Water-Saving Irrigation and N Reduction Increased the Rice Harvest Index, Enhanced Yield and Resource Use Efficiency in Northeast China

Abstract

For agricultural production, improving the rice harvest index (HI) through agricultural management practices is a major means to enhance water and N utilization efficiency and yield. Both irrigation regimes and nitrogen (N) rates are important aspects of agricultural management practices. However, it is unclear how the rice HI is affected by water and N. This study aimed to clarify the mechanism underlying the response of the rice HI to water and N, and to explore the most suitable water-saving and N reduction management practices to ensure yield. A two-year (2021~2022) field experiment was conducted on Mollisols in Northeast China. In this experiment, nine treatments were performed, involving three irrigation regimes (flooded irrigation, controlled irrigation, and “thin-shallow-wet-dry” irrigation) and three N rates (110, 99, and 88 kg/ha). The rice agronomic traits and transfer of photoassimilates under different water and N management practices were observed and studied; rice HI, WUE, and the NUE of rice was calculated and analyzed. The highest HI was achieved under controlled irrigation with a 99 kg/ha N rate, at values of 0.622 and 0.621 in 2021 and 2022, respectively. Controlled irrigation (CI) with an appropriate reduction in the N rate increased the proportion of productive tillers, the transfer rate of dry matter and non-structural carbohydrates (NSCs), the sugar–spikelet ratio, the grain–leaf ratio, and the leaf area index (LAI) during the heading–flowering stage. A subsequent analysis indicated that the main reason for the increase in the HI was the increase in the sugar–spikelet ratio during the heading–flowering stage. A high HI increased the rice yield by increasing the thousand-grain weight. The present study suggested that water-saving irrigation regimes and appropriate N reduction not only led to water and fertilizer resource savings but also improved agronomic characteristics during rice growth and enhanced transport capacity. Thus, these practices improved the rice HI and have enormous potential for increasing yield. Therefore, regulating the rice HI through water and N management methods should be considered an important strategy for improving rice yield.

1. Introduction

Rice is one of the most important food crops in the world [1] and provides a food source for one-third of the world’s population [2]. Given the current situation of increasingly scarce water resources and a growing population, achieving further increases in rice yield is an important goal in global agriculture [3]. To increase rice output above the current level, productive capacity must be increased [4,5]. The harvest index (HI) is an important indicator that determines the production potential of crops. It is closely related to the source–sink relationship of crops [6,7]. The balance between source and sink is crucial for ensuring rice yield [8]. Regarding the improvement of rice HI, previous research has mostly focused on the improvement of rice varieties, and the response mechanism of a single variety to agricultural management practices is not yet clear. Therefore, it is crucial to clarify the response mechanism of rice HI to agricultural management practices for developing scientific rice management strategies [9,10].

To improve the rice HI, the dry matter transport capacity should be enhanced, and the photosynthetic product distribution should be optimized [11]. In addition, the inherent objective is to improve the efficiency of the transfer of photosynthetic products to the panicle after the rice heading–flowering stage. Non-structural carbohydrates (NSCs) stored in the stem sheath are important grain-filling substances that can stimulate the activity of “sinks” in the early stage of grain filling and initiate and promote grain filling [12]. According to previous studies, the rice HI has improved mainly in the following ways. First, the proportion of productive tillers increases to reduce redundant vegetative growth [13,14]. Secondly, as the economic organ of rice, spikelets can actively extract photosynthetic assimilates from photosynthetic production organs. Therefore, increasing the “sinks” on the basis of an appropriate LAI, that is, increasing the grain–leaf ratio of rice, can promote the transport of photosynthetic assimilates to grains [15]. Moreover, increasing the sugar–spikelet ratio during the heading–flowering stage of rice may be beneficial for improving the rice HI. A higher sugar–spikelet ratio during the heading stage may increase the accumulation of NSCs in the rice stem and sheath before heading and is beneficial for enhancing the transport of assimilates from the stem and sheath to the grain during the heading to maturity stage, promoting rice grain filling [16].

It is of great significance to study the response mechanism of rice HI to agricultural management practices, which is important for reducing resource waste, improving rice production potential, and establishing suitable agricultural management modes. Specifically, the rice HI is influenced by many factors, such as rice variety, environmental conditions, irrigation modes, and N application levels [17,18,19,20]. Several studies have shown that water and N management have a significant impact on the source–sink balance of rice, causing differences in rice HI by affecting the distribution and transportation of photoassimilates [21,22]. Adopting appropriate irrigation regimes based on the region and climate can improve agronomic traits such as root morphology [23], dry matter distribution [24], and tillering growth dynamics [25], thereby increasing the production potential of rice. Wang et al. [26] showed that water-saving irrigation could improve biomass accumulation in rice while promoting dry matter and nutrient transport in the later stages of rice growth. Compared to conventional irrigation, a water-saving irrigation regime could effectively reduce non-productive tillering of rice and optimize its source–sink balance [27]. A water-saving irrigation regime could enhance the rice HI due to inhibiting ineffective tillering can increase the reactivation of NSCs stored in the stem towards the grains [28]. Several studies have indicated that a moderately dry irrigation regime can promote the redistribution of accumulated assimilates in rice and enhance the activity of amylase and sucrose-phosphate synthase (SPS) in the stems of rice [29,30]. In terms of N fertilizer management, a study in Southwest China showed that the dry matter transport capacity of rice was significantly greater at moderate N application rates than at high or low N application rates [31]. Moreover, a reasonable N input can promote the accumulation and transport of NSCs in rice stems and sheaths [32]. According to previous reports, the nutrients and soluble sugars required for grain development mainly originate from plant nutrient remobilization. High N application rates enhance the allocation of photosynthetic products to structural carbon, which cannot be transferred to crop grains in the later stages of plant growth, leading to a decrease in the HI [33]. At present, the source–sink relationship and distribution of photosynthetic products in rice plants can be regulated by selecting appropriate water or N fertilizer management practices, but there is relatively little information on the response mechanism of rice HI to the interaction between water-saving and N reduction measures.

As one of the four major black soil regions in the world, Northeast China is an important grain production base in China. As of 2021, the rice planting area in Northeast China was 5.26 × 10⁶ ha [34]. However, excessive N fertilizer input during rice cultivation not only fails to increase yield, but also leads to a large amount of residual NH₄⁺-N and NO₃⁻-N in the soil. The residual NH₄⁺-N and NO₃⁻-N enter the water and atmosphere through leaching, nitrification, and denitrification, posing a threat to the ecological environment [35,36]. A study from ¹⁵N indicated that controlled irrigation could not only supplement surface soil fertility, but also reduce fertilizer N leaching, ensuring crop N utilization [37]. Moreover, previous studies have shown that appropriate N fertilizer management measures and a moderate degree of soil desiccation could prevent the activity of nitrification and denitrification, reduce nitrogen fertilizer loss, and ensure the effectiveness of N fertilizer and grain yield [38,39]. Therefore, a moderate reduction in N fertilizer and water-saving irrigation can ensure rice yield while reducing potential environmental risks caused by excess N fertilizer [40,41]. In recent years, water-saving irrigation regimes have gradually been promoted in Northeast China to address the issue of increased agricultural water consumption [42]. In this context, determining whether reducing N fertilizer application under different water-saving irrigation regimes can improve rice HI and revealing the mechanism by which water-N interactions enhance rice HI are highly important for practicing sustainable agriculture.

