WEI Hai-yan, ZHU Ying, QIU Shi, HAN Chao, HU Lei, XU Dong, ZHOU Nian-bing, XING Zhi-peng, HU Ya-jie, CUI Pei-yuan, DAI Qi-gen, ZHANG Hong-cheng

Jiangsu Key Laboratory of Crop Genetics and Physiology/Jiangsu Co-Innovation Center for Modern Production Technology of Grain Crops/Agricultural College, Yangzhou University, Yangzhou 225009, P.R.China

Abstract There is limited information about the combined effect of shading time and nitrogen (N) on grain filling and quality of rice.Therefore, two japonica super rice cultivars, Nanjing 44 and Ningjing 3, were used to study the effect of shading time and N level on the characteristics of rice panicle and grain filling as well as the corresponding yield and quality. At a low N level(150 kg N ha-1, 150N), grain yield decreased (by 21.07-26.07%) under the treatment of 20 days of shading before heading(BH) compared with the no shading (NS) treatment. These decreases occurred because of shortened panicle length,decreased number of primary and secondary branches, as well as the grain number and weight per panicle. At 150N, in the treatment of 20 days of shading after heading (AH), grain yield also decreased (by 9.46-10.60%) due to the lower grain weight per panicle. The interaction of shading and N level had a significant effect on the number of primary and secondary branches. A high level of N (300 kg N ha-1, 300N) could offset the negative effect of shading on the number of secondary branches and grain weight per panicle, and consequently increased the grain yield in both shading treatments. In superior grains, compared with 150N NS, the time to reach 99% of the grain weight (T99) was shortened by 1.6 to 1.7 days, and the grain weight was decreased by 4.18-5.91% in 150N BH. In 150N AH, the grain weight was 13.39-13.92% lower than that in 150N NS due to the slow mean and the maximum grain-filling rate (GRmean and GRmax). In inferior grains, grain weight and GRmean had a tendency of 150N NS＞150N BH＞150N AH. Under shaded conditions, 300N decreased the grain weight due to lower GRmean both in superior and inferior grains. Compared with 150N NS, the milling and appearance qualities as well as eating and cooking quality were all decreased in 150N BH and 150N AH. Shading with the high level of 300N improved the milling quality and decreased the number of chalky rice kernels, but the eating and cooking quality was reduced with increased chalky area and overall chalkiness. Therefore, in the case of short term shading, appropriate N fertilizer could be used to improve the yield and milling quality of rice, but limited application of N fertilizer is recommended to achieve good eating and cooking quality of rice.

Keywords: shading time, N levels, grain filling, rice quality, japonica super rice

1. Introduction

Rice (Oryza sativa L.) is one of the most important staple foods throughout the world, providing much of the energy,protein, and nutrients for more than half of the world’s population. With population increases and improved living standards, more and more rice varieties with high yield and quality are needed (Peng et al. 2009; Calingacion et al.2014). To meet the food requirements all over the world,“super rice” began to be bred by several major international and national institutions in 1980 (Higashi 1987; Chen et al.2017). So far, 130 super rice cultivars have been bred in China and 21.5% are inbred japonica super rice with high yield and relatively good quality (http://www.super-rice.com/index.asp).

Previous studies have revealed that both light and N play an important role in the growth of rice. Plenty of light increases dry matter accumulation (Wang L et al. 2015)and grain yield (Singh 2005). Lack of light (shading) lowers the photosynthetic rate of leaves (Jiang and Cheng 2015;Wang Q et al. 2015) and reduces the activities of enzymes participating in photosynthesis and nutrient metabolism(Fu et al. 2009; Ishibashi et al. 2014). In low light, the tiller number per square meter, leaf area index, root activity,carbohydrate accumulation, and even the grain yield (Wang et al. 2014a, b) were decreased while the rate of white immature kernels was increased (Yoshida and Hara 1977).N application before panicle initiation not only increases the spikelet number (Ding et al. 2014) per panicle but also helps to increase the grain-filling rate, shorten the filling time, and consequently increase the grain weight and yield. According to the results of previous studies, the effects of N on the indexes of rice quality differed. The application of N could increase the milling recovery (Leesawatwong et al. 2005),ratio of head rice, and the content of amino acids (Ning et al.2010) and protein (Ning et al. 2009) in brown and milled rice. For physical appearance, an appropriate amount of N fertilizer decreased the rate of chalky kernels (Zhou et al.2015) and overall chalkiness (Perez et al. 1996), but overuse of N increased the rate of chalky grains and thus undesirable grain appearance. The apparent amylose content and gel consistency were mostly decreased with the increase of N level (Zhu et al. 2017), which in turn degraded the eating and cooking quality of rice (Champagne et al. 2009).

