Vascular Bundle Characteristics of Different Rice Variety Treated with Nitrogen Fertilizers and Its Relation to Stem Assimilates Allocation and Grain Yield

Abstract

The vascular bundle plays a vital role in photoassimilates transportation in rice. However, the vascular bundle characteristics of basal internode and its relationship with stem assimilates allocation and yield remain unclear. In this study, experiments with four different types of rice varieties subjected to three nitrogen application levels were conducted. The results showed that large vascular bundles (LVB) and small vascular bundles (SVB)-related traits of basal internode exhibited genotypic differences. Indica-japonica hybrid varieties had the highest number of cross sectional areas and phloem areas of LVB and SVB, thus the highest stem nonstructural carbohydrates (NSC) translocation, grain filling percentage and grain yield, followed by indica hybrid varieties and indica conventional varieties, and those were the lowest in japonica conventional varieties. The LVB and SVB related traits were significantly and positively correlated with stem NSC translocation, grain filling percentage and grain yield, respectively These results suggested that improving the characteristics of basal internodes was beneficial to enhance stem NSC translocation and consequently increase grain yield. Nitrogen application increased LVB and SVB related traits. Therefore, varieties with developed vascular bundles of basal internodes and cultivation techniques for improving vascular bundle related traits should be considered as the effective route for increasing grain yield.

1. Introduction

Rice (Oryza sativa L.) is one of the most important food crops in China, the annual production is about 206 million tons, accounting for 28% of global rice supply [1]. Increasing rice grain yield plays a vital role in ensuring food security in China and even the world. Grain filling is the final stage of rice yield formation and determines the grain weight and yield [2]. The assimilates required for grain filling in rice come from the leaf photosynthesis after heading and the nonstructural carbohydrates (NSC) stored in stems and sheaths before heading [3]. The transport rate and amount of stem NSC affect grain filling and yield [4]. The apparent contribution rate of NSC transported from stems and sheaths to rice yield has been suggested to be up to 28% [5], and its amount depends on growth conditions and nitrogen supply level [6,7]. Therefore, promoting the translocation of stem NSC to grain is a promising way to improve rice yield.

In rice, prior to heading, NSC is stored in stems in the form of starch, and the lower the internode, the higher the NSC content; after heading, starch is converted back to sucrose that is transported through a continuous mature vascular bundle network to the filling grains [8,9]. The vascular bundle from the leaf is continuous with that in leaf sheath, and then connects to the stem. The vascular bundles in the internode are continuous and continue into the panicle, finally entering the rachilla at the base of the filling grains [10]. The assimilates move out of the stem and throughout the long-distance transport vascular bundle to the filling grains, and the vascular bundles-related traits, such as number, size, and capacity, influence the transport efficiency. These are limiting factors for grain yield potential realization [11,12,13].

Vascular bundles contain two tissues with different functions: the xylem is used for transporting soluble mineral nutrients and water, and the phloem plays the role in transporting photoassimilates [14]. There are two kinds of vascular bundles observed in rice stems—large vascular bundles (LVB) and small vascular bundles (SVB) [15]. For a long time, the study of the botanical characteristics of rice vascular bundles of peduncle has been paid attention to by many researchers. It is known that the vascular bundles of peduncle control the pathway of assimilates to the grains. The number of LVB and SVB of peduncle is closely related to NSC transport rate, seed setting rate, harvest index and grain yield [5,16]. Notably, the contribution of the LVB and SVB of peduncle to NSC transport and yield formation is different; the role of SVB is greater than that of LVB [5], and a significant and negative correlation is even observed between the number of LVB and grain filling percentage, and grain yield [17]. The different amount of ABERRANT PANICLE ORGANIZATION1 (a gene related to spikelet numbers in rice) expression in LVB and SVB, and also the functional difference of phloem between LVB and SVB in NSC transport, may be the main reasons [15,16].