For this purpose, in this study, a 2-year field experiment was conducted with different irrigation regimes and N rates. Our vision was to improve the related agronomic traits of photoassimilates’ transport and distribution of rice by combining the management practices of water-saving irrigation and reducing N application, achieving the goal of conserving water and fertilizer resources while ensuring yield. The rice yield and its components, the accumulation and transport of aboveground biomass, the accumulation and transport of NSCs, the proportion of productive tillers, the leaf area index (LAI), the sugar–spikelet ratio, and the grain–leaf ratio during the heading–flowering stage were observed, and the impact of relevant indicators on the rice HI was analyzed. This study aimed to reveal the underlying mechanism of and the impact of water and N management strategies on rice HI in Northeast China. The findings of this study will provide useful information for further enhancing the production potential of rice and will provide new insights into the impact of water and N management practices on rice growth.

2. Materials and Methods

2.1. Site Description

This experiment was conducted at the Rice Irrigation Test Station (46°57′58″ N, 127°39′39″ E) in Heilongjiang Province, China. The station is located in the Heping Irrigation District, Qing’an County, Songnen Plain, Northeast China (Figure 1). The meteorological temperature and precipitation data collected during the 2021 and 2022 rice growing seasons are shown in Figure 2. The region has an average annual precipitation of 500–600 mm, a frost-free period of 128 days and is characterized by a continental monsoon climate in the cold temperate zone. According to the U.S. Department of Agriculture (USDA) soil taxonomy, the soil in the experimental field was classified as Mollisol, which is the main soil type on the Songnen Plain in Northeast China. Rice has been cultivated in the experimental area for more than 20 years. The physical and chemical properties of the 0–20 cm soil layer are shown in

Table 1. Properties of the tested soil.

Soil Lay-ers/(cm)	Organic Matter/(g/kg)	Alkaline N/(mg/kg)	Available P/(mg/kg)	Available K/(mg/kg)	pH Value	Particle Composition/(%)			C/N Ratio	Soil Texture
						2.0–0.02	0.02–0.002	<0.002
0–20	36.8	181.8	31.1	115.8	6.4	37.2	32.1	30.1	13.91	Sand clay
20–40	32.8	147.6	26.50	105.9	6.9	31.9	37.2	30.4	10.92	Sand clay

.

2.2. Experimental Design

Three irrigation regimes were established (

Table 2. Different water management regimes during the rice growth stages.

Irrigation Regime	Regreening	Early Tillering	Middle Tillering	Late Tillering	Jointing–Booting	Heading–Flowering	Grain Filling
CI	0~30	70%θs~30	70%θs~30	Drainage	80%θs~30	80%θs~30	70%θs~0
WI	0~30	0~10	90%θs~10	Drainage	10~20	5~15	0~10
FI	0~30	30~50	30~50	Drainage	30~50	30~50	30~50

Note: θs is the soil-saturated water content mass fraction in the root layer. The data before “~” are the lower limit of moisture control, and the data after “~” are the upper limit of moisture control. Depth of the field water layer unit: mm.

): flooded irrigation (FI), controlled irrigation (CI), and thin-shallow-wet-dry irrigation (WI). All the irrigation regimes involved natural drying during the yellow–ripening stage. Under CI, the irrigation duration and amount were determined by the soil moisture content of the root layer. The soil moisture content corresponding to the upper limit of irrigation was the saturated water content of the soil, and the lower limit was a different percentage of the saturated moisture content in each growth stage. At 7:00 and 18:00 every day, a TPIME-PICO64/32 soil moisture tester was used to measure the soil moisture content of each experimental plot. When the soil water content was close to or below the lower limit of irrigation, artificial irrigation was carried out until the upper limit of irrigation was reached. The soil water content was kept between the upper and lower limits of irrigation at the corresponding growth stage every day, and the irrigation amounts for each treatment were recorded. Under FI, a layer of water (30–50 mm) was maintained in the field during rice growth. Natural drying occurred during the yellow–ripening stage, while proper drainage and sun-drying were implemented only during the late tillering stage. Under WI, a thin water layer of less than 10 mm was maintained after transplanting, and the soil moisture was kept in a saturated state. Drainage and sun-drying were performed at the late tillering stage; a shallow water layer of 10~20 mm was maintained during the jointing–booting stage; a thin water layer of 5~15 mm was maintained during the heading–flowering stage; and a thin water layer of 10 mm was maintained during the grain-filling stage. The depth of the water layer in the paddy field was measured in the test plot every day using a buried ruler at 8:00 to determine whether irrigation was needed. The water depth data of the field from transplanting to harvesting under three irrigation regimes are shown in Figure 3, where the data for each irrigation regime are taken as the average of the three N rates under that irrigation regime. The N rate (pure N) was set at three levels: a conventional N rate (N, 110 kg/ha N rate), a 10% reduction (N1, 99 kg/ha N rate), and a 20% reduction (N2, 88 kg/ha N rate). The experimental treatment design is presented in

Table 3. Experimental treatment design.

Irrigation regime	Controlled irrigation
N rate (kg/ha)	110	99	88
Treatment name	CN	CN1	CN2
Irrigation regime	“Thin-shallow-wet-dry” irrigation
N rate (kg/ha)	110	99	88
Treatment name	WN	WN1	WN2
Irrigation regime	Flooded irrigation
N rate (kg/ha)	110	99	88
Treatment name	FN	FN1	FN2

. The field design was based on a two factor random split plot design, with a total of 9 treatments, each with 3 replicates, for a total of 27 experimental plots, each with an area of 100 m² (10 m × 10 m). The width and height of the ridges between the plots were about 30 and 10 cm, respectively. To prevent the exchange of water and N between each plot, plastic plates with a depth of 40 cm were inserted on both sides of the ridges. The tested rice variety was Suijing 18, the main cultivated variety in Heilongjiang Province. The planting density was 24 hills/m², with 3 plants per hill. N, P and K fertilizers were selected from commonly used varieties by local farmers. Urea (containing 46.4% N) was used as the N fertilizer, which was applied at a basal: tillering: panicle fertilizer ratio of 5:2:3. Basal fertilizer was applied 1 day before rice transplantation, and the tillering and panicle N fertilizer were applied 14 and 54 days after transplanting, respectively. Calcium superphosphate (containing 12% P₂O₅) was used as the phosphorous fertilizer and potassium sulfate (containing 60% K₂O) was used as the potassium fertilizer. The amounts of phosphorus and potassium fertilizers applied in each treatment were the same: 45 kg/ha P₂O₅ and 80 kg/ha K₂O. Phosphorus fertilizer was applied once before transplanting, and potassium fertilizer was applied twice, once before transplanting and once at the age corresponding to 8.5 rice leaves, at a ratio of 1:1. The management of pesticides was consistent with the conventional management of local farmers.

