Optimizing Water and Nitrogen Management for Green Pepper (Capsicum annuum L.) under Drip Irrigation in Sub-Tropical Monsoon Climate Regions

Abstract

Green pepper (Capsicum annuum L.) is one of the major vegetables cultivated in sub-tropical monsoon climate regions. However, with the unreasonable use of water and nitrogen (N) fertilizer, efficient water and N fertilizer management systems need to be identified. The goal of this project was to investigate the coupling effects of different amounts of water and N on green pepper yield, water use efficiency (WUE), as well as N use efficiency (NUE) in sub-tropical monsoon climate regions. The optimum combination of water and N inputs was determined for multi-objective optimization through the multiple regression analysis and the combinations of likelihood functions. The pot experiment was conducted during the green pepper growing seasons (May–September) of 2019 and 2020 in a greenhouse at Nanchang, Jiangxi of China that included three water deficit levels, i.e., mild deficit (W1: 95~80%θFC, %θ field capacity simplified as %θFC), moderate deficit (W2: 80~65%θFC), and severe deficit (W3: 65~50%θFC), and four levels of nitrogen application (N_app) rate, i.e., 6.0 (N1), 3.0 (N2), 1.5 (N3), and 0.0 g plant⁻¹ (N4), for a total of twelve treatments (i.e., 3 × 4) with six replications. Results show that water levels have an extremely significant effect (p < 0.01) on green pepper yield and WUE, but no effect on NUE (p > 0.05). N treatments have significant effects on green pepper yield, WUE, and NUE. Meanwhile, the effects of water levels and N treatment interaction on WUE and NUE were extremely significant (p < 0.01), but varied on yield between the two years. The maximum yields (576.26 and 619.00 g plant⁻¹) occurred when the water level and N_app rate were 80~65%θFC and 6.0 g plant⁻¹. While the water level and N_app rate were 80~65%θFC and 3.0 g plant⁻¹, the WUEs and NUEs reached the maximum, which were 20.14 and 17.71 g L⁻¹, 76.54, and 77.73 g⁻¹ in 2019 and 2020. The dualistic and quadric regression equations of irrigation amount and N_app rate indicated that the yield, WUE and NUE cannot reach the maximum at the same time. By establishing a multiobjective optimization model using combinations of likelihood functions, it was concluded that the water level shall be controlled in 80~65%θFC and the N_app rate is 3.78 g plant⁻¹, which can be used as the suitable strategy of water and N management for the maximum comprehensive benefits of yield, WUE, and NUE for green pepper. The obtained optimum combination of water and N inputs can provide a scientific basis for irrigation and fertilization optimization and management in sub-tropical monsoon climate regions.

1. Introduction

Green pepper (Capsicum annuum L.) is widely cultivated in sub-tropical monsoon climate regions due to its high economic benefits and became an important source of income for local farmers [1]. Due to the high temperature and little rainfall in the growing season of green pepper and the abundant surface water in this region, farmers often applied excessive water and nitrogen (N) to ensure adequate yield and profitability, which not only hampers increasing yield, but also causes a series of problems, such as pollution by nitrate leaching and nitrous oxide emission, as well as ecological imbalance [2,3]. Therefore, optimizing water and N management for green pepper is critical to the sustainable development of agriculture in this region.

Water is one of the most limiting factors influencing the nutrition absorption, growth and production of vegetables, and reasonable irrigation methods could improve the water use efficiency (WUE) [4]. Drip irrigation (DI) is the most efficient method of irrigation for vegetable cultivation. The delivery of low amounts of water at a high frequency usually limits water evaporation and drainage, thus representing a high WUE, which was confirmed experimentally [5,6]. Some studies of plant water physiology showed that a certain degree of water deficit could be purposefully caused to prevent the redundant growth of roots, stems, and leaves, adjust the distribution of photosynthetic products to different tissues and organs to increase yield, abandon the growth of vegetative organs and the total amount of organic synthetic substances, and finally realize the goal of increasing economic yield and WUE while saving a lot of water [7,8]. Therefore, a reasonable water level should be considered in irrigation.