The middle and lower reaches of the Yangtze River are one of the most important production regions of japonica rice in China; about 50% of inbred japonica super rice is bred and used in this area (Wang et al. 2017). However, to obtain very high rice yield, overuse of N fertilizer has become widespread in this region. The average rate of N fertilizer even reached 387 kg ha-1during the period of 2004-2008 in Jiangsu Province of China, which is almost two times of the world average (Chen et al. 2011; Peng et al. 2011).Meanwhile, more than 6% reduction of solar radiation per decade was also observed in this region due to precipitation,the number of clouds, and air pollution (Liang and Xia 2005;Deng et al. 2012). Therefore, it is urgent to study the effects of high N and low light (shading) on the growth of rice to effectively manage rice growth under these conditions.

The combined effect of shading and N on rice has been the focus of a limited number of studies. Results showed that long-term shading combined with a high level of N resulted in taller plants and greener leaves (Samarajeewa et al. 2005) with larger area (Sims et al. 2012), and unfavorable delayed senescence and yield loss (Huang et al. 2013). Short-term shading also decreased nutrient absorption and grain yield (Wang et al. 2014b), but a proper application of N could compensate for some adverse effects of short-term shading on rice (Wang et al. 2014a; Pan et al.2016). However, most previous studies focused on the effect of shading and N on the growth and yield of rice, but little is known about their combined effect on panicle size and grain-filling of rice, which are important factors influencing grain yield and quality. In this study, we used two japonica super rice cultivars that are popular in the middle and lower reaches of the Yangtze River. Treatments included shading before and after heading combined with N application rates of 150 and 300 kg ha-1. The objective of this study was to investigate the combined effect of shading time and N application rate on the characteristics of the rice panicle and grain filling as well as the corresponding yield and quality.

2. Materials and methods

2.1. Plant materials and growth conditions

Nanjing 44 and Ningjing 3 cultivars, which have been certified as super japonica rice by Ministry of Agriculture of China, were adopted as materials in this study. Field experiments were conducted on a research farm of Yangzhou University in Jiangsu Province, China (32°30′N(xiāo),119°25′E) during the rice growing season (May-October)of 2012 and repeated in 2013. The weather data during the growing season of rice measured at a weather station close to the experimental site are shown in Table 1.The soil of the field was sandy loam with 0.13% total N,87.45 mg kg-1alkali hydrolysable N, 32.8 mg kg-1Olsen-P,88.3 mg kg-1exchangeable K, 2.09% soil organic matter,and pH value of 6.88.

The sowing date was 17 May in both years. Seedlings were raised in the seedbed and transplanted on 14 June in both years. The hill spacing of rice was 14.4 cm×26.0 cm with two seedlings per hill. The experiment was laid out in a split-split plot design with three replications. The main plots were two rice cultivars, and the subplots consisted of three shading treatments including no shading (NS),20 days of shading before heading (BH), and 20 days of shading after heading (AH). A black net was used to cover the rice field in shading treatments and the light intensitiesof shading treatment at 08:00, 10:00, 12:00, 14:00, and 16:00 in 1 day were about 50% of the unshaded treatment.The sub-subplots were two N application levels, including 150 kg ha-1(150N) and 300 kg ha-1(300N). N was applied as urea in four splits: 25% on 13 June as basal fertilizer,25% at 7 days after transplanting, 25% when the rice has four leaves that have not appeared, and 25% when the rice has two leaves that have not appeared. The plot area was 15 m2(3 m×5 m). Each plot was separated by a ridge that was 35 cm wide and 20 cm high covered with plastic film.Basal fertilizers of 150 kg P ha-1(as super-phosphate) and 150 kg K ha-1(as KCl) were also applied on 13 June in both years. The rice field was watered after transplanting until 7 days before maturity. Diseases, insects, and weeds were controlled to avoid yield loss and decreased quality.