2.3. Laboratory Analysis and Data Analysis

2.3.1. Stems and Tillers

On the 5th day after transplantation, 5 evenly growing hills in each treatment area were marked, and the number of tillers (spikes) was measured every 5 days until the maturity stage.

$$\mathrm{P}\mathrm{r}\mathrm{o}\mathrm{p}\mathrm{o}\mathrm{r}\mathrm{t}\mathrm{i}\mathrm{o}\mathrm{n} \mathrm{o}\mathrm{f} \mathrm{p}\mathrm{r}\mathrm{o}\mathrm{d}\mathrm{u}\mathrm{c}\mathrm{t}\mathrm{i}\mathrm{v}\mathrm{e} \mathrm{t}\mathrm{i}\mathrm{l}\mathrm{l}\mathrm{e}\mathrm{r}\mathrm{s} (\%)=\frac{\mathrm{N}\mathrm{u}\mathrm{m}\mathrm{b}\mathrm{e}\mathrm{r} \mathrm{o}\mathrm{f} \mathrm{s}\mathrm{p}\mathrm{i}\mathrm{k}\mathrm{e}\mathrm{s} \mathrm{a}\mathrm{t} \mathrm{t}\mathrm{h}\mathrm{e} \mathrm{y}\mathrm{e}\mathrm{l}\mathrm{l}\mathrm{o}\mathrm{w}-\mathrm{r}\mathrm{i}\mathrm{p}\mathrm{e}\mathrm{n}\mathrm{i}\mathrm{n}\mathrm{g} \mathrm{s}\mathrm{t}\mathrm{a}\mathrm{g}\mathrm{e}}{\mathrm{M}\mathrm{a}\mathrm{x}\mathrm{i}\mathrm{m}\mathrm{u}\mathrm{m} \mathrm{n}\mathrm{u}\mathrm{m}\mathrm{b}\mathrm{e}\mathrm{r} \mathrm{o}\mathrm{f} \mathrm{t}\mathrm{i}\mathrm{l}\mathrm{l}\mathrm{e}\mathrm{r}\mathrm{s}}$$

(1)

2.3.2. Aboveground Biomass

The dry matter accumulation in each aboveground organ in each treatment was determined at the heading–flowering and maturity stages. Plant samples were collected by the five-point sampling method and washed with an agricultural compression sprayer. The organs were separated, put into a sample bag, and taken back to the laboratory, where the samples were put into an oven, heated to 105 °C, and sterilized under air blast conditions for 30 min. The samples were subsequently dried at 85 °C to a constant mass, after which the dry matter weights of the different parts were determined.

2.3.3. Content of Non-Structural Carbohydrates (NSCs)

The stems and sheaths of the plant samples were dried and crushed to measure the aboveground biomass, after which the plants were passed through a 100-mesh sieve. A 0.1 g dry sample was weighed into a 10 mL centrifuge tube, and Yoshida’s method [43] was used to measure the concentrations of soluble sugars and starch. The sum of the two contents was defined as the content of NSCs.

2.3.4. N Content

The grains, stems, leaves and roots were washed using a pressurized water gun, placed in a drying oven at 105 °C for 30 min, dried at 85 °C to a constant mass and weighed. The weighed rice plant was crushed into various portions corresponding to the plant organs with a pulverizer, screened through an 80 mesh (0.18 mm) screen and digested by the H₂SO₄–H₂O₂ method, and the solution was analyzed with a continuous flow analyzer (autoanalyzer-3, Bran + Luebbe, Norderstedt, Germany) for N content [40].

2.3.5. Yield and Yield Components

After harvest, the rice yield in each treatment was measured using the five-point sampling method. After drying, the number of effective panicles, grains per panicle, grain setting rate, and thousand-grain weight of the rice plants were measured to obtain the rice yield.

2.3.6. Leaf Area

Three hills were selected in each plot, and the effective leaf area was manually measured using a steel ruler with millimeter-scale accuracy. The leaf area index (LAI) was calculated using the area (coefficient) method, with a coefficient of 0.75.

2.4. Calculation Methods for Relevant Indicators

2.4.1. Harvest Index ($HI)$

$$HI=\frac{Yield}{Dry matter accumulation}$$

(2)

2.4.2. Transfer of Dry Matter

Transfer amount of dry matter = the amount of dry matter in the aboveground vegetation at the heading–flowering stage − the amount of dry matter in the aboveground vegetation at the maturity stage.

Transfer rate of dry matter (DMTR) = the transfer amount of dry matter/the amount of dry matter in the aboveground vegetation at the heading stage, where the transfer amount of dry matter is the amount of apparent dry matter transferred from vegetative organs after heading–flowering, kg/ha; the transfer rate of dry matter is the transfer rate of dry matter from vegetative organs after heading–flowering, % [35].

2.4.3. Transfer of NSCs

$$Content of NSCs=Content of soluble sugar+Content of starch$$

(3)

where $Content of NSCs$ is the content of non-structural carbohydrates in the vegetative organs of rice, g/m²; $Content of soluble sugar$ is the content of soluble sugar in the vegetative organs of rice, g/m²; and $Content of starch$ is the content of starch in the vegetative organs of rice, g/m².

The transfer amount of NSCs (g/m²) = the amount of NSCs in the aboveground vegetation at the heading–flowering stage − the amount of NSCs in the aboveground vegetation at the maturity stage.

The transfer rate of NSCs (%) = the amount of NSCs transferred after heading–flowering/the amount of NSCs transferred from the aboveground vegetation at heading–flowering [44].

2.4.4. Nitrogen Use Efficiency (NUE) for Grain Production [45]

$$NUE=\frac{Grain yield per unit area}{Total N uptake per unit area}$$

(4)

2.4.5. Water Use Efficiency $\left(WUE\right)$

$$\begin{gathered} WUE=\frac{Yield}{ET} \\ ET=P+I+G+W1-R-D-W2 \end{gathered}$$

(5)

where $ET$ is the water consumption, mm; $P$ is the rainfall, mm; $I$ is the irrigation amount, mm; $G$ is the amount of groundwater recharge, mm; $W1$ is the soil water storage capacity of the 0~60 cm soil layer after rice transplanting, mm; $R$ is the amount of drainage during rice growth, mm; $D$ is the amount of deep soil leakage, mm; and $W2$ is the soil water storage capacity during rice harvest, mm [20,46].

