Effects of Water-Fertilizer-Air-Coupling Drip Irrigation on Soil Health Status: Soil Aeration, Enzyme Activities and Microbial Biomass

: In order to investigate the effects of water-fertilizer-air-coupling drip irrigation on soil health status, including soil aeration (SA), enzyme activity (EA) and microbial biomass (MB), and its response relationship, this glasshouse experiment was conducted using tomato as the test crop


Introduction
Water-fertilizer-air-coupling drip irrigation involves the loss of less water to evaporation, resulting in higher water-use efficiency and nutrient-use efficiency [1]. In recent years, there have been many studies on water-fertilizer-air-coupling irrigation to save water and increase crop yields in China and throughout the world [2][3][4][5]. However, under drip irrigation, the infiltration of irrigation water will displace the air in soil pores, resulting in a decrease in soil permeability [6]. Moreover, long-term water-fertilizer-integrated drip irrigation can easily lead to soil compaction and acidification, inhibit the absorption of nutrients by the roots, and cause other problems. This can cause a decrease in soil aeration and permeability and the physiological metabolism of roots [7], which restricts the potential of crops to produce high yields [8]. Therefore, in addition to focusing on the crop water and fertilizer conditions, the degree of aeration should not be ignored.
Based on the development of drip irrigation systems, water-fertilizer-air-coupling drip irrigation technology is an agricultural engineering technology that saves water and fertilizer, produces high yields and is highly efficient, which enables the synchronization

Experimental Site
The experiment was conducted in a modern glasshouse at the North China University of Water Resources and Electric Power (Zhengzhou, China) ( Figure 1). This test station (34 • 47 5.91 N, 113 • 47 20.15 E) is located in the sub-humid warm temperate zone, with an average annual temperature of 14.3 • C, an average annual rainfall of 632 mm, a frost-free period of 220 days and annual sunshine of approximately 2400 h. The total construction area of the glasshouse was 537.6 m 2 , the span was 9.6 m, and the open room was 4 m. In the glasshouse, there were fans and wet curtains in the south and north, respectively. The average temperature during the tomato growth period was 13.2~23.4 • C, and the relative humidity was 34.5% to 78.9%. The tomato variety used for the experiment was Jinpeng 8. The soil type was loamy clay, and the organic matter content was 1.36%. The contents of available nitrogen, phosphorus and potassium were 73.22, 15.31 and 89.65 mg·kg −1 , respectively. The field water holding rate was 28% by mass, and the pH value was 6.5. The soil bulk density of the 0-30 cm soil layer was 1.26, 1.48 and 1.50 g·cm −3 , respectively. The contents of soil sand, silt and clay were 32.99%, 34.03% and 32.98%, respectively.
In this experiment, tomato was used as the test crop, and common subsurface drip irrigation was used as the control. The soil aeration, enzyme activity and microbial biomass were systematically monitored under different increments of fertilization, irrigation and oxygen, and the effects of the factors and their interaction with soil aeration, enzyme activity and microbial biomass under water-fertilizer-air-coupling drip irrigation were studied. The partial least squares structural equation model (PLS-SEM) analysis of waterfertilizer-air-coupling drip irrigation, soil aeration, soil enzyme activity and soil microbial biomass was constructed to reveal the relationship between water-fertilizer-air-coupling drip irrigation and their response. These results provide a theoretical basis for the improvement of soil health status by water-fertilizer-air-coupling drip irrigation.

Experimental Site
The experiment was conducted in a modern glasshouse at the North China University of Water Resources and Electric Power (Zhengzhou, China) ( Figure 1). This test station (34 ° 47′5.91′′ N, E113 ° 47′20.15') is located in the sub-humid warm temperate zone, with an average annual temperature of 14.3 °C, an average annual rainfall of 632 mm, a frost-free period of 220 days and annual sunshine of approximately 2400 h. The total construction area of the glasshouse was 537.6 m 2 , the span was 9.6 m, and the open room was 4 m. In the glasshouse, there were fans and wet curtains in the south and north, respectively. The average temperature during the tomato growth period was 13.2~23.4 °C, and the relative humidity was 34.5% to 78.9%. The tomato variety used for the experiment was Jinpeng 8. The soil type was loamy clay, and the organic matter content was 1.36%. The contents of available nitrogen, phosphorus and potassium were 73.22, 15.31 and 89.65 mg·kg −1 , respectively. The field water holding rate was 28% by mass, and the pH value was 6.5. The soil bulk density of the 0-30 cm soil layer was 1.26, 1.48 and 1.50 g·cm −3 , respectively. The contents of soil sand, silt and clay were 32.99%, 34.03% and 32.98%, respectively.