In addition, nitrogen (N) is a necessary nutrient element, which is considered as the maximum quantity of all the minerals for most vegetable growth [9], and the proper use of N fertilizer can compensate a part of the water stress on the plant and promote the plant growing condition [10]. Agricultural production needs a lot of N fertilizer. In encountering drought stress conditions, the use of N fertilizer enhances plant resistance, and therefore the yield and WUE [11]. Even though applying the N fertilizer has some crop production benefits, there are notable problems in the fertilization by N fertilizer [12]. For example, it contaminates the subsurface and subsurface water and causes the leaching of nitrate in the soil [13]. Wan et al. [14] found that compared with 2016, the ammonia N and total N concentrations in major rivers in Jiangxi Province increased by 33.3% and 26.0%, respectively, in 2018. Hence, managing N fertilizer can assure the maximum yield and sustainability of the environment.

Modern production practices involve optimization of both water and N usage, in order to optimize crop production and minimize the risk of N leaching toward underground water [15], and undersupply or oversupply of either may offset the benefits of the other, reducing the overall production and nutrient gains [16]. Careful water and N management is required for green pepper to ensure regular growth, high dry matter content, and marketable fruits. The common recommendation is that N fertilizer should be applied in accordance with the amount of extractable soil water, residual N, and the amount of nitrate in the irrigation water, and the effects of water and N interaction on cucumber [17], tomato [18], and sugar beet [19] all showed that there is an optimal level of N application (N_app) rate for each water level.

However, previous studies on green pepper mainly focused on genotype and deep processing [20,21], but little on water and N management. Suitable water and N management will not only help ensure crop production, but also avoid wasting water and fertilizer. It is often difficult to balance the multiple objectives of high efficiency, yields, and economic benefits through a single treatment. Studies on determining the multi-objective optimization of water and N management based on green pepper production and environmental benefits remain scarce. Therefore, preventing environmental pollution, conserving agricultural inputs, increasing green pepper yield, and reducing production costs mandate further research to examine the response of green pepper to water and N inputs under a DI system in sub-tropical monsoon climate regions. The objectives of this study were: (I) to quantify the response of green pepper yield, WUE, and NUE to variable soil water content (SWC) levels and N application (N_app) rates; (II) to determine the effects of variable SWC levels and N_app rates and their interaction on yield, WUE and NUE of green pepper; and (III) to determine an optimal water and N management strategy that can comprehensively improve the yield, WUE and NUE of green pepper under DI in sub-tropical monsoon climate regions.

2. Materials and Methods

2.1. Experimental Site

The pot experiment was conducted during May to September in 2019 and 2020 in a greenhouse at Nanchang Institute of Technology, Jiangxi province, China. This region experiences a subtropical monsoon climate, with annual average temperature, frost-free period, and rainfall of 17~17.7 °C, 251~272 days, and 1600~1700 mm, respectively (Figure 1).

The experiment soil was taken from the 0~40 cm layer in a field in Jinxian Country (sampling date 20th April 2019). After the soil was air-dried naturally, it was screened through a 2 mm sieve for use. The particle size composition of the soil was determined by the Mastersizer 2000 laser particle size analyzer. The proportions of clay particles (0 < d ≤ 002 mm), powder particles (0.002 < d ≤ 0.02 mm), and sand particles (0.02 < d ≤ 2 mm) were 9.14%, 40.24%, and 50.62% (volume fraction), respectively, which was judged as loam according to international classification standards. The physical and chemical properties of the soil were determined from three composite samples (

Table 1. Physical and chemical properties of the soil at the experiment location.

