Effects of Sowing Date and Nitrogen (N) Application Rate on Grain Yield, Nitrogen Use Efficiency and 2-Acetyl-1-Pyrroline Formation in Fragrant Rice

Abstract

Purpose: This study aimed to assess the effects of the sowing date and nitrogen application rate on the grain yield, nitrogen use efficiency (NUE), 2-acetyl-△1-pyrroline (2-AP) contents and biochemical parameters related to 2-AP formation in fragrant rice. Methods: A factorial split-plot arrangement of treatments was set up in a split-zone experiment with two factors and three levels carried out for 3 years. The main plots included two sowing dates: April 1st and May 1st. Subplots contained three different nitrogen levels i.e., 0 kg N ha⁻¹ (N0), 120 kg N ha⁻¹ (N1) and 180 kg N ha⁻¹ (N2). Results: The results indicated that compared with April 1st, the delay in the sowing date (May 1st) decreased the number of grains per panicle, 1000-grain weight, grain yield, NUE and contents of 2-AP, △1-pyrroline, proline and pyrroline-5-carboxylic acid (P5C), as well as the proline dehydrogenase (ProDH) activity. Furthermore, compared with N0, higher-N (N1 and N2) applications increased the panicle number, the number of grains per panicle, grain yield and contents of 2-AP, △1-pyrroline, proline and P5C, as well as the activities of ProDH and ornithine aminotransferase (OAT). The seed-setting rate, 1000-grain weight and NUE were decreased under N2 compared to N1, but the 2-AP content and yield were higher under the high-N application N2 (180 kg ha⁻¹) compared to N1 (120 kg ha⁻¹). Conclusions: Sowing on April 1st and the N2 (180 kg ha⁻¹) application improved the yield and promoted 2-AP biosynthesis, while sowing on May 1st reduced the yield and 2-AP content. Therefore, sowing on April 1st with 180 kg ha⁻¹ of nitrogen is the best, which can not only ensure the yield but also improve the fragrant quality of rice.

1. Introduction

Rice (Oryza sativa L.), one of the most important food crops in the world, is the main source of food for more than half of the world’s population [1,2]. China is the largest rice producer in the world and produces approximately 28% of the total rice yield worldwide [3,4]. As commodity economies have developed and residents’ dietary habits and quality of life have improved, the rice consumption level and structure have undergone fundamental changes [5,6]. Strong consumer demand exists for medium- and high-end rice, particularly high-quality rice [5,7]. Fragrant rice has a rich aroma, which is regarded as a treasure in rice. Fragrant rice has long been favored by consumers for its fragrant and delicious flavors and excellent taste [8]. Apart from its aromatic qualities, fragrant rice also has a high nutritional value and good market prospects [9]. Further strengthening fragrant rice cultivation research and continuously improving planting efficiency will further promote the stable and efficient development of fragrant rice production in the future [10].

The sowing date is an important factor affecting the yield and quality of rice. Bonelli et al. [11] reported that changes in the planting date could alter the crop growth rate and the length of crop phenological phases, affecting the potential grain yield and its components. The sowing date will affect the accumulated temperature of rice at different growth stages, and the appropriate sowing date can make rice use the temperature and light resources effectively and give full play to the potential yield and quality of rice varieties [12]. The growth process of rice is affected by the weather (temperature, light, rainfall and wind speed), which may be changed by adjusting the sowing date [13]. Weather affects the phenological period of rice, the establishment of the nutrient structure and the accumulation and transportation of dry matter [14]. Proper sowing dates will increase dry matter accumulation and yield and optimize the rice quality [15]. In addition, adjusting the sowing date can avoid extreme weather (high temperature, low temperature, heavy rainfall and drought) and the stress of weeds and pests on rice [16]. The yield of fragrant rice and the formation of its aroma are closely related to differences in the weather, soil and rice cultivars, but there are few reports on the effect of the sowing date.

Nitrogen (N) is essential to the growth and development of rice, and it participates in many metabolic processes, such as protein hydrolysis and amino acid metabolism [17]. As food demand increases, agricultural production uses a large amount of nitrogen fertilizer to achieve greater yields [18,19]. However, the excessive application of N fertilizer has resulted in a decrease in the N use efficiency (NUE) of rice [20]. At the field level, the average recovery efficiency of N in fertilizer is only 33% [21]. N fertilizer that is not absorbed by rice will be lost through surface runoff, infiltrate into groundwater and volatilize into the atmosphere, causing a series of environmental problems, such as water eutrophication, soil compaction, acid rain and greenhouse effects [22,23]. In addition, excess N increases the susceptibility of crops to lodging and disease and limits the opportunities for increased rice production [24]. Improving NUE is one of the most effective means of increasing crop productivity while decreasing environmental degradation and planting costs. Additionally, N application affected 2-AP formation in fragrant rice [25]. Mo et al. [26] reported that different N application levels affected the 2-AP content in fragrant rice. Therefore, moderate and timely N application improves the 2-AP content in fragrant rice.