Irrigation System Description
A glasshouse plot experiment was used to establish a completely randomized block deign with 3 factors (fertilizer rate, irrigation volume and oxygenation capacity) and 2 levels, respectively. There were eight treatments (Table 1), and each treatment was repeated four times. Each plot was 2 m long and 1 m wide, and there were 32 plots in total. Each plot was repeated once. John Deere underground drip irrigation tubing (Moline, IL,

Irrigation System Description
A glasshouse plot experiment was used to establish a completely randomized block deign with 3 factors (fertilizer rate, irrigation volume and oxygenation capacity) and 2 levels, respectively. There were eight treatments (Table 1), and each treatment was repeated four times. Each plot was 2 m long and 1 m wide, and there were 32 plots in total. Each plot was repeated once. John Deere underground drip irrigation tubing (Moline, IL, USA) was used to supply the water. The drip irrigation belt was 16 mm in diameter; the wall thickness was 0.6 mm; the designed flow rate of the dripper was 1.2 L·h −1 ; the distance between emitters was 33 cm, and the maximum working pressure was 202.65 kPa. The ridge height was 10 cm above the soil surface, and the depth of the drip irrigation belt was 10 cm. In each plot, five tomato seedlings were transplanted on the ridge, and the distance between the Agronomy 2022, 12, 2674 4 of 16 plants was 33 cm. Each distinct water supply pipe was controlled separately and equipped with a water meter to record the amount of irrigation water.

Experimental Design
The tomato plants were transplanted on 14 September 2018 and harvested on 4 January 2019. The crop was irrigated on the day of transplanting and covered with plastic film on the 7th day after transplanting, irrigated treatment occurred on the 14th day and the crop was irrigated every 5-7 days. When the plant was 30-40 cm high, the vine was hung, and tomatoes were topped after three-panicle fruit. The whole growth period of the tomato is 110 days. The growth period was divided into four stages, i.e., the seedling stage , the flowering and fruit-setting stage (12)(13)(14)(15)(16)(17)(18)(19)(20)(21)(22)(23)(24)(25)(26)(27)(28)(29), the fruit expansion stage (30 October-3 December) and the maturation stage (4 December-3 January). A water meter was used to measure the quantity of water supply in each block. Nonaeration drip irrigation treatment used well water for irrigation with an independent water supply device, while aeration drip irrigation used aerated water by the venturi air injector (Mazzei air injector 684; Mazzei Injector Corporation, Bakersfield, CA, USA). Among them, the basic PH value of irrigation water as well water was 7.83, the chemical oxygen demand (COD) was 8.7 mg·L −1 , the five-day biochemical oxygen demand (BOD 5 ) was 4.2 mg·L −1 , the conductivity was 663 µs·cm −1 , the total nitrogen was 0.55 mg·L −1 , the total phosphorus was 0.02 mg·L −1 , the total potassium was 5.02 mg·L −1 , the total calcium was 16.3 mg·L −1 , the total magnesium was 1.32 mg·L −1 , and the total sodium was 35.4 mg·L −1 . In this experiment, aerated water (aeration for 20 min) with an air entrainment ratio of approximately 15% was prepared using a water storage pipeline, circulation pump and the venturi air jet device. The water supply pressure of each district was 0.10 MPa according to the drip meter that metered the irrigation. The amount of irrigation water was calculated as follows [29]: where I is the irrigation amount per time of irrigation event (L); A is the plot area (m 2 ), which was 2 m 2 in this study; E P is the water depth evaporated from the E-601 pan during the interval of two irrigation events (mm); K P is the evaporation pan coefficient of crops; the low rate is 0.6, and the high rate is 1.0. The evaporation and water probe readings of the E-601 standard evaporating pan were measured every day at 8:00-9:00. The irrigation period was from 5-7 days, and the plots were irrigated from 8:00-15:00.
Water-soluble fertilizer (20% N; 20% P 2 O 5 , and 20% K 2 O) was applied twice during the entire growth period. Each application utilized half of the fertilizer. The plants were fertilized on the 48th day and the 67th day after transplanting. The water-soluble fertilizer was mixed into the water flow with the fertilizer applicator and mixed in the water tank. The dissolved oxygen in the soil was measured using an optical fiber oxygen sensor (OXY4-mini; PreSens, Regensburg, Germany). The probe was installed in the radical 5 cm from the crop stem and in the vertical 15 cm from the soil surface. The concentrations of dissolved oxygen were measured at 9:00 and 15:00 on the 1st day after the flowering and fruit-setting stage (on the 41st-45th day after transplanting), fruit expansion stage (on the 61st-65th day after transplanting) and maturation stage (on the 81st-85th day after transplanting).