Table 1 The weather data during the growing season of rice in 2012 and 2013

2.2. Sampling and measurements

Determination of grain fillingAbout 450 panicles that headed on the same day were labeled and recorded in each plot. The flowering date and position of each spikelet on labeled panicles were also recorded. 10-15 labeled panicles of each plot were sampled every 5 days from anthesis to maturity stages. Superior spikelets that flowered on the first 2 days (located on apical primary branches) within a panicle and inferior spikelets that flowered on the last 2 days(located on proximal secondary branches) within a panicle were separated from the sampled panicles according to the method of Wang Z Q et al. (2015). About 100 sampled spikelets (grains) were dried at 70°C to constant weight,dehulled, and weighed.

The processes of grain filling were fitted by Richards’growth equation (Richards 1959) as described by Zhu et al. (1988):

Where, W is the grain weight, A is the maximum grain weight (mg per grain), t is the time after anthesis (days),and B, K, and N are coefficients determined by regression.

The grain filling rate (GR), the average grain filling rate(GRmean), the maximum grain filling rate (GRmax), the time to reach the maximum grain filling rate (Tmax), the grain weight with the maximum grain filling rate (Wmax), and the time to reach 99% of the maximum grain weight (T99) were calculated as the derivative of Richards’ growth equation as follows:

Determination of panicle characteristicAt maturity stage,five hills containing plants were sampled randomly from each plot to determine the panicle characteristics including panicle length, grain number, and weight per panicle, and number of primary and secondary branches per panicle.

Rice qualityFifty hills of rice plants in each plot were harvested at maturity stage. Quality traits, including brown rice, milled rice, and head rice rates, amylose content and gel consistency were measured according to GB/T17891-1999 (1999).

Rice grains were dried to a standard moisture content of 14% and stored for more than 3 months. Grain samples (120 g)with three replications from each plot were collected for grain quality analysis. Rice grains were passed through a dehusker to obtain brown rice. The brown rice was polished for milled rice. Head rice with grain length greater than or equal to 3/4 of its total length was separated from milled rice manually. The brown rice rate, milled rice rate, and head rice rate were expressed as the percentages of total rough rice weight.

One hundred milled grains per plot were selected randomly to test the appearance quality manually according to China’s National Standards (GB/T17891-1999 1999).Grains containing white belly, white center, and white back or a combination of these were considered chalky kernels.The chalky kernel rate, chalky area, and overall chalkiness were calculated using the following formulas:

Chalky kernel rate (%)=Number of chalky kernels/100 milled grains×100

Chalky area (%)=Area of chalky/Total area of rice kernel×100

Chalkiness (%)=Chalky kernel rate×Chalky area

Milled rice was ground into flour with a stainless-steel grinder (SCINO CT410, Foss Analytical A/S Denmark). Flour sifted with 0.25-mm sieves was prepared to test the amylose content using the amylose-iodine reaction. The protein content of the rice flour samples was determined by an automatic azotometer with the Kjeldahl Method (Kjeltec 8400,Foss Analytical A/S Denmark), and the measured value was multiplied by 5.95 to convert to crude protein content. The gel consistency was analyzed according to China’s National Standards (GB/T17891-1999).

Pasting viscosityRice pasting properties were determined using a Rapid Visco Analyser (RVA, Super3, Newport Scientific, Australia), following the procedure of the American Association of Cereal Chemists. A total of 3 g of flour sifted with 0.15-mm sieves was mixed with 25 g deionised water in the RVA sample can. The peak viscosity, hot viscosity,cool viscosity in cP (centipoise) units and their derivative parameters breakdown (peak viscosity-hot viscosity),setback (cool viscosity-peak viscosity), and consistency(cool viscosity-hot viscosity) were recorded with matching Software of Thermal Cline for Windows (TCW).

Rice yieldRice yield was determined from a harvest area of 6.0 m2in each plot and adjusted to the standard moisture content of 0.14 g H2O g-1.