2.4.6. Leaf Area Index ($LAI)$

$$LAI=\frac{\mathrm{T}\mathrm{o}\mathrm{t}\mathrm{a}\mathrm{l} \mathrm{l}\mathrm{e}\mathrm{a}\mathrm{f} \mathrm{a}\mathrm{r}\mathrm{e}\mathrm{a}}{\mathrm{L}\mathrm{a}\mathrm{n}\mathrm{d} \mathrm{a}\mathrm{r}\mathrm{e}\mathrm{a}}$$

(6)

2.4.7. Sugar–Spikelet Ratio and Grain–Leaf Ratio at the Heading–Flowering Stage

$$Sugar-spikelet ratio=\frac{Accumulation of NSCs}{Visible number of spikelets}$$

(7)

$$Grain-leaf ratio=\frac{Visible number of spikelets}{LAI}$$

(8)

where $Sugar-spikelet ratio$ is the sugar–spikelet ratio at the heading–flowering stage, mg/spikelet; $Visible number of spikelets$ is the visible number of spikelets per square meter at the heading–flowering stage, grain/m²; and $Grain-leaf ratio$ is the grain–leaf ratio at the heading–flowering stage, grain/cm² [16,47].

2.5. Data Processing Methods

All indicator values were calculated as averages. Microsoft Excel 2019 was used to conduct a preliminary statistical processing of the experimental data. Origin 2021 was used for plotting and correlation analysis, Pearson’s correlation was used at p = 0.05 and p = 0.01 for correlation analysis. SPSS 22.0 was used for significance analysis, Duncan’s test was used for multiple comparisons at p = 0.05, and a two-way analysis of variance (ANOVA) (p = 0.05) was used to test the main and interactive effects of irrigation and N management on the related indicators of rice growth.

3. Results

3.1. Aboveground Dry Matter Accumulation, Yield, and Components

According to

Table 4. Response of aboveground dry matter accumulation and yield to water and N practices and analysis of variance.

	Treatment	Aboveground Dry Matter Accumulation (kg/ha)		Yield (kg/ha)
		2021	2022	2021	2022
	FN	13,377.33 d	13,380 d	7559.83 e	7238.21 e
	FN1	15,168.8 a	15,225.6 a	9219.16 a	9154.85 a
	FN2	13,552 d	13,639.2 cd	7717.62 de	7583.34 de
	CN	13,781.2 cd	13,797.6 cd	8191.97 c	8320.72 bc
	CN1	14,938.4 ab	14,940 ab	9285.77 a	9277.9 a
	CN2	13,818.4 cd	13,821.6 cd	8149.15 c	7892.8 cd
	WN	14,312.8 bc	14,311.2 bc	8681.98 b	8472.93 b
	WN1	14,840 ab	14,908.8 ab	9220.83 a	9213.32 a
	WN2	13,806.4 cd	13,744.8 cd	7986.15 cd	7683.62 de
Source of variation
Irrigation regime (I)	F value	21.772		21.993
	p value	<0.001		<0.001
N rate (N)	F value	56.895		178.120
	p value	<0.001		<0.001
I × N	F value	22.498		10.853
	p value	<0.001		<0.001

Note: the different letters of the data indicate that the difference level between treatments was significant (p < 0.05), compared to data within the same year, same as below. F, C and W represent flooded irrigation, controlled irrigation and “thin-shallow-wet-dry” irrigation, respectively. N, N1 and N2 represent the 110, 99 and 88 kg/ha N rates, respectively.

, under the same irrigation regime, compared with those at the 110 and 88 kg/ha N rates, the dry matter accumulation at the 99 kg/ha N rate significantly increased by 11.63~13.79%, 8.09~8.40%, and 3.68~8.47% compared to those under FI, CI and WI, respectively. At the same N rate, the dry matter accumulation in the WN treatment was 6.96~6.99% and 3.72~3.86% greater than that in the FN and CN treatments, respectively. The values of FN1 were 1.54~1.91% and 2.12~2.21% greater than those of CN1 and WN1, respectively, while the values of CN2 were 1.34~1.97% and 0.09~0.56% greater than those of FN2 and WN2, respectively.

Under the same irrigation regime, the yield was greatest at the 99 kg/ha N rate and significantly greater than that at the 110 and 88 kg/ha N rates (p < 0.05). Compared with those at the 110 and 88 kg/ha N rates, the yields at the 99 kg/ha N rate increased by 19.46~26.48%, 11.50~17.55% and 6.21~19.91% under FI, CI and WI, respectively. At the same N rate, the values under WN were 14.84~17.06% and 1.83~5.98% greater than those under FN and CN, respectively. At the 99 and 88 kg/ha N rates, the yield was greatest under CI. CN1 had values 0.70~1.34% and 0.70~0.70% higher than those of FN1 and WN1, respectively, while CN2 had values 4.08~5.59% and 2.04~2.72% higher than those of FN2 and WN2, respectively.

The results indicated that irrigation, N, and their interactions had a significant impact on dry matter accumulation and rice grain yield (p < 0.001) (Table 4). At the same N rate, FI reduced the yield. Under the same irrigation regime, the 99 kg/ha N rate not only significantly increased dry matter accumulation but also improved yield. Therefore, an appropriate water and N management practice is crucial for improving biomass accumulation and yield.

The yield components are shown in

Table 5. Components of rice yield to water and N practices.

	Number of Effective Panicles (106/ha)		Thousand-Grain Weight (g)		Seed Setting Rate		Number of Grains Per Panicle
	2021	2022	2021	2022	2021	2022	2021	2022
FN	4.08 ± 0.02 d	4.06 ± 0.03 c	23.93 ± 0.21 e	22.92 ± 0.44 f	0.87 ± 0.005 e	0.84 ± 0.01 d	89 ± 1.51 abc	92.6 ± 2.42 abc
FN1	4.24 ± 0.03 b	4.14 ± 0.02 b	27.07 ± 0.12 bc	27.02 ± 0.49 bc	0.89 ± 0.005 cd	0.88 ± 0.01 bc	90.25 ± 1.01 ab	93 ± 2 ab
FN2	4.16 ± 0.01 c	3.98 ± 0.02 d	24.42 ± 1.0 e	24.24 ± 0.40 e	0.88 ± 0.005 de	0.86 ± 0.03 bcd	86.33 ± 0.85 bc	91.4 ± 1.51 bc
CN	4.16 ± 0.01 c	4.04 ± 0.02 c	24.47 ± 0.32 e	25.24 ± 0.38 d	0.87 ± 0.01 e	0.85 ± 0.02 cd	92.5 ± 1.99 a	96 ± 2 a
CN1	4.08 ± 0.01 d	3.9 ± 0.03 e	28.02 ± 0.19a	27.96 ± 0.29 a	0.9 ± 0.011 bc	0.89 ± 0.02 ab	90.25 ± 2.90 ab	95.6 ± 2.25 b
CN2	4.08 ± 0.02 d	3.88 ± 0.02 e	25.61 ± 0.53 d	25.36 ± 0.45 d	0.89 ± 0.002 cd	0.87 ± 0.01 bcd	87.63 ± 4.19 bc	92.2 ± 2.03 abc
WN	4.32 ± 0.02 a	4.22 ± 0.03 a	27.8 ± 0.37 ab	26.68 ± 0.38 c	0.88 ± 0.01 de	0.87 ± 0.03 bcd	82.15 ± 0.69 d	86.5 ± 2.18 d
WN1	4.16 ± 0.02 c	4.06 ± 0.03 c	27.95 ± 0.21 a	27.56 ± 0.44 ab	0.92 ± 0.005 a	0.92 ± 0.01 a	86.2 ± 0.81 c	89.5 ± 2.18 bcd
WN2	3.84 ± 0.02 e	3.73 ± 0.03 f	26.73 ± 0.48 c	25.72 ± 0.43 d	0.91 ± 0.01 ab	0.89 ± 0.01 ab	85.5 ± 2.00 cd	88.8 ± 2.43 cd