The difference of vascular bundle traits depends not only on genotype, but also on environment. The number of LVB and SVB ranges from 2.89 to 7.35 (×103, number m⁻²) and 3.71 to 9.34 (×103, number m⁻²), respectively, through investigation of 46 recombinant inbred lines from Zhenshan 97 × Minghui 63 [5]. Similarly, it is reported that erect-panicle varieties have more LVB and SVB than the curved-panicle type [18]. The number of LVB and SVB of peduncle in subspecific hybrid rice was greater than that in interspecific hybrid rice and conventional rice [19].

Nitrogen has an important effect on the development of vascular bundles and stem NSC translocation. The amount of nitrogen application affects the number and cross-sectional area of vascular bundles of peduncle. Plants grown under low nitrogen conditions exhibited decreased numbers of LVB and SVB [5,15,19,20]. The translocation of NSC in stems was higher under low nitrogen application than under high nitrogen application [5].

In rice plants, the basal internodes function as the main place for assimilate storage before heading and they are also closely related to stem lodging resistance. The vascular bundle system within the basal internodes also carries out essential functions such as the delivery of assimilates and the provision of mechanical support [9,21,22]. However, at present, the characteristics of a vascular bundle of basal internode and its relationships with stem NSC translocation and yield formation are still not fully understood. This study aimed to clarify this problem, and will be able to obtain new knowledge for improving grain yield in rice breeding practice and to provide theoretical guidance for high-yielding cultivation through the improvement of the translocation efficiency of the vascular bundle system.

In this study, we investigated the botanical characteristics of LVB and SVB in the third internode from the top using different types of rice varieties under different nitrogen application levels, and their correlations with stem NSC translocation and yield formation were analyzed to understand the role of the vascular bundle of basal internode in grain filling and yield formation.

2. Materials and Methods

2.1. Rice Varieties and Growth Conditions

A pot experiment was conducted during the rice growing season in 2020 from May to November at Yangzhou University Farm, Yangzhou, Jiangsu Province, China (32°24′ N, 119°26′ E). Eight rice varieties, including four types of varieties, two indica conventional rice: Huanghuazhan (HHZ) and Yangdao 6 (YD6); two japonica conventional rice: Wuyujing 3 (WYJ3) and Zhendao 88 (ZD88); two indica hybrid rice: Liangyoupeijiu (LYPJ) and Shanyou 63 (SY63); and two indica-japonica hybrid rice: Yongyou 25 (YY25) and Yongyou 8050 (YY8050), were used in this study. The seeds were soaked for 48 h and then sown in nursery plates containing a seedling matrix on 21 May 2020. Seedlings were transplanted 25 days after sowing into pots (25 cm in diameter and 30 cm high) filled with 15 kg air-dried and crushed soil. Four hills of plant were grown in each pot with two seedlings per hill. Three N treatments were set; the amounts of nitrogen application were 24 kg N ha⁻¹ (low nitrogen, LN), 80 kg N ha⁻¹ (medium nitrogen, MN) and 240 kg N ha⁻¹ (high nitrogen, HN) in the form of urea. The base fertilizer, tillering fertilizer and panicle fertilizer were applied at the ratio of 4:2:4 at 1 day before transplanting, 10 days after transplanting and at 3.5 leaf-age remainders (36 days after transplanting), respectively. Phosphate and potassium fertilizers were 1000 kg ha⁻¹ (P₂O₅) and 400 kg ha⁻¹ (K₂O), respectively. Basal fertilizers were mixed into the soil before transplanting, and top-dressed fertilizers were dissolved in water and irrigated into pots. The type of soil was a sandy loam that was prepared from the local paddy field. Before transplanting, the soil had organic matter 24.0 g kg⁻¹, alkali-hydrolysable N 122.1 mg kg⁻¹, Olsen-P 34.5 mg kg⁻¹, exchangeable K 68.4 mg kg⁻¹, while at maturity, the soil had organic matter at 24.3, 23.9, 24.5 mg kg⁻¹, alkali-hydrolysable N 102.3, 136.1, 147.9 mg kg⁻¹, Olsen-P 34.5, 34.1, 33.8 mg kg⁻¹, exchangeable K 68.4, 68.9, 68.3 mg kg⁻¹ under low nitrogen, medium nitrogen and high nitrogen treatments, respectively. All the pots were arranged in a split block design with the main plot being varieties and the subplot being nitrogen levels. The plants were watered well during the duration of their growth, and pests, diseases and weeds were intensively controlled using chemical pesticides.