Aroma is one of the important quality traits of fragrant rice, and 2-AP is an important characteristic substance of aroma in fragrant rice. 2-AP biosynthesis in fragrant rice is a very complex phenomenon. Numerous previous studies have identified some precursors and enzymes in the pathways of 2-AP biosynthesis. In brief, proline, glutamic acid, ornithine, 1-pyrroline-5-carboxylate (P5C) and 1-pyrroline have been detected as the key precursors for 2-AP biosynthesis, among which proline is the most important precursor and may be directly involved in the formation of 2-AP [27,28,29]. In addition, enzymes such as proline dehydrogenase (ProDH), Δ1-pyrroline-5-carboxylate synthetase (P5CS), ornithine aminotransferase (OAT) and diamine oxidase (DAO) were suggested to be highly related to 2-AP formation [4,30,31,32]. The biochemical parameters and several important steps involved in 2-AP biosynthesis in fragrant rice can be interpreted from previous studies, and one possible biosynthetic pathway of 2-AP is shown in Figure 1.

2-AP is the main component of the aroma of fragrant brown rice, and its biosynthesis and regulation are affected by not only genetic factors but also the external environment and crop management factors [33,34]. Therefore, the selection of the sowing time and the management of the N application rate are key strategies to improve the 2-AP content of fragrant rice. However, the effects of different sowing dates with different N application levels on the yield, NUE and 2-AP content of fragrant rice have yet to be studied in detail. In the present study, Meixiangzhan 2 (bred by the Rice Research Institute, Guangdong Academy of Agricultural Sciences, China) was used as a material for a three-year field experiment with two sowing dates and three N levels to evaluate: (1) the effects of the sowing date and N application rate on the yield formation of fragrant rice; (2) the effects of different sowing dates and nitrogen application levels on NUE; (3) the relationship between the sowing date and N application rate on the 2-AP content and its formation-related biochemical parameters.

2. Materials and Methods

2.1. Experimental Site Description

In 2017–2019, experiments were conducted at the Yangtze University farm in Jingzhou County (30.32° N, 112.04° E), Hubei Province, China. There is a subtropical agricultural climate in the area. The annual average temperature was 16.5 °C, the accumulated temperature ≥10 °C was 5094.9 to 5204.3 °C, the annual average precipitation was 1095 mm, and the annual average sunshine time was 1718 h. Meteorological data during rice growth in 2017–2019 are shown in Figure 2. The soils were sandy loam (three intact soil cores 0–20 cm from each subplot were collected using a 12 cm wide sampling soil sampler and then mixed to form one composite sample), and the soil characteristics’ average values (n = 3) are shown in

Table 1. Growth stage and soil characteristics’ average values (n = 3) of experimental field for two sowing dates.

Year	Sowing Date	Transplanting Date	Jointing Date	Filling Date	Harvest Date	Total N	Available N	Total P	Total K	Organic Matter	pH
						g kg−1	mg kg−1	g kg−1	g kg−1	g kg−1
2017	Apr 1	May 2	June 9	July 3	Aug 2	2.24	83.94	0.57	3.49	21.61	6.8
	May 1	May 31	July 8	July 30	Aug 27	2.31	89.17	0.53	3.42	20.19	6.4
2018	Apr 1	May 2	June 9	July 4	Aug 2	2.27	87.16	0.53	3.55	21.41	6.7
	May 1	May 30	July 5	July 28	Aug 25	2.28	85.46	0.55	3.51	20.34	6.5
2019	Apr 1	May 3	June 10	July 4	Aug 2	2.25	86.88	0.59	3.67	21.39	6.7
	May 1	May 3	July 7	July 30	Aug 29	2.26	83.49	0.57	3.71	20.81	6.6

Note: The experiment was conducted at Yangtze University farm, Jingzhou County, Hubei Province, China (2017–2019).

. The experimental variety is Meixiangzhan 2 (O. sativa L), which is a conventional temperature-sensitive rice variety bred by the Rice Research Institute of Guangdong Academy of Agricultural Sciences in China and possesses excellent rice quality with a growth duration of 110–115 days.

2.2. Experimental Treatments and Design

A split-plot design was used for the experiment, with the sowing date as the main plot and N application as the subplot. A total of 18 subplots (4.5 m × 6.5 m) were established in the field, with two planting dates, three N rates and three replicates. The sowing dates were 1 April and 1 May in each year of 2017–2019. The N application rates were N0: 0 kg N ha⁻¹; N1: 120 kg N ha⁻¹; and N2: 180 kg N ha⁻¹. N in the form of urea was split-applied at the basal (40%), mid-tillering (30%) and flowering stages (30%).

Seedlings were transplanted to the field after growing for 1 month in a seedbed. The planting density was 25 hills m⁻², and two seedlings were transplanted per hill. Plastic film was used to cover ridges that were 35 cm wide. Three days before transplanting, plastic film was installed 20 cm below the soil surface. In order to avoid yield loss, weeds, pests and diseases were tightly controlled [35,36].

2.3. Measurement Items and Methods

2.3.1. Yield and Yield Components

At maturity, 5 m² plant samples were taken at each plot’s center to measure grain yields. We separated the grains from the rachis, sorted the filled and unfilled grains and determined the weight of the filled grains. Grain yield was calculated at 13.5% moisture content after grain filling and drying them at 70 °C to a stable weight. We collected plant samples from 12 hills (0.48 m²) inside the harvest area to determine yield components (panicle density, number of spikelets per panicle, 1000-grain weight and seed-setting rate).