Soil Oxygen Diffusion Rate
The oxygen diffusion rate was measured synchronously with the dissolved oxygen measurements. The platinum electrode, reference electrode and copper pair electrode were buried separately in each plot. The electrode was installed in the radical 5 cm from the crop stem and in the vertical 15 cm from the soil surface. The diffusion rate of oxygen in the soil was determined by a depolarization method with an automatic instrument (Shanghai Instrument & Electric Instrument Co., Ltd., Shanghai, China).

Soil Respiration Rate
The soil respiration rate was measured using an ADC LCi-SD soil respiration system (Delta-T Devices, Ltd., Cambridge, UK). The measured time was consistent with that of the dissolved oxygen. The base of the soil respiration chamber was buried in the soil to be sampled before the measurement. The reading was measured after the chamber had stabilized (2~3 min).

Soil Enzyme Activity
Soil samples were collected on the 42nd day during the flowering and fruit-setting stage, on the 62nd day during the fruit expansion stage and on the 111st day during the maturation stage after transplanting to determine the activities of urease, catalase and phosphatase in tomato rhizosphere soil, respectively. One plant was randomly selected from each plot, and the loose soil on the roots was shaken off, and the root topsoil that was bound to the roots was gently brushed off with a soft brush. It was then placed in a sterilized plastic bag as the rhizosphere soil sample. These were taken three times in the plot and mixed as the result of the sampling, and then immediately taken back to the laboratory. Plant residues were removed from the fresh soil, and the soil samples were passed through a 2 mm sieve and stored at 4 • C to determine the activities of catalase, urease and phosphatase. The soil catalase activity was determined by the KMnO 4 titration method, and the soil moisture content was determined by the drying method. The enzyme activity in each gram of dry soil was calculated, expressed by the milliliters of KMnO 4 solution consumed by 1 g of soil, and the unit was mL·g −1 . The soil urease activity was measured by the phenol-sodium hypochlorite colorimetric method and expressed as the mass of NH 4 + −N produced within 24 h in 1 g of soil, and the unit was mg·g −1 ·d −1 . The soil phosphatase activity was measured by phenyl disodium phosphate colorimetry and expressed as the mass of phenol diluted from 1 g of soil within 24 h, and the unit was mg·g −1 ·d −1 [30].

Soil Microbial Biomass
In the preparatory experiment, it was found that the effect of water-fertilizer-aircoupling drip irrigation on soil microbial biomass during the flowering and fruit-setting stage and fruit expansion stage of the tomato plant was weak, so the cumulative value of soil microbial biomass on the 111st day after tomato transplanting was selected for measurement. The numbers of bacteria, fungi and actinomycetes were determined by the dilution plate spreading method. The soil in the root zone was collected and placed into sterilized plastic bags, and the soil suspension was prepared with sterile water. The same soil sample was inoculated with three consecutive dilutions, and each dilution was repeated three times. Beef extract peptone agar medium was used for the bacteria, Martin Bengal red agar medium for the fungi and modified Coles 1 medium for the actinomycetes as previously described [31].

Statistical Analysis
The data were analyzed and drawn using Microsoft Excel 2021 (Redmond, WA, USA), and the interactive analysis of variance (ANOVA) and Tukey methods were used for the significance tests (SPSS v. 22.0; IBM, Inc., Armonk, NY, USA). The partial least squaresstructural equation model (PLS-SEM) analysis was conducted using SmartPLS GmbH (v. 3.0; Oststeinbek, Germany).