2.2. The Vascular Bundle Related Traits Determination

At the heading stage, the six main stems that emerged on the same day were sampled from each treatment. The third internode from the top was collected, and 1–2 cm length of the middle of the internode was collected and fixed in formalin-acetic acid-alcohol for making paraffin sections. After that, the transverse sections were observed and photographed under an inverted microscope (Nikon Corporation, Inc., Tokyo, Japan). The vascular bundle traits, such as number of LVB and SVB, cross sectional area and cross sectional phloem area, and so on, were measured using Photoshop CS5 Software package (Adobe Systems, Inc., San Jose, CA, USA).

2.3. The Carbohydrates Content Determination

The main stems that emerged on the same day were tagged at heading. Several tagged plants were sampled at heading, 10 days after heading (DAH), 20 DAH, 30 DAH and maturity. The stems were collected and dried at 80 °C in the oven for 72 h to constant weight, and were then ground into powder. The determination of soluble sugars and starch was adapted from Li et al. [23]. Briefly, the powder sample was extracted using 80% ethanol; the extract was used to determine the contents of soluble sugars and sucrose. The above residue was extracted with 9.2 mol L⁻¹ perchloric acid (HClO₄) and 4.6 mol L⁻¹ HClO₄, successively. The extract was used to determine the starch content.

The contents of soluble sugars and starch were measured at 620 nm on a microplate reader (Nano Quant, infinite M200, Tecan Trading AG, Männedorf, Switzerland) using a colorimetric method with the anthrone reagent as described previously [24].

The determination of sucrose content was performed essentially as described previously [25]. The extract was decolorized using activated carbon and was purified by filtering. Next, 0.4 mL of a filtered solution was added to 0.2 mL NaOH, then the mixture was boiled for 5 min at 100 °C. After cooling, 0.8 mL of 0.1% resorcinol and 2.8 mL of 9 mol L⁻¹ HCl were added. The reaction mixture was incubated for 10 min at 80 °C, and then the absorbance was measured at 480 nm.

The stem NSC content (mg g⁻¹) was calculated as the sum of the soluble sugar content and the starch content of stems. The apparent transferred mass (ATM, mg g⁻¹) of stem NSC is the NSC content at heading minus that at maturity. The apparent transferred ratio (AR, %) of stem NSC was calculated as the ratio of ATM to the NSC content at heading.

2.4. The Grain Yield and Its Components Determination

At maturity, three hills of plants were selected from each treatment based on the average number of effective panicles per plant. The sampled plants were separated into leaves, stems and panicles. All parts were oven-dried at 80 °C to a constant weight. The spikelets of the panicles were threshed by hand, and the grains were divided into filled and unfilled grains by submersing in tap water. The numbers of filled and unfilled grains were counted. Then, spikelet number, grain filling percentage, 1000-grain weight, harvest index, biomass and grain yield were calculated. Spikelet number (No. panicle⁻¹) = the sum of number of filled and unfilled grains/effective panicle number; grain yield (g plant⁻¹) = filled grains weight per plant; biomass (g plant⁻¹) = the summation of dry weights of aboveground plant parts; grain filling percentage (%) = number of filled grains/the sum of number of filled and unfilled grains; harvest index = grain yield/biomass.

2.5. Statistical Analysis

Two-way analysis of variance (ANOVA) was performed using the Statistix 9 software package (Analytical software, Tallahassee, FL, USA). The least significant difference test (LSD test) was used to estimate the significant differences of the mean values of investigated traits among varieties exposed to the different nitrogen conditions at a 5% probability level. Pearson correlation analysis was performed using the SigmaPlot 10.0 software package (SPSS Inc., Chicago, IL, USA).