2.3. Statistical analysis

Data were analyzed using analysis of variance (ANOVA)with SPSS 22.0 for Windows. The means were compared by the least significant difference (LSD) test at the 0.05 probability levels.

3. Results

3.1. panicle characteristics

Panicle characteristics and grain yield of both japonica super rice cultivars varied with different shading and N level treatments (Table 2).

Compared with 150N NS, the panicle length, numbers of primary and secondary branches, grain number and weight per panicle, and the grain yield of rice were reduced by 8.10-8.70%, 7.13-9.15%, 30.24-33.34%, 17.75-29.46%,20.06-24.61%, and 21.07-26.07%, respectively, in the treatment of 150N BH. In BH, 300N increased grain yield by improving panicle characteristics.

In 150N AH, the decreased grain weight per panicle resulted in 9.46-10.60% reduction of grain yield. Compared with 150N AH, the panicle length, the numbers of primary and secondary branches, the grain number and weight per panicle, and grain yield were increased by 3.76-5.67%,6.30-10.98%, 19.59-23.79%, 10.11-20.79%, 9.27-10.29%,and 30.29-36.38%, respectively, in the treatment of 300N AH.

Both N level and shading treatment had significant effects on the grain yield and panicle characteristics of rice. In addition, an interaction effect of N level and shading treatment on numbers of primary and secondary branches with grain yield was observed.

3.2. Dynamics of grain weight and grain-filling rate

The weight of superior grains increased rapidly after anthesis in all treatments (Fig. 1). From 0 to 20 days after anthesis, the superior grain weight of 150N BH was almost the same as that of 150N NS, but 150N BH had a lower grain weight after the 20th day. The grain weight in 300N BH was decreased at 25 days after anthesis stage compared with 150N BH. In 150N AH, the superior grain weight was significantly lower than that of 150N NS or 150N BH. The grain weight of 300N AH was lower than that of 150N AH.

In the inferior rice grains, the grain weight increased slightly from 0 to 20 days after anthesis stage, therefore,there was little difference in the grain weight among 150N NS, 150N BH, and 150N AH treatments during this period. When the superior grain had nearly attained its maximum weight, the weight of inferior grain increased rapidly. Compared with 150N NS, the inferior grain weight of 150N BH was lower at 40 days after anthesis stage. In 150N AH, the inferior grain weight was lower than 150N NS at 25 days after anthesis stage. Regardless of shading,300N treatment decreased the weight of inferior grains.

The determination coefficients of Richards’ growth equation in different treatments were about 0.99 (Table 3),which means the Richards’ growth equation described the dynamics of grain weight and grain filling rate very well.

The grain filling of superior spikelets started before that of inferior spikelets (Fig. 2). Compared with 150N NS, in 150N BH, the time to reach 99% of the grain weight (T99)was shortened by 1.6-1.7 days, and the grain weight (A) was decreased by 4.18-5.91% (Table 3). In 150N AH, although T99was longer, the grain weight was still 13.39-13.92% lower than that in 150N NS due to the slow mean and maximum grainfilling rates (GRmeanand GRmax). Compared with the treatments of shading under 150N,the interaction between shading time and 300N decreased the grain weight with lower GRmax, GRmean, and Wmax.

The grain-filling rate of inferior grain increased rapidly after the superior grain reached its maximum grain-filling rate. The grain weight (A), GRmax, GRmean, and Wmaxof inferior grains showed a decreasing tendency of 150N NS＞150N BH＞150N AH. With shading treatment, the high level of 300N decreased grain weight with a slow GRmean.

3.3. Milling quality

The milling quality of rice was decreased due to shading but was improved through the appropriate application of N fertilizer(Table 4).

Compared with 150N NS, only the rate of head rice was significantly decreased in the treatment of 150N BH. In BH, 300N increased the milling quality, particularly for the percentage of head rice.

In the treatment of 150N AH, the rates of brown rice, milled rice, and head rice were 4.47-5.62%, 14.63-15.62%, and 21.42-23.66%, respectively lower than those of 150N NS. In AH, the milling quality was also improved through the application of more N fertilizer.

Both N level and shading treatment had a significant effect on the milling quality of rice. An interaction effect between N level and shading treatment on the rate of milled rice was observed.

