Optimal Drip Fertigation Regimes Improved Soil Micro-Environment, Root Growth and Grain Yield of Spring Maize in Arid Northwest China

Abstract

Understanding the spatial distributions of soil water, temperature and nutrients as well as their effects on maize growth and grain yield is vital for optimizing drip fertigation regimes. In this study, a 2 year field experiment was conducted on drip-fertigated spring maize with plastic mulching in arid northwestern China in 2015 and 2016. Four irrigation levels were set: as I₆₀ (60% ET_c; ET_c is crop evapotranspiration), I₇₅ (75% ET_c), I₉₀ (90% ET_c) and I₁₀₅ (105% ET_c) in 2015; and as I₆₀ (60% ET_c), I₈₀ (80% ET_c), I₁₀₀ (100% ET_c) and I₁₂₀ (120% ET_c) in 2016. Two fertilization rates of N-P₂O₅-K₂O were set: as F₁₈₀ (180-90-90) and F₂₄₀ (240-120-120). The results showed that the average soil water content in the deeper soil layer (80–120 cm) increased with the increase in irrigation level, and the lowest average soil water content in the 0–80 cm soil layer occurred under I₉₅ in 2015 and under I₁₀₀ in 2016. The irrigation level more significantly influenced the soil temperature at 5 cm than at the other depths. With the decrease in the irrigation level and progression of the growth period, the soil temperature increased. The soil nitrate nitrogen content in the root zone decreased with increasing irrigation level. The largest soil nitrate nitrogen content at the 0–100 cm depth occurred under I₆₀ in both 2015 and 2016. Significant differences were observed for root length density in the 0–20 cm soil layer at various lateral locations. In deeper (60–100 cm) soil layers, the root length density under I₇₅ (2015) and I₈₀ (2016) was greater than at other depths. Grain yield, water use efficiency (WUE) and partial factor productivity (PFP) increased with the increase in irrigation level in 2015, while it increased and then decreased in 2016. I₁₀₅F₁₈₀ achieved the maximum grain yield (18.81 t ha⁻¹), WUE (3.32 kg m⁻³), and PFP (52.26 kg kg⁻¹) in 2015, while I₁₀₀F₁₈₀ achieved the maximum grain yield (20.51 t ha⁻¹), WUE (3.99 kg m⁻³), and PFP (57.02 kg kg⁻¹) in 2016. The optimal drip fertigation regimes for spring maize in arid northwest China were recommended as 90–100% ET_c and 180-90-90 (N-P₂O₅-K₂O) kg hm⁻².

1. Introduction

Water scarcity is a major factor limiting the growth and yield of crops in arid and semi-arid agricultural areas around the world [1]. The Hexi region of Gansu Province is an important maize-producing area in northwestern China. There is a severe shortage of water resources in this region, with the average annual surface evaporation (>2000 mm) greatly exceeding average annual precipitation (<150 mm) [2]. As a consequence, the majority of agriculture is irrigated with water from inland rivers in this region. Water conservation practices have already been applied in many arid and semi-arid parts of the world [3,4,5,6]. However, local farmers typically use flood irrigation with spread fertilization, which often results in serious waste of both water and fertilizer resources, as well as deterioration of the fragile desert oasis environment. Therefore, exploring water conservation measures to improve water and fertilizer use efficiency is important to achieving the sustainability of agricultural productivity in this region [7]. The availability of soil water and nutrients, together with an enhanced soil temperature, are the most critical factors affecting crop growth and yield. The three factors often interact with each other so that an optimal combination can vastly influence the development of a healthy plant root system, resulting in high water use efficiency and yields [8,9,10,11].

Drip irrigation under film mulching has been widely used in arid and semi-arid regions with the acceleration of China’s agricultural modernization [12,13]. The combination of drip irrigation with film mulching, which prevents soil evaporation of the sparse water, has been proven useful in conserving both water and fertilizers [14]. In addition, the film mulching can effectively increase soil temperature [15] and decrease deep drainage [16], thus improving water use efficiency and grain yield [17]. Various studies have investigated the effects of irrigation frequency [18], irrigation patterns [19] and irrigation water amounts [20], whereby drip irrigation resulted in enhanced water and fertilizer use efficiency [21]. However, as a point-source irrigation method, drip irrigation can lead to the non-uniform distribution of water and nutrients in the soil [22]. There are few systematic studies on how water and fertilizer supply affects soil water, heat, and nutrient distributions in the whole soil profile or in the root zone.

Researchers have studied the distributions of soil water and nutrients under drip irrigation and fertilization with film mulching [23]. Nevertheless, the influence is rather complicated, since the spatial distribution of soil water and nutrients is also affected by irrigation, fertilization and the different growth stages of crops [24,25], as well as by crop type and irrigation method [26]. For crops that are sensitive to high temperatures, increasing the soil temperature can suppress the crop growth. Wang et al. [27] found that drip irrigation under film mulching could significantly reduce soil evaporation and preserve high soil water content at the early growth stage of potatoes. However, film mulching at the middle and late stages can cause excessively high soil temperature. In addition, it has been demonstrated that the spatial distribution of soil temperature is influenced by the shadow projected on the soil surface from crops [28]. Soil nitrate nitrogen content is considered as the main index in field nutrient management [29]. Irrigation may result in significant losses of nitrate nitrogen in soils [30] and influence the distribution and accumulation of nitrate nitrogen in the soil profile [31], while an appropriate supply of water and fertilizer can elevate water and fertilizer use efficiency [32], thus leading to a reduction in the risk of groundwater pollution and water and fertilizer waste [33].