3. Results

3.1. Grain Yield and Its Components

Grain yield, panicle plant⁻¹ and biomass of all varieties were significantly increased with nitrogen application level; spikelets per panicle of indica conventional rice (IC) and indica hybrid rice (IH) were increased with nitrogen application level; however, no similar trend was observed in japonica conventional rice (JC) and indica-japonica hybrid (IJH). The effect of nitrogen on grain filling percentage had no consistent trend among different varieties. Nitrogen levels had no or little effect on the 1000-grain weight and harvest index. Among different types of varieties, IJH had the highest grain yield, followed by IH and IC, and JC was the lowest. Both high spikelets plant⁻¹ and grain filling percentage contributed to the high yield of IJH (

Table 1. Grain yield and yield components of different rice varieties under different nitrogen application levels.

Traits	N Treatment	Indica Conventional Rice			Japonica Conventional Rice			Indica Hybrid Rice			Indica-Japonica Hybrid Rice
		HHZ	YD6	Mean	ZD88	WYJ3	Mean	LYPJ	SY63	Mean	YY25	YY8050	Mean
Grain yield	LN	9.0 c	10.6 c	9.8 c	9.1 c	8.0 c	8.6 c	10.7 c	9.6 c	10.1 c	12.8 c	14.0 c	13.5 c
(g plant−1)	MN	17.4 b	22.6 b	20.0 b	21.9 b	14.0 b	18.0 b	30.7 b	26.5 b	28.6 b	26.7 b	22.6 b	24.6 b
	HN	33.1 a	45.0 a	39.1 a	31.0 a	29.1 a	30.0 a	42.7 a	46.5 a	44.6 a	39.1 a	50.6 a	44.9 a
Panicle	LN	4.0 c	3.0 c	3.5 c	4.0 c	4.3 c	4.2 c	4.0 c	3.0 c	3.5 c	3.0 c	3.0 c	3.0 c
(No. plant−1)	MN	6.3 b	5.0 b	5.7 b	7.0 b	7.0 b	7.0 b	8.3 b	7.0 b	7.7 b	5.0 b	4.7 b	4.8 b
	HN	11.3 a	10.0 a	10.7 a	11.0 a	13.7 a	12.3 a	9.0 a	11.0 a	10.0 a	7.7 a	9.7 a	8.7 a
Spikelets	LN	136.9 a	136.6 b	136.7 b	111.2 a	78.9 a	95.1 a	151.9 b	128.3 c	140.1 b	201.6 b	209.7 a	205.6 b
(No. panicle−1)	MN	150.8 a	168.7 a	159.8 a	137.1 a	82.1 a	109.6 a	177.3 b	145.2 b	161.2 b	268.3 a	216.8 a	242.6 a
	HN	164.9 a	179.3 a	172.1 a	124.7 a	88.3 a	106.5 a	228.0 a	160.6 a	194.3 a	248.5 a	217.0 a	232.8 a
Grain filling percentage	LN	85.5 b	93.8 a	89.6 b	85.3 b	89.4 b	87.4 b	77.7 b	93.5 b	85.6 b	91.6 a	95.4 a	93.5 a
(%)	MN	92.3 a	94.3a	93.3 a	92.1 ab	91.9 b	92.0 ab	85.2 a	94.6 a	89.9 a	89.8 a	95.9 a	92.8 a
	HN	92.8 a	90.8 b	91.8 ab	95.8 a	94.6 a	95.2 a	86.6 a	95.3 a	91.0 a	91.7 a	91.7 b	91.7 a
1000-grain weight	LN	19.1 a	27.6 a	23.3 a	24.0 a	26.7 a	25.3 a	22.7 a	26.6 a	24.6 a	23.1 a	23.3 a	23.2 a
(g)	MN	19.8 a	28.3 a	24.1 a	24.9 a	26.6 a	25.7 a	24.2 a	27.6 a	25.9 a	22.2 a	23.2 a	22.7 a
	HN	19.5 a	27.6 a	23.6 a	23.8 a	25.5 a	24.6 a	24.1 a	27.7 a	25.9 a	22.3 a	23.4 a	22.9 a
Biomass	LN	17.3 c	20.0 c	18.7 c	18.4 c	16.4 c	17.4 c	20.2 c	17.2 c	18.7 c	25.0 c	26.1 c	25.5 c
(g plant−1)	MN	32.2 b	42.1 b	37.2 b	40.0 b	28.1 b	34.0 b	56.1 b	50.2 b	53.2 b	54.5 b	45.4 b	50.0 b
	HN	67.3 a	89.6 a	78.5 a	53.7 a	54.3 a	54.0 a	80.2 a	88.9 a	84.5 a	80.0 a	90.7 a	85.3 a
Harvest index	LN	0.52 ab	0.53 a	0.52 a	0.54 b	0.53 a	0.54 a	0.53 a	0.55 a	0.54 a	0.50 a	0.46 c	0.48 b
	MN	0.54 a	0.53 a	0.54 a	0.55 b	0.50 b	0.52 b	0.54 a	0.53 b	0.54 a	0.49 a	0.50 b	0.49 b
	HN	0.49 b	0.50 b	0.50 b	0.58 a	0.54 a	0.56 a	0.53 a	0.52 b	0.53 a	0.49 a	0.56 a	0.52 a

