Effects of Different Irrigation and Drainage Modes on Lodging Resistance of Super Rice Japonica 9108

Abstract

In order to determine the optimal irrigation and drainage mode for the anti-lodging cultivation of Super Rice Japonica 9108, barrel loading tests of different irrigation and drainage modes were carried out in the Water-Saving Park of Hohai University in Nanjing from June to October in 2019 and 2020. Three treatments were set up: Frequent and Shallow Irrigation (FSI), Rain-catching and Controlled Irrigation (RC-CI) and Drought Planting with Straw mulching (DPS). In each mode, the growth index, stem morphology index, material production index and stem mechanical index of rice at yellow maturity period were measured, and their relationship with the lodging index was analyzed. The results showed that compared with FSI, the lodging index of RC-CI was reduced by an average of 24.0%. RC-CI can promote the lateral development of the base internodes, increase the accumulation of stem sheath dry matter and the internode fullness and enhance mechanical strength and anti-lodging ability of the stem. Meanwhile, RC-CI can appropriately reduce the plant height, so as to reduce the bending moment of the base internodes. As a consequence, the lodging risk was effectively reduced. The lodging index of DPS was reduced by an average of 16.0% compared with FSI. Because DPS was subject to severe water deficit, its internodes thickness and stem wall thickness were inferior to that of FSI, leading to the weakening of the mechanical strength of stem, and the morphological characteristics and mechanical characteristics of the stem were not improved. Despite this, DPS still had a strong resistance to lodging. The output rate and conversion rate of the stem sheath were reduced, and while the plant height and center of gravity height were significantly reduced, the bending moment remained low. Thus, DPS can still reduce the lodging risk of rice. Compared with FSI, the average yield of RC-CI increased by 5.8% in two years, and the average yield of DPS was reduced by 4.4% in two years. DPS under severe water deficit reduced the accumulation of dry matter in the panicle and the yield index of rice, which was not conducive to a high yield. Considering the yield and lodging resistance of the super rice “Japonica 9108”, RC-CI is the best irrigation and drainage mode, followed by DPS. This study can provide data support and theoretical support for regulating the lodging resistance of super rice through irrigation measures.

1. Introduction

Rice, one of the most important food crops in the world, is the staple food for more than half of the world’s population [1]. The International Rice Research Institute predicts that rice production needs to increase by 30% to meet the growing food demand of mankind by 2050 [2]. Lodging, the main factor affecting rice yield [3,4,5], will damage the population structure and increase the difficulty of mechanical harvesting. In addition, overlapping leaves caused by lodging reduces photosynthetic capacity and inhibits the accumulation of dry matter. Furthermore, it will lead to the rupture of stem vascular bundles and hinder the transportation of carbon and nitrogen organics to grains. Yield decreases by 1% for every 2% increase in the lodging rate [6,7,8,9]. Previous studies have established that apart from factors such as rice varieties, cultivation methods and fertilizer management [7,10,11,12,13,14], water management is a key controllable factor that significantly affects rice yield and stem lodging. Rice needs abundant irrigation water, which accounts for more than 65% of Chinese agricultural water consumption. Traditional flooding irrigation not only aggravates agricultural water stress and limits the potential for high rice yield, but it also increases the length of internodes, reduces the mechanical strength of stems and raises the risk of lodging due to continuous flooding [15].

The output of crop straw in China is nearly 1.04 billion tons, but the straw utilization rate is only 33% [16]. Under the dual-carbon target, the Chinese government plans to increase the comprehensive utilization rate of straw to over 86%. Straw is rich in cellulose, hemicellulose and lignin [17], which can be converted into organic matter under the action of microorganisms and release nitrogen, phosphorus and potassium, so it is of great significance for promoting rice plant growth and improving yield [17,18,19,20]. However, the straws returned to the field will decompose under the flooded anaerobic environment to reduce harmful substances, such as H₂S, F²⁺, Mn²⁺, etc., which will make the roots blackened and inhibit rice growth [21]. Considering that straw returning to the field can effectively reduce the evaporation between trees by improving the surface coverage and create favorable conditions for dry farming [22,23,24], scholars have proposed Drought Planting with Straw mulching (DPS) mode. The research on lodging resistance of straw dry farming mainly focuses on dry crops such as wheat [25,26] and corn [27], but few on rice. Soil water, fertilizer, gas, heat, etc., are affected after the straw returning [28,29]. Especially after the rice field adopts the dry farming method, the field changes from continuous flooding to a long-term anhydrous layer, and the soil permeability and redox are changed. With the continuous increase in straw returning area, exploring the effect of DPS on rice lodging resistance and yield is conducive to the healthy development of rice cultivation agriculture.

Rice is a semi-aquatic plant and adapts to drought and flooding stress to a certain extent. Combined with the characteristics of frequent rainstorms during flood season, scholars have proposed the Rain-catching and Controlled Irrigation (RC-CI) mode, which can reduce the lower limit of irrigation while increasing the upper limit of rain storage and therefore maximize the use of rainfall resources and reduce irrigation quotas and nitrogen and phosphorus emissions [30]. RC-CI can significantly increase stem thickness, stem wall thickness, stem fullness and transverse rupture strength compared with traditional flooding irrigation, so the rice can obtain higher lodging resistance [31]. Alternating stress of drought and flood during the elongation period can cut down the fresh quality of aboveground parts, enhance the stem fullness and effectively reduce the lodging of rice stems [9]. In other words, RC-CI can keep great lodging resistance on the basis of maintaining high yield.