. Under FI, the number of effective panicles was highest at the 99 kg/ha N rate, while under CI and WI, it was greatest at the 110 kg/ha N rate. The thousand-grain weight of the rice was the highest at the 99 kg/ha N rate under the same irrigation regime. At the 110 and 88 kg/ha N rates, the thousand-grain weights were both highest under WI, while at the 99 kg/ha N rate, the thousand-grain weight was highest under CI. The seed setting rate of rice was the highest at the 99 kg/ha N rate under the same irrigation regime, and it was the highest under WI at the same N rate. The number of grains per panicle was greatest at the 99 kg/ha N rate under both FI and WI, and decreased with a decreasing N rate under CI. The number of grains per panicle was greatest under CI at the same N rate. From the above analysis, there was an offset balance between the four yield component factors.

As shown in

Table 6. Analysis of variance of components of rice yield under different water and N management practices.

Source of Variation	Number of Effective Panicles		Thousand-Grain Weight		Seed Setting Rate		Number of Grains Per Panicle
	F Value	p Value	F Value	p Value	F Value	p Value	F Value	p Value
I	6.2	0.004	70.859	<0.001	16.435	<0.001	19.292	<0.001
N	33.186	<0.001	112.733	<0.001	26.124	<0.001	2.464	0.097
I × N	18.430	<0.001	14.626	<0.001	0.395	0.811	2.444	0.06

, the differences in the four indicators were significant under the different irrigation or N rates (p < 0.05), but the effect of water and N interaction on the seed setting rate and the number of grains per panicle was not statistically significant (p > 0.05). The results showed that CI was not conducive to the accumulation of effective panicles, but had the greatest number of grains per panicle. Under FI, the thousand-grain weight was low, and WI was most conducive to improving the seed setting rate. Reducing the N rate appropriately improved the thousand-grain weight and seed setting rate.

3.2. Harvest Index (HI)

According to

Table 7. Response of harvest index to water and N practices and analysis of variance.

	Treatment	Harvest Index
		2021	2022
	FN	0.565 e	0.541 f
	FN1	0.608 a	0.601 bc
	FN2	0.569 de	0.556 e
	CN	0.594 c	0.603 b
	CN1	0.622 a	0.621 a
	CN2	0.59 c	0.571 d
	WN	0.607 b	0.592 c
	WN1	0.621 a	0.618 a
	WN2	0.578 d	0.559 e
Source of variation
I	F value	42.67
	p value	<0.001
N	F value	108.426
	p value	<0.001
I × N	F value	8.712
	p value	<0.001

, the HI of CN1 was the highest, reaching 0.622 (2021) and 0.621 (2022). Under the same irrigation regime, the rice HI was greatest at the 99 kg/ha N rate and was significantly greater than those at the 110 and 88 kg/ha N rates (p < 0.05). Compared with those at the 110 and 88 kg/ha N rates, the HI at the 99 kg/ha N rate increased by 6.85~11.09% under FI, by 2.99~8.76% under CI and by 2.31~10.55% under WI. In the two-year experiment, the rice HI under FI was the lowest at the same N rates. Similarly, the rice HI in CN and WN were significantly greater than that in FN (5.13~11.46% and 7.43~9.43%, respectively); those in CN1 and WN1 were significantly greater than that in FN1 (2.30~3.33% and 2.14~2.83%, respectively; p < 0.05); and those in CN2 and WN2 were 2.70~3.69% and 0.54~1.58% greater than that in FN2.

As shown in Table 7, irrigation regime, N and their interactions had a significant impact on the HI (p < 0.001). The research showed that the 99 kg/ha N rate was the most beneficial for improving the rice HI, while flooded irrigation could reduce it.

3.3. NUE and WUE

There was a significant difference in WUE among different water and N management practices (p < 0.05) (

Table 8. Response of water use efficiency (WUE) and nitrogen use efficiency (NUE) to water and N practices and analysis of variance.

	Treatment	WUE (kg/m3)		NUE
		2021	2022	2021	2022
	FN	1.18 h	1.14 h	62.50 f	59.99 d
	FN1	1.20 g	1.17 g	72.83 abc	68.99 ab
	FN2	1.15 i	1.11 i	68.32 cde	65.98 bc
	CN	1.82 b	1.78 b	66.13 def	63.98 bcd
	CN1	1.94 a	1.91 a	77.61 a	74.03 a
	CN2	1.72 c	1.68 c	71.51 bcd	68.97 ab
	WN	1.68 d	1.64 d	64.88 ed	61.98 cd
	WN1	1.60 e	1.58 e	76.17 ab	73.77 a
	WN2	1.43 f	1.40 f	70.35 cd	68.01 b
Source of variation
I	F value	65.762		60.472
	p value	0.001		0.001
N	F value	4.000		474.504
	p value	0.111		<0.001
I × N	F value	67.013		0.067
	p value	<0.001		0.991

). Under WI, WUE decreased with decreasing N rate, and WUE at the 110 kg/ha N rate was increased by 3.80~17.26% compared to that at the 99 and 88 kg/ha N rates. Under FI and CI, the WUE at the 99 kg/ha N rate was the highest. Compared with those at the 110 and 88 kg/ha N rates, the WUE under FI and CI increased by 1.69~5.41% and 6.59~13.69%, respectively. In a two-year study, CI had the highest WUE at the same N rate. CN had values 54.24~56.14% and 8.54~8.54% greater than those of FN and WN, respectively; CN1 had values 61.67~63.25% and 20.89~21.25% greater than those of FN1 and WN1, respectively; and CN2 had values 49.57~51.35% and 20.00~20.28% greater than those of FN2 and WN2, respectively.