LN, low nitrogen; MN, medium nitrogen; HN, high nitrogen; HHZ, Huanghuazhan; YD6, Yangdao 6; ZD88, Zhendao 88; WYJ3, Wuyujing 3; LYPJ, Liangyoupeijiu; SY63, Shanyou 63; YY25, Yongyou 25; YY8050, Yongyou 8050. Different letters following means indicate significant difference among the three nitrogen treatment for each trait p < 0.05.

).

3.2. Carbohydrate Contents in Stems

The contents of sucrose and soluble sugars in the stems of all types of varieties increased from heading (HD) to 20 days after heading (DAH) and then decreased until maturity under all the three nitrogen treatments. The contents of starch and NSC in stems showed a downward trend from HD to maturity except starch and NSC contents at 10 DAH under HN condition in IJH. The contents of starch and NSC at HD were higher in IJH than the other types of varieties, and JC and IC had low contents of starch and NSC during HD to 30 DAH; however, the starch and NSC contents in IJH decreased to a level that was lower than JC at maturity. The carbohydrates contents (sucrose, soluble sugars, starch, and NSC) were higher under LN than those under MN, and higher under MN than those under HN (Figure 1).

3.3. The NSC Translocation in Stems

The stem NSC translocation was described by ATM and AR, which decreased with nitrogen levels. IJH had the highest ATM and AR under the three nitrogen treatments; the ATM and AR of IJH were significantly higher than those of the other types of varieties, followed by IH and IC; JC was the lowest (Figure 2).

3.4. Relationships of Stem NSC Contents at Heading and at Maturity, and Stem NSC Translocation with Grain Yield-Related Traits

Grain yield, grain filling percentage and harvest index were positively correlated with NSC content at HD. These correlations were significant, except between NSC content and grain yield under HN. Grain yield, grain filling percentage and harvest index were negatively correlated with NSC content at maturity; however, the correlations between NSC content at maturity and grain filling percentage, and between NSC content at maturity and harvest under LN were not significant. No significant correlations were observed between NSC content at HD and 1000-grain weight, and between NSC content and 1000-grian weight (Figure 3).

Grain yield, grain filling percentage and harvest index were significantly and positively correlated with ATM and AR of stem NSC under all the three nitrogen application levels. However, 1000-grain weight was not significantly correlated with ATM and AR (Figure 4).