Spring maize roots are significantly affected by irrigation and fertilization [34]. The distribution of crop roots is limited by the extent of wetting, and the fine roots are generally concentrated in the wet volume [35]. At the same time, the soil temperature on the side far away from the drip irrigation belt is higher and the soil is dryer. As the growing stage progresses, it is not conducive to root growth and nutrient absorption in summer [36]. Liu et al. [37] found that a large root/shoot ratio in summer maize led to premature aging and reduced yield. Therefore, exploring the effects of irrigation and fertilization on root growth under plastic film is also important for increasing yield and water use efficiency of spring maize.

Recently, Sui et al. [38] found that drip irrigation significantly promoted environmental factors in northeastern China. Many research studies have been conducted on the positive effects of drip irrigation scheduling on the yield of cotton [39], rape [33], tomato [40] and potato [41] in arid area of northwestern China. However, quantitative analyses of theoretical yield increases and optimal irrigation and fertilization scheduling for drip-fertigated spring maize with plastic mulching in this area have been limited. In order to investigate the coupling effects of water and fertilizers on the growth and grain yield of spring maize under drip fertigation, it is necessary to explore the distributions of soil water, temperature and nitrate nitrogen as well as root growth characteristics.

Therefore, the present study was conducted to explore (1) the coupling effects of various irrigation and fertilization regimes on soil water, temperature and nutrient spatial distributions, and (2) the optimal irrigation and fertilization scheduling for drip-fertigated spring maize with plastic mulching in the arid regions of northwestern China.

2. Materials and Methods

2.1. Description of Experimental Plots

The field experiments were conducted during the growing season for spring maize (April to September) in 2015 and 2016 at the Agricultural and Ecological Water Saving Experimental Station of China Agricultural University, on the edge of the Tengger Desert at an altitude of 1584 m (Figure 1). The local arid climate is continental with an annual average temperature of 8 °C. In 2015, the precipitation at the test field during the spring maize growth season was 141 mm in total, and it was 115 mm in 2016. The soil profile (0–120 cm) was characterized as clay soil with an average soil bulk density of 1.52 g cm⁻³, a water holding capacity of 32.9% (volumetric, the same below), a saturated water content of 42.6%, and a wilting point of 12%. The soil contained 50.3 mg kg⁻¹ alkaline hydrolysable nitrogen, 3.82 mg kg⁻¹ effective phosphorus, 8.9 g kg⁻¹ organic matter and 0.35 g kg⁻¹ salt, with a pH of 8.22. The underground water used for irrigation was extracted from a depth of 40 m, with a mineralization degree of 0.71 g L⁻¹, and the mass concentrations of Na⁺, K⁺, Mg²⁺, Ca²⁺, SO₄²⁻, Cl⁻, and HCO₃⁻ were 1297.58, 1297.58, 457.15, 319.25, 2, 962.25, 1, 501.92, 411.94 mg L⁻¹, respectively.

2.2. Experimental Design

The spring maize variety “Qiangsheng51” (Shanxi Qiangsheng seed company, Taiyuan, China) was tested in the present study. The applied fertilizers were 46% N in the form of urea, 18% N and 46% P₂O₅ in the form of diammonium phosphate, and 52% K₂O in the form of potassium sulphate according to local practice. Polyethylene film was used as cover film with a width of 1.2 m and a thickness of 8 µm. The drip fertilization equipment consisted of water source, water pump, proportional fertilizer pump, and pipeline system for water transport and distribution. The drip irrigation belt was patch-type with drippers spaced at 20 cm, a dripper flow rate of 2.3 L h⁻¹, and a working pressure of 0.2 MPa. Four irrigation levels were selected: 60%, 75%, 90% and 105% of ET_c (I₆₀, I₇₅, I₉₀, and I₁₀₅, respectively) in 2015. In 2016, the four levels were increased to I₆₀, I₈₀, I₁₀₀, and I₁₂₀, because it was found that maize yield continued increasing with the irrigation amount in 2015. ET_c is crop evapotranspiration (mm m⁻² d⁻¹) and calculated as follows:

$$E{T}_c=E{T}_0·K_c$$

(1)

where K_c is maize crop coefficient, which was selected as 0.7 at seedling stage, 1.2 at jointing stage to filling stage, and 0.6 at milk to maturity stage [42], and ET₀ is the reference crop evapotranspiration (mm m⁻² d⁻¹), calculated according to the Penman–Monteith equation recommended by FAO-56 [42]. The monthly variations in meteorological variables during the two growing seasons are shown in Figure 2.

The plots had a length of 6 m and a width of 5 m. Maize plants were planted in rows with a plant distance of 20 cm, and wide (80 cm) and narrow rows (40 cm) were alternated. One drip irrigation belt, located in the middle of the narrow rows with a spacing of 20 cm to each row, was used to irrigate two rows. Each dripper corresponded to two maize plants, and one drip irrigation belt controlled 46 maize plants. The maize plant density was 76,500 plants per ha. More planting details are shown in Figure 3.

The maize was irrigated at 10 day intervals, but in case of rainfall, the date of irrigation was postponed to the next day. The ET_c used to calculate the irrigation amount each time was the sum of ET_c within 10 days before the irrigation day. The irrigation efficiency was 0.95, and the water distribution uniformity was 0.93 [41].