Under the same irrigation regime, NUE was highest at the 99 kg/ha N rate (Table 8). Compared with those at the 110 and 88 kg/ha N rates, the NUE at the 99 kg/ha N rate increased by 4.56~16.53%, 7.34~17.36% and 8.27~19.02% under FI, CI, and WI, respectively. In the two-year study, at the same N rate, NUE was greatest under CI. The values of CN were 5.80~6.65% and 1.93~3.23% greater than those of FN and WN, respectively. CN1 had values 6.56~7.30% and 0.36~1.89% higher than those of FN1 and WN1, respectively, while CN2 had values 4.53~4.67% and 1.40~1.65% higher than those of FN2 and WN2, respectively.

As shown in Table 8, the irrigation regime had a significant impact on WUE and NUE (p < 0.05), the N rate had a significant impact on NUE (p < 0.001), and the interaction between irrigation and N had a significant impact on WUE (p < 0.001). Research has shown that appropriate water and N management practices could improve water and N fertilizer utilization efficiency. FI results in the wastage of water and fertilizer resources.

3.4. LAI at Heading–Flowering and the Proportion of Productive Tillers

According to

Table 9. Response of LAI at heading–flowering and proportion of productive tillers to water and N practices and analysis of variance.

	Treatment	LAI at Heading–Flowering		Proportion of Productive Tillers (%)
		2021	2022	2021	2022
	FN	4.21 c	4.19 b	62.96 e	62.67 e
	FN1	4.32 ab	4.31 a	79.13 cd	75 d
	FN2	3.86 e	3.83 d	78.77 d	74.25 d
	CN	4.24 c	4.21 b	83.84 abc	80.14 bc
	CN1	4.35 a	4.33 a	86.43 ab	81.25 ab
	CN2	3.98 d	3.95 c	87.95 a	85.11 a
	WN	4.25 bc	4.22 b	83.06b cd	79.91 bc
	WN1	4.33 a	4.32 a	83.84 abc	80.57 bc
	WN2	3.91 de	3.88 cd	78.82 d	76.2 cd
Source of variation
I	F value	8.391		78.608
	p value	0.001		<0.001
N	F value	508.68		19.047
	p value	<0.001		<0.001
I × N	F value	2.943		17.176
	p value	0.03		<0.001

, under the same irrigation regime, the LAI at the heading–flowering stage was highest at the 99 kg/ha N rate and was significantly greater than those at the 110 and 88 kg/ha N rates (p < 0.05). Compared with those at the 110 and 88 kg/ha N rates, the LAI at the 99 kg/ha N rate increased by 2.61~12.53% under FI, 2.59~9.62% under CI, and 1.80~11.34% under WI. In addition, the LAI was the lowest under FI at the same N rate. CN and WN had 0.48~0.71% and 0.72~0.95% higher values, respectively, than FN. CN1 and WN1 had 0.46~0.69% and 0.23% higher values, respectively, than FN1. CN2 and WN2 had 3.11~3.13% and 1.30~1.31% higher values, respectively, than FN2.

The proportions of the productive tillers of FN1 were 0.46~25.68% (2021) and 1.01~19.67% (2022) greater than those of FN and FN2, respectively (Table 9). Similarly, the proportions of productive tillers in WN1 were 0.94~6.37% (2021) and 0.83~5.73% (2022) greater than those in WN and WN2, respectively. Under CI, the proportion of productive tillers increased with decreasing N rate. The proportions in CN2 were 1.76~4.90% (2021) and 4.75~6.20% (2022) greater than those in CN and CN1, respectively. However, the proportion of productive tillers was the lowest under FI at the same N rate. CN and WN had values 27.89~33.16% and 27.51~31.93% higher than that of FN, respectively; CN1 and WN1 had values 8.33~9.22% and 5.95~7.43% higher than that of FN1, respectively; and CN2 and WN2 had values 11.64~14.62% and 0.06~2.62% higher than that of FN2, respectively.

As shown in Table 9, the irrigation regime, N rate, and their interaction had a significant impact on the LAI at heading–flowering and the proportion of productive tillers (p < 0.05). The results indicated that reducing the N rate increased the LAI at the heading–flowering stage, and FI was not conducive to increasing the LAI at the heading–flowering stage. In addition, both FI and a higher N rate could reduce the proportion of productive tillers.

3.5. Sugar–Spikelet Ratio and Grain–Leaf Ratio at Heading–Flowering

According to

Table 10. Response of sugar–spikelet ratio and grain–leaf ratio at heading–flowering to water and N practices and analysis of variance.

	Treatment	Sugar–Spikelet Ratio at Heading–Flowering (mg/Spikelet)		Grain–Leaf Ratio at Heading–Flowering (Grain/cm2)
		2021	2022	2021	2022
	FN	5.03 e	4.89 d	0.93 c	0.94 c
	FN1	5.44 ab	5.32 b	1.03 ab	1.05 ab
	FN2	5.13 de	4.94 d	0.93 c	0.95 c
	CN	5.31 bc	5.33 b	1.01 b	1.03 b
	CN1	5.52 a	5.49 a	1.04 a	1.06 a
	CN2	5.27 c	5.15 c	0.95 c	0.96 c
	WN	5.34 bc	5.3 b	1.02 ab	1.04 ab
	WN1	5.48 a	5.47 a	1.03 ab	1.05 ab
	WN2	5.23 cd	4.98 d	0.94 c	0.95 c
Source of variation
I	F value	25.038		32.013
	p value	<0.001		<0.001
N	F value	56.986		181.912
	p value	<0.001		<0.001
I × N	F value	4.539		20.549
	p value	0.004		<0.001

, during the two-year study, under the same irrigation regime, the sugar–spikelet ratio at heading–flowering was the highest at the 99 kg/ha N rate and was significantly higher than those at the 110 and 88 kg/ha N rates (p < 0.05). Compared with those at 110 and 88 kg/ha N, the sugar–spikelet ratio at 99 kg/ha was 3.00~6.60% higher under CI, 6.04~8.79% higher under FI and 2.62~9.84% higher under WI. In addition, the sugar–spikelet ratio was the lowest under FI at the same N rate. The values of CN and WN were significantly higher (p < 0.05) than those of FN (5.57~9.00% and 6.16~8.38% higher, respectively). CN1 and WN1 had values that were 1.47~3.20% and 0.74~2.82% higher than that of FN1, and CN2 and WN2 had values that were 2.73~4.25% and 0.81~1.95% higher, respectively, than that of FN2.

Under the same irrigation regime, the grain–leaf ratio at the heading–flowering stage was greatest at the 99 kg/ha N rate. Compared with those at the 110 and 88 kg/ha N rates, the grain–leaf ratio at the heading–flowering stage at the 99 kg/ha N rate significantly increased by 2.91~10.42% under CI and by 10.53~11.70% under FI (p < 0.05). Compared to those of WN and WN2, the value of WN1 was 0.96~10.53% higher. In addition, the grain–leaf ratio at the heading–flowering stage was the lowest under FI at the same N rate. CN and WN had values that were 8.60~9.57% and 9.68~10.64% greater, respectively, than that of FN (p < 0.05); CN1 had values that were 0.95~0.97% greater than that of FN1, while WN1 had a value comparable to that of FN1; and CN2 and WN2 had values that were 1.05~2.15% and 0~1.08% higher, respectively, than that of FN2.