Physical Properties		Chemical Properties
Bulk Density (g cm−3)	Field Capacity(%)	pH	EC (dS m−1)	Orgnic C (mg m−3)	Available N (mg kg−1)	Available P (mg kg−1)	Available K (mg kg−1)
1.3	29.2	5.9	0.91	26.2	26.79	41.53	174.34

).

2.2. Experimental Design

Factorial design was adopted in this experiment, including two factors of soil water content (SWC) and nitrogen application (N_app) rate. The SWC was divided into three levels, which were slight deficit (W1: 95~80%θFC, %θ field capacity reduced to %θFC), moderate deficit (W2: 80~65%θFC), and heavy deficit (W3: 65~50%θFC), while the N_app rate was divided into four levels: high N (N1: 6.0), medium N (N2: 3.0), low N (N3: 1.5), and no N (N4: 0.0 g plant⁻¹), for a total of 12 treatments (i.e., 3 × 4), and each treatment was repeated 6 times, equaling 72 total pots. A row of 12 pots was randomly arranged along the east–west direction. In order to eliminate the influence of ventilation or light at different positions in the greenhouse on the experiment results, the positions of pots were moved once a week during the experiment.

The SWCs were maintained at 95%θFC before the transplanting for all treatments. One uniformity green pepper seedling for pots was transplanted on 1 May, and only controlled SWC during the flowering and fruit setting stage, and the full fruit stage. Green peppers were irrigated to 95%θFC to ensure plant survival in the seedling stage, and no water was irrigated in the late fruiting stage for all treatments. The type of green pepper used in this experiment was ‘sweet pepper F1 hybrids’. The experiment was carried out in a plastic bucket (32 cm in depth and 26 cm in diameter), filled with 19.32 kg of soil. A plastic saucer was placed under each pot to prevent water loss by leaching. The SWC and N_app rate applied in each growth stage of green pepper was shown in

Table 2. The soil water content and urea applied in each growth stage of green pepper.

Treatment	SWC (%θFC)				Urea (g plant−1)
	Seedling Stage	Flowering and Fruit Setting Stage	Full Fruit Stage	Late Fruiting Stage
W1N1	~95	80~95	80~95	not irrigated	6.0
W1N2	~95	80~95	80~95		3.0
W1N3	~95	80~95	80~95		1.5
W1N4	~95	80~95	80~95		0.0
W2N1	~95	65~80	65~80		6.0
W2N2	~95	65~80	65~80		3.0
W2N3	~95	65~80	65~80		1.5
W2N4	~95	65~80	65~80		0.0
W3N1	~95	50~65	50~65		6.0
W3N2	~95	50~65	50~65		3.0
W3N3	~95	50~65	50~65		1.5
W3N4	~95	50~65	50~65		0.0

Note: %θ FC means %θ field capacity.

.

On the 5th day after transplanting, 6.0 g plant⁻¹ KH2PO5 and 30% of each treated N fertilizer were evenly mixed as a base fertilizer and applied as a fertilizer solution with irrigation water, and the remaining 70% N fertilizer was applied twice in the form of fertilizer solution during the flowering and fruit setting stage, as well as the full fruit stage. When N fertilizer was not applied, the needed amount of irrigation water was done via independent household dripper, with flow rate of 2.0 L h⁻¹, wetting ratio of 0.75 for each dripper (volume ratio). When N fertilizer was applied, urea was firstly dissolved in irrigation water and then fertilized, as indicated above. In order to avoid the influence of the pot boundary on the experiment, the pots were placed in soil pits with the same size as the pots in this experiment. For the two-year experiment, the green pepper types were the same. Green peppers were all transplanted on 1 May and off-test on 30 September. During the experiment, measures such as rain shading and pest control were all the same. In addition, when the temperature out of the green greenhouse was higher than 30° in two years, both sides of the greenhouse would be opened to weaken the growth inhibition of green pepper by high temperature.

The irrigation water was taken from the nearby river, and the physical and chemical indicators of irrigation water are shown in

Table 3. The physical and chemical indicators of irrigation water (sampling date: 2019.04.25).