In 2015, the initial soil water content of the study field was high, so no seeding water was applied. The irrigation began when the plants had three leaves and one heart, and ended 15 to 20 days before harvest. Two fertilization (N-P₂O₅-K₂O) levels were set, as F₁₈₀ (180-90-90) and F₂₄₀ (240-120-120). Therefore, eight combinations were tested in eight plots, with each combination tested in triplicate. The plants were fertilized on four occasions during the entire growth stage: 20% of the fertilizers at seedling stage, 30% at trumpet stage, 30% at tasseling stage and 20% at filling stage.

2.3. Measurements and Methods

2.3.1. Soil Water Content

Soil samples were collected before irrigation and 2 days after irrigation using a soil auger. From the middle point of two drippers on one drip irrigation belt in a block, soil samples were obtained at 0, 10, 20, 30, and 60 cm distances to the drip irrigation belt. The soil layers were 0–20, 20–40, 40–60, 60–80, 80–100 and 100–120 cm. The detailed soil collection is shown in Figure 3. The soil samples were dried at 105 °C until constant weight, and the gravimetric soil water content (SWC) was determined by the drying method.

2.3.2. Soil Temperature

Using a MicroLite-U disk temperature recorder (Fourtec, Rosh Haayin, Israel), the soil temperature was recorded for sample F₁₈₀ at depths of 5, 15 and 25 cm every 15 min during the entire growth stage. The temperature at 25 cm depth was recorded at a distance of 0 and 40 cm to the drip irrigation belt. The detailed sensor locations are shown in Figure 3.

2.3.3. Soil Nitrate Nitrogen Content

The nitrate nitrogen content in the soil was measured for samples collected at the same location as soil water measurements, as shown in Figure 3. Soil samples were left to dry without extra heat, after which they were ground and passed through a 2 mm sieve. For each sample, 5 g of sieved soil was extracted with 50 mL 2 mol L⁻¹ KCL solution and the nitrate nitrogen content (NO₃⁻-N) in the extraction solution was measured by a continuous flow analyzer (AutoAnalyzer-III, Bran+Luebbe, Norderstedt, Germany).

2.3.4. Root Sampling

Root samples at various lateral locations and soil layers were collected using a soil drill (inner diameter 10 cm) at the filling stage. Two plants in each block were selected, located at point B as shown in Figure 3 representing plant base, at a location 10 cm closer to the drip irrigation belt (point A) and 10 cm further away from the drip irrigation belt (point C). Above-ground plant parts were removed, and soil layers were collected from surface to 100 cm depth at 20 cm intervals. The live plant roots were cleaned by a root washing machine and scanned with a scanner (Epson Perfection V700 photo, Seiko Epson Corp., Nagano, Japan) under 500 megapixels, and the black/white JPG image was recorded. WinRHIZO image analyzer software (Regent Instrument Inc., Québec City, QC, Canada) was adopted to acquire the root length. Using the obtained root length, the root length density (RLD, cm cm⁻³) could be calculated according to Equation (2):

$$RLD \left(\mathrm{cm} {\mathrm{cm}}^{-3}\right)=root length \left(\mathrm{cm}\right) / soil volume \left({\mathrm{cm}}^3\right)$$

(2)

2.3.5. Grain Yield

When maize plants entered the ripening stage, 20 plants were randomly selected per plot, and the maize ears were harvested after the appearance of completely white bracts. The kernels were bagged, naturally dried, and weighed. The 100 grain weight was obtained, and the individual plant yield and yield per ha were calculated.

2.3.6. Crop Water Consumption

Under different fertigation conditions, the water consumption of spring maize during the whole growth stage can be calculated using the water balance equation:

$$ET=P+U+I-D-R-\mathsf{\Delta }W$$

(3)

where ET is crop water consumption (mm), P is the precipitation (mm), U is underground water recharge (mm), I is irrigation amount (mm), R is runoff amount (mm), D is deep leakage amount (mm), and ΔW is soil water change at different growth stages (mm). The soil water content (%) of the plot is the average value of ET. The groundwater table was deep (>40 m) in the test area, and the terrain was flat with little rainfall. In addition, the wetting depth of drip irrigation was shallow. Thus, U, R, and D can be ignored. This simplifies Equation (4) to:

$$ET=P+I -\mathsf{\Delta }W$$

(4)

2.3.7. Water Use Efficiency (WUE)

The water use efficiency can be calculated using the following equation:

$$WUE=Y/ET$$

(5)

where Y is grain yield (kg hm⁻²).

2.3.8. Partial Factor Productivity (PFP)

The partial factor productivity can be calculated using the following equation:

$$PFP=Y/FT$$

(6)

where FT is the total amount of N, P₂O₅, and K₂O applied (kg hm⁻²).

2.4. Data Analysis

Data processing and analysis of variance were conducted using Excel 2010 and SPSS12.0 software. Multiple comparisons were adopted using the SSR method. The significant difference was 0.05 (α = 0.05). The figures were plotted by Sigmaplot10.0.

3. Results

3.1. Soil Water Content

The movement of soil water is the combined effect of irrigation, water upward evapotranspiration, downward permeation, precipitation and crop root uptake and directly affects soil temperature and soil nutrient transport, thus affecting the crop growth and yield formation. During the spring maize growing seasons of 2015 and 2016, the average water content in the 0–120 cm soil layer at different horizontal positions in eight irrigation and fertilization treatments was observed during the whole growth stage, and the average water content of each soil layer before and after irrigation (Figure 4 and Figure 5) and the average soil water before irrigation were analyzed (

Table 1. Average SWC at different locations in different irrigation and fertigation treatments.

