3.1 Effects of PASP on soil water storage
The soil water storage data measured at the different PASP concentrations in the cotton field are shown in Fig. 2. The soil water content of the field soil treated with PASP was higher than that of CK1, and PASP could keep the cotton root system at a higher water content level for a long time. The average water content for a PASP treatment was maintained between 40.2% and 51.9%. During the whole physiological period of cotton, the average soil water content increased by more than 4% with the application of appropriate PASP. After the cotton field was drop-irrigated, the soil moisture peaked. PASP increased soil water-holding capacity, reduced soil water loss and prolonged soil water holding period after each irrigation. The soil water content decreased a few days after irrigation, and the decrease speed of PASP treatments was slower than that of CK1. The soil water content was related to the concentration of PASP, with the highest values at 75 kg ha-1 and 150 kg ha-1. The application of 15 kg ha-1 PASP also increased the soil water content, and the effect was much better at the budding stage and flowering stage. Among three concentration, 75 kg ha-1 was the best for maintaining available water content around cotton rhizosphere. At different physiological periods of cotton, PASP had different effects on soil moisture retention. At the seedling stage and boll opening stage of cotton, light had a strong effect on soil water content, and this effect depended on the PASP content. At the budding stage and flowering stage, the water content of the cotton field was higher than that of the other two physiological periods.
3.2 Effect of PASP on microbial α and β-diversity
After the cutting and filtering of reads, the number of bacterial sequences varied from 101118 to 189828 per sample (mean = 124603), whereas the number of fungal sequences varied from 101687 to 151372 per sample (mean = 117631). According to the 16S rRNA (bacteria) and ITS (fungi) genes of soil samples, the characteristic sequences (ASVs) were obtained after dereplication (Tables S1, S2). Sequences were annotated by the silva database and unite database.
The comparison of Chao1, Observed species and the Shannon index showed the effects of PASP on bacterial and fungal diversity (Table S3). The richness and diversity of bacterial and fungal communities were significantly increased by applying PASP. At the seedling stage and bolling stage of cotton, the Chao1, Shannon and Observed species index of bacteria were the highest when the PASP concentration was 75kg ha− 1. At the flowering period, the Shannon and Observed species index of bacteria were the highest with PASP at 15kg ha− 1. At the seedling and bolling stages, of cotton treated with 15kg ha− 1 and 75kg ha− 1, the Chao1, Shannon and Observed species index of fungal communities were higher than those of CK1. At the budding stage, the effect on the fungal communities was the highest when the concentration of PASP was 75kg ha− 1. At the higher concentration of the bottled experiment, the richness and diversity of soil bacterial and fungal communities significantly increased at the flowering and bolling stages, when PASP ratio was 1:1000 and 1:10000. However, PASP significantly reduced the richness and diversity of soil microbial communities, when the concentration of PASP in soils exceeded 1:1000.
Changes in the bacterial and fungal community composition among the different soil samples were estimated by nonmetric multidimensional scaling analysis (NMDS) (Fig. 3). Bacterial communities of samples were far away from each other in the same physiological period of cotton based on the Bray-Curtis distance and greatly affected by the PASP concentration (Fig. 3A). Fungal communities at lower PASP concentrations were separated from those at higher concentrations (Fig. 3B). At the seedling stage of cotton, the similarity of microbial communities of soil samples with adjacent concentrations was higher than those of soil samples treated with markedly different PASP concentrations, but the differences of microbial communities gradually increased with time (Fig. 3). The differences in bacterial communities of PASP4, PASP5 and PASP6 samples were small, but they were significantly different from those of other samples. The aggregation degree of soil fungal communities of PASP4.t1, PASP5.t4, PASP6.t4 and PASP7.t1 samples were relatively discrete, and the composition of fungal communities was quite different from each other. Analysis of NMDS revealed that the influence of the PASP concentration on bacterial β-diversity was higher than the influence on fungal β-diversity.