Indicators	COD mg L−1	TN mg L−1	NH4-N mg L−1	TP mg L−1	TK mg L−1	pH
Content	23.66	3.18	3.13	0.11	7.31	6.6

Note: COD means chemical oxygen demand; TN means total nitrogen; NH4-N means ammonium nitrogen; TP means total phosphorus; and TK means total potassium.

, which met the standard for irrigation water quality (GB5084-2021).

2.3. Measurements and Calculations

2.3.1. Yield

After the fruits were mature, they were picked once every 3 days. An electronic scale with an accuracy of 0.01 g was used to record the yield per plant, and the total yield per plant was the sum of each value.

2.3.2. ET, WUE and NUE

(1)

Evapotranspiration (ET) was calculated using water balance method.

ET = P + U + I − F − R − ΔW

(1)

where P indicates the precipitation (mm); U indicates the groundwater recharge (mm); R indicates the runoff (mm); F indicates the deep seepage (mm); and ΔW indicates the change in SWC from the beginning to the end of the experiment (mm).

According to the actual conditions during the experiment, the contributions of P, U, F, R, and ΔW were negligible, so ET was equal to I in this experiment.

(2)

Water use efficiency (WUE)

WUE = Y/ET

(2)

where WUE indicates the water use efficiency (g L⁻¹); Y indicates the green pepper yield (g plant⁻¹); and ET indicates the evapotranspiration (L plant⁻¹).

(3)

Nitrogen use efficiency (NUE)

NUE = (Y_Ni−Y_N4)/Ni

(3)

where Y_Ni indicates the yield in the nitrogen application area (g plant⁻¹), Y_N4 indicates the yield of N4 of the corresponding SWC treatment (g plant⁻¹), and Ni indicates the amount of urea applied (g plant⁻¹).

2.3.3. Statistical Analysis

The equations were established using the cftool in MATLAB R2021a, which is a commercial mathematics software produced by MathWorks of the Portola Valley, CA, USA and can be used in data analysis, image processing, and other fields. The data were analyzed using the analysis of variance (ANOVA) to evaluate the effects of different irrigation amounts and nitrogen levels on green pepper yield, WUE, and NUE. Differences between the means were determined using the least significant difference (LSD) test and reported to be significant at p < 0.05 or extremely significant at p < 0.01.

3. Results

3.1. Yield

In the 2019 and 2020 growth seasons, the green pepper yield was significantly influenced by soil water content (SWC), nitrogen (N), and the interaction between SWC and N (SWC×N) (

Table 4. Significant levels of various treatment effects on ET, yield, water, and nitrogen use efficiency.

	2019				2020
	ET	Yield	WUE	NUE	ET	Yield	WUE	NUE
SWC	**	**	**	**	**	**	**	*
N	**	**	**	**	**	**	**	**
SWC×N	ns	**	**	**	ns	**	**	**

Note: SWC means soil water content; N means nitrogen; ET means evapotranspiration; * means significant difference (p < 0.05); ** means extremely significant difference (p < 0.01); while ns means no significant difference (p > 0.05).

).

Green pepper yield varied widely under different SWC through DI at different levels of N_app rates (Figure 2). When the water level was W1, the yields first increased and then decreased with the increase in N_app rates, and when the water levels were W2 and W3, the yields increased gradually with the increase in N_app rates, and that of N1 and N2 were larger than that of N3 and N4 significantly (p < 0.05), while there were no significant differences between N1 and N2 (p > 0.05). When the N_app rate was 6.0 g plant⁻¹, the yields first increased and then decreased with the increase in water levels, and that of W2 was significantly larger than that of W3 (p < 0.05), and when the N_app rate was less than 3.0 g plant⁻¹, the yields increased with the increase in water levels and that of W1 was significantly larger than that of W3 (p < 0.05)(Figure 2). During the two years, the treatments with the largest yields were all W2N1, with yields of 576.26 and 619.00 g plant⁻¹, respectively.