Effects on Soil Aeration
Preliminary research showed that there was no significant difference between the dissolved oxygen concentration, oxygen diffusion rate and soil respiration rate among the different treatments of fertilization rates. Thus, F2 fertilization rates were selected for analysis, which took place at the flowering and fruit-setting stage, fruit expansion stage and maturation stage ( Figure 2).
The soil dissolved oxygen decreased immediately after irrigation and then increased gradually during the flowering and fruit-setting stage, fruit expansion stage and maturation stage in all the treatments (Figure 2a-c). Comparing the mean soil dissolved oxygen values of each irrigation cycle, it was found that the increase in irrigation water significantly reduced the soil dissolved oxygen; under O1 and O2 levels, the soil dissolved oxygen of W2 treatment decreased by 1.4% and 1.9% on average compared with the W1 treatment, respectively (p < 0.05). The increase in oxygen content significantly increased the soil dissolved oxygen, under W1 and W2 levels, the soil dissolved oxygen of O2 treatment increased by 9.8% and 9.3% on average compared with the O1 treatment, respectively (p < 0.05).
Similar to the dissolved oxygen, the oxygen diffusion rate values of each treatment decreased to their lowest values and then gradually increased (Figure 2d-f). The increase in oxygen content significantly increased the soil oxygen diffusion rate, under W1 and W2 levels; the soil oxygen diffusion rate of O2 treatment increased by 11.5% and 20.0% on average compared with the O1 treatment, respectively (p < 0.05). The effects of different irrigation water on the oxygen diffusion rate was not significant (p ≥ 0.05).
The soil respiration values were highest in W2O2 whereas as it was lowest in W2O1, and the soil respiration rate increased with the increase in soil temperature (Figure 2g-i). The rate of soil respiration was significantly enhanced by the aeration treatment. The effect was most significant on the afternoon of the second day after irrigation, in the flowering and fruit-setting stage, W1O2 and W2O2 increased by 41.63% and 59.84%, respectively, compared with W1O1 and W2O1. In the fruit expansion stage, it was increased by 36.31% and 31.52%, respectively. In the maturation stage, it increased by 79.40% and 53.74%, respectively (p < 0.05). The irrigation water had no significant effect on the rate of soil respiration (p ≥ 0.05).

Effects on Soil Urease Activity of Tomato in Glasshouse Plots
At the flowering and fruit-setting stage and fruit expansion stage, F, W, O single factor and O, W interaction had significant effects on soil urease activity (p < 0.05). At the maturity stage, F, O, W single factor, O, W interaction and F, O, W interaction had significant effects on soil urease activity (p < 0.05) (Figure 3).
The activity of tomato rhizosphere soil urease increased first and then decreased with the advancement of the growth period. At flowering and fruit-setting stage and fruit expansion stage, the urease activity was highest with F2W2O2 treatment, which were 0.33 and 0.36 mg·g −1 ·d −1 , respectively, and at the maturation stage, F2W102 treatment was the highest at 0.28 mg·g −1 ·d −1 . At the flowering and fruit-setting stage and fruit expansion stage, the overall change trend of urease activity was manifested as the increase in urease activity with the increase in fertilization amount, irrigation water and dissolved oxygen volume, but the change trend was different in the maturation stage, with the increase in fertilization amount and dissolved oxygen, the activity of urease gradually increased, and gradually decreased with the increase in irrigation water, which indicates that reasonable irrigation water can improve soil enzyme activity, and excessive soil moisture was not conducive to the growth and reproduction of soil microorganisms, resulting in a decrease in soil urease activity.
agar medium for the fungi and modified Coles 1 medium for the actinomycetes as previously described [31].

Statistical Analysis
The data were analyzed and drawn using Microsoft Excel 2021 (Redmond, WA, USA), and the interactive analysis of variance (ANOVA) and Tukey methods were used for the significance tests (SPSS v. 22.0; IBM, Inc., Armonk, NY, USA). The partial least squares-structural equation model (PLS-SEM) analysis was conducted using SmartPLS (v. 3.0).

Effects on Soil Aeration
Preliminary research showed that there was no significant difference between the dissolved oxygen concentration, oxygen diffusion rate and soil respiration rate among the different treatments of fertilization rates. Thus, F2 fertilization rates were selected for analysis, which took place at the flowering and fruit-setting stage, fruit expansion stage and maturation stage (Figure 2).  The soil dissolved oxygen decreased immediately after irrigation and then increased gradually during the flowering and fruit-setting stage, fruit expansion stage and maturation stage in all the treatments (Figure 2a-c). Comparing the mean soil dissolved oxygen values of each irrigation cycle, it was found that the increase in irrigation water significantly reduced the soil dissolved oxygen; under O1 and O2 levels, the soil dissolved oxygen of W2 treatment decreased by 1.4% and 1.9% on average compared with the W1 treat-

Effects on Soil Urease Activity of Tomato in Glasshouse Plots
At the flowering and fruit-setting stage and fruit expansion stage, F, W, O single The activity of tomato rhizosphere soil urease increased first and then decreased with the advancement of the growth period. At flowering and fruit-setting stage and fruit expansion stage, the urease activity was highest with F2W2O2 treatment, which were 0.33 and 0.36 mg·g −1 ·d −1 , respectively, and at the maturation stage, F2W102 treatment was the highest at 0.28 mg·g −1 ·d −1 . At the flowering and fruit-setting stage and fruit expansion stage, the overall change trend of urease activity was manifested as the increase in urease activity with the increase in fertilization amount, irrigation water and dissolved oxygen volume, but the change trend was different in the maturation stage, with the increase in fertilization amount and dissolved oxygen, the activity of urease gradually increased, and gradually decreased with the increase in irrigation water, which indicates that reasonable irrigation water can improve soil enzyme activity, and excessive soil moisture was not conducive to the growth and reproduction of soil microorganisms, resulting in a decrease in soil urease activity.