3.3 Bacterial and fungal community composition
Comparisons of the relative abundance of the 20 dominant bacterial and fungal genera are showed in Fig. 4 and Fig.S2. Bacteria were more sensitive to the application of PASP than fungi. Applying PASP to the cotton field, relative abundances of Pseudomonas, Aeromicrobium, Rokubacteriales, and Nitrospira did not change too much, while those of Methylophaga, Cupriavidus, Panacagrimonas, Sphingomonas, Bacillus, Achromobacter, and Candidimonas changed greatly. The low amount of applied PASP increased the relative abundance of Methylophaga, Sphingopyxis, Panacagrimonas, and Candidimonas. Methylophaga, Panacagrimonas, Pseudomonas and Aeromicrobium were more abundant at the low PASP concentration and in CK soil samples than at high PASP concentrations, while Cupriavidus, Sphingomonas and Acinetobacter were the most abundant bacterial genera at the higher PASP concentration. Most of the abundant genera in all samples were Gram-negative bacteria, aerobic and facultative anaerobic bacteria. The relative abundance of Methylophaga decreased significantly, while the relative abundance of Aeromicrobium increased at the flowering stage compared with the other three cotton physiological periods (Fig. 4A, Fig.S2A).
PASP application did not change too much the relative abundances of Llyonectria and Mycosphaerella, while those of Pseudeurotium, Fusarium, Nectria, Alternaria, Tricharina, Cephalotrichum, and Mortierella changed greatly. Nectria, Pseudeurotium, Mortierella and Cephalotrichum were more abundant fungal genera at low than at high concentrations of PASP, whereas, Fusarium, Aspergillus and Pseudogymnoascus were more abundant at the high PASP concentrations. Applying PASP increased the relative abundance of Fusarium but reduced those of Pseudeurotium and Nectria. At the bolling stage, the relative abundance of Alternaria was significantly higher than at the other three physiological periods (Fig. 4B, Fig.S2B). Therefore, PASP significantly reduced the relative abundance of Methylophaga, Panacagrimonas, Pseudeurotium and Nectria, while increased that of Sphingomonas, Cupriavidus and Fusarium. At the bottling test, with the highest concentration of PASP, the soil microbial changed markedly (Fig. 4). In addition, due to the large span of PASP concentrations in the experiment with bottles, the composition of soil microbial community varied greatly among different PASP concentrations.
3.4 Relations between microbial community composition and soil physiochemical properties
The relationships between soil physiochemical properties and microbial community composition were investigated by Redundancy analysis (RDA). The two-dimensional diagrams of RDA explained 39.54% and 15.4% of the whole bacterial and fungal variance, respectively (Fig. 5). SOM, AN and OP significantly increased by increasing the applied PASP rates, while the effect of PASP on AK was non-significant (Fig. 5, Table S4). SOM (r = 0.474, p = 0.015), AN (r = 0.868, p = 0.001) and OP (r = 0.501, p = 0.004) exhibited a significantly effect on the bacterial community composition with AN having the greatest impact on bacterial communities. Relative abundances of Pseudomonas, Aeromicrobium, and Sphingomonas were correlated with soil physiochemical properties strongly. For example, the relative abundances of Pseudomonas and Aeromicrobium had the strongest negative correlation with AN, SOM, OP and AK, while that of Sphingomonas had a strong positive correlation with AN, AK, OP and SOM. Relative abundances of Panacagrimonas and Cupriavidus had positive correlation with AK and SOM, but had negative correlation with AN, and had a low correlation with OP. The application of PASP changed the composition of microbial communities with greater effects at higher application rates. Samples with similar PASP concentrations had similar composition both PASP4.t1 and PASP5.t1 samples showed the best similarity (Fig. 5A).
