treatments | Panicle length(cm) | TKW(g) | Yield(kg/hm2) |
Jigu 38 | Jigu 19 | Jigu 38 | Jigu 19 | Jigu 38 | Jigu 19 |
CK | 18.70 ± 0.43a | 17.97 ± 0.69ab | 2.60 ± 0.07a | 2.63 ± 0.05a | 4950 ± 48ab | 5483 ± 217a |
Se1 | 19.23 ± 1.00a | 19.54 ± 0.61a | 2.56 ± 0.08a | 2.64 ± 0.07a | 5157 ± 183a | 5521 ± 219a |
Se2 | 20.57 ± 0.90a | 19.67 ± 0.63a | 2.64 ± 0.08a | 2.65 ± 0.03a | 4850 ± 76ab | 5587 ± 271a |
Se3 | 18.37 ± 0.78a | 17.30 ± 0.60b | 2.56 ± 0.03a | 2.62 ± 0.05a | 4735 ± 30b | 4720 ± 110b |
Note: different letters in the table represent statistical difference among different selenium application treatments (P ≤ 0.05). |
Table 4
Selenium contents in aboveground parts of millet under different selenium fertilizer levels
Treatments | Root (mg/kg) | Stem/leaf (mg/kg) | Panicle (mg/kg) |
Jigu 38 | Jigu 19 | Jigu 38 | Jigu 19 | Jigu 38 | Jigu 19 |
CK | 0.19 ± 0.01c | 0.23 ± 0.02c | 0.23 ± 0.02b | 0.26 ± 0.02c | 0.03 ± 0.00c | 0.04 ± 0.00d |
Se1 | 0.46 ± 0.06b | 0.45 ± 0.06b | 0.36 ± 0.02ab | 0.36 ± 0.04b | 0.13 ± 0.01b | 0.15 ± 0.01c |
Se2 | 0.69 ± 0.06a | 0.63 ± 0.03a | 0.48 ± 0.07a | 0.45 ± 0.02a | 0.25 ± 0.01a | 0.26 ± 0.01a |
Se3 | 0.64 ± 0.03a | 0.57 ± 0.09ab | 0.43 ± 0.05a | 0.43 ± 0.02ab | 0.23 ± 0.01a | 0.23 ± 0.01b |
Note: different letters in the table represent statistical difference among different selenium application treatments (P ≤ 0.05). |
3.2 The effects of selenium on physiochemical properties of rhizosphere soil
For physiochemical properties of rhizosphere soil, selenium fertilization showed limited impact on pH, actual selenium content of rhizosphere soil as well as urease activity (Tables S1&S2). Se2 significantly increased acid phosphatase and sucrase activity in soil of Jigu 19. These results suggested that pH and urease activity may not be related to selenium absorption and transportation. The unaffected selenium content of rhizosphere soil indicated that external selenium fertilizer primarily functioned on millet plants therefore did not accumulate and remain in soil.
3.3 The effects of selenium on microbial community of rhizosphere soil
To explore the effects of selenium fertilization on the microbial community located in rhizosphere soil, different analysis methods were applied from different aspects. The Venn analysis shown in Figure S1 indicated that the number of bacterial OTU (sequences with similarity > 98% were regarded as the same OTU) enhanced with the increase of selenium fertilization level for both Jigu 38 and Jigu 19. The richness and diversity of bacterial community were shown by sobs index and Shannon index respectively. As shown in Table S3, selenium fertilization slightly improved both indices compared to CK. Se3 exhibited the highest level of bacterial richness and diversity.
Species with relative abundance > 1% were selected for further analysis of bacterial community composition (Fig. 2). At the phylum level, Actinobacteriota was observed as the dominant phylum, followed by Proteobacteria, Acidobacteriota, Chloroflexi, and Firmicutes. These dominant phyla accounted for more than 80% of relative abundance. Compared to CK, selenium fertilization decreased the relative abundance of Actinobacteriota and Proteobacteria. However, no significant differences were observed among groups with different fertilization levels. At the genus level, various undefined genera appeared for both Jigu 38 and Jigu 19. Bacillus, Blastococcus, Rubrobacter, Microvirga, and Sphingomonas were observed as the dominant genera. In addition, norank_f__JG30-KF-CM45, norank_f__Vicinamibacteraceae, and norank_f__norank_o__Vicinamibacterales also had high relative abundance. The relative abundance of those dominant genera was lower in Se2 compared to Se1 and Se3.
For fungi, the OTU number increased with the elevation of selenium level for Jigu 38, while increased first and then decreased for Jigu 19 (Figs. 2c&2d). Se2 and Se3 also improved the richness and diversity of fungal community for both Jigu 38 and Jigu 19, while the differences were not statistically significant.
At the phylum level, Ascomycota, Basidiomycota and Mortierellomycota were observed as the dominant phyla, constituting more than 90% of relative abundance (Fig. 3). For Jigu 38, selenium fertilization decreased the relative abundance of Ascomycota while increased the relative abundance of Mortierellomycota. However for Jigu 19, such trends were the opposite. The relative abundance of Basidiomycota was improved for both Jigu 38 and Jigu 19. At the genus level, Jigu 38 and Jigu 19 showed similar fungal composition, with Cladosporium, Mortierella, Gibberella, Chaetomium, Solicoccozyma, Cephaliophora, and Acremonium as the dominant genera. Se3 decreased the relative abundance of Cephaliophora while increased the relative abundance of Cladosporium.
3.4 The correlations between microbial community and physiochemical properties of crops and rhizosphere soil
The correlations between microbial community and physiochemical properties of soil were studied with RDA analysis (Fig. 4). The results showed different correlations for different crops. For Jigu 38, the environmental factors contributed to microbial community to different extents, and the order of environmental factors was similar for bacteria and fungi. Sucrase activity and pH were significantly correlated with the change of both microbial communities. For Jigu 19, environmental factors posed different effects on bacterial and fungal community variation. Only urease activity had significant correlations with fungal community variation.