3.2. WUE

The data on WUE as estimated from total yield per ET varied widely under different SWC and N levels (Figure 3). Within two years, the SWC and N showed extremely significant individual and interactive effects (p < 0.01) on WUE of green peppers (Table 4).

When the water level was the same, the WUEs showed a trend of first increasing and then decreasing with the increase in the N_app rates, and that of N2 was significantly larger than that of N3 and N4 (p < 0.05), but the difference between N3 and N4 did not reach a significant difference level (p > 0.05). When the N_app rate was the same, the WUEs also showed a trend of first increasing and then decreasing as the water levels increased, and that of W2 was significantly larger than that of W3 (p < 0.05), while the difference between W1 and W3 was significant between different N treatments (Figure 3).

The highest WUE values, 20.14 and 17.71 g L⁻¹, respectively, were obtained in the W2N2 treatments, while the lowest values, 7.38 and 7.95 g L⁻¹, respectively, were obtained in the W3N4 treatments for both two years.

3.3. NUE

Nitrogen use efficiency (NUE) is a comprehensive effect index reflecting the local nitrogen fertilizer application rate. Both N and the interaction between SWC and N (SWC×N) had extremely significant effects on NUE (p < 0.01), while the effects of SWC on NUE were different in two years (Table 4).

The NUEs of the green peppers in 2019 and 2020 were shown in Figure 4. At the same water levels, with the increase in N_app rates, the NUEs first increased and then decreased. Taking W1 in 2019 as an example, there was a significant difference in NUE between the three nitrogen levels, and when the N_app rates were N3, N2, and N1, the NUEs were 18.39, 48.60, and 37.55 g g⁻¹, respectively, with the increase of 29.43% and −62.16%, and the difference was not significant. When the N_app rate was the same, the NUEs also showed a trend of first increasing and then decreasing with the increase in water levels. Taking N1 in 2019 as an example, when the water levels were W3, W2, and W1, the NUEs were 34.63, 39.45, and 18.39 g g⁻¹, respectively, with an increase of 13.92% and −53.38%, and the difference did not reach a significant level (p > 0.05). Within the two years, the NUEs treated with W2N2 were the largest, with values of 76.54 and 77.63 g g⁻¹, respectively (Figure 4).

3.4. Model Development and Evaluation

The amounts of irrigation (I) and N_app rates were taken as the independent variables, and the yield, WUE, and NUE were used as the dependent variables. The binary quadratic regression equation was established based on the principle of the least squares method, and the data were analyzed by using MATLAB R2021a. The predicted and measured data also had a signification correlation (R2 > 0.76, p < 0.01). Hence, these models were able to predict the yield, WUE, and NUE with an acceptable accuracy.

According to the formulas in Figure 5, the relationship models of yield, WUE, and NUE with irrigation amount and N_app rate could be obtained. When the irrigation amount and N_app rate were about 38~44 L plant⁻¹ and 3.8~5.4 g plant⁻¹, 27–40 L plant⁻¹ and 2.4~6.0 g plant⁻¹, 27–37 L plant⁻¹ and 2.3~4, 4 g plant⁻¹, the yield, WUE, and NUE were at a high level (Figure 5).

It was found to be difficult to obtain the maximum yield, WUE, and NUE simultaneously. The maximum yield of 609.34 g plant⁻¹ was found under the combination of I 41.37 L plant⁻¹ and N_app rate of 4.50 g plant⁻¹. The maximum WUE of 14.70 g L⁻¹ was achieved when 30.65 L plant⁻¹ of I and 4.21 g plant⁻¹ of N_app rate. The largest NUE of 74.34 was obtained with 30.54 L plant⁻¹ of I and 3.37 g plant⁻¹ of N_app rate (

Table 5. Maximum yield (g plant⁻¹), water and nitrogen use efficiency of green pepper, and their corresponding irrigation amount and N_app.