Effect on Soil Catalase Activity of Tomato in A Glasshouse Plot
At the flowering and fruit-setting stage, F and O single factor and their interaction had significant effects on the soil catalase activity (p < 0.05). At the fruit expanding stage, F and O single factor and their interaction, and O and W interaction had significant effects on the soil catalase activity (p < 0.05). At the maturity stage, F, W, O single factor and O, W interaction had significant effects on the soil catalase activity (p < 0.05) (Figure 4).   Figure 3. Analysis of variance of soil urease activity in the tomato root zone under different treatments. Note: F represents the fertilizer amount, O represents the oxygen amount, and W represents the irrigation amount. Different letters indicate significant differences at the level of p < 0.05. * p < 0.05. ** p < 0.01. ns, not significant at p < 0.05.

Effect on Soil Catalase Activity of Tomato in A Glasshouse Plot
At the flowering and fruit-setting stage, F and O single factor and their interaction had significant effects on the soil catalase activity (p < 0.05). At the fruit expanding stage, F and O single factor and their interaction, and O and W interaction had significant effects on the soil catalase activity (p < 0.05). At the maturity stage, F, W, O single factor and O, W interaction had significant effects on the soil catalase activity (p < 0.05) (Figure 4).
The activity of soil catalase increased first and then decreased with the advancement of the growth period. At the same growth stage, catalase activity gradually increased with the increase in fertilization rate and dissolved oxygen amount, while the increase in irrigation water had different effects in different growth periods. At the flowering and fruit-setting stage, the effect of irrigation water on soil catalase activity was not significant, and at the fruit expansion stage, under O2 levels, the increase in irrigation water volume would increase soil catalase activity; F1W2O2 and F2W2O2 increased by 15.35% and 11.33% compared with the W1 treatment, respectively (p < 0.05), but under O1 level, the corresponding treatment reduced the soil catalase activity and F1W2O1 decreased by 20.19% compared with F1W1O1, respectively (p < 0.05). At the maturation stage, the increase in irrigation water volume reduced the activity of soil catalase, F1W2O1 and F2W2O1 decreased by 42.88% and 32.00% compared with the W1 treatment, respectively (p < 0.05). There was no significant difference in the rest (p ≥ 0.05).

Effect on the Activity of Soil Phosphatase of Tomato in the Glasshouse Plot
At the flowering and fruit-setting stage, F and O single factors had significant effects on soil phosphatase activity (p < 0.05), and at the fruit expansion and maturity stage, F, O and W single factors had significant effects on soil phosphatase activity (p < 0.05) ( Figure 5).
The activity of soil phosphatase increased first and then decreased with the advancement of the growth period, and the F2W2O2 treatment was the highest in the three growth periods, which were 0.07, 0.09 and 0.06 mg·g −1 ·d −1 , respectively. At the same growth stage, the phosphatase activity increased with the increase in fertilization, irrigation and dissolved oxygen, and the oxygenation treatment had a more substantial impact on the activity of soil phosphatase. The activity of soil catalase increased first and then decreased with the advancement of the growth period. At the same growth stage, catalase activity gradually increased with the increase in fertilization rate and dissolved oxygen amount, while the increase in irrigation water had different effects in different growth periods. At the flowering and fruitsetting stage, the effect of irrigation water on soil catalase activity was not significant, and at the fruit expansion stage, under O2 levels, the increase in irrigation water volume would increase soil catalase activity; F1W2O2 and F2W2O2 increased by 15.35% and 11.33% compared with the W1 treatment, respectively(p < 0.05), but under O1 level, the corresponding treatment reduced the soil catalase activity and F1W2O1 decreased by 20.19% compared with F1W1O1, respectively(p < 0.05). At the maturation stage, the increase in irrigation water volume reduced the activity of soil catalase, F1W2O1 and F2W2O1 decreased by 42.88% and 32.00% compared with the W1 treatment, respectively (p < 0.05). There was no significant difference in the rest (p ≥ 0.05).