As shown in Table 10, the irrigation regime, N rate, and their interaction had a significant impact on the sugar–spikelet ratio and grain–leaf ratio at heading–flowering (p < 0.05). The present study showed that a moderate reduction in the N rate effectively improved the sugar–spikelet ratio and grain–leaf ratio at heading–flowering, which is beneficial for enhancing the transport of assimilates from stem sheaths to grains during the heading–flowering to maturity stages, promoting rice grain filling. FI was not conducive to improving the sugar–spikelet ratio at heading–flowering or the grain–leaf ratio, which led to a decrease in grain-filling capacity.

3.6. Transfer Rate of Dry Matter and NSCs

According to

Table 11. Response of transfer rates of dry matter and NSCs to water and N practices and analysis of variance.

	Treatment	Transfer Rates of Dry Matter (%)		Transfer Rates of NSCs (%)
		2021	2022	2021	2022
	FN	13.85 d	12.25 e	36.4 f	36.4 d
	FN1	16.02 ab	14.89 ab	40.8 abc	40.7 ab
	FN2	14.53 cd	13.18 d	36.5 f	36.9 cd
	CN	15.02 c	13.73 cd	38.8 cde	37.6 cd
	CN1	16.58 a	15.48 a	42.5 a	41.7 a
	CN2	14.89 c	13.59 cd	37.9 def	37.7 cd
	WN	15.5 bc	14.29 bc	39.7 bcd	38.9 bc
	WN1	16.27 ab	15.18 ab	41.6 ab	40.6 ab
	WN2	14.71 cd	13.39 cd	37.1 ef	37.9 cd
Source of variation
I	F value	25.038		32.013
	p value	<0.001		<0.001
N	F value	56.986		181.912
	p value	<0.001		<0.001
I × N	F value	4.539		20.549
	p value	0.004		<0.001

, under the same irrigation regime, the transfer rate of dry matter was the highest at the 99 kg/ha N rate. Compared with those at the 110 and 88 kg/ha N rates, the transfer rate of dry matter at the 99 kg/ha N rate significantly increased by 10.25~21.55%, 10.39~13.91%, and 4.97%~13.37% under FI, CI, and WI, respectively. This study also showed that the transfer rate of dry matter was the lowest under FI. The transfer rates of dry matter of CI and WI were 8.45~12.08% and 11.91~16.65% higher than that of FI at the 110 kg/ha N rate, 3.50~3.96% and 1.56~1.95% higher at the 99 kg/ha N rate, and 2.48~3.11% and 1.23~1.59% higher at the 88 kg/ha N rate, respectively.

Under the same irrigation regime, the transfer rate of NSCs was the highest at the 99 kg/ha N rate. Compared with those at the 110 and 88 kg/ha N rates, the transfer rates of NSCs at the 99 kg/ha N rate significantly increased by 9.54~12.14% under CI and by 10.30~12.09% under FI. Compared to those of WN and WN2, the values of WN1 were 4.37~12.13% higher. During the two-year study, at the 110 kg/ha N rate, the transfer rates of NSCs of WN were 6.87~9.07% and 2.32~3.46% higher than those of FN and CN, respectively. At the 99 kg/ha N rate, the transfer rates of NSCs of CN1 were 2.46~4.17% and 2.16~2.71% higher than those of CN and CN2, respectively. At 88 kg/ha N, the transfer rate of NSCs was the lowest under FI. The transfer rates of the NSCs of CN2 and WN2 were 2.17~3.84% and 1.64~2.71% higher, respectively, than those of FN2.

As shown in Table 11, the transfer rates of dry matter and NSCs were significantly impacted by the irrigation regime (p < 0.05) and N rate (p < 0.001), but the interaction of irrigation and N had no significant effect on the transfer rate of dry matter or NSCs (p > 0.05). The results indicated that the 99 kg/ha N rate had the greatest effect on promoting the transfer of assimilates and promoted grain filling. In contrast, FI had a negative impact on the transfer of assimilates.

4. Discussion

Improving rice productivity in the Mollisols of Northeast China is crucial for ensuring food security, as this region is one of the main grain-producing regions in the country. As an important indicator for evaluating crop productivity, the rice harvest index is considered to improve resource utilization efficiency while ensuring yield [48,49]. The results of the present study supported that there was a highly significant positive correlation between the HI, WUE, NUE, and yield (p < 0.01) (Figure 4). In general, it is believed that HI is regulated by the environment, variety and field management practices of the plant growth [50,51]. However, there were few reports on the regulatory mechanisms of water and N practices on HI. The results showed that water-saving irrigation and moderate N reduction significantly improved rice HI. Previous studies have shown that the irrigation regime affects the transport and residue of N fertilizer in the soil, and the regulation of N fertilizer on rice growth benefits from the most suitable root zone soil moisture [37,52]. Therefore, it is feasible to improve rice productivity while saving water and fertilizer resources through water and N management practices [53,54,55]. Our study revealed the response mechanism of HI to water and N management practices, which improved the source–sink relationship of rice and promoted the transportation of photoassimilates.

Water-saving irrigation and N reduction have increased the proportion of productive tillers and the LAI during the heading–flowering period, optimizing the canopy structure and population quality of rice, improving environmental conditions such as light, temperature, and CO₂ concentration within the canopy, and optimizing the distribution of crop nutrients [56,57]. In the present study, the proportion of productive tillers significantly increased under the water-saving irrigation regimes because these regimes controlled the water content and quickly eliminated non-productive tillers during the late tillering stage. Studies have shown that reducing nutrient competition between non-productive and productive tillers is beneficial for improving the rice HI [58]. Research on the LAI in different regions and varieties has shown that it is influenced by factors such as genes [59], environment [60], and field management measures [61], among which N input is one of the most important regulatory measures [62]. Moreover, a study based on comparing different varieties of rice showed a strong positive correlation between LAI and HI [63]. A higher LAI is beneficial for nutrient absorption and promotes rice production. However, there was a study also suggested that an excessive LAI may lead to an excessive accumulation of aboveground biomass, resulting in an imbalance in the source–sink relationship. Therefore, in the research, the maximum point of the quadratic curve between LAI and HI is LAI equal to 8.0 [64]. In this study, the LAI might not reach 8.0 due to planting density and variety, and there was a positive correlation between LAI and HI. The higher LAI in this study benefited from appropriate water and N management, which may be due to the threshold tolerance of rice to ammonium [65]. The higher N input was applied in the form of urea, resulting in excessive ammonium concentration in local soil and toxic effects on rice [66], mainly manifested as the inhibition of rice root development and leaf growth [67,68]. On the other hand, studies have shown that rice growth in low potassium soils is more susceptible to the negative effects of high ammonium [69]. In the same depth soil layer, the potassium ion content of flooded irrigation is often lower than that of water-saving irrigation. This explains that under 110 kg/ha N rate, the LAI under flooded irrigation was not as good as that under water-saving irrigation. Moreover, the growth of spikelet is closely related to the dry matter production and nutrient accumulation during its differentiation period [70]. Due to the negative impact of high ammonium on rice under the 110 kg/ha N rate or flooded irrigation, this may hinder root nutrient absorption and leaf photosynthesis, while under a 88 kg/ha N rate, it may lead to insufficient nutrient supply. Both of these are not conducive to the transfer of nutrients to the reproductive organs, resulting in a decrease in the growth of spikelets. The grain–leaf ratio is also an important indicator reflecting the source–sink relationship, often used to characterize the quality of the source and the transport ability of sink to source. Studies have shown that increasing the grain–leaf ratio within an appropriate LAI range is beneficial for achieving high yield [15]. In this study, it was observed that the grain–leaf ratio was highest at the 99 kg/ha N rate under FI, CI and WI. This may be due to the fact that under the same experimental variety, exceeding a certain N application level will to some extent inhibit the transport from source to sink, leading to a phenomenon of “luxurious absorption” and a decrease in grain–leaf ratio. The results indicated that appropriate water and N management could accelerate the growth rate of spikelets per unit of land area compared to the growth rate of leaf area [71]. Therefore, appropriate water and N regulation is the foundation for high-yield cultivation and an important strategy guiding the establishment of ideal rice populations.