Target	I	Napp	Yield	WUE	NUE
Maximum yield	41.37	4.50	609.34	11.23	55.38
Maximum WUE	30.65	4.21	516.16	14.70	70.57
Maximum NUE	30.54	3.37	507.99	14.46	74.34

Note: I means irrigation amount (L plant⁻¹); N_app means nitrogen application rate (g plant⁻¹); WUE means water use efficiency (g L⁻¹); and NUE means nitrogen use efficiency (g g⁻¹).

). Therefore, it was necessary to further analyze and explore the optimal combination of water and nitrogen to obtain the best indicators.

3.5. The Strategy for SWC-N Management Based on Comprehensive Benefits

Increasing yield can increase the income of farmers and improve WUE and NUE, which could not only reduce the investment, but also promote the sustainable development of agriculture. However, the coupling effects of water and N inputs on the yield, WUE, and NUE exhibited a downward convex shape, and it is impossible to maximize yield, WUE, and NUE at the same time (Table 5). So the suitable strategy of SWC-N management for maximizing the comprehensive benefits of yield, WUE, and NUE for green pepper must be determined under DI, which could achieve the purpose of high yield, and irrigation water and N saving in red soil regions.

Combinations of likelihood functions were used for parameter estimation [22], of which C1 was the addition combination, C2 was the multiplication combination, and C3 was the mean square root combination, and they were calculated as:

$$C1=\frac{1}{K}({\displaystyle \sum _{\mathrm{i}=1}^K{y}_i})$$

(4)

$$C2={\displaystyle \prod _{i=1}^K{y}_i}$$

(5)

$$C3={(\frac{1}{K}{\displaystyle \sum _{i=1}^K{y}_i^2})}^{1/2}$$

(6)

where y_i is relative yield, relative WUE, or relative NUE, and K is the number of targets (three for this test).

Yield, WUE, and NUE have different units that could not be compared directly, therefore needed to be normalized; that is, to use the yield, WUE, and NUE of respective treatment to divide the maximum value to get the relative yield, the relative WUE, and the relative NUE.

Assume that the yield, WUE, and NUE have the same weights for the three combinations (Equations (4)–(6)), then the targets of C1, C2, and C3 and the corresponding I and N_app rate could be obtained (Figure 6). These three combinations could be used for target optimization. Although the WUE and NUE achieved the maximum value in C1 combination, the yield decreased by 7.63% compared with the maximum value. Compared with C3, when the irrigation amount of C2 increased from 34.63 to 36.09, the yield increased only by 15.02 g, and the WUE and NUE of the C2 combination were also higher than those of C3 combination, so it can be considered that the C2 combination was the best (

Table 6. Relationship between the investments of irrigation amount and N_app and the targets of C1, C2, and C3 for green peppers in 2019 and 2020.

Target	I	Napp	Yield	WUE	NUE
C1	33.96	3.82	562.83	14.32	72.09
C2	34.63	3.78	569.95	14.17	71.69
C3	36.09	3.91	584.97	13.79	69.55

Note: I means irrigation amount (L plant⁻¹); N_app means nitrogen application rate (g plant⁻¹); WUE means water use efficiency (g L⁻¹); and NUE means nitrogen use efficiency(g g⁻¹).

).

Through comparison, it was found that the I and N_app rate of C2 combination were the closest to those of W2N1 within two years. Assuming the irrigating frequency of optimal I was the same as W2, the corresponding SWC of optimal I were 79.52~65%θFC and 81.03~65%θFC in 2019 and 2020, respectively, which does not exceed ±2.00% of the difference of W2. Therefore, it is reasonable that the irrigating frequency of optimal I was the same as W2.

Integrate the yield, WUE, and NUE analyses for green pepper under DI, the suitable strategy of water and N management, which is applicable for DI of green pepper in sub-tropical monsoon climate regions of China is to control the SWC within 80~65%θFC, and the N_app rate of 3.78 g plant⁻¹, and it may achieve the multiple targets of water saving, fertilizer saving, and yield increase.