AN (r = 0.700, p = 0.001), OP (r = 0.743, p = 0.001) and SOM (r = 0.588, p = 0.005) had a significantly impact on the fungal community composition with OP having the greatest impact on fungal communities. The relative abundances of Tricharina and Cephalotrichum had strong negative correlation with soil physiochemical properties, while those of Panacagrimonas, Cupriavidus, Methylophaga and Sphingopyxis had the weakest correlation with soil physiochemical properties. SOM, OP and AN were positively correlated with the relative abundances of Fusarium, but negatively correlated with that of Pseudeurotium. Nectria and Llyonectria had the strongest positive correlation with AK, while Acremonium and Pseudogymnoascus had the strongest negative correlation with AK. The negative correlation between the relative abundance of Pseudorotium and SOM was the strongest one. The higher PASP application rates had greater effect in fungal diversity than lower PASP application rates. The distance of each sample in the fungal community was close with PASP4.t1, PASP4.t3 and PASP5.t3 having low species composition similarity with the other soil samples.
3.5 Effects of PASP on cotton yield and quality
Cotton yield and quality were affected by PASP under field conditions (Table 1). Seed cotton yield and single boll weight were higher in PASP treatments than those of CK1. The application of PASP increased the boll number and fruit branch number of cotton at flowering stage, and the reflectance degree. The most significant (p < 0.05) effects by PASP were on Micronaire and short fiber index of cotton, and the effect on the reflectance degree. Compared to CK1, single boll weight increased by 2.1%, 5.48% and 10.14%, and seed cotton yield increased by 3.94%, 8.31% and 7.71% at the PASP concentrations of 15 kg ha− 1, 75 kg ha− 1 and 150 kg ha− 1, respectively. The maximum seed cotton yield of 6977.38kg ha− 1 was achieved with 75 kg ha− 1 of applied PASP.
Table 1
Effects of different PASP application rates on cotton yield and fiber quality
PASP rate
(kg ha− 1)
|
Fiber quality
|
Cotton yield
|
Rd
|
MIC
|
LEN(mm)
|
SFI (%)
|
UI (%)
|
STR(cN/tex)
|
BN
|
FBN
|
SBW(g)
|
SCY(kg ha− 1)
|
15
|
78.37 ± 0.27c
|
4.81 ± 0.15a
|
27.37 ± 0.36a
|
11.67 ± 0.62ab
|
83.33 ± 0.35ab
|
27.09 ± 0.66b
|
10.17 ± 0.39a
|
6.87 ± 0.28a
|
6.34 ± 0.08bc
|
6696.27 ± 218.62a
|
75
|
79.10 ± 0.12a
|
4.97 ± 0.07a
|
27.66 ± 0.40a
|
13.07 ± 1.31a
|
82.97 ± 0.23b
|
26.51 ± 0.64b
|
10.37 ± 0.43a
|
7.03 ± 0.24a
|
6.55 ± 0.08b
|
6977.38 ± 155.86a
|
150
|
79.07 ± 0.17ab
|
4.99 ± 0.13a
|
27.92 ± 0.43a
|
12.13 ± 0.28ab
|
83.23 ± 0.48ab
|
27.19 ± 0.02ab
|
10.23 ± 0.41a
|
6.87 ± 0.28a
|
6.84 ± 0.09a
|
6939.34 ± 260.12a
|
CK1
|
78.53 ± 0.09bc
|
4.89 ± 0.11a
|
28.62 ± 0.44a
|
9.70 ± 0.78b
|
84.03 ± 0.09a
|
28.70 ± 0.24a
|
9.10 ± 0.35b
|
6.53 ± 0.27a
|
6.21 ± 0.09c
|
6442.33 ± 193.80a
|
Values of cotton yield and quality were mean of three replicates ± SE; n = 30. a, b and c indicate significant differences (p < 0.05; LSD test) in the same column. Rd, Reflectance degree; MIC, Micronaire; LEN, Length; SFI, Short fiber index; UI, Uniformity index; STR, Strength; BN, Boll numbers per plant; FBN, Fruit branches numbers; SBW, Single boll weight; SCY, seed cotton yield. |