3.5. Vascular Bundle Traits

The number of LVB and SVB of the third internode from the top increased with nitrogen levels; however, no significant differences were observed among different nitrogen application levels (Figure 5). There was no significant effects of nitrogen application levels on cross sectional area per LVB and SVB, phloem area per LVB and SVB, total cross sectional area of LVB and SVB, total phloem area of LVB and SVB, total cross section of all vascular bundles, and total phloem area of all vascular bundles in the third internode from the top (Figure 5, Figure 6 and Figure 7). When mentioning the vascular bundle traits above, among different types of varieties the IJH was the highest, followed by IH and IC, and JC is the lowest. Significant differences were observed between IJH and JC under all the three nitrogen application levels (Figure 5, Figure 6 and Figure 7).

3.6. Relationships of Stem NSC Translocation with Vascular Bundle Traits

The ATM and AR of stem NSC were significantly and positively correlated with the number of LVB and SVB, the cross sectional area per LVB and SVB, the phloem area per LVB and SVB, the total cross sectional area of LVB and SVB, and the total phloem area of LVB and SVB in the third internode from the top under all three nitrogen application levels (Figure 8 and Figure 9).

3.7. Relationships between Vascular Bundle Traits and Grain Yield-Related Traits

Grain yield and grain filling percentage were significantly and positively correlated with all vascular bundle related traits that are shown in

Table 2. Correlations of vascular bundle traits of basal internode with yield-related traits (n = 24).

Traits	Treatment	Grain Yield	GFP	GW	HI
Number of LVB	LN	0.72 **	0.51 *	−0.10	−0.39
	MN	0.42 *	0.69 **	−0.39	−0.26
	HN	0.47 *	0.46 *	−0.24	−0.24
Average cross sectional area of LVB	LN	0.63 **	0.48 *	0.21	−0.38
	MN	0.57 **	0.43 *	0.02	−0.25
	HN	0.66 **	0.54 **	−0.10	−0.28
Average phloem area of LVB	LN	0.61 **	0.58 **	−0.38	−0.37
	HN	0.59 **	0.49 *	−0.01	−0.18
	HN	0.71 **	0.56 **	−0.19	−0.20
Total cross sectional area of LVB	LN	0.68 **	0.39	−0.03	−0.37
	HN	0.57 **	0.57 **	−0.18	−0.32
	HN	0.73 **	0.67 **	−0.22	−0.18
Total phloem area of LVB	LN	0.66 **	0.74 **	0.01	−0.38
	HN	0.55 **	0.56 **	−0.14	−0.27
	HN	0.65 **	0.55 **	−0.27	−0.23
Number of SVB	LN	0.55 **	0.43 *	−0.35	−0.32
	HN	0.44 *	0.47 *	−0.36	−0.26
	HN	0.76 **	0.50 *	−0.16	−0.24
Average cross sectional area of SVB	LN	0.70 **	0.60 **	−0.07	−0.36
	HN	0.60 **	0.82 **	−0.39	−0.25
	HN	0.52 **	0.59 **	−0.25	−0.31
Average phloem area of SVB	LN	0.58 **	0.51 *	−0.05	−0.31
	HN	0.67 **	0.81 **	−0.30	−0.10
	HN	0.65 **	0.48 *	−0.21	−0.20
Total cross sectional area of SVB	LN	0.66 **	0.65 **	−0.28	−0.40
	HN	0.41 *	0.56 **	−0.36	−0.35
	HN	0.53 **	0.43 *	−0.30	−0.39
Total phloem area of SVB	LN	0.61 **	0.50 *	−0.11	−0.33
	HN	0.43 *	0.50 *	−0.26	−0.31
	HN	0.64 **	0.50 *	−0.25	−0.28
Total cross sectional area of LVB and SVB	LN	0.68 **	0.42 *	−0.03	−0.34
	HN	0.51 *	0.64 **	−0.32	−0.32
	HN	0.51*	0.43 *	−0.18	−0.33
Total phloem area of LVB and SVB	LN	0.63 **	0.46 *	0.15	−0.28
	HN	0.53 **	0.62 **	−0.24	−0.28
	HN	0.67 **	0.41 *	−0.15	−0.24

LVB, large vascular bundle; SVB, small vascular bundle; LN, low nitrogen; MN, medium nitrogen; HN, high nitrogen; GFP, grain filling percentage; GW, grain weight; HI, harvest index. *, p < 0.05; **, p < 0.01.