Effect on the Activity of Soil Phosphatase of Tomato in the Glasshouse Plot
At the flowering and fruit-setting stage, F and O single factors had significant effects on soil phosphatase activity (p < 0.05), and at the fruit expansion and maturity stage, F, O and W single factors had significant effects on soil phosphatase activity (p < 0.05) ( Figure  5).
The activity of soil phosphatase increased first and then decreased with the advancement of the growth period, and the F2W2O2 treatment was the highest in the three growth periods, which were 0.07, 0.09 and 0.06 mg·g −1 ·d −1 , respectively. At the same growth stage, the phosphatase activity increased with the increase in fertilization, irrigation and dissolved oxygen, and the oxygenation treatment had a more substantial impact on the activity of soil phosphatase.     Figure 5. Analysis of variance of soil phosphatase activity in the tomato root zone under different treatments. Note: F represents the fertilizer amount, O represents the oxygen amount, W represents the irrigation amount. Different letters indicate significant differences at the level of p < 0.05. ** p < 0.01. ns, not significant at p < 0.05.

Effect on Soil Microbial Biomass
F, W and O single factors have significant effects on the number of bacteria (p < 0.05). F, O, W single factor, and O, W interaction had significant effects on the number of fungi (p < 0.05). O, W single factor and their interaction, and F, W interaction had significant effects on the number of actinomycetes (p < 0.05) (Figure 6).

Improvement Analysis of Soil Health Status Based on PLS-SEM
The structural equation model can visualize the relationship between soil aeration, enzyme activities and microbial biomass, and calculate the correlation coefficient of each relationship. The generalized least squares method was used to construct the water-fertilizer-air-coupling drip irrigation, soil aeration, soil enzyme activity, soil microbial biomass model, and the model was revised and tested to eliminate the paths with no significant effect and no important practical significance. The optimized adaptation model is shown in Figure 7. The model checking indicators are as follows: The goodness of fit (GoF) was calculated as follows: where ̅̅̅̅̅̅ is the average value of the extracted mean variance, which is 0.824, and 2 ̅̅̅̅ is the average value of the coefficient of determination, which is 0.625. The GoF is considered as the most important fitness index information in PLS-SEM, which is between 0 and 1. A higher value indicates a better fitness of the model. Typically   Figure 6. Analysis of variance of soil microbial biomass in the tomato root zone under different treatments. Note: F represents the fertilizer amount, O represents the oxygen amount, W represents the irrigation amount. Different letters indicate significant differences at the level of p < 0.05. * p < 0.05. ** p < 0.01. ns, not significant at p < 0.05.
The number of bacteria, fungi and actinomycetes was the highest in the F2W2O2 treatment, which were 2.63 × 10 6 , 13.67 × 10 6 and 4.13 × 10 6 ·g −1 , respectively, and their number gradually increased with the increase in irrigation amount and dissolved oxygen. The effect of fertilization amount on microbial biomass was not significant (p ≥ 0.05).