In this experiment, super rice Japonica 9108 was used as the trial material. We integrated and compared RC-CI and DPS, with FSI as control. The changes in rice growth index, stem morphology index, material production index and stem mechanical index under different treatments were determined; the relationship between them and the lodging index was analyzed; and the response law of the rice lodging resistance trait to different irrigation and drainage modes was elucidated to determine the suitable irrigation and drainage mode for the lodging resistance cultivation of super rice.

2. Materials and Methods

2.1. Description of the Trial Area

The trial was conducted in June–October 2019 and 2020 in the Water-Saving Park of Hohai University (31°57′ N, 118°50′ E), Nanjing, China. The trial area belongs to the north subtropical monsoon climate area, with an average annual precipitation of 1073 mm. During the 2019 growth period, the highest daily relative humidity was 100%, while the lowest was 56%; the highest temperature was 33.9 °C, while the lowest was 17.2 °C. During the 2020 growth period, the highest daily relative humidity was 100%, while the lowest was 54%; the highest temperature was 32.7 °C, while the lowest was 14.4 °C. The precipitation during the rice growing period in 2019 and 2020 was 270 mm (P = 95%, which is an extraordinary dry year; P means the percentage of the number of years that a certain precipitation is guaranteed in the total number of years calculated and is used to choose typical years for irrigation project design) and 475 mm (P = 65%, which is a typical dry year), respectively.

The soil is clay loam taken from the ploughing layer of the paddy field in the Water-Saving Park, with a pH value of 7.2, dry bulk density of 1.31 g cm⁻³ and field water holding capacity and saturated water content of 25.3% and 38.7%, respectively (weight water content). The total nitrogen, total phosphorus, available nitrogen, available phosphorus and available potassium of the soil were 62.9 mg kg⁻¹, 33.0 mg kg⁻¹, 47.5 mg kg⁻¹, 10.5 mg kg⁻¹ and 90.0 mg kg⁻¹, respectively.

2.2. Trial Design

Adopting the barrel planting trial, the length, width and height of the square barrel were measured as 40 × 40 × 100 cm. The bottom of the barrel was laid with a sand and gravel anti-filter layer. After the soil was air-dried, crushed and sieved over 5 mm, it was packed into the barrel in layers at a depth of 60 cm, the soil surface was 20 cm from the top for water storage.

The trialed rice variety was the super rice “Japonica 9108”. The seedlings were raised on June 3 in 2019 and on May 25 in 2020. Division of growth period of rice is shown in

Table 1. Division of growth period of rice.

Year	Rejuvenation	Early Tillering	Late Tillering	Elongate	Heading	Milk Maturity	Yellow Maturity	Total Days
2019	6.24~7.2	7.3~7.23	7.24~7.31	8.1~8.21	8.22~9.1	9.2~9.25	9.26~10.16	115 d
2020	6.30~7.6	7.7~7.31	8.1~8.11	8.12~8.28	8.29~9.23	9.24~10.15	10.16~10.26	120 d

Transplanted rice is used in this experiment, so the seedling raising period is not included in the whole growth period of rice, and the growth period is recorded from the rejuvenation period.

. There are 4 holes per barrel, 3 seedlings per hole and the row spacing is 20 cm. The fertilizers are urea (containing 46.7% N), superphosphate (containing 6.3% P) and potassium chloride (containing 52.4% K). Total pure nitrogen application is 255 kg hm⁻² (N:P:K = 1:0.09:0.25). Nitrogen fertilizer was applied 3 times according to the ratio of base fertilizer: tiller fertilizer: panicle fertilizer = 4:3:3; phosphorus fertilizer was applied as base fertilizer at one time; and potassium fertilizer was applied 2 times according to the ratio of base fertilizer: panicle fertilizer = 7:3.

The trial set up 3 treatments of Frequent and Shallow Irrigation (FSI, as control), Rain-catching and Controlled Irrigation (RC-CI) and Drought Planting with Straw mulching (DPS), with 5 repetitions, and the irrigation and drainage standards are shown in

Table 2. Irrigation and drainage standards for different irrigation and drainage modes.

Treatment	Control Index	Rejuvenation	Tillering		Elongate	Heading	Milk Maturity	Yellow Maturity
			Early	Late
FSI	Lower limit of irrigation	10 1	10	60%θ 2	10	10	10	Natural drying
	Upper limit of irrigation	30	30	0	40	40	40	0
	Upper limit of rain storage	40	100	0	150	150	150	0
RC-CI	Lower limit of irrigation	10	70%θ	60%θ	70%θ	80%θ	70%θ	Natural drying
	Upper limit of irrigation	30	0	0	0	0	0	80%θ
	Upper limit of rain storage	80	150	0	200	200	200	0
DPS	Lower limit of irrigation	80%θ	60%θ	50%θ	60%θ	60%θ	50%θ	Natural drying
	Upper limit of irrigation	0	0	0	0	0	150	80%θ
	Upper limit of rain storage	40	60	0	80	80	80	0

¹ The number unit in the table is “mm”, which means the depth of field surface water; ² θ is the saturated water content for the 20 cm soli and “%” means the percentage of soil moisture content of 20 cm to θ.

. The irrigation water was obtained from the tap water system of the lab. When the surface water depth or soil moisture reaches the lower limit of irrigation, rice is irrigated until the water level reaches the upper limit of irrigation. When the surface water depth exceeds the upper limit of precipitation storage, it is drained to the upper limit of rain storage. Precipitation in each growth period is shown in Figure 1. With the exception of irrigation and drainage measures, all other agricultural technical measures were the same. DPS covered the soil surface with the shredded straw that had been soaked and composted for 10 days at a depth of 2 cm, and the amount of grass used was 6000 kg hm⁻² (dry matter mass).