Treatment	Soil Layer in 2015 (cm)						Treatment	Soil Layer in 2016 (cm)
	0–20	20–40	40–60	60–80	80–100	100–120		0–20	20–40	40–60	60–80	80–100	100–120
Irrigation							Irrigation
I60	0.193 a	0.188 a	0.213 a	0.211 ab	0.229 c	0.225 c	I60	0.187 ab	0.192 ab	0.212 b	0.198 c	0.214 d	0.213 d
I75	0.186 ab	0.182 ab	0.206 ab	0.216 ab	0.241 b	0.245 b	I80	0.185 ab	0.187 b	0.204 bc	0.21 b	0.226 c	0.226 c
I90	0.177 b	0.173 b	0.196 b	0.205 b	0.243 b	0.247 b	I100	0.182 b	0.173 c	0.195 c	0.204 bc	0.236 b	0.234 b
I105	0.179 b	0.18 ab	0.198 ab	0.221 a	0.252 a	0.268 a	I120	0.194 a	0.201 a	0.229 a	0.228 a	0.246 a	0.246 a
Location							Location
P1	0.183 b	0.181 b	0.206 b	0.215 b	0.242 a	0.247 a	P1	0.183 b	0.183 b	0.209 b	0.209 b	0.232 a	0.232 a
P2	0.18 b	0.175 b	0.196 b	0.21 b	0.242 a	0.245 a	P2	0.18 b	0.182 b	0.208 b	0.208 b	0.231 a	0.231 a
P3	0.176 b	0.178 b	0.198 b	0.209 b	0.241 a	0.245 a	P3	0.181 b	0.184 b	0.204 b	0.204 b	0.228 a	0.23 a
P4	0.181 b	0.179 b	0.205 b	0.21 b	0.244 a	0.247 a	P4	0.185 b	0.188 b	0.21 b	0.21 b	0.229 a	0.229 a
P5	0.196 a	0.191 a	0.212 a	0.223 a	0.239 a	0.248 a	P5	0.206 a	0.205 a	0.219 a	0.219 a	0.232 a	0.227 b

Different lower-case letters denote significant differences at p < 0.05 among treatments.

).

The temporal and spatial distributions of soil water content in the root zone of spring maize under plastic film mulching were very complicated due to various amounts of irrigation. It can be seen from Table 1 that the average soil water contents in the 0–80 cm soil layer in the I₉₅ treatment in 2015 and the I₁₀₀ treatment in 2016 were the lowest, with average values of 0.187 and 0.189 cm³ cm⁻³, respectively. Compared to other irrigation treatments, all treatments had different degrees of reduction. The possible reasons are that reasonable irrigation amounts can promote crop growth, and transpiration consumes more soil water, whereas low-water treatment (I₆₀) caused unhealthy maize growth and thus consumed less soil water, such that the soil water before irrigation was not too low. In deeper soil layer (80–120 cm), the higher the irrigation amount, the higher the average soil water content, which was due to the deeper groundwater level (>40 m) in the experimental area. Therefore, the deep soil water content was mainly affected by irrigation and root systems. Therefore, in the process of drip irrigation and fertilization, the frequency of irrigation should be increased to reduce the amount of irrigation, so as to ensure that water and fertilizers do not leach beyond the crop root zone and improve the use efficiency of water and fertilizers.

The fluctuations in soil water content gradually decreased with the increase in soil depth in the 2 year spring maize growing season (Figure 4 and Figure 5). The average soil water content before irrigation in P1, P2, P3 and P4 was not significantly different between 0–80 cm soil layers (p > 0.05), but significantly lower than that of P5 (p < 0.05) (Table 1). The opposite trend was observed after irrigation, with significantly lower soil water content in P5 compared with the average water content at the other locations (Figure 4 and Figure 5). There was no significant difference in average soil water before irrigation at the 80–120 cm soil depth (Table 1), and no significant fluctuation in average soil water content after irrigation (Figure 4 and Figure 5).

During the whole spring growing season, the average soil water content in all soil layers decreased gradually before irrigation in 2015. In 2016, there was a significant decreasing trend in the average soil water content before irrigation at the middle growth stage (77–122 days after sowing), but it gradually increased at the later growth stage. This may be related to the generally higher temperature in June and July 2016 as compared to 2015 (Figure 1). High temperatures made the maize more water-intensive, causing the root system to consume more deep soil water. This was also evidenced by the significant fluctuations in deep soils (80–120 cm soil layer) at the mid-growth stages in 2016 (Figure 5).

3.2. Soil Temperature

Soil temperature is closely related to crop growth and development. Soil temperature was recorded for the spring maize in the fertilization level F₁₈₀ at various irrigation levels in 2015 (Figure 6) and 2016 (Figure 7). Part A in Figure 7 and Figure 8 illustrates that the soil temperature at a depth of 5 cm below the soil surface did not significantly differ among different irrigation treatments, at least at the seedling stage and middle jointing stage. With the advance of the season, the difference for various irrigation levels became noticeable, leading to higher soil temperature with decreasing irrigation. For the entire growth stage in 2015 (Figure 6C), the average soil temperature of treatment I₆₀ was 20.9 °C, and this decreased by 3.2%, 4.4% and 5.9% for treatments I₇₅, I₉₀ and I₁₀₅, respectively. In 2016 (Figure 7C) the average temperature of I₆₀ was 23.7 °C, and this decreased by 3.3%, 4.4%, and 8.7% for I₈₀, I₁₀₀, and I₁₂₀, respectively. Increasing irrigation amounts led to an increased leaf area index, which blocked and absorbed more solar radiation and reduced the radiation reaching the soil surface.