. However, vascular bundle related traits had no significant correlations with 1000-grain weight and harvest index (Table 2).

4. Discussion

The rice yield is mainly determined by the source, sink, and flow systems of the plant. Among them, great improvement has been achieved for the source and sink systems; however, little research progress has been made on the flow system. Generally, the increase of the number of vascular bundles is consistent with higher efficiency of photoassimilate transport and grain yield [12,15]. It is reported that the number of vascular bundles in the stem is significantly and positively correlated with branches of panicle, and each large vascular bundle is directly connected with a primary branch [26]. There are genotypic differences in the number of vascular bundles of peduncle in rice. It is reported that the difference in vascular bundles is one of the important parameters for the differentiation between indica and japonica [27]. In general, indica varieties tend to have a larger number of LVB in peduncle and higher V/R (number of large vascular bundles/number of primary branches) than those in japonica varieties, though the number of SVB between them is not significantly different, the V/R of indica is 1.6 to 2.0, whereas it is around 1.0 in japonica—a high V/R ratio in japonica may enhance grain yield [28,29]. However, these results are based on the studies of vascular bundle characteristics in peduncle.

In this study, there were significant phenotypic differences in vascular bundle related traits in basal internode among different types of rice varieties. Compared with the japonica variety, the number of both LVB and SVB in basal internode was significantly higher than that in indica variety, which was different from the difference in the number of peduncle LVB and SVB between indica and japonica varieties [26]. In addition, it was found that the indica variety have larger area and phloem area of single LVB and SVB than those in japonica variety, resulting in their larger total area and total phloem area of LVB and SVB than japonica variety.

Among the four types of varieties used in this study, indica-japonica hybrid variety showed the highest number of LVB and SVB, cross sectional area and phloem area of single LVB and SVB, and total area and total phloem area of LVB and SVB, those were the lowest in japonica conventional variety. Compared with the japonica conventional variety, the indica-japonica hybrid variety and indica hybrid variety showed more spikelets per panicle and thus a higher yield, and higher ATM and AR, which indicated that heterosis showed in the transport efficiency of photoassimilating to the filling grains, and the hybrid within subspecies had stronger heterosis than that in hybrid within interspecific. It was reported that the truncated dep1 could significantly increase the number of both LVB and SVB [26], and Narrow leaf1 (NAL1) and newly identified QTL (qLVN6, qSVN7, and qSVA8.1) were found to influence peduncle vascular bundle related traits in rice [12,30], and also several genes related to vascular development were reported in maize [31], which would be used to improve vascular bundle system in rice breeding. Therefore, physiological analysis and dissection of the genetic basis of vascular bundle variations in basal internode and also in peduncle and caryspsis between indica and japonica are needed in future research, which may provide a technical route to breed high-yield hybrid rice. High yield and lodging are contradictory; the characteristics of basal internodes are closely related to lodging [21]. Vascular bundles provide mechanical support for the stem. Increasing the number of vascular bundles in basal internodes could improve stem NSC translocation based on the results of this study, and thus increase grain yield. On the other hand, a great number of vascular bundles could enhance the mechanical strength of basal internodes [32], and thus improve the lodging resistance of stems. Therefore, increasing the number of basal vascular bundles could coordinate the contradiction between high yield and lodging.