2.3. Determination Index and Methods

2.3.1. Growth index

At the yellow maturity period, 2 holes of different barrels were randomly selected for each treatment, and 3 main stems were taken from each hole to measure the following indices:

(1)

Plant height and panicle position height: measure the height from the stem base to the top of the panicle, record as H_P, cm; measure the height from the stem base to the panicle position, record as H_E, cm;

(2)

Height of center of gravity and ratio of center of gravity: cut the rice stalk from the base, place it horizontally on the fulcrum, move it to the balance of the stem, mark the center of gravity and measure the length from it to the bottom, which is the height of the center of gravity, recorded as H_G, cm; the ratio of center of gravity is the ratio of the height of the center of gravity to the plant height, recorded as R, %;

(3)

Fresh mass of panicle and fresh mass per plant: measure the fresh mass of panicle and per plant with a balance with induction of 0.001 g, recorded as M₁ and M₂, respectively, g;

(4)

The yield indices: select one hole randomly from each barrel for a total of 5 holes for each treatment at harvest; measure effective panicles per hole and the number of solid grains and total grains per panicle, and calculate the ratio of solid grains per panicle to total grains per panicle as the seed setting rate, recorded as R_S, %; measure thousand-grain weight, recorded as W, g; measure yield per hole, recorded as Y_M, g; calculate product of effective panicles per hole, solid grains per panicle and thousand-grain weight as theoretical yield per hole, recorded as Y_T, g.

2.3.2. Stem Morphological Index

(1)

Internode length: measure the length of the I~III internode according to the node, recorded as L, cm;

(2)

Internode thickness and stem wall thickness: retain the stem and sheath, measure the outer diameter of the long axis and short axis of each internode with a vernier caliper. The average is the internode thickness, recorded as δ₁, cm. Measure the wall thickness of the middle long axis and short axis of each internode; the average is the stem wall thickness, recorded as δ₂, cm;

(3)

Internode fullness: Internode fullness = Internode dry matter mass/Internode length, recorded as F, g/cm.

2.3.3. Material Production Index

At the heading period and the yellow maturity period, select one hole randomly from each barrel for a total of 5 holes for each treatment:

(1)

Aboveground dry matter mass: The aboveground parts of the rice were divided into stem sheaths, leaves and panicles and put into an oven at 105 °C for 30 min; after drying at 80 °C to constant weight, measure the dry mass of each part, g;

(2)

Stem sheath dry matter output rate: stem sheath dry matter output rate = (stem sheath dry matter mass at heading stage—stem sheath dry matter mass at yellow maturity period)/stem sheath dry matter mass at heading period, %;

(3)

Stem sheath dry matter conversion rate: stem sheath dry matter conversion rate = (stem sheath dry matter mass at heading period—stem sheath dry matter mass at yellow maturity period)/stem sheath dry matter mass at yellow maturity period, %.

2.3.4. Stem Mechanical Index

At the yellow maturity period, 2 holes were randomly selected for each treatment, and 3 main stems were selected from each hole. The following indices were measured for the base section II and III.

(1)

Transverse rupture strength: use CMT6104 electronic universal testing machine (Measure Test Simulate Industrial Systems Co., Ltd., Shanghai, China) to determine transverse rupture strength. Fix the stem horizontally on two fulcrums with a distance of 8 cm and place the pressure strain sensor in the center of the stem, and continuously decline at a speed of 0.1 mm s⁻¹ until the stem is broken. The resulting pressure is the transverse rupture strength, recorded as F_max, N;

(2)

Bending moment: Bending moment = Length from the base of the internode to the panicle top (cm) × Fresh weight from the base of the internode to the panicle top (g) × 0.001 × 9.8, recorded as B_M, N·cm [32];

(3)

Bending moment at break: M = (F_max × D)/4. In the formula, M is the bending moment at break, N·cm; D is the distance between the two fulcrums, cm [32];

(4)

Lodging index: Lodging index = B_M/M, recorded as LI [32].

2.4. Data Analysis

Excel was used for original data analysis and chart drawing. One-way analysis of variance (ANOVA) with Duncan’s multiple range test at the 0.05 probability level was performed using the general linear model-univariate procedure with SPSS 26.0 software (SPSS, Chicago, IL, USA). In addition, the Pearson correlation coefficient was calculated using the SPSS 26.0 software, too. The ENTROPY TOPSIS model was used for the selection of the optimal irrigation and drainage mode. Matrix Laboratory (MATLAB, MathWorks, Natick, MA, USA) was used to write the m file of the ENTROPY TOPSIS method.

3. Results

3.1. Water Consumption of Rice in Whole Growth Period under Different Irrigation and Drainage Modes

The water consumption of rice in the whole growth period under different irrigation and drainage modes is shown in

Table 3. Water consumption of rice in whole growth period under different irrigation and drainage modes.