Improvement Analysis of Soil Health Status Based on PLS-SEM
The structural equation model can visualize the relationship between soil aeration, enzyme activities and microbial biomass, and calculate the correlation coefficient of each relationship. The generalized least squares method was used to construct the waterfertilizer-air-coupling drip irrigation, soil aeration, soil enzyme activity, soil microbial biomass model, and the model was revised and tested to eliminate the paths with no significant effect and no important practical significance. The optimized adaptation model is shown in Figure 7. The model checking indicators are as follows: The goodness of fit (GoF) was calculated as follows: where AVE is the average value of the extracted mean variance, which is 0.824, and R 2 is the average value of the coefficient of determination, which is 0.625.
12, x FOR PEER REVIEW 12 of 17 Figure 7. Influence path diagram of water-fertilizer-air-coupling drip irrigation on soil health status. Note: The value in the ellipse represents the determination coefficient R 2 value, the ellipse represents the latent variable, the box represents the observed variable, the red arrow indicates a significant positive effect, the blue arrow indicates a significant negative effect, the arrow line thickness indicates the size of the path coefficient in the model, and the dotted line indicates no significant effect. The value on the real line between the latent variable and the observed variable represents the external load value, * p < 0.05. ** p < 0.01.
The adaptation index of the PLS-SEM GoF = 0.718 indicates a better adaptation effect of the model. The estimated value of the outer loadings was >0.7, all of which reach the acceptable level. To ensure the validity of data, composite reliability (CR) was used to test the reliability of each latent variable. The CR values of EA, MB and SA were 0.902, 0.941 and 0.953, respectively, and their absolute values were all >0.7, indicating that the data used were highly reliable. Cronbach's Alpha (CA) values of EA, MB and SA were 0.842, 0.907 and 0.927, respectively, and their absolute values were all >0.7 [33], which indicated that they had a high convergence validity. Average variance extracted (AVE) of the four latent variables were all >0.5 [34], and the square root value of each AVE on the diagonal was greater than the correlation of the correlation dimension (Table 2). This model passed the reliability and validity tests and meets the needs of modeling water-fertilizer-air-coupling drip irrigation soil aeration, enzyme activities and microbial biomass.
The results in Figure 7 showed that the explanatory degree of endogenous variables R 2 was 0.25, 0.50 and 0.75 indicate weak, medium and high prediction accuracy. In this study, the R 2 values of SA, EA and MB were 0.807, 0.701 and 0.368, respectively. Environmental factors could explain SA and EA well, while MB could be explained weakly, indicating that there were other factors influencing MB. WFA had significant positive effects on MB (1.510), EA (0.903) and SA (0.898), while SA had significant negative effects on MB (−0.696), and EA had significant negative effects on MB (−0.518). Comprehensive analysis showed that water-fertilizer-air-coupling drip irrigation could effectively improve soil health status, and had the greatest effect on MB, followed by EA and SA.   The GoF is considered as the most important fitness index information in PLS-SEM, which is between 0 and 1. A higher value indicates a better fitness of the model. Typically, 0.1, 0.25 and 0.36 are used to indicate poor, good and excellent model adaptation, respectively [32].
The adaptation index of the PLS-SEM GoF = 0.718 indicates a better adaptation effect of the model. The estimated value of the outer loadings was >0.7, all of which reach the acceptable level. To ensure the validity of data, composite reliability (CR) was used to test the reliability of each latent variable. The CR values of EA, MB and SA were 0.902, 0.941 and 0.953, respectively, and their absolute values were all >0.7, indicating that the data used were highly reliable. Cronbach's Alpha (CA) values of EA, MB and SA were 0.842, 0.907 and 0.927, respectively, and their absolute values were all >0.7 [33], which indicated that they had a high convergence validity. Average variance extracted (AVE) of the four latent variables were all >0.5 [34], and the square root value of each AVE on the diagonal was greater than the correlation of the correlation dimension (Table 2). This model passed the reliability and validity tests and meets the needs of modeling water-fertilizerair-coupling drip irrigation soil aeration, enzyme activities and microbial biomass. The results in Figure 7 showed that the explanatory degree of endogenous variables R 2 was 0.25, 0.50 and 0.75 indicate weak, medium and high prediction accuracy. In this study, the R 2 values of SA, EA and MB were 0.807, 0.701 and 0.368, respectively. Environmental factors could explain SA and EA well, while MB could be explained weakly, indicating that there were other factors influencing MB. WFA had significant positive effects on MB (1.510), EA (0.903) and SA (0.898), while SA had significant negative effects on MB (−0.696), and EA had significant negative effects on MB (−0.518). Comprehensive analysis showed that water-fertilizer-air-coupling drip irrigation could effectively improve soil health status, and had the greatest effect on MB, followed by EA and SA.

Effects on Soil Aeration in the Root Zone of Crops
Gaseous oxygen in soil pores and liquid oxygen in the soil solution are the main components of oxygen in soil. The supply of oxygen to roots in the root zone can be divided into three processes: first, oxygen diffuses from the atmosphere to the air-filled pores of the soil. Secondly, oxygen diffuses to the root surface in the form of soil dissolved oxygen through water films. Finally, oxygen enters the plant from the surface of root system. The dissolved oxygen concentration in the soil solution was greatly affected by water-fertilizer-air-coupling drip irrigation. The dissolved oxygen concentration of the soil solution enriched with oxygen was significantly higher than that of the control (Figure 2a-c). Bhattarai et al. studied the characteristics of dissolved oxygen in the rhizosphere soil of tomatoes grown in clayey soil under increasing amounts of oxygenated irrigation water and showed that the concentration of dissolved oxygen in the soil increased by 12%, which was generally consistent with the conclusion of this experiment [35].
The oxygen diffusion rate can reflect the availability of oxygen to plants and is the most representative index of soil ventilation. It is generally considered that the oxygen diffusion rate cannot meet the normal demand of crops when it is below the threshold value of 0.4 µg·cm −2 ·min −1 . Severe dyspnea occurs below 0.2 µg·cm −2 ·min −1 [36]. The oxygenated irrigation water contains many oxygen-containing substances, which can maintain a good soil oxygen environment and then enhance the oxygen diffusion rate, and the improvement effect lasts at least 24 h (Figure 2d-f). Aerobic treatment, which uses a two-phase flow of water and air to transport the tiny air or oxygen bubbles and water to the root zone of the crop, can alleviate the anoxic environment of the soil caused by irrigation, which could be the reason for the significant increase in aerobic oxygen diffusion rate compared with the control.
Soil respiration rate is the primary manner of gas exchange between the soil and the atmosphere, which primarily originates from the autotrophic respiration of crop roots and heterotrophic respiration of soil microorganisms. Aerobic irrigation improved the oxygen diffusion rate and dissolved oxygen in the soil and enhanced the autotrophic respiration of the roots [13], increased the growth of aerobic microorganisms and soil enzyme activity, and improved heterotrophic respiration. In the current study, the soil respiration rate of the aeration treatment after irrigation was higher than that of the control treatment, and it was higher in the afternoon. This parameter correlated with temperature (Figure 2g-i), but the effects of the amount of irrigation on the soil respiration rate was not readily apparent; the reason was that oxygen was limited when the soil pores were filled with water, interfering with the ability of soil organisms to respire.