4. Discussion

Reasonable water and N inputs can improve crop yield, WUE, and NUE. Based on previous studies, this paper discussed the effects of different SWC and N levels on the yield, WUE, and NUE of green peppers, evaluated the appropriate parameters of irrigation water and N fertilizer, and obtained the appropriate water–N management strategy for the green peppers in the experimental area.

4.1. Effects of Different Water Level and N_app Rate on the Yield

Water and N are essential elements for crop development and yield [23], and the effects of different water and N conditions on crop yield are the results of the single factor of water or N and the interaction of the two [24,25].

A previous study reported that a certain level of SWC can increase green pepper yield, although dependent on the amount of Napp rate to different extents [26]. The different water and Napp rate in our study significantly affected final yield. The lowest yield was in W3, perhaps because the water stress accelerated flowering, which decreased the number of fruits and eventually the yield. Water stress also decreased the accumulation of water in the fruit, fruit weight, and yield [17]. Yavuz et al. [27] reported that crop yield increased with irrigation amount, but the highest yield in our study was in W2 treatment rather than that at the W1 treatment. When the water level was the same, the effect of water level on yield increased with the increase in N_app rates, but the effect of N on crop yield was not unlimited. When the N_app rate was less than 3.0 g plant⁻¹, the yields increased with the increase in water levels, while when the N_app rate was 6.0 g plant⁻¹, the yields first increased and then decreased with the increase in water levels, which is consistent with previous studies [26].

The relationship between crop yield and N_app rate is well described by quadratic models [28]. Higher N_app rates can effectively increase crop yields, but excessive N_app rates do not increase crop yields significantly. Reasonable N_app rate can increase crop absorption of soil water and increase crop transpiration, thereby increasing the crop yield, while when the N_app rate exceeds a certain range, the physiological activity of the root system begins to weaken, and the ability of the root system to absorb nutrients is also weakened [29]. In order to maintain a larger biomass, the root system needs to compete for more photosynthetic products in the above-ground part and consume a lot of nutrients, which results in the reduction in crop yield [30].

In addition, the optimum N_app rate amplified the effect of SWC. The N_app rate of 3.0 g plant⁻¹ (N2) produced the highest yield, but the application above this level did not continue to increase yield, which was likely because the higher N_app rate extended the crop vegetative growth and decreased the transport of photosynthates to the fruit [17].

4.2. Effects of Different Water Level and N_app Rate on the WUE and NUE

How to improve WUE under the condition of ensuring a high crop yield is a hot topic in academic research. Appropriate water and N_app rate are helpful in improving the WUE of green pepper [31]. This study showed that the interaction of SWC and N had a significant impact on the WUE, and the WUE was the highest under the condition of moderate water level and N_app rate. Within a certain range of N_app rate, WUE increased with the increase in N_app rate under the same SWC level. When the N_app rate was 3.0 g plant ⁻¹, the WUE were at the maximum in two years. With the increase in SWC, the yields of green pepper increased gradually, but the yield per unit crop water consumption was reduced, so the WUE decreased. In addition, when the water level was moderate, the permeability of the soil was better, so that the root morphology of the crop was optimized, and the physiological activity of the root system was strengthened [32]. Moderate water stress could promote the growth of the crop root system to the lower layer of the soil, improve the water absorption capacity of the root system, and increase the yield, which in turn improved the WUE [33]. When the water level and N_app rate exceeded a certain range, it would cause redundant plant growth, lead to self-consumption, and reduce crop yield, which resulted in the reduction in WUE [34].