Year	Treatment	Irrigation	Precipitation	Deep Leakage	Surface Drainage	Water Storage Reduction	Evapotranspiration 3	Water Consumption 4
2019	FSI	924.8 1 ± 3.3 a 2	270	105.6 ± 2.5 a	0	53.7 ± 1.7 b	1035.5 ± 1.6 a	1141.1 ± 2.9 a
	RC-CI	829.0 ± 5.4 b	270	79.9 ± 2.8 b	0	56.9 ± 1.6 a	962.2 ± 7.9 b	1042.1 ± 10.3 b
	DPS	747.2 ± 10.0 c	270	70.6 ± 2.8 c	0	55.6 ± 1.9 ab	891.7 ± 7.4 c	961.6 ± 9.8 c
2020	FSI	572.5 ± 3.1 a	475	270.0 ± 1.6 a	31.7 ± 0.6 b	51.2 ± 0.8 c	694.6 ± 1.7 a	996.3 ± 3.6 a
	RC-CI	453.6 ± 2.9 b	475	183.0 ± 1.0 b	12.9 ± 0.7 c	56.6 ± 0.5 a	676.1 ± 1.2 b	872.1 ± 2.6 b
	DPS	356.7 ± 3.4 c	475	92.7 ± 3.0 c	62.7 ± 0.9 a	54.4 ± 2.4 b	621.8 ± 2.6 c	777.3 ± 3.2 c

¹ The number unit in the table is “mm”. ² Different letters indicate significant differences between treatments at the 0.05 level. ³ Evapotranspiration = (Irrigation + Precipitation) − (Deep leakage + Surface drainage + Water storage reduction). ⁴ Water consumption = Evapotranspiration + Deep leakage.

. In 2019, drainage was all deep leakage caused by large amount of irrigation; in 2020, drainage also included surface drainage after precipitation. Except for the water storage reduction, all the other water consumption under different irrigation and drainage modes were in the order of FSI > RC-CI > DPS, with significant differences. It might be due to the fact that the upper and lower limits of irrigation of RC-CI and DPS were lower than FSI, so their field surface could maintain a water free state, while FSI always had a water layer in growth periods except for the later tillering period. Therefore, water was easily consumed by evapotranspiration in FSI, and needed to be supplemented by irrigation. Excessive irrigation led to deep leakage, thus increasing the drainage. As the ground is covered with straw, the deep leakage and evapotranspiration of DPS are significantly reduced by 49.4% and 12.2%, respectively, compared with FSI.

3.2. Effects of Irrigation and Drainage Modes on Growth Indices of Rice

The growth indices of rice in different irrigation and drainage modes are shown in

Table 4. Growth index of rice in different irrigation and drainage modes.

Year	Treatment	HP 1/cm	HG/cm	R/%	HE/cm	M1/g	M2/g	RS/%	W/g	YM/g	YT/g
2019	FSI	84.0 ± 1.3 a 2	39.6 ± 1.1 a	47.1 ± 0.6 a	69.3 ± 0.8 a	5.3 ± 0.3 a	19.3 ± 1.1 a	84.4 ± 0.5 ab	27.6 ± 0.4 ab	51.2 ± 2.7 ab	58.8 ± 1.3 ab
	RC-CI	81.0 ± 1.4 b	38.5 ± 0.7 ab	47.0 ± 1.1 a	64.6 ± 2.5 b	5.3 ± 0.3 a	18.1 ± 1.0 ab	85.0 ± 0.3 a	28.2 ± 0.5 a	53.3 ± 4.1 a	60.4 ± 1.7 a
	DPS	79.8 ± 1.1 b	37.4 ± 0.8 b	47.9 ± 0.5 a	61.5 ± 2.1 c	4.7 ± 0.2 b	17.8 ± 1.1 b	83.8 ± 0.3 b	27.0 ± 0.3 b	46.9 ± 2.6 b	56.3 ± 0.9 b
2020	FSI	73. 3 ± 1.7 a	32.2 ± 0.6 a	43.9 ± 0.6 a	57.0 ± 1.2 a	4.50.5 a	19.7 ± 1.4 a	89.0 ± 1.3 b	25.6 ± 0.4 b	41.4 ± 0.7 b	47.4 ± 2.0 b
	RC-CI	70.3 ± 2.0 b	30.3 ± 2.0 ab	43.1 ± 1.7 ab	54.0 ± 1.2 b	4.7 ± 0.5 a	17.8 ± 2.0 ab	92.7 ± 1.6 a	26.1 ± 0.2 a	44.5 ± 1.7 a	53.3 ± 1.8 a
	DPS	69.0 ± 1.6 b	28.4 ± 1.9 b	41.1 ± 2.3 b	51.7 ± 1.3 c	3.8 ± 0.3 b	16.8 ± 2.0 b	88.5 ± 1.8 b	25.4 ± 0.3 b	41.3 ± 1.8 b	45.7 ± 1.9 b

¹ H_P means plant height; H_G means height of center of gravity; R means ratio of center of gravity; H_E means panicle position height; M₁ means fresh mass of panicle; M₂ means fresh mass per plant; W means thousand-grain weight; R_S means seed setting rate; Y_M means the measured yield per hole; Y_T means the theoretical yield per hole. ² Different letters indicate significant differences between treatments at the 0.05 level.

. As can be seen from the table, there were similar laws between the two years’ data. Compared with FSI, RC-CI significantly reduced plant height and panicle position height. Except for the yield indices, all the other growth indices of DPS were significantly lower than FSI, and the average plant height and height of center in two years decreased by 5.4% and 8.7%, respectively.

No significant difference in ratio of center of gravity among these treatments was evident in 2019, but the DPS’s was significantly smaller than the FSI’s in 2020, and a possible explanation for this might be the large difference in precipitation between the two years. It can be found that the significant decrease in the DPS’s ratio of center of gravity in 2020 was caused by the greater reduction in the height of the center of gravity. Although there was more precipitation in 2020, less precipitation during the elongation period inhibited plant elongation, caused dry matter accumulating at the base, and the stem nodes became thicker, resulting in a decrease in the ratio of the center of gravity. Overall, RC-CI and DPS can effectively reduce plant height, height of center of gravity and panicle position height under drought stress and inhibit plant elongation. Meanwhile, water deficit affected material accumulation, resulting in a decrease in the fresh quality of panicle per plant, and DPS is particularly outstanding due to excessive water shortage.