2.3.2. Nitrogen Use Efficiency

Following constant-weight drying at 80 °C of grain and straw harvested at maturity, tissue N was analyzed. The N content of the grains and straw was measured using a Kjeldahl Apparatus (FOSS-8400, Lund, Sweden). The internal N use efficiency (IE_N, grain yield over the total amount of N uptake in plants at maturity), apparent recovery efficiency of N fertilizer (RE_N, the percentage of fertilizer N recovered in aboveground plant biomass at the end of the cropping season), agronomic N use efficiency (AE_N, the yield increase that results from N application in comparison with no N application), physiological nitrogen efficiency (PE_N, the yield increase over the N uptake increase that results from N application in comparison with no N application), nitrogen partial factor productivity (PFP_N, the ratio of grain yield to the amount of N applied) and N harvest index (HI_N, the ratio of the amount of N in grains to the amount of N in aboveground plant biomass at maturity) were calculated according to the method in [37].

2.3.3. Estimation of 2-Acetyl-△1-Pyrroline (2-AP) Concentration in Grains

Based on Mo et al.’s report, the synchronization distillation and extraction (SDE) of grain samples was combined with the GCMS-QP 2010 Plus (Shimadzu Corporation, Japan) to determine 2-AP concentrations [38]. In our experiment, we used collidine (2, 4, 6-trimethylpyridine) (Sigma, Switzerland) as an internal standard. First, 145 mL of pure water was added to a 500 mL round-bottom flask containing 10 g of finely ground grain; then, 5 mL of 0.914 g per milliliter internal standard was added. The flask was attached to a steam distillation continuous extraction head, heated at 150 °C in an oil bath pot. In a 500 mL round-bottom flask attached to the other head of the steam distillation continuous extraction instrument, diethyl ether (35 mL) was used as a solvent, heated at 42 °C in the HH-2 water bath pot (Jiangsu, China). A 10 °C temperature was maintained during the steam distillation continuous extraction by a cold-water circulation machine (YKKY-LX-300, Beijing, China). Isolation lasted 35 min. The ether extract was dried over sodium sulfate and filtered (0.22 μm filter paper, Shimadzu, Japan), and the GCMS-QP 2010 Plus was used to measure 2-AP concentrations. This GCMS-QP 2010 Plus was configured as follows: gas chromatograph, automatic injector, Restek Rxi-5 ms column (30 m × 0.32 mm × 0.25 m), Shimadzu AOC/20i and SPL1. The carrier gas was 99.999% helium gas, supplied by Guangzhou Gases Co., LTD, Guangzhou, China, at a flow rate of 2.0 mL∙min⁻¹. In the GC oven, the temperature was 40 °C (1 min), then increased to 65 °C at 2 °C per minute, held at 65 °C for 1 min, then increased to 220 °C at 10 °C per minute and held at 220 °C for 10 min. The ion source temperature was 200 °C. The retention time of 2-AP was 7.5 min under these conditions. The 2-AP content was expressed as g∙kg⁻¹. Based on their mass spectra, the relative contents of the aroma compounds were determined by comparing them with NIST records. The 2-AP quantitative chromatogram is shown in Figure 3.

2.3.4. Measurements of △1-Pyrroline, Proline and Pyrroline-5-Carboxylic Acid (P5C) Contents in Grains

The △1-pyrroline content in grains was measured using the methods of Okpala et al. [39]. The △1-pyrroline content was determined after 30 min of reaction with 1, 4-diaminobutane at 30 °C and 30 min rest at room temperature. The proline content in grains was measured using the methods of Mo et al. [26]. The absorbance was recorded at 520 nm after the reaction with ninhydrin. The pyrroline-5-carboxylic acid (P5C) content in grains was measured according to the method described by Mo et al. [40]. After the reaction with trichloroacetic acid (TCA) and 2-aminobenzaldehyde, the absorbance was recorded at 440 nm. All of the above measurements were performed with three biological replications, and the average data were used for data analysis.

2.3.5. Determination of the Activity of Proline Dehydrogenase (ProDH), △1-Pyrroline-5-Carboxylic Acid Synthetase (P5CS), Ornithine Aminotransferase (OAT) and Diamine oxidase (DAO) in Grains

ProDH activity was determined according to the methods of Bao et al. [33]. The post-reaction absorbance was recorded at 440 nm. Based on the methods of Bao et al. [33], P5CS activity was measured. The reaction system consisted of 50 mM Tris-HCL buffer, 20.0 mM MgCl₂, 50 mM sodium glutamate, 10 mM ATP and 100 mM hydroxamate-HCl, and 0.5 mL of enzyme extract was added. According to Luo et al. [41], OAT activity was measured. After the reaction, the absorbance was recorded at 440 nm, and the activity was calculated using an extinction coefficient of 2.68 mM⁻¹ cm⁻¹. The DAO activity was measured by the methods of Bao et al. [33]. The reaction was initiated by the addition of 0.1 mL 20 mM Put. After the reaction, the absorbance was recorded at 555 nm.

2.4. Statistical Analyses

Statistical data were analyzed using analysis of variance (SPSS 21.0), and differences between treatments were separated using a least significant difference (LSD) test at a 5% probability level. The figures in this paper were drawn with Origin 9.1.

3. Results

3.1. Effect of Sowing Date on Temperature and Precipitation during the Rice Growing Period

Meteorological data during the rice growing period in 2017–2019 are shown in

Table 2. Meteorological data of two sowing dates in 2017–2019.