3.4. Appearance quality

Compared with 150N NS, the chalky kernel rate, chalky area, and overall chalkiness of japonica super rice in 150N BH were increased by 16.89-32.27%, 37.72-67.03%,and 61.16-146.97%, respectively (Table 5).When rice was shaded before heading,300N decreased the rate of chalky kernels.Compared with 150N BH, the chalky area and chalkiness of rice were increased by 38.37-48.69% and 27.13-35.18%, respectively, in the treatment of 300N BH.

Fig. 1 Grain weight of superior and inferior spikelets of rice. 150N, 150 g N ha-1; 300N, 300 g N ha-1. NS, no shading; BH, shading at 20 days before heading stage; AH, shading at 20 days after heading stage; SG and IG, superior grain and inferior grain, respectively.

In the treatment of 150N AH, the chalky kernel rate,chalky area, and chalkiness of rice were 55.48-71.27%,100.85-119.93%, and 212.28-276.56%, respectively higher than those of 150N NS. Compared with 150N AH,the rate of chalky kernel was decreased by 7.92-9.15%,while the chalky area and chalkiness of rice were increased by 25.25-29.28% and 14.27-18.33%, respectively,in 300N BH.

Both N level and shading treatment had significant effects on the appearance quality of rice. An interaction effect of N level and shading treatment on the chalky area of rice was observed.

3.5. Eating and cooking quality

Compared with 150N NS, rice from the 150N BH treatment had lower amylose content and gel consistency and higher protein content (Table 6). When rice was shaded before heading, the amylose content and gel consistency were decreased by 7.63-9.44% and 11.70-17.28%, respectively,while the protein content was increased by 9.45-14.19%with the application of 300N.

In the treatment of 150N AH, the amylose content and gel consistency were 7.85-11.16% and 13.33-15.39%lower, respectively, and the protein content was 14.08-23.33% higher than the treatment of 150N NS. When rice was shaded after heading, 300N decreased the amylose content and gel consistency and increased protein content compared with 150N.

Both N level and shading treatment had significant effects on the content of amylose, protein, and gel consistency. An interaction effect of N level and shading treatment on the gel consistency of rice was observed.

3.6. Starch viscosity

The starch viscosities were measured in 2012 and 2013 with good repeatability and stability. Only the data from 2012 are described here.

Compared with 150N NS, the peak viscosity, hot viscosity,cool viscosity, breakdown, and consistency were decreased while the setback was increased in the treatment of 150N BH (Table 7). These effects were further enhanced in the treatment of 300N BH.

In the treatment of 150N AH, the peak viscosity, hot viscosity, cool viscosity, breakdown, and consistency were lower than those in 150N NS. When plants were shaded after heading, starch viscosities, breakdown, and consistency had larger decreases at 300N.

Fig. 2 Grain filling rate of superior (SG) and inferior (IG) spikelets of rice. The grain filling rate was calculated according to Richards’equation (Richards 1959). 150N, 150 g N ha-1; 300N, 300 g N ha-1. NS, no shading; BH, shading at 20 days before heading stage;AH, shading at 20 days after heading stage.

Significant effects of N level or shading treatment on the starch viscosities, breakdown, setback, and consistency were observed. Meanwhile, the interaction of N level and shading treatment had significant effects on the peak viscosity, hot viscosity, cool viscosity, and setback.