RC-CI had obvious increased yield, for which the measured yield and theoretical yield were 5.8% and 7.5% higher than FSI, respectively. With lowest surface drainage, we can find that RC-CI can make full use of precipitation resources, and the rice yield can be significantly increased in general dry years, and there was also a trend of increasing yield in extraordinarily dry years. The measured yield and theoretical yield of DPS were 4.8% and 4.0% lower, respectively, than FSI, and the water consumption was 18.6% lower than FSI. From these results, it was clear that severe water deficit was unfavorable to the stable high yield of rice and should be avoided as much as possible to reduce the yield reduction.

3.3. Effects of Irrigation and Drainage Modes on Stem Morphological Indices of Rice

The morphological indices of rice stem nodes in different irrigation and drainage modes are shown in

Table 5. Morphological index of rice stem nodes in different irrigation and drainage modes.

2019	Treatment	L 1/cm			δ1/cm
		I 2	II	III	I	II	III
	FSI	7.0 ± 0.6 a 3	9.4 ± 0.3 a	12.8 ± 0.7 a	8.3 ± 0.8 a	7.0 ± 0.2 a	6.5 ± 0.2 a
	RC-CI	5.6 ± 0.6 b	9.1 ± 0.2 a	12.3 ± 0.5 a	8.9 ± 0.6 a	7.2 ± 0.3 a	6.6 ± 0.5 a
	DPS	4.4 ± 0.5 c	8.6 ± 0.4 b	11.1 ± 0.3 b	7.4 ± 0.6 b	6.6 ± 0.3 b	6.0 ± 0.5 b
	Treatment	δ2/cm			F/(g·cm−1)
		I	II	III	I	II	III
	FSI	2.7 ± 0. 2 a	2.1 ± 0.2 a	1.8 ± 0.1 a	0.162 ± 0.022 a	0.079 ± 0.004 a	0.056 ± 0.003 a
	RC-CI	2.8 ± 0.2 a	2.2 ± 0.2 a	1.9 ± 0.2 a	0.176 ± 0.021 a	0.084 ± 0.005 a	0.060 ± 0.007 a
	DPS	2.4 ± 0.2 b	1.9 ± 0.1 b	1.6 ± 0.1 b	0.134 ± 0.016 b	0.064 ± 0.003 b	0.049 ± 0.003 b
2020	Treatment	L/cm			δ1/cm
		I	II	III	I	II	III
	FSI	6.1 ± 0.6 a	9.2 ± 0.3 a	9.5 ± 0.2 a	9.1 ± 0.2 b	7.5 ± 0.2 b	6.9 ± 0.3 b
	RC-CI	5.3 ± 0.8 b	8.4 ± 1.1 ab	9.1 ± 0.3 b	9.7 ± 0.3 a	8.0 ± 0.3 a	7.3 ± 0.2 a
	DPS	4.2 ± 0.4 c	7.9 ± 0.7 b	8.7 ± 0.3 b	8.2 ± 0.2 c	7.1 ± 0.1 c	6.5 ± 0.4 c
	Treatment	δ2/cm			F/(g·cm−1)
		I	II	III	I	II	III
	FSI	2.6 ± 0.2 b	2.2 ± 0.1 b	1.8 ± 0.0 b	0.201 ± 0.035 b	0.149 ± 0.011 b	0.095 ± 0.010 b
	RC-CI	2.9 ± 0.1 a	2.4 ± 0.1 a	1.9 ± 0.2 a	0.236 ± 0.017 a	0.171 ± 0.020 a	0.115 ± 0.004 a
	DPS	2.6 ± 0.2 b	2.1 ± 0.0 b	1.8 ± 0.1 b	0.172 ± 0.007 c	0.114 ± 0.002 c	0.066 ± 0.002 c

¹ L means internode length; δ₁ means internode thickness; δ₂ means stem wall thickness; F means internode fullness. ² I, II and III represent the first, second and third internode from the base. ³ Different letters indicate significant differences between treatments at the 0.05 level.

. The length of each internode in the two-year experiment was in the order of FSI > RC-CI > DPS, showing that water stress can inhibit the elongation of rice plants. The internode thickness, stem wall thickness and internode fullness were all in the order of RC-CI > FSI > DPS. There were no significant differences found between FSI and RC-CI in 2019, but there were in 2020. It seems possible that these results were due to the precipitation during the elongation period, which was the key period for basal internode elongation, only accounting for 5.3% of the total growth period in 2020. Furthermore, suffering from drought at this time inhibited elongation, and accumulation of dry matter was more used for the lateral growth of the stem, bringing out the thicker stem wall of RC-CI. Due to severe water deficit, DPS had obtained significantly inferior stem node morphology indices than RC-CI and FSI.

3.4. Effects of Irrigation and Drainage Modes on Material Production Indices of Rice

The quality and ratio of dry matter in the upper part of rice in different modes is shown in Figure 2. The accumulation of dry matter was mainly concentrated in the stem sheaths at heading stage and in the panicles at the yellow maturity period. This was due to the material stored in stem sheath transferring to the panicle after the panicle was filled with the grain grout, leading to a reduction in stem sheath dry matter. While the dry matter mass of stem sheaths at each growth stage of RC-CI was greater than other treatments, the dry matter mass of panicle at yellow maturity period was still significantly the greatest, indicating that alternating mild drought–flood stress could not only promote dry matter accumulation in stem sheaths and improve internode fullness but also had no damage on the transfer of stem sheath dry matter to the grain in the later stage. This demonstrated that RC-CI could create conditions for high and stable yields under the condition of ensuring lodging resistance. Under severe drought, the accumulation of dry matter in DPS mode decreased, leading to a stem shape inferior to other treatments, while reducing stem sheath dry matter output and inhibiting grain filling, finally decreasing panicle dry matter at yellow maturity, which was not conducive to a high yield. It is notable that trends of fresh weight and dry weight were inconsistent: in the yellow maturity, the fresh weight of FSI was greater than RC-CI, but the dry weight was less than RC-CI. It may be related to the higher water content of plants under the full irrigation system of FSI. The difference in water content of plants seemed to be caused by the higher limits of irrigation of FSI during the whole growth period, which made the soil water content remain higher than RC-CI after drying. Accordingly, the plant roots continued obtaining water from the soil.