According to Figure 6B and Figure 7B, in non-raining days during the entire growth stage, the soil temperature decreased with increasing soil depth, and the temperature in different soil layers showed great variations at the seedling stage. The average temperature at 5 cm depth was highest. With the advance of the growth stage, the temperature difference between various layers was reduced. During the whole growth stage, the average temperature at 5 cm depth below the soil surface in 2015 and 2016 was 0.38 °C, 0.55 °C (2015) and 0.79 °C, 0.96 °C (2016), respectively (Figure 6C and Figure 6C).

The average soil temperature on the day of irrigation decreased with increasing irrigation amounts (Figure 6(A1) and Figure 7(A1)), which was affected by solar radiation. The soil temperature increased with solar radiation, and the lower the irrigation amount, the more significantly the soil temperature increased (Figure 6(A2) and Figure 7(A2)). From the influence of irrigation on different soil layers, the soil temperature at 5 cm below the surface was more significant than that at 15 cm and 25 cm depths (Figure 6(B1) and Figure 7(B1)). Similarly, as seen from Figure 6(B2) and Figure 7(B2), the increase in soil temperature at 5 cm depth below the surface was also faster than at 15 cm and 25 cm depths at late growth stage.

The difference in soil temperature at different locations (T3, T4 and T5) in the same soil layer was not significant, so only the daily average soil temperature in different soil layers from different irrigation treatments was statistically analyzed (Figure 6D and Figure 7D). From the statistical analysis of the results, the degree of dispersion of soil temperature at point T1 was larger than elsewhere. With the increase in soil depth, the degree of dispersion was reduced, and the median value lowered with increasing soil depth. The median of soil temperature for various irrigation treatments at the same location decreased with increasing irrigation amounts. In the 2 year growing season, compared to 2015, the fluctuation of discrete points was larger in 2016, with average median increased by 3.4 °C.

3.3. Soil Nitrate Nitrogen

Water movement is the main natural factor affecting the distribution of profile soil nitrate nitrogen. Irrigation increases profile soil water content, which is a key human factor to modify soil nitrate nitrogen distribution. For spring maize under different irrigation and fertigation conditions in 2015 and 2016, the distribution of soil nitrate nitrogen content in the root layer at harvest and the related analysis results are shown in Figure 8 and

Table 2. Average NO₃⁻-N content at different locations in different irrigation and fertigation treatments.

Treatment	Soil Layer in 2015 (cm)						Treatment	Soil Layer in 2016 (cm)
	0–20	20–40	40–60	60–80	80–100	100–120		0–20	20–40	40–60	60–80	80–100	100–120
Irrigation							Irrigation
I60	17.74 a	14.99 a	35.6 a	29.3 a	37.34 a	25.66 c	I60	19.15 a	24.94 a	39.7 a	37.81 a	28.74 a	27.9 b
I75	14.32 b	13.26 b	26.83 b	24.38 b	32.29 b	32.82 b	I80	9.97 c	18.74 b	29.52 b	35.74 b	28.06 a	23.5 c
I90	9.58 c	12.38 c	9.86 c	22.62 b	26.13 c	37.36 a	I100	8.88 d	7.23 d	15.65 c	22.34 c	13.12 c	33.95 a
I105	9.66 c	13.38 b	8.72 c	12.82 c	20.8 d	27.87 c	I120	14.59 b	10.69 c	12.3 d	15.66 d	19.62 b	25.66 bc
Location							Location
P1	10.03 d	15.38 a	19.06 bc	23.14 c	30.96 a	34.29 a	P1	9.89 d	13.97 c	21.87 cd	27.98 a	21.92 abc	30.98 a
P2	6.61 e	11.06 c	16.06 c	20.64 d	31.09 a	33.58 a	P2	9.25 d	13.33 c	20.7 d	27.82 ab	20.86 c	27.28 bc
P3	12.99 c	11.71 c	18.37 bc	16.29 e	27.52 b	27.06 c	P3	11.1 c	13.97 c	23.09 bc	28.4 a	21.58 bc	28.35 b
P4	14.96 b	12.9 b	19.87 b	26.43 a	29.84 ab	31.04 b	P4	15.3 b	16.94 b	24.58 b	26.16 b	23.58 ab	24.86 c
P5	19.55 a	16.48 a	27.49 a	24.88 b	26.3 b	28.69 c	P5	20.21 a	18.78 a	31.23 a	29.06 a	23.97 a	27.31 bc

Different lower-case letters denote significant differences at p < 0.05 among treatments.