Year	Sowing Date	Sowing–Transplanting				Transplanting–Jointing				Jointing–Filling				Flowering (4 d)	Filling–Maturity
		P	Tmax	Tmin	Tmean	P	Tmax	Tmin	Tmean	P	Tmax	Tmin	Tmean	Tmax	P	Tmax	Tmin	Tmean
		mm	°C	°C	°C	mm	°C	°C	°C	mm	°C	°C	°C	°C	mm	°C	°C	°C
2017	1-April	161.9	23.29	13.77	18.14	242.8	28.57	18.76	23.2	53.2	29.37	22.78	25.58	32.70	55.6	34.24	26.51	29.86
	1-May	93.9	28.62	18.35	22.97	202.7	29.51	22.66	25.62	54.5	34.60	26.55	30.07	34.90	140.4	33.68	25.88	28.89
2018	1-April	115.5	24.64	15.11	19.33	155.3	28.34	20.23	23.74	90.2	31.74	24.15	27.48	29.08	96.6	33.93	26.40	29.75
	1-May	130.7	28.25	20.30	23.62	146.9	30.80	23.01	26.48	51.5	34.01	26.67	30.09	34.35	17.1	34.61	26.25	29.96
2019	1-April	84.2	23.34	13.51	17.86	206.0	27.53	19.03	22.92	152.3	30.35	22.43	25.75	31.40	88.3	33.44	25.00	28.74
	1-May	169.5	26.07	17.28	21.28	188.8	30.93	22.87	26.39	82.1	33.23	24.96	28.63	34.88	23.3	35.41	26.53	30.28

Note: P, T_max, T_min and T_mean present precipitation, maximum temperature, lowest temperature and average temperature, respectively.

. The temperature at the jointing to filling stages of late-sown rice dropped by 2.61–4.49 °C compared to that of early-sown rice. The highest temperature at the flowering stage of late-sown rice is 2.2–5.3 °C higher than that of early-sown rice, and the temperature at the heading to maturity stage of late-sown rice rose by 0.21–1.54 °C compared to that of early-sown rice. The consistent performance in 2017–2019 showed obvious differences in precipitation during the two sowing dates. There were obvious differences in rainfall and temperature between April 1 and May 1, and rainfall and temperature were important factors affecting the yield and quality characteristics of rice. Adjusting the sowing date could avoid the adverse effects of precipitation and temperature on rice.

3.2. Effect of Sowing Dates and Nitrogen Application Rate on Yield and Yield Components

The panicle number, number of grains per panicle, seed-setting rate, 1000-grain weight and yield were significantly affected by the sowing date and N application rate, among which the number of grains per panicle, 1000-grain weight and grain yield were significantly affected by the interaction between the sowing date and nitrogen application rate (

Table 3. Effect of sowing date and nitrogen application on yield and yield components.

Year	Sowing Date	Nitrogen	Panicle Number	Grain Number Per Panicle	Seed-Setting Rate	1000-Grain Weight	Yield
			m2		%	g	kg ha−1
2017	1-April	0	184.11 ± 0.72 c	100.17 ± 0.86 c	90.67 ± 2.14 a	21.23 ± 0.14 bc	3309.42 ± 68.69 d
	1-April	120	235.66 ± 11.62 ab	121.83 ± 6.36 a	89.00 ± 0.92 a	22.50 ± 0.33 a	5092.88 ± 212.87 b
	1-April	180	249.59 ± 2.24 a	128.09 ± 4.19 a	84.39 ± 3.25 c	20.43 ± 0.60 bc	5660.14 ± 209.96 a
	1-May	0	188.01 ± 6.26 c	89.78 ± 4.25 d	85.73 ± 3.94 bc	21.06 ± 0.86 bc	3001.23 ± 279.55 d
	1-May	120	231.50 ± 10.33 b	105.49 ± 1.92 bc	82.92 ± 1.80 c	21.42 ± 0.66 b	4393.13 ± 27.87 c
	1-May	180	252.02 ± 13.89 a	109.51 ± 1.60 b	81.79 ± 1.16 c	20.22 ± 0.54 c	4704.56 ± 223.13 c
2018	1-April	0	199.12 ± 14.85 c	105.74 ± 2.33 c	85.36 ± 2.54 a	22.97 ± 0.33 a	3732.12 ± 29.28 d
	1-April	120	249.81 ± 2.54 b	118.97 ± 6.13 b	88.75 ± 3.70 a	23.20 ± 1.01 a	5558.20 ± 247.70 b
	1-April	180	267.76 ± 11.82 b	133.86 ± 5.21 a	83.82 ± 3.61 a	21.39 ± 0.72 b	6156.74 ± 96.30 a
	1-May	0	211.42 ± 8.92 c	94.21 ± 5.42 d	84.24 ± 4.52 a	21.17 ± 0.84 b	3267.46 ± 115.73 e
	1-May	120	251.32 ± 10.75 b	101.32 ± 2.07 cd	86.02 ± 3.52 a	21.19 ± 0.59 b	4989.78 ± 160.90 c
	1-May	180	293.21 ± 14.57 a	108.98 ± 2.14 c	82.61 ± 5.14 a	20.49 ± 0.98 b	5126.73 ± 347.08 c
2019	1-April	0	180.20 ± 8.74 e	104.45 ± 3.72 c	84.98 ± 1.28 a	21.22 ± 0.42 bc	3510.24 ± 168.53 d
	1-April	120	220.71 ± 4.83 c	120.90 ± 2.91 ab	86.65 ± 1.07 a	23.30 ± 0.69 a	4833.24 ± 49.74 b
	1-April	180	241.22 ± 4.86 b	127.99 ± 10.60 a	81.77 ± 5.22 a	21.83 ± 0.88 b	5509.15 ± 154.73 a
	1-May	0	196.09 ± 12.80 d	93.03 ± 4.38 d	84.13 ± 4.10 a	21.87 ± 0.18 b	3086.15 ± 51.58 e
	1-May	120	232.64 ± 6.42 bc	106.62 ± 3.80 c	85.51 ± 3.53 a	21.31 ± 0.62 bc	4336.59 ± 202.42 c
	1-May	180	269.12 ± 2.54 a	112.57 ± 2.20 bc	82.48 ± 4.95 a	20.49 ± 0.56 c	4662.81 ± 95.35 b
Analysis of variance
Y			**	ns	ns	*	**
S			**	**	*	**	**
N			**	**	**	**	**
Y×S			*	ns	ns	ns	ns
Y×N			ns	ns	ns	ns	*
S×N			ns	*	ns	*	**
Y×S×N			ns	ns	ns	ns	ns