Effects on Soil Enzyme Activity and Microbial Biomass in the Root Zone
The soil enzyme activity reflects the ability of soil nutrient transformation and is a potential index to maintain soil fertility. In this study, the activities of urease, catalase and phosphatase in the soil were highest at the fruit expansion stage and lowest at the maturation stage. These parameters showed a trend of first increasing and then decreasing during the entire growth stage. This could be owing to the sufficient soil nutrients in the early stage of crop growth, the constant increase in rhizosphere soil enzyme activity, and medium-term fertilization. The activities of soil enzyme reached their maximum. The later soil nutrients were gradually consumed, and the activity of soil enzymes decreased. This was not consistent with the study by Li et al. that showed that soil enzyme activities decreased first and then increased under the condition of muskmelon planting under aerated irrigation [37]. The activity was also affected by the soil nutrients, seasons, management practices and other conditions [38,39]. The soil urease activity of soil treated with oxygen improved significantly during the fruit expansion stage and maturation stage, and the soil urease activity of soil treated with water was also positively affected during the flowering and fruit-setting stage and fruit expansion stage, but was opposite in the maturation stage. The effects of the fertilization amount on the soil urease activity was significant only under the low water control. Li et al. pointed out that aeration in the root zone could improve soil urease activity, and the effect of improvement was most significant near the buried depth of the drip irrigation zone, which was consistent with the effects of aeration on soil urease activity in this study [37]. The results showed that the activity of catalase in the soil increased significantly during the three periods of aeration compared with the control, which was consistent with the conclusion of Zhang et al. that the soil catalase activity increased after root zone aeration [40]. The effect of irrigation varied at different growth stages. The activity of catalase increased with the increase in oxygenated water during the fruit expansion stage, and the activity of catalase in the soil decreased with the increase in irrigation water during the control period. Appropriate water can increase the activities of soil enzymes, but too much water will reduce the activity of catalase in the soil as shown by Zhu et al. [41]. The increase in irrigation water in the control treatment still decreased the activity of soil catalase during the mature period, but the oxygen enrichment treatment improved the anoxic condition of the root zone soil caused by irrigation. Thus, the activity of catalase in the soil treated with a high water content was not significantly lower than that of the control. The effect of the irrigation amounts on soil phosphatase activity was significantly weaker than that of oxygenation, and there was a certain difference only in the maturity stage, and the amount of phosphatase activity in the oxygenation treatment of high water quantity was increased. The effect of fertilization amount on soil phosphatase activity was significant in the fruit expansion and maturity stage, and the high fertilization amount treatment increased the soil phosphatase activity.
Microbial composition, growth and functional microbial community dynamics are affected by the soil environment but also by crop root growth [42,43]. The numbers of bacteria and fungi in oxygen-enriched treatment was significantly higher than that in the control, and the number of bacteria and fungi in high irrigation treatment was also higher than that in low irrigation treatment. The number of actinomycetes under water-fertilizerair-coupling drip irrigation was not significantly affected by bacteria and fungi. There was no significant difference in the numbers of microorganisms in the soil between the two fertilization treatments.

Responses of Soil Health Status to Water-Fertilizer-Air-Coupling Drip Irrigation
The lack of aeration in the root zone leads to a decrease in the content of oxygen in the soil, and the increase in content of hydrogen peroxide in root cells under hypoxia stress leads to the inhibition of cellular activity [44]. The level of enzyme activity in the soil can reflect its biological activity, soil quality and health conditions [45]. In soil crop systems, the root zone soil contains many beneficial bacterial communities, including plant growth hormone producing bacteria and nitrogen-fixing bacteria. Fungi improve soil structure by breaking down lignocellulose. Actinomycetes are the primary biocontrol microorganisms in the soil, and they can produce specific antibiotics, which have a significant role in maintaining soil health status [46][47][48].
In this study, the soil health conditions primarily included the influence of SA, EA and MB. The SA was reflected by the dissolved oxygen concentration, oxygen diffusion rate and soil respiration rate, which could be interpreted as 80.7% of the SA. The EA was reflected by the soil catalase activity, phosphatase activity and urease activity, which could