. According to Figure 8, the average soil nitrate nitrogen content in the root layer for treatment I₆₀ demonstrated two peaks. The content was reduced with increasing irrigation amounts. Compared with treatment I₆₀, the average soil nitrate nitrogen content under treatments I₇₅, I₉₀, and I₁₀₅ in 2015 was decreased by 16.1%, 20.4%, and 38.8%, respectively. The average soil nitrate nitrogen content under treatments I₈₀, I₁₀₀, and I₁₂₀ in 2016 was decreased by 18.3%, 43.3%, and 44.8%, respectively. In the 2 year growing season, the nitrate nitrogen content under treatment I₆₀ at 0–100 cm depth was significantly higher than that under the other irrigation treatments (Table 2). In terms of lateral nitrate nitrogen distribution, the profile soil nitrate nitrogen content decreased along the far side of the drip irrigation belt with increasing irrigation amounts. It can be seen from the results (Table 2) that the difference in soil nitrate nitrogen content was extremely significant (p < 0.01) for various soil layers at different irrigation levels. For various lateral locations, the difference in soil nitrate nitrogen content in 0–60 cm soil layers was also extremely significant (p < 0.01). In the 2 year growing season, the nitrate nitrogen content was highest at location P5 for 0–60 cm soil layers, reaching 19.5 (2015) and 21.5 (2016) mg kg⁻¹. The lowest values appeared at locations P2 and P3, and P4 also had lower nitrate nitrogen content. In deep soil layers (60–120 cm), the average soil nitrate nitrogen content was highest at location P1, followed by locations P4 (2015) and P5 (2016). Therefore, the soil nitrate nitrogen content was affected by both irrigation level and root growth.

3.4. Root Length Density

The root length density (RLD, cm cm⁻³) of maize plants at the filling stage at each location is shown for four depths in Figure 9. Significant differences were observed for RLD of plant roots at 0–20 cm depth collected at various lateral locations under the same irrigation level. Plants grown under lower irrigation levels (I₆₀, I₇₅, and I₈₀) produced much lower RLD on the far side of drip irrigation belt (point C locations) than elsewhere. With increased irrigation amounts, the RLD of plants at point C positions gradually increased. Comparing samples taken under the same fertilization regime, the RLD of spring maize taken at a depth of 0–60 cm at various lateral locations varied, following different trends. The RLD on the near side of the drip irrigation belt (point A) decreased with increasing irrigation amounts, and the RLD on the far side of the drip irrigation belt (point C) increased with increasing irrigation. In 2015, the RLD of point B locations increased with increasing irrigation, whereas in 2016, it first increased and then decreased with increasing irrigation, peaking at I₁₀₀. In 2015, the RLD in the 0–60 cm soil layers of treatments I₆₀, I₇₅, I₉₀, and I₁₀₅ accounted for 80.1%, 76.2%, 77.0% and 78.7% of the total RLD, respectively. In 2016, these percentages were 79.0%, 75.9%, 76.2% and 77.0% for I₆₀, I₈₀, I₁₀₀, and I₁₂₀, respectively. In deeper (60–100 cm) soil layers and at various locations, the RLD of treatments I₇₅ (2015) and I₈₀ (2016) was higher than elsewhere under the same fertilization level.

3.5. Yield, Water Use Efficiency and Partial Factor Productivity

The influence of different irrigation and fertigation treatments on 100-grain weight, yield, water consumption and partial factor productivity of spring maize is shown in

Table 3. The 100-grain weight, yield, water consumption, WUE and PFP under different irrigation and fertigation treatments in 2015 and 2016.

In 2015						In 2016
Treatment	100-Grain Weight (g)	Yield /(t ha−1)	Water Consumption (mm)	WUE /(kg m−3)	PFP /(kg kg−1)	Treatment	100-Grain Weight (g)	Yield /(t ha−1)	Water Consumption (mm)	WUE (kg m−3)	PFP (kg kg−1)
I60F180	34.4 b	13.08 de	458.81 bc	2.85 bc	36.33 bc	I60F180	33.6 c	14.31 d	437.31 c	3.27 b	39.76 c
I75F180	35.8 ab	15.92 bc	497.15 b	3.2 ab	44.23 ab	I80F180	34.8 bc	17.77 bc	505.37 bc	3.52 ab	49.36 b
I90F180	37.1 ab	18.26 a	549.6 a	3.32 a	50.73 a	I100F180	36.3 ab	20.53 a	513.88 ab	3.99 a	57.02 a
I105F180	39.1 a	18.81 a	595.72 a	3.16 ab	52.26 a	I120F180	36.7 ab	18.78 abc	567.06 a	3.31 b	52.18 ab
I60F240	34.45 b	11.71 e	446.28 c	2.62 c	24.4 d	I60F240	34.28 bc	13.04 d	431.73 c	3.02 c	27.17 d
I75F240	36.13 ab	15.22 cd	491.08 bc	3.1 b	31.7 cd	I80F240	35.12 b	17.17 c	492.13 bc	3.49 ab	35.77 c
I90F240	37.57 ab	17.47 ab	541.09 a	3.23 ab	36.4 bc	I100F240	36.66 ab	19.6 ab	505.81 ab	3.88 ab	40.84 c
I105F240	38.75 ab	17.95 ab	591.44 a	3.03 bc	37.39 bc	I120F240	37.07 a	17.81 bc	564.15 a	3.16 bc	37.1 c

Different lower-case letters denote significant differences at p < 0.05 among different treatments.