Note: Different lowercase letters in the same column of values represent significant differences (p  <  0.05) in the same year according to LSD. Values are the mean ± standard error (n = 3). * and **, significances at p < 0.05 and p ≤ 0.01, respectively; ns, not significant at the p = 0.05 level.

). Compared with early-sown rice, late-sown rice showed an increase in the panicle number of 4.79% and decreases in the number of grains per panicle, 1000-grain weight and yield of 13.23%, 4.47% and 13.36%, respectively. Compared with N0, N1 increased the panicle number, the number of grains per panicle and yield by 22.67%, 14.97% and 46.70%, and N2 increased the panicle number, the number of grains per panicle and yield by 35.72%, 22.75% and 59.85%, respectively. The seed-setting rate increased by 0.73% in N1 and decreased by 3.54% in N2, and the 1000-grain weight increased by 2.63% in N1 and decreased by 3.61% in N2 as compared with N0. The above results showed that the yield decreased when delaying the sowing date and increased with an increasing N application rate.

3.3. Effect of Sowing Date and Nitrogen Application Rate on N Use Efficiency

Generally, the sowing date significantly affected the apparent recovery efficiency of N (RE_N), agronomic N use efficiency (AE_N), partial factor productivity of applied N (PFP_N) and internal N use efficiency (IE_N), and N application rate significantly affected RE_N, AE_N, PFP_N, physiological N use efficiency (PE_N) and the N harvest index (HI_N). The sowing date and nitrogen application rate had no significant interaction effect on RE_N, AE_N, PFP_N, PE_N, IE_N and Hi_N (

Table 4. Effect of sowing date and nitrogen application on N use efficiency.

Year	Sowing Date	Nitrogen	REN	AEN	PFPN	PEN	IEN	HIN
			%	kg kg−1	kg kg−1	kg kg−1	kg kg−1	%
2017	1-April	0						64.04 ± 6.26 a
	1-April	120	31.48 ± 3.21 a	14.86 ± 2.29 a	42.44 ± 1.77 a	67.36 ± 8.27 a	70.93 ± 2.19 a	67.28 ± 5.69 a
	1-April	180	25.68 ± 2.27 b	13.06 ± 0.80 ab	31.45 ± 1.17 c	50.97 ± 2.04 ab	70.47 ± 0.43 a	64.96 ± 4.47 a
	1-May	0						65.63 ± 4.28 a
	1-May	120	27.69 ± 2.36 ab	11.6 ± 2.11 bc	36.61 ± 0.23 b	62.43 ± 16.61 ab	66.84 ± 2.08 a	66.80 ± 2.78 a
	1-May	180	20.71 ± 0.74 c	9.46 ± 0.32 c	26.13 ± 1.24d	45.72 ± 2.05b	67.39 ± 3.63 a	62.84 ± 3.37 a
2018	1-April	0						61.69 ± 2.42 b
	1-April	120	30.56 ± 3.06 a	13.83 ± 1.89 a	42.11 ± 1.88a	58.81 ± 6.95ab	71.19 ± 7.02 a	69.70 ± 4.90 a
	1-April	180	25.23 ± 0.78 bc	12.24 ± 0.63 a	31.09 ± 0.49 c	48.54 ± 1.92 b	69.99 ± 2.07 a	63.74 ± 3.65 ab
	1-May	0						66.91 ± 4.46ab
	1-May	120	27.42 ± 1.86 ab	13.05 ± 0.61 a	37.80 ± 1.22 b	71.00 ± 1.45 a	68.89 ± 1.90 a	66.90 ± 1.70 ab
	1-May	180	21.40 ± 2.86 c	9.39 ± 1.83 b	25.89 ± 1.75 d	44.68 ± 12.77 b	65.36 ± 6.11 a	65.21 ± 4.41 ab
2019	1-April	0						64.09 ± 6.44 a
	1-April	120	29.78 ± 1.52 a	11.03 ± 1.14 a	40.28 ± 0.41 a	50.77 ± 11.48 ab	68.68 ± 2.83 ab	69.33 ± 2.88 a
	1-April	180	22.26 ± 0.23 c	11.11 ± 0.24 a	30.61 ± 0.86 c	49.88 ± 1.56 ab	73.69 ± 2.75 a	69.09 ± 3.63 a
	1-May	0						63.22 ± 0.98 a
	1-May	120	25.42 ± 1.04 b	10.42 ± 2.11 a	36.14 ± 1.69 b	56.30 ± 9.00 a	67.66 ± 3.69 ab	67.66 ± 2.75 a
	1-May	180	21.59 ± 0.55 c	8.76 ± 0.40 a	25.91 ± 0.53d	40.62 ± 2.63 b	64.39 ± 2.97 b	65.93 ± 1.83 a
Analysis of variance
Y			ns	**	ns	ns	ns	ns
S			**	**	**	ns	**	ns
N			**	**	**	**	ns	*
Y×S			ns	ns	ns	ns	ns	ns
Y×N			ns	ns	ns	ns	ns	ns
S×N			ns	ns	ns	ns	ns	ns
Y×S×N			ns	ns	ns	ns	ns	ns