Researchers had noticed that vascular bundle related traits were closely correlated with yield formation for years [33]. Evans et al. [34] showed that increasing the phloem area of vascular bundle was conducive to increasing the transport rate of assimilates. However, some researchers also believe that “flow” is not a limiting factor of yield. For example, studies in wheat and sorghum showed that the reduction of the vascular bundle or transportation capacity does not affect yield [35,36]. In recent years, genetic studies have found that the rice dep1 (dense and erect panicle1) gene controls the number of vascular bundles of peduncle and the panicle type, compared with wild-type dep1 rice plants, the number of LVB and SVB at the panicle neck of dep1 mutant rice plants has increased significantly, and the harvest index and yield have increased significantly [18,37]. These studies show that “flow” is still one of the important factors affecting rice yield. In the present study, basal internode vascular bundle related traits were positively and significantly correlated with ATM, AR, grain filling percentage, harvest index and grain yield. The varieties with a large area and a high number of vascular bundles had high ATM and AR, and thus high grain yield. Therefore, more vascular bundles in basal internode are conducive to the transport of stem NSC to grains, and consequently increase grain yield.

It is reported that the role of large and small vascular bundles of peduncle in stem NSC translocation is different; the effect of SVB on stem NSC translocation and yield formation is greater than the effect of LVB [5,15,38]. However, similar results were not observed in basal internode in this study. The phenotypes of basal internode and peduncle are different in the same plant; the basal internodes have a greater stem diameter than that of peduncle, the transport efficiency of LVB and SVB may be different. More evidence is needed to obtain a full understanding of the difference between LVB and SVB in both basal internode and peduncle.

Low nitrogen treatment increased stem NSC translocation and its contribution to yield [5,39]. Under the condition of low nitrogen application, rice accelerates the start of senescence process and promotes the translocation of stem assimilates to filling grains [40]. Under the condition of high nitrogen application, the plant respiratory consumption and the demand for carbon and nitrogen during plant establishment increase, which reduces the contribution of stem NSC to grain yield [41]. However, Deng et al. [42] showed that nitrogen application increased the translocation of stem NSC at the late filling stage of rice using polyaspartate urea. In this study, reducing fertilizer nitrogen could increase ATM and AC. These results suggested that nitrogen supply and cultivation management have different effects on NSC transport and distribution in rice stems, which may be related to rice genotypes and their nitrogen response characteristics [39,43]. The number and the cross-section area of vascular bundles increased with the level of nitrogen application in rice [20], and similar results have been reported, that nitrogen application increases the number of vascular bundles in the peduncle and the cob of the ear system in maize [44]. In addition, compared with the application of all nitrogen fertilizer as a base fertilizer, multiple applications of nitrogen fertilizer at tillering, panicle initiation and heading stage, respectively, increased the number and cross-sectional area of large and small vascular bundles in peduncle, and different application ratios had different effects on the increasing of the cross-sectional area of LVB [45]. In this study, it was observed that increasing the application of nitrogen fertilizer increased the number, cross-sectional area and phloem area of large and small vascular bundles, which was consistent with the previous results [20]. In addition, melatonin could increase rice grain yield through dual regulation of vascular development [46]. Therefore, optimizing nitrogen management and applying a growth regulator can improve vascular bundle properties and promote the NSC transport of stems and consequently increase grain yield.

5. Conclusions

The vascular bundle of basal internode is an important trait in rice production. A highly significant difference was found among different types of rice varieties in this study. The indica-japonica hybrid type had the highest number and size of large and small vascular bundles in the third internode from the top, followed by indica hybrid and indica conventional types; the japonica conventional type was the lowest. The vascular bundle related traits were significant and positively correlated with stem assimilates translocation, grain filling percentage and grain yield. Varieties with a high number, cross sectional area and phloem area of large and small vascular bundles had high stem assimilates’ translocation, grain filling percentage and grain yield. Nitrogen application improved the vascular bundle related traits of basal internode. Reducing nitrogen fertilizer application increased stem assimilates translocation. Overall, this study provides important clues for increasing rice production by improving vascular bundle system of basal internode in rice breeding practice and cultivation regulation.