This study showed that NUE values first increased, and then decreased as the amount of N_app rates increased for green pepper for the same SWC levels, and the coupling of moderate water level and N_app rate was conducive to the formation of yield and promoted the effective absorption of N by crop roots. The SWC had an extremely significant effect on NUE in 2019, and a significant effect on NUE in 2020. The results indicate that SWC can promote the ability of the crop to absorb nutrients within a certain range in this planting mode [23]. The absorption and assimilation of N fertilizer by crops must undergo a series of N metabolizing enzymes [35]. The suitable water level could create suitable soil aeration, which could improve the activity of N metabolizing enzymes significantly. When the water level was too low or too high, the vitality of crop roots would decrease, while the roots’ ability to absorb nutrients would reduce, which was not conducive to the maintenance of N metabolism enzyme activity and the absorption of N by crop roots [18]. In our study, we found that NUE became a relatively stable lower value once N_app rate exceeded the critical amount. This result was mainly due to both the yield and N uptake by crops reaching their highest values at this critical level of N_app rate. When vegetative and reproductive growth are balanced, high N_app rate can achieve a higher NUE, but excessive N_app rate disrupts the balance between vegetative and reproductive growth in green pepper, leading to overly strong nutrient uptake, postponed maturity, and reduced production [36]. When the N_app rate was insufficient, crop yield would be affected, and as the N_app rate increased, the effect of N on crop yield began to weaken, so too much or too small of a N_app rate was not conducive to the increase in NUE. So, moderate water level and N_app rate coupled with water stress could increase N activity [37], which was conducive to yield, which was mainly due to both the yield and N uptake by crops reaching their highest values at this critical level of N_app rate.

Therefore, proper regulation of water and N can significantly improve the effect of “adjusting water with fertilizer”, coordinate the growth of above-ground and underground parts, maintain high root physiological activity, improve the activity of key enzymes in N metabolism, promote plant N metabolism, and avoid self-consumption caused by redundant growth, which improves the WUE and NUE of green pepper [38].

4.3. Determination of Optimal Combination of SWC and N_app Rate

In agricultural production, it is difficult to achieve optimal values for different goals at the same time. For instance, increasing water level and N_app rate increase crop yield but reduce WUE and NUE, while increasing costs. In the present study, regression analysis was used to determine the relationship between green pepper yield, WUE, NUE, irrigation amount, and N_app rate.

At present, there are many methods available for multi-objective optimization, such as the maximum likelihood method [39] and the spatial analysis method [40]. In this study, it was verified that the maximum likelihood method could be used to solve the problem of comprehensive benefits, which was confirmed by this study. Cheng et al. [31] studied the water–N coupling effect of green peppers in a greenhouse under DI and showed that the combined treatment of 70%θFC irrigation water lower limit and 120 kg ha⁻¹ pure N was the optimal water and N treatment for green peppers, which is consistent with this study. It can be considered that the combination of water level within 80~65%θFC and the N_app rate of 3.78 g plant⁻¹ may achieve the multiple targets of water saving, fertilizer saving, and yield increase.

5. Conclusions

When the water level was W1, with the increase in N_app rates, the yields increased first and then decreased, and when the water level was W2 and W3, the yields increased gradually with the increase in N_app rates. With the increase in N_app rates, the WUEs showed a trend of first increasing and then decreasing, and that of N2 was significantly larger than that of N3 and N4. With the increase in water levels, the WUEs also showed a trend of first increasing and then decreasing, and that of W2 was significantly larger than that of W3. With the increase in water levels and N_app rates, the NUEs showed a trend of first increasing and then decreasing. The regression equations showed controlling the water level within 80~65%θFC and a N_app rate of 3.78 g plant⁻¹ can be considered as the best strategy for water and N management for green peppers under DI in the sub-tropical monsoon climate regions. This will result in maximum comprehensive benefits of yield, WUE, and NUE, which can enhance crop production in sub-tropical monsoon climate regions of China.

This experiment only studied the effects of different water–N couplings on ET, yield, WUE, and NUE of green peppers, while the migration of N in soil–plant systems has a positive impact on green pepper, which will be the focus in future studies.