The transfer rates of photoassimilates are crucial for the distribution of nutrients in rice after flowering [72,73]. Dry matter accumulation and translocation are prerequisites for crop organ differentiation and yield [74,75]. Dry matter accumulation was the highest at the 99 kg/ha N rate in the study, there was no significant difference between the FN1, CN1 and WN1 treatments, indicating that there was a certain threshold for dry matter accumulation in response to water and N management. In our study, a high capacity for dry matter translocation enhances the rice HI (Figure 5), the reason for obtaining this result is that the increasing dry matter translocation rates are beneficial for the transport of nutrients from stems and leaves to panicles [76]. The study also showed that implementing water-saving irrigation and reducing the N rate by 10% significantly improved the transfer of dry matter. Several scholars have come to similar conclusions, suggesting that appropriate water and N management can enhance the transfer capacity of dry matter in the later stages of crop growth [77,78], thereby improving HI and yield of rice. This may be because more N input can shorten the nutritional growth period of rice, advance the reproductive growth period, and thus reduce the transfer of dry matter to grains [79]. Some studies also suggest that inappropriate N application rates or methods are not conducive to building an excellent rice population quality [78], and exacerbate nutrient competition within plants, thereby reducing the transfer of nutrients to grains in the later stages of growth [80]. Moreover, water-saving irrigation can improve soil permeability and enhance root vitality in the later stage compared to flooded irrigation, which is beneficial for the synergistic absorption of water and nitrogen by rice roots, thereby promoting the transportation of aboveground substances in rice during the late growth stage [81]. The yield of rice grains originates from the photosynthesis of leaves, and grain filling is a biochemical process involving carbohydrate metabolism; NSCs are important products of photosynthesis, and their distribution from source to sink is a decisive factor in grain filling. Studies have shown that during the grain filling, the NSCs stored in vegetative tissues are remobilized and subsequently transferred to the grain, and the duration and speed of the process determine the weight of the grain [82], which is closely related to HI [83]. In this study, it was observed that water-saving irrigation and an appropriate reduction in N could improve the NSC transfer rate. It may be due to water-saving irrigation, which improves the utilization rate of NSC after heading–flowering and promotes grain filling by regulating the activity of key enzymes involved in carbon metabolism [84]. Moreover, an appropriate reduction in N fertilizer and an appropriate irrigation regime could induce premature senescence during the filling period, shorten the filling period, and promote the remobilization of NSCs from vegetative tissues to grains, which is beneficial for improving the rice HI [28,85]. The HI, thousand-grain weight, and seed setting rate of rice were significantly improved under water-saving irrigation and a 10% N reduction. This indicates that appropriate water and N management in Mollisols was beneficial for promoting the filling of weak spikelets and improving rice HI [86,87]. The sugar–spikelet ratio during the heading stage is closely related to the physiological activity of grain sinks (hormone content, enzyme activity, etc.) [88]. An increase in the sugar–spikelet ratio during the heading stage promotes the transport of NSCs stored in rice vegetative tissues to grains [89], and it was found that the correlation coefficient between the sugar–flower ratio and HI was the highest in our study (Figure 5). However, there are currently few reports on the response of this physiological mechanism to water and N. In this study, the sugar–flower ratio was significantly improved under appropriate water and N management practices. This may be attributed to the use of appropriate water and N management practices that promote the accumulation of pre-flowering photoassimilates while improving the N utilization efficiency [90]. Ren et al. [91] have found, in their research on the sugar–flower ratio of different rice varieties, that N efficient varieties have a higher activity of starch synthesis-related enzymes and a stronger sink activity during the heading stage, often resulting in higher sugar–flower ratios. This study demonstrated the feasibility of regulating the remobilization ability of NSCs through water and N management, through sinks in vegetative tissues to increase HI.

In conclusion, the results indicated that water and N management practices regulated the rice population quality and the grain-filling process during rice growth, thereby improving the rice HI. These results could provide a theoretical basis for further improving the production potential of rice while conserving water and fertilizer resources.

5. Conclusions

The challenge in improving resource utilization efficiency while ensuring yield is to increase the harvest index. Our study aimed to reveal the response mechanism of the rice harvest index in relation to water and N management practices. The results of the 2-year field experiment showed that the HI of rice was the highest under the controlled irrigation regime at a 99 kg/ha N rate, reaching values of 0.622 (2021) and 0.621 (2022). The study revealed the mechanism to improve HI, as such practice could increase the proportion of productive tillers and the LAI during the heading–flowering stage, achieving the goals of reducing redundant vegetative growth and optimizing population quality. Moreover, with the water and N management practices, the sugar–spikelet ratio and grain–leaf ratio improved during the heading–flowering stage, which was conducive to promoting the transport of photosynthetic assimilates in the later growth stage, thereby improving the HI of rice. In this study, there was a highly significant positive correlation between HI and the thousand-grain weight and seed setting rate of rice, making increasing the HI an important strategy for further improving rice grain yield.

Therefore, regulating the harvest index of rice through irrigation and N management practices may be an important strategy to achieve high yields and efficient utilization of agricultural production. In Northeast China, controlled irrigation can effectively improve rice harvest index compared to flooded irrigation. Meanwhile, we suggest reducing the N rate by 10% on the basis of the existing conventional N rate in sustainable agricultural practices. Our research provides data support and reference significance for relevant research in other regions, especially semi-arid areas.