. The effect of different irrigation and fertigation treatments on yield, water consumption, and partial factor productivity was extremely significant (p < 0.01) and significant (p < 0.05) for water use efficiency. The 100-grain weight and yield were in the range of 33.6–39.1 g and 13.1–20.5 t ha⁻¹, respectively. At the same fertilization level, the yield increased with irrigation amounts in 2015, while it increased and then decreased with increasing irrigation amounts in 2016. The partial factor productivity and yield showed similar variation trends. For water consumption, the consumption in 2015 was higher than that in 2016 by 19.2 mm. The possible reason is that the soil water content was high before seeding, and the main consumption came from soil evaporation at the seedling stage, leading to ineffective water consumption. As a result, the average WUE in 2015 was lower than that in 2016 by 0.38 kg m⁻³. The treatments with maximum WUE in 2015 and 2016 were I₉₀F₁₈₀ and I₁₀₀F₁₈₀ with values of 3.32 and 3.99 kg m⁻³, respectively. I₁₀₅F₁₈₀ achieved the maximum PFP (52.26 kg kg⁻¹) in 2015, while I₁₀₀F₁₈₀ achieved the maximum PFP (57.02 kg kg⁻¹) in 2016.

4. Discussion

4.1. Effects of Irrigation and Fertilization on Soil Water Content and Soil Temperature

In agricultural production, the proper distribution of soil water, nutrients and temperature in the crop root zone is an important requirement for crop growth and high yield. Under drip irrigation conditions, the soil water content in 0–40 cm soil layers is mainly affected by irrigation. High irrigation levels often lead to higher water content [43] and deep water peaks in the soil profile [44]. In areas with shallow groundwater tables, the water content in the 60–100 cm soil layers mainly rises by capillary effects [45]. However, in the dry areas of northwestern China with no ground water recharges, the deep soil water content is mainly affected by irrigation and root absorption (Table 1). Using drip irrigation under film mulching, differences in the flow rate of drippers may cause large variations in soil water content at different lateral locations [26]. In addition, research studies have showed that differences in irrigation level and root absorption are the main reason for large soil water content differences [11,29]. The wetting ratio of drip irrigation and distribution of soil nitrate nitrogen can affect the root distribution [34,46,47,48]. In this study, under the double effect of irrigation level and root absorption, the SWC before irrigation was significantly different at different locations in various irrigation treatments (Table 1). The average soil water content in the 60–120 cm soil layer before irrigation showed a significant difference (p < 0.05). Although the water content was high near the drip irrigation side, the highest water content appeared at location P5 after maize root uptake (Table 1). Therefore, in actual production, small-amount/high-frequency irrigation systems should be used as much as possible to prevent the water and fertilizers from being transported beyond the root zone and to avoid the waste of water and fertilizers. Furthermore, the soil temperature can also be affected by the irrigation process, soil water content, soil nutrients and crop growth stages [25]. Because dry soils have lower heat capacity than wet soils, soils with low water content are easier to heat up during the day [48] and cool down at night. This explains why the soil temperature under different irrigation and fertigation treatments showed no apparent difference at the early growth stage. From the seedling to jointing stage, the soil temperature was mainly affected by solar radiation and temperature. The soil temperature at different locations in the same soil layer was similar. With the growth of spring maize, particularly after the 12 collars stage, the maize plants have a larger leaf area index and darken the ground. On the other hand, with lower irrigation levels, the solar radiation can easily reach the soil surface, creating higher soil temperatures than that at higher irrigation levels [49]. This study found that the average daily soil temperature at 5 cm soil depth under the I₆₀ treatment increased by 1.2 °C compared to that under the I₁₀₅ treatment in 2015. The average daily soil temperature at 5 cm soil depth under the I₆₀ treatment in 2016 increased by 2.1 °C compared to that under the I₁₂₀ treatment.

4.2. Effects of Irrigation and Fertilization on Soil Nitrate Nitrogen Content

The soil nitrate content is significantly affected by the water and fertilizer supply. Irrigation can cause soil nitrate nitrogen to migrate towards the deep soil layer. With the increase in irrigation amounts, the leaching level and depth are also increased. Under low water supply, most of nitrogen is concentrated in the upper soil layer. Lower irrigation levels correspond to higher remaining nitrate nitrogen content in the upper soil layer [14]. A high irrigation level can transfer a large amount of soil nitrate nitrogen downward into soil layers below the maize root zone. This then becomes an ineffective nitrogen supply [50]. Jia et al. [50] studied nitrate nitrogen leaching for summer maize under different water and nitrogen treatments. They found that the leaching was most serious in the 0–200 cm soil layer under the high irrigation and nitrogen treatments. Qiu et al. [51] studied the influence of irrigation depth, irrigation level and nitrogen fertilizer level on nitrate nitrogen leaching. It was found that nitrate nitrogen leaching increased with increasing irrigation level and nitrogen application. This study showed that the nitrate nitrogen content under the I₆₀ treatment was significantly higher than that under the other treatments in 0–120 cm soil layers under various fertilization levels at harvest. Compared with the I₆₀ treatment, the nitrate nitrogen content under the I₁₂₀ treatment in 2016 was decreased by 44.7%. Similarly, Wang et al. [52] concluded that nitrate nitrogen leaching can easily occur when the lateral uniformity is below 60% under drip irrigation conditions, and it is extremely easy for the soil nitrate nitrogen to migrate laterally with soil water away from the root zone, leading to reduced crop absorption. In this work, the variations in soil nitrate nitrogen were analyzed for different irrigation and fertigation levels, lateral locations and soil layers. In the surface soil layers (0–40 cm), the average soil nitrate nitrogen content at various fertilization levels had the minimum value under the I₉₀ (2015) and I₁₀₀ (2016) treatments. In terms of lateral locations, the nitrate nitrogen content in different soil layers was significantly different, with a maximum value at location P5 in the 0–60 cm soil layer. In deeper soil layers (60–120 cm), the average soil nitrate nitrogen content peaked at location P1, followed by locations P4 (2015) and P5 (2016). The second highest location, at P5 (2016), may be related to the increased irrigation level. Therefore, an excessively high irrigation level can transfer nitrate nitrogen beyond the root zone. Considering water conservation and nutrient migration depth, the irrigation level under drip irrigation conditions should not be too high.