3.5. Effects of Irrigation and Drainage Modes on Stem Mechanical Indices of Rice

The mechanical indices of rice stems in different irrigation and drainage modes are shown in

Table 6. Mechanical index of rice stems in different irrigation and drainage modes.

2019	Treatment	Fmax 1/N		BM/N·cm		M/N·cm		LI
		II	III	II	III	II	III	II	III
	FSI	10.89 ± 0.61 b 2	8.59 ± 0.47 b	13.53 ± 0.75 a	9.71 ± 0.48 a	21.79 ± 1.22 b	17.19 ± 0.94 b	0.62 ± 0.05 a	0.57 ± 0.03 a
	RC-CI	12.32 ± 0.14 a	9.25 ± 0.66 a	11.67 ± 0.60 b	8.49 ± 0.46 b	24.64 ± 0.29 a	18.50 ± 1.33 a	0.47 ± 0.03 c	0.46 ± 0.04 b
	DPS	10.41 ± 0.32 b	8.05 ± 0.28 b	10.96 ± 0.62 b	8.16 ± 0.86 b	20.81 ± 0.64 b	16.11 ± 0.56 b	0.53 ± 0.03 b	0.51 ± 0.05 b
2020	Treatment	Fmax/N		BM/N·cm		M/N·cm		LI
		II	III	II	III	II	III	II	III
	FSI	11.47 ± 0.41 b	8.88 ± 0.36 b	12.95 ± 0.40 a	8.97 ± 0.81 a	22.94 ± 0.82 b	17.77 ± 0.72 b	0.57 ± 0.02 a	0.51 ± 0.04 a
	RC-CI	12.98 ± 0.15 a	9.67 ± 0.26 a	10.26 ± 0.77 b	7.468 ± 0.56 b	25.96 ± 0.30 a	19.35 ± 0.52 a	0.40 ± 0.02 c	0.39 ± 0.03 b
	DPS	11.09 ± 0.99 b	8.48 ± 0.42 b	9.86 ± 0.64 b	7.116 ± 0.80 b	22.19 ± 1.98 b	16.97 ± 0.83 b	0.47 ± 0.03 b	0.42 ± 0.05 b

¹ F_max means transverse rupture strength; B_M means bending moment; M means bending moment at break; LI means lodging index. ² Different letters indicate significant differences between treatments at the 0.05 level.

. Overall, RC-CI had superior lodging resistance than FSI. Rice plants in RC-CI tend to grow low, and the improvement in stem shape such as stem thickness and wall thickness increased the transverse rupture strength, reduced the bending moment and the lodging index and finally made rice obtain strong lodging resistance. Although the dry matter accumulation of DPS is not as good as other treatments, due to the plant height and the fresh weight of the aboveground part being reduced at the same time, the bending moment turned out the lowest and the lodging index was reduced, which can still effectively reduce the lodging risk. The lodging index of the second node of RC-CI was significantly lower than that of DPS, and that of the third node had no significant difference with DPS. The reason for this may be that alternating mild drought–flood stress can promote the accumulation of dry matter in the internodes. Precipitation in this trial mainly occurred in the early growth period, and at this time, alternating mild drought–flood stress had great impact on the basal internodes. Consequently, the rice’s stem node morphology was shorter, thicker and firmer, and the mechanical strength and lodging resistance was better.

3.6. Correlation Analysis between Lodging Index and Various Indices of Rice

The correlation between the rice lodging index and various indices of the stem is shown in

Table 7. Correlation between rice lodging index and stem indices.

Relevant Indices of Stems	Project	Lodging Index
		II	III
Growth index	Plant height (HP)	0.568 ** 1	0.636 **
	Panicle position height (HE)	0.504 **	0.624 **
	Height of center of gravity (HG)	0.402 *	0.575 **
	Ratio of center of gravity (R)	0.604 **	0.650 **
	Fresh mass of panicle (M1)	0.272	0.533 **
	Fresh mass per plant (M2)	0.478 **	0.345 *
Stem morphological indices	Internode length (L)	0.515 **	0.538 **
	Internode thickness (δ1)	−0.395 *	−0.347 *
	Stem wall thickness (δ2)	−0.302	−0.160
	Internode fullness (F)	−0.417 *	−0.497 **
Material production indices	Total dry matter mass in yellow maturity period	−0.354 *	−0.256
	Stem dry matter mass in yellow maturity period	−0.483 **	−0.484 **
	Stem sheath dry matter output rate	0.479 **	0.456 **
	Stem sheath dry matter conversion rate	0. 481 **	0.455 **
Stem mechanical indices	Transverse rupture strength (Fmax)	−0.630 **	−0.494 *
	Bending moment (BM)	0.859 **	0.869 **
	Bending moment at break (M)	−0.630 **	−0.494 *

¹ The numbers in the table indicate the correlation coefficient between the lodging index and each index, * indicates a significant correlation at the 0.05 level and ** indicates an extremely significant correlation at the 0.01 level.