Note: Different lowercase letters in the same column of values represent significant differences (p <  0.05) in the same year according to LSD. Values are the mean ± standard error (n = 3). * and **, significances at p < 0.05 and p ≤ 0.01, respectively; ns, not significant at the p = 0.05 level. RE_N: Apparent recovery efficiency of N; AE_N: agronomic N use efficiency; PFP_N: partial factor productivity of applied N; PE_N: physiological N use efficiency; IE_N: internal N use efficiency; and HI_N: N harvest index.

). Compared with early-sown rice, late-sown rice showed decreases in RE_N, AE_N, PFP_N and IE_N of 41.72%, 45.11%, 42.36% and 37.16%, respectively. Compared with N1, N2 decreased RE_N, AE_N, PFP_N and PE_N by 20.59%, 14.40%, 27.32% and 23.56%, respectively. Compared with N0, N1 increased HI_N by 5.73%. There was no difference in HI_N between N0 and N2. The above results showed that N use efficiency decreased when delaying the sowing date and increasing the N application rate. However, when the N application rate exceeded 180 kg ha⁻¹, the seed-setting rate and 1000-grain weight of the yield components decreased.

3.4. Effect of Sowing Date and Nitrogen Application Rate on 2-AP, Synthetic Precursors and Enzymes Involved in 2-AP Biosynthesis

Overall, the contents of 2-acetyl-△1-pyrroline (2-AP), △1-pyrroline, proline and pyrroline-5-carboxylic acid (P5C), as well as the activities of proline dehydrogenase (ProDH), diamine oxidase (DAO) and ornithine aminotransferase (OAT), were significantly affected by the sowing date and N application rate, except that the sowing date had no significant effect on △1-pyrroline-5-carboxylic acid synthetase (P5CS) and OAT activities, and the sowing date and nitrogen application had significantly different effects on 2-AP, △1-pyrroline, ProDH and P5C (

Table 5. Effect of sowing date and nitrogen application on 2-AP, synthetic precursors and enzymes involved in 2-AP biosynthesis in grains.