4.3. Effects of Irrigation and Fertilization on Root Growth of Spring Maize

Root is a plant organ that absorbs water and nutrients from soils. Root growth is directly affected by the distribution of water and nutrients in the soil profile [23], as well as by soil temperature [24]. Zhang et al. [53] concluded that an optimal irrigation level in drip irrigation under film mulching can produce a good water and heat environment, promoting maize root growth in the 20–50 cm soil layers and enhancing the yield. In this paper, using the root distribution of spring maize at the filling stage as an index, the RLDs at different distances to the drip irrigation belt were studied. The research showed that different irrigation and fertigation levels affected RLD significantly. In the 2 year growing season, shallow soil layers (0–20 cm) demonstrated root growth towards the near side of the drip irrigation belt under low irrigation levels. With increasing irrigation, the root growth on the far side of the drip irrigation belt was enhanced (Figure 4). This can be attributed to the differences in distribution of soil water and nitrate nitrogen under drip irrigation conditions with different irrigation and fertigation levels (Table 1 and Table 2). The RLD of deep soil layers at various locations peaked under the I₇₅ (2015) and I₈₀ (2016) treatments. The main reason is that the ratio of root in deep soil layers is increased at the early growth stage with moderate water shortage [54]. Generally, the influences of soil water and nitrate nitrogen on root growth are not isolated but interact with each other [26].

4.4. Effects of Irrigation and Fertigation on Yield, Water Use Efficiency and Partial Factor Productivity

Appropriate water and nitrogen applications are helpful to improve the water use efficiency of crops [28]. Chilundo et al. [23] revealed that the WUE was not affected by the quantity of water, but a decrease in the rate of nitrogen application caused a decrease in WUE at all quantities of water applied. This study found no significant impact of fertilization (F180 and F240) on WUE. As for the irrigation amount, mild water stress showed no significant effect, but severe water stress significantly improved water use efficiency. Jin et al. [55] concluded that drip irrigation under film mulching can create better soil water and heat environments, enhancing nutrient absorption at the early growth stage of maize and promoting vegetative growth of maize. In addition, more nutrients are accumulated in the late growth stage, increasing the number of spikes and mass of 100 grains. The booting stage to milk ripening stage of maize are the key stages to form grains. During these stages, the temperature is high, and the oxygen required by spring maize plants increases. The respiration of roots can be suppressed by overly low soil water content or overly high soil nutrients, leading to premature failure of the root system and a decrease in root activity, which in turn results in reduced water and nutrient use efficiency. Appropriate irrigation and fertilization levels can increase soil permeability. Moreover, fewer nutrients are transferred beyond the root zone, and the soil temperature is appropriate, leading to good root respiration. As a result, more water and nutrients are produced from plant photosynthesis. Irmak et al. [56] found that the yield from 75% full irrigation under drip irrigation conditions was statistically similar to that from full irrigation treatment with maximum water use efficiency, which was different from our results. In this study, the maximum yield and water use efficiency were achieved at the irrigation level of 90–100% ET_c under various fertilization levels in 2015 and 2016. A possible reason could be that the precipitation was only 150 mm in the study area, while the precipitation in Irmak’s test area was above 400 mm. The extra precipitation provided required water for maize growth in their study.

5. Conclusions

With increases in irrigation amounts, maize grain yield, PFP, and WUE increased in 2015, while they first increased and then decreased in 2016. Increasing fertilization rates had no significant impact on yield and WUE when irrigation amounts were increased, but it significantly reduced PFP. In general, there was irrigation and fertilization had a significant effect on water content in the 0–80 cm soil layer and average daily temperature at the 5, 10, and 15 cm depths from the 6 collars stage to maturity stage of spring maize in 2015 and 2016. The soil water content between the wide rows was significantly different (p < 0.05) from that at other locations, and the RLD of maize on the near side of the drip irrigation belt decreased with increasing irrigation amounts, while that on the far side increased with increasing irrigation amounts in the root zone (0~60cm). The highest water use efficiency occurred under the I₉₀F₁₈₀ treatment at 3.32 kg m⁻³ in 2015, with only a 2.9% decrease in grain yield compared to the highest yield treatment (I₁₀₅F₁₈₀) and 65 mm of irrigation water savings. I₁₀₀F₁₈₀ obtained the highest grain yield (20.5 t ha⁻¹) and water use efficiency (3.99 kg m⁻³) in 2016. Comprehensively considering the grain yields and water scarcity in this region, the optimal water-fertilizer coupling strategy for spring maize in arid northwestern China includes 90–100% ET_c and 180-90-90 (N-P₂O₅-K₂O) kg hm⁻². The results of this study can provide guidance for the optimal design and operation of irrigation fertilization systems for spring maize under drip irrigation with film mulching in arid regions of northwestern China.