. Among the growth indices, the lodging index of the second and third nodes was extremely significantly positively correlated with plant height, the height of the center of gravity, panicle position height and fresh mass per plant; the second node’s lodging index was significantly positively correlated with ratio of center of gravity, and positively correlated with fresh mass of panicle but not significant; the third node’s lodging index was extremely significantly positively correlated with the ratio of the center of gravity and significantly positively correlated with the fresh mass of the panicle.

Among the stem morphological indices, the lodging index of the second and third nodes was extremely significantly positively correlated with internode length, extremely significantly negatively correlated with internode fullness, significantly negatively correlated with internode thickness and negatively correlated with stem wall thickness, but not significantly.

Among the material production indices, the lodging index of the second and third nodes was extremely significantly negatively correlated with stem dry matter mass in yellow maturity period and extremely significantly positively correlated with the output rate and conversion rate of the stem sheath; the lodging index of the second nodes was significantly negatively correlated with the total dry matter mass, and the lodging index of the third nodes was negatively correlated, but not significant.

Among the stem mechanical indices, the lodging index of the second node was extremely significantly negatively correlated with transverse rupture strength and bending moment at break and extremely significantly positively correlated with the bending moment; the lodging index of the third node was significantly negatively correlated with transverse rupture strength and bending moment at break and extremely significantly positively correlated with the bending moment.

In conclusion, the lodging index has the highest correlation with the bending moment.

4. Discussion

4.1. Effects of Stem Indices on Lodging Resistance of Rice

The lodging index is the main index to evaluate the lodging resistance of rice. The lower the lodging index, the stronger the lodging resistance [33,34]. Extensive research has shown that there is a certain correlation between the lodging index and stem traits such as plant height, height of the center of gravity, panicle height, internode length, internode thickness and stem wall thickness [35,36,37,38]. The lodging index has positive correlation with plant height and the height of the center of gravity, and as plant height increases and center of gravity shifts upwards, the lodging index increases and the lodging resistance of stems decreases [35,36,37]. The same conclusion was reached in this trial. Therefore, the morphology of rice stems and their lodging resistance could be enhanced by appropriately reducing plant height, the height of the center of gravity and the panicle height. The main reasons for the stronger lodging resistance of super rice were the shorter internode, the thicker stem wall and the higher internode fullness [36]. The short and thick basal internodes and the thick and full stem wall are conducive to enhancing the lodging resistance of rice [38]. In addition, improving internode fullness is more important for rice lodging resistance [34]. This trial showed that the lodging index was extremely significantly positively correlated with internode length, and negatively correlated with internode thickness, stem wall thickness and internode fullness. The shortening of internode length and the increasing internode thickness and stem wall thickness makes the stem develop laterally and enhances the internode fullness, which is the key to improve the lodging resistance of rice.

Another important factor affecting rice lodging is material production. The aboveground material accumulates more, the biomass at the base stem is larger and the internode fullness is higher, thereby enhancing the lodging resistance [39]. This trial found that the lodging index was negatively correlated with the total dry matter mass at yellow maturity period, but the third node did not reach a significant level. Under certain assumptions, it could be construed that the internode fullness, which was determined by dry matter mass and internode length of the internode, was significantly negatively correlated with the lodging index. The partial correlation analysis of the lodging index and the total dry matter mass was performed by controlling the internode length, finding a negative but not significant correlation between the two (r = −0.330). We came to the conclusion that when the stem nodes are short enough, the relationship between stem node morphology and lodging resistance is more closely relative to dry matter accumulation, which has little effect on lodging resistance [35]. The accumulation of stem dry matter was extremely significantly negatively correlated with the lodging index, indicating that the increase in stem dry matter accumulation could promote the thickening of stem, increase the mechanical strength and reduce the lodging index, so as to ensure the lodging resistance of rice [31]. This trial had shown that the greater the stem material output and transformation of the stem, the less dry matter stored in the stem and the worse the lodging resistance of rice was. Overall, this finding was in accordance with the finding reported by Yang et al. [40] that “after full panicle, as the grain filling, the stem material is transferred to the panicle. Gradually, the dry matter and transverse rupture strength of stem is reduced, and the lodging index is improved”.

The lodging index is directly related to the mechanical properties of the stem. This trial found that the lodging index was significantly negatively correlated with the transverse rupture strength and bending moment at break and was extremely significantly positively correlated with the bending moment, which was consistent with the result Wei et al. [41] found. Some scholars believe that the contribution of transverse rupture strength to the reduction in lodging index is the greatest, and the greater the transverse rupture strength, the stronger the lodging resistance of stems [42]. In this trial, partial correlation analysis was carried out by controlling transverse rupture strength and bending moment. Consequently, there was a positive correlation between the lodging index and bending moment (r = 0.990) and a negative correlation between the lodging index and transverse rupture strength (r = −0.982), and the former was more relevant. It suggested that the bending moment had the highest contribution to the lodging index. The bending moment had a very significant positive correlation with internode length (r = 0.639) and fresh mass per plant (r = 0.610). As the two increased, the bending moment increased and the lodging index increased, resulting in a weakening of the stem’s lodging resistance [38].