Year	Sowing Date	Nitrogen	2-AP	△1-Pyrroline	Proline	ProDH	P5CS	OAT	DAO	P5C
			μg kg−1 FW	mmol g−1 h−1 FW	μg g−1 FW	μmol g−1 h−1 FW	μmol g−1 h−1 FW	μmol g−1 h−1 FW	U g−1 FW	μmol g−1 FW
2017	1-April	0	24.68 ± 0.61 d	1.48 ± 0.05 d	14.70 ± 0.48c	2.36 ± 0.06d	3.88 ± 0.13 a	18.51 ± 0.89 d	2.00 ± 0.03 d	1.22 ± 0.07 b
	1-April	120	32.49 ± 0.19 c	1.74 ± 0.02 b	15.26 ± 0.46b	3.33 ± 0.14c	3.94 ± 0.10 a	19.97 ± 0.34 bc	2.83 ± 0.01 b	1.39 ± 0.08 ab
	1-April	180	36.38 ± 2.96 ab	1.85 ± 0.05 b	16.42 ± 0.64 b	3.69 ± 0.20 c	3.98 ± 0.11 a	21.88 ± 0.85 ab	2.86 ± 0.04 b	1.42 ± 0.04 a
	1-May	0	24.06 ± 0.40 d	1.23 ± 0.07 c	13.46 ± 0.09 b	2.15 ± 0.06 d	3.70 ± 0.08 a	18.42 ± 0.44 d	1.97 ± 0.03 d	1.26 ± 0.09 b
	1-May	120	30.73 ± 0.34 bc	1.72 ± 0.03 b	14.89 ± 0.47 b	2.88 ± 0.25 b	3.82 ± 0.22 a	20.31 ± 0.43 c	2.57 ± 0.03 c	1.34 ± 0.06 a
	1-May	180	34.37 ± 1.08 a	1.74 ± 0.03 a	15.09 ± 1.09 a	3.02 ± 0.08 a	3.89 ± 0.19 a	21.38 ± 0.76 a	2.99 ± 0.02 a	1.44 ± 0.05 a
2018	1-April	0	25.44 ± 1.67 e	1.40 ± 0.07 c	14.25 ± 0.81 c	2.48 ± 0.15 c	3.89 ± 0.15 a	18.75 ± 0.35 d	2.14 ± 0.02e	1.26 ± 0.03 cd
	1-April	120	31.07 ± 1.13 c	1.73 ± 0.05 b	15.47 ± 0.2 ab	3.25 ± 0.06 c	3.82 ± 0.19 a	20.33 ± 0.39 b	2.42 ± 0.03 d	1.36 ± 0.01 bcd
	1-April	180	38.20 ± 1.13 b	1.89 ± 0.09 b	15.66 ± 0.63 a	3.41 ± 0.18 b	4.00 ± 0.1 a	21.27 ± 0.49 a	3.41 ± 0.06 a	1.51 ± 0.09 b
	1-May	0	22.39 ± 0.38 d	1.28 ± 0.01 c	13.26 ± 0.59 b	2.10 ± 0.15 b	3.79 ± 0.30 a	17.89 ± 0.59 c	1.80 ± 0.02 f	1.28 ± 0.02 d
	1-May	120	30.98 ± 1.50 c	1.66 ± 0.02 b	14.87 ± 0.49 a	2.87 ± 0.06 a	3.95 ± 0.04 a	19.95 ± 0.59 b	2.76 ± 0.06 c	1.33 ± 0.02 bc
	1-May	180	34.36 ± 0.18 a	1.74 ± 0.11 a	15.60 ± 0.28 a	2.92 ± 0.11 a	4.13 ± 0.16 a	21.99 ± 0.17 a	3.10 ± 0.02 b	1.39 ± 0.07 a
2019	1-April	0	25.11 ± 0.88 c	1.42 ± 0.06 d	14.23 ± 0.40 c	2.33 ± 0.07 e	3.87 ± 0.06 a	18.21 ± 0.66 c	1.91 ± 0.02 f	1.20 ± 0.05 c
	1-April	120	31.93 ± 1.87b	1.70 ± 0.03b	14.91 ± 0.30 a	3.20 ± 0.09 c	3.99 ± 0.08 a	20.09 ± 0.27 b	2.68 ± 0.01 c	1.40 ± 0.01 c
	1-April	180	36.62 ± 0.54 a	1.85 ± 0.12 b	15.80 ± 0.67 ab	3.68 ± 0.02 b	4.04 ± 0.16 a	21.36 ± 0.98 ab	3.28 ± 0.02 a	1.53 ± 0.05 b
	1-May	0	23.23 ± 1.00 c	1.25 ± 0.01 c	13.07 ± 0.34 b	2.03 ± 0.10 d	3.87 ± 0.06 a	17.80 ± 0.72 c	1.99 ± 0.02 e	1.24 ± 0.07 c
	1-May	120	30.35 ± 1.06 b	1.69 ± 0.03 b	15.21 ± 0.52 ab	2.78 ± 0.13 b	3.98 ± 0.15 a	20.23 ± 0.41b	2.36 ± 0.05 d	1.28 ± 0.04 b
	1-May	180	35.68 ± 1.02 a	1.64 ± 0.08 a	14.91 ± 0.58 a	3.01 ± 0.18 a	4.03 ± 0.10 a	20.81 ± 0.16 a	3.17 ± 0.06 b	1.40 ± 0.07 a
Analysis of variance
Y			ns	ns	ns	ns	ns	ns	***	ns
S			**	**	**	**	ns	ns	***	*
N			**	**	**	**	**	**	***	**
Y×S			ns	ns	ns	ns	ns	ns	*	ns
Y×N			ns	ns	ns	ns	ns	ns	***	ns
S×N			ns	**	ns	**	ns	ns	ns	**
Y×S×N			ns	ns	ns	ns	ns	ns	***	ns

Note: Different lowercase letters in the same column of values represent significant differences (p  < 0.05) in the same year according to LSD. Values are the mean ± standard error (n = 3). 2-AP: 2-acetyl-△1-pyrroline; OAT: ornithine aminotransferase; DAO: diamine oxidase; P5C: pyrroline-5-carboxylic acid; P5CS: △1-pyrroline-5-carboxylic acid synthetase; ProDH: proline dehydrogenase. * and **, significances at p  < 0.05 and p  ≤ 0.01, respectively; ns, not significant at the p  = 0.05 level.

). Compared with early-sown rice, late-sown rice showed decreases in the contents of 2-AP, △1-pyrroline, proline, P5C and ProDH and DAO activity of 5.59%, 7.37%, 4.64%, 14.32%, 2.69% and 3.61%, respectively. Compared with N0, the contents of 2-AP, △1-pyrroline, proline and P5C and the activities of ProDH, OAT and DAO increased by 29.43%, 27.05%, 9.21%, 36.13%, 10.31%, 8.58% and 32.26% in N1 and increased by 48.79%, 32.88%, 12.67%, 46.39%, 17.44%, 16.49% and 59.27% in N2, respectively. The above results showed that the 2-AP content, synthetic precursors and important enzyme activities decreased with a delay in the sowing date and increasing N application.

4. Discussion

In this study, we explored the effects of the sowing date and nitrogen application rate on the yield, nitrogen use efficiency and aroma quality of aromatic rice as cultivation measures through a 3-year experiment from 2017 to 2019. Generally, the time of sowing has a marked effect on the growth and yield of most crops. When the sowing delay exceeds the optimal time, it usually leads to a decline in yield [42]. Late sowing increased the panicle number per unit but decreased the number of grains per panicle, 1000-grain weight and yield per unit compared with early sowing (Table 3). The temperature, both from the transplanting to heading stage and from the heading to maturity stage, of early-sown rice was lower than that of late-sown rice (Table 2). Temperature is a necessary condition for selecting a suitable sowing date. Increasing the temperature positively affected dry matter accumulation when the temperature rose slightly above the optimal level but did not reach the threshold [43]. Therefore, a higher temperature promotes nutrient absorption and dry matter accumulation in rice, which may be why the panicle number per unit of late-sown rice was higher. However, the number of grains per panicle, 1000-grain weight and yield per unit of early-sown rice were higher for late-sown rice, although the temperature of late-sown rice was higher. This result showed that the temperature at the heading to maturity stage of early-sown rice had yet to reach the critical point that reduces rice growth [43,44]. The higher temperature at the heading to the maturity stage of late-sown rice may be the reason for the lower number of grains per panicle, 1000-grain weight and yield per unit. The highest temperature at the flowering stage of late-sown rice is close to 35 °C, which is higher than that of early-sown rice, and a high temperature increases the risk of pollen abortion, thereby reducing the seed-setting rate of rice, which explains the decrease in yield per unit of late-sown rice. The panicle number per unit, number of grains per panicle and yield per unit rose with increasing N application (Table 2). Similar results have been repeatedly documented [45,46,47,48].