4. Discussion

4.1. Combined effect of shading time and nitrogen level on rice yield and grain filling

The effect of shading on rice panicle size and grain filling depends on the initial time and its duration (Samarajeewa et al. 2005). Twenty days before heading stage is the critical stage of panicle initiation (Ling et al. 1994). In this study,shading during this period resulted in poor development and reduced the size of rice panicles (Table 2). Meanwhile,the area of three leaves from the top at heading stage were also significantly decreased because of shading, which was reported by our previous study (Wang et al. 2014a).The restrained leaf source reduced the average and the maximum grain filling rate and final grain weight of BH compared to NS (Table 3), particularly in inferior grains. After heading stage, the panicle size was no longer influenced by shading. However, shading after heading stage decreased the photosynthetic rate of leaves and constrained the available carbohydrate for grain development (Jiang and Cheng 2015). Therefore, in our study, the average or the maximum grain filling rate and final grain weight of AH were significantly lower than that of NS both in superior and inferior grain (Fig. 1 and Table 3). Jiang et al. (2013)found that when rice was shaded at tilling, elongating, and after heading stages, the greatest yield loss was observed in the treatment of shading after heading stage; similar results were also obtained by Cai and Luo (1999). However, our results revealed that the treatment of BH suffered a heavier yield loss in comparison with the treatment of AH (Table 2).The discrepant results of these studies may be due to the different durations of shading and types of rice cultivars(Li et al. 1999; Okawa et al. 2003). In this experiment, the influence of BH on the number of branches and spikelets per panicle was irreparable. The reduced sink capacity will inevitably decrease the yield of rice even with an adequate supply of carbohydrate after heading. The duration of AH treatment was only 20 days in the present study and the yield loss could be compensated to a certain extent by extending the time of grain filling with high light levels andhigh temperature at later growth stages (Tables 1 and 3,Wang et al. 2014a).

Table 4 Combined effect of shading time and nitrogen (N) application on milling quality of japonica super rice

Table 5 Combined effect of shading time and nitrogen (N) application on appearance quality of japonica super rice

The interaction of shading and N levels has significant effects on the yield formation of rice (Table 2). A high level of N could offset the negative effect of shading on the number of secondary branches and grain number per panicle and increase the yield of rice. Compared with superior grains,the maximum weight and mean filling rate of inferior grains were much more impacted by the interaction of shading and high N level (Table 3), which may be becausephotoassimilation to grains varies largely with their position on a panicle (Yoshida and Hara 1977). Inferior grains usually get less assimilated carbohydrate due to later flowering and grain filling and therefore grain filling and weight of inferior grains are more easily influenced by stressed conditions(Lisle et al. 2000; Zhang et al. 2012).

Table 6 Combined effect of shading time and nitrogen (N) application on eating and nutritional quality of japonica super rice

Table 7 Combined effect of shading time and nitrogen (N) application on starch viscosity of japonica super rice (2012)

4.2. Combined effect of shading time and nitrogen level on grain quality of rice

Our previous study showed that both the limited leaf area in BH treatment and the weakened light intensity in AH treatment reduced photosynthesis during heading to maturity stages (Wang et al. 2014a). The inadequate supply of carbohydrate resulted in a slow grain-filling rate, less plump grains (Table 3) and poor milling quality(Table 4). The appearance qualities of shading treatments were also worse (Table 5); these results are in agreement with previous studies reporting that poor grain filling due to stressed conditions was usually accompanied by more chalky grains because of the abnormal development and loosely packed starch granules in grains (Patindol and Wang 2003). Eating and cooking quality of rice is determined by several factors including apparent amylose content, protein content, gel consistency, and starch viscosity (Ramesh et al. 2000; Aluko et al. 2004; Tian et al. 2005; Bao et al.2006a, b). Cooked rice with low amylose content, high gel consistency, high peak viscosity, and small value of setback(Beckles and Thitisaksakul 2014) tended to be sticky, soft,glossy, with slow retrogradation and good taste (Bao 2012),while high content of protein usually leads to firmer texture of rice because starch is wrapped with more protein and could not easily absorb enough water for complete gelatinization(Martin and Fitzgerald 2002). In this study, compared with the treatment of NS, the eating and cooking quality of rice decreased in BH and AH treatments due to the increased protein content, lower gel consistency, and decreased starch viscosity.

The interaction of shading time and N levels had significant effects on the quality of rice. Under shaded conditions, with the application of 300 N, the milling quality of rice was improved. The rate of chalky kernels was decreased while the chalky area and chalkiness of kernels were increased(Table 5), which means the chalky area of individual chalky kernels was dramatically enlarged. In general, cooked rice with good taste is usually low in amylose content. In our study, although a high level of N combined with shading decreased the amylose contents of rice, results from the protein content, gel consistency, and starch viscosity all led to degraded eating and cooking quality. The possible reason could be the unbalanced relationship between carbon and N metabolism (Lawlor 2002; Xi et al. 2016). During the grain filling stage, the metabolisms of carbon and N are two major physiological processes with end-products of starch and protein, respectively (Lawlor 