4.2. Effects of Different Irrigation and Drainage Modes on Lodging Resistance of Rice

Studies have shown that rice has strong compensatory growth ability under alternating drought and flood stress [43,44]. Drought stress inhibits stem extension growth and thickens basal stem, and timely rehydration promotes rice elongation and antagonizes the inhibitory effect caused by water loss [45,46,47]. Alternating drought and flood stress can not only reduce the plant height and internode length of rice, but it can also increase the dry matter accumulation and transformation rate of the stem [47,48]. This trial showed that RC-CI effectively inhibited the extension of the stem through alternating mild drought and flood stress, so the basal internodes could grow laterally. Compared with FSI, the internodes were significantly shortened, and the stem thickness, stem wall thickness and internode fullness were significantly increased, which made the stem have higher mechanical strength and transverse rupture strength [49]. At the same time, the plant height was reduced, which significantly reduced the bending moment, so the lodging index had been effectively reduced. The combined effect of the above factors had significantly improved the lodging resistance of rice and reduced the lodging risk.

Through the comparative experiments of drip irrigation drought direct broadcast, flood irrigation drought direct broadcast and conventional flooding irrigation, some scholars found that dry direct irrigation can significantly lower the plant height and internode length of rice than conventional flooding irrigation, and the stem fullness and lodging resistance were significantly improved [41]. Conversely, others argued that the internode fullness of dry cultivated rice is lower than that of conventional irrigation, and the lodging rate is still higher [50]. The long-term anhydrous layer in the paddy field under DPS made the rice undergo severe water deficit. The plant height and the center of gravity became lower, and the internode length became shorter, which was beneficial to reducing the risk of lodging. However, due to long-term drought and failure to obtain more water compensation in time, stem thickness and stem wall thickness were inhibited. Moreover, the dry matter accumulation in the stem decreased [51], resulting in a decrease in internode fullness. Finally, the transverse rupture strength of stems was inferior to that of FSI. In this trial, rice under DPS still had good lodging resistance. A possible reason was that the reduction in plant height and aboveground dry matter significantly reduced the bending moment of the stem caused by the weight of the plant itself. The effect of reduced internode fullness on rice lodging resistance was weakened, the lodging index remained low and the rice lodging resistance was strong.

4.3. Effects of Different Irrigation and Drainage Modes on Rice Yield

The difference in irrigation water volume had a certain impact on rice yield, thousand-grain weight and seed setting rate. Moderate stress had no damage to yield, but severe stress could cause the yield and water use efficiency to decrease [52]. Alternating dry and wet stress can increase rice yield by increasing the number of grains per panicle and effective panicle [53]. Using “Japonica 44” as the trial material, it was found that, compared with FSI, both CI (Controlled Irrigation) and RC-CI significantly increased rice yield and that the rice yield factor was at a better level [54]. Similar conclusions were obtained in this trial. Under different irrigation and drainage modes, concerning the measured and theoretical yield per hole in two years, RC-CI was the highest, and DPS was the lowest. Compared with FSI, the seed setting rate and thousand-grain weight of RC-CI were 2.5% and 2.1% higher, indicating that alternating mild drought–flood stress can promote grain filling, improve the thousand-grain weight and seed setting rate of super rice and thus increase yield [5]. The water control of DPS affected the normal development of rice. Water stress reduced rice grain filling, leading to 6.3% lower seed setting rate than FSI, which may be the main reason for low final yield [50].

4.4. Optimal Irrigation and Drainage Mode by Entropy Weight TOPSIS Model

According to the correlation analysis between the lodging index and various indices of rice, the growth index, stem morphological index and stem mechanical index were selected as the first-level indices. The corresponding second-level indices, the types of each index and the evaluation results are shown in

Table 8. Evaluation indices of the ENTROPY TOPSIS model.

The First-Level Index	The Second-Level Index	Type of Indices
growth index	Plant height/cm	Negative
	Ratio of center of gravity/%	Negative
	The measured yield per hole/g	Positive
stem morphological index	Internode length/cm	Negative
	Internode thickness/cm	Positive
	Stem wall thickness/cm	Positive
	Internode fullness/(g·cm−1)	Positive
stem mechanical index	Transverse rupture strength/N	Positive
	Bending moment/N·cm	Negative
	Bending moment at break/N·cm	Positive
	Lodging index	Negative

and

Table 9. Evaluation results of the ENTROPY TOPSIS model.

Treatment	Distance from the Optimal Solution	Distance from the Worst Solution	Normalized Score	Ranking
FSI	0.0872	0.0559	0.2370	3
RC-CI	0.0184	0.1111	0.5205	1
DPS	0.0889	0.0591	0.2425	2

. From the evaluation results, it was concluded that RC-CI was the best model for the cultivation of lodging-resistant rice, followed by DPS.

5. Conclusions

(1)

Compared with FSI, RC-CI promoted the lateral development of the basal internode, increased the accumulation of dry matter in the stem and internode fullness, enhanced the mechanical strength of the stem and improved the lodging resistance of the stem. At the same time, it appropriately reduced the plant height, so as to reduce the bending moment of the base internodes. As a consequent, the lodging risk was effectively reduced.

(2)

DPS was subject to severe water deficit, so the internode thickness and stem wall thickness were inferior to that of FSI, leading to a weakening of the mechanical strength of the stem, and the morphological properties and mechanical properties were not improved. This did not impair the lodging resistance of DPS. For decrease in the output rate and conversion rate of the stem, coupled with the significant reduction in plant height, the bending moment remained low. Therefore, DPS can reduce the lodging risk of rice.

(3)

RC-CI can significantly increase the thousand-grain weight, seed setting rate and yield of rice under alternating mild drought–flood stress. DPS under severe water deficit reduced the accumulation of dry matter in the panicle and the yield index of rice, which was not conducive to a high yield.

(4)

Considering the yield and lodging resistance of the super rice “Japonica 9108”, the best irrigation and drainage mode is RC-CI, followed by DPS.