However, N2 (180 kg ha⁻¹) decreased the seed-setting rate and 1000-grain weight and significantly increased the panicle number per m⁻², which agrees with previous reports [49]. Although increasing the N application can increase the spikelet number per unit, the assimilates for grain filling were limited and largely depended on the quantity that was transferred directly to the grains after flowering and redistributed from reserves in the stems and leaves either pre- or post-anthesis, resulting in a decrease in the seed-setting rate and 1000-grain weight under high nitrogen [50].

Our research shows that late sowing decreased RE_N, AE_N, PFP_N and PF_N, illustrating that the N use efficiency of late-sown rice was lower (Table 3). The study by Qin et al. [51] also supported our findings that the RE_N, AE_N and PF_N of late-sown rice are lower than those of early-sown rice. In this study, 70% of the N was applied to the field at the transplanting to heading stage, in which the temperature of late-sown rice was higher. Reportedly, higher temperatures increase the loss of N in the form of ammonia volatilization [52]. This may be part of the reason for the low N use efficiency of late-sown rice. In addition, a higher environmental temperature results in poor rice growth and lower yield and can also reduce N use efficiency, as observed in this experiment. Increasing the N application decreased RE_N, AE_N, PFP_N and PF_N, which is consistent with the observations by Wang et al. [53]. This result shows that rice had a threshold for N demand, and the excess N applied was not absorbed and used by the rice plants and may have remained in the soil or been lost to the surrounding environment. With the increase in the N application rate, HI_N first increased and then decreased, which is consistent with the conclusion drawn by Pan et al. [54].

The formation of rice fragrance is a complex and changeable physiological process [55]. Various environmental factors (light, water and temperature) and cultivation measures (the regulation of soil fertility, mineral elements and exogenous aroma-enhancing compounds) will also directly affect the synthesis of aroma substances in rice [55]. 2-AP is considered the most important volatile compound in fragrant rice [26]. A vast number of studies have reported that precursors (proline, glutamic acid, ornithine, P5C and 1-pyrroline) and enzymes (ProDH, P5CS, OAT and DAO) were related to 2-AP formation in fragrant rice [27,28,29,30,32,56,57,58,59,60,61]. In this study, late sowing decreased the contents of 2-AP, △1-pyrroline and P5C and the ProDH activity in grains. This might be because late-sown rice in the Jianghan Plain from the jointing to the heading stage encounters high temperatures, which is not conducive to the synthesis and accumulation of 2-AP. Prodhan et al. [62] reported that the maximum downregulation of the badh2 gene and the highest concentration of 2-AP was also observed at 25 °C. These results support our conclusion that a higher temperature at the heading to the maturity stage of late-sown rice limits the ProDH activity and the synthesis of 2-AP, △1-pyrroline and P5C. Our research showed that the contents of 2-AP in the grains improved with increasing N application, in agreement with previous studies reporting that increasing N application improved the 2-AP content [26,40,63,64], further confirming that increasing N application increases the 2-AP content in grains by increasing the contents of △1-pyrroline, proline and P5C and the activities of ProDH, OAT and DAO. In recent years, ProDH, P5CS, OAT and DAO have generally been considered to play a key role in the process of 2-AP biosynthesis in fragrant rice. Previous studies have shown that increasing the amount of nitrogen application can significantly improve the aroma of fragrant rice to a certain extent [44,64], but whether the sowing date has an effect on the aroma of fragrant rice has not been studied. Our study showed that the adjustment of the sowing date has a significant regulatory effect on the aroma of fragrant rice, while the interaction between the sowing date and the amount of nitrogen application has no significant effect on the content of 2-AP but has a significant effect on the activities of precursors (△1-pyrroline and P5C) and enzymes (ProDH) related to 2-AP synthesis. Without delaying the sowing date, the regulation of fragrant rice 2-AP biosynthesis by appropriately increasing the nitrogen application rate (180kg ha⁻¹) could occur through the mechanism shown in Figure 4.

5. Conclusions

From our findings, it can be concluded that the sowing date and nitrogen (N) application significantly affected the grain yield, N use efficiency and 2-acetyl-△1-pyrroline (2-AP) content. Rice planted on May 1st encountered high temperatures during the grain-filling stage in Jianghan Plain, which reduced the synthesis of 2-AP. Therefore, sowing on Apr 1st and increasing the N application (180 kg ha⁻¹) could be a better agronomic practice for increasing the grain yield and 2-AP content of fragrant rice. From the perspective of improving N use efficiency, the recommended N application rate is 120 kg ha⁻¹. However, in terms of comprehensive yield traits, fragrant quality and nitrogen use efficiency, the most appropriate N application of nitrogen should be 180 kg ha⁻¹, which can be guaranteed to improve grain yield and 2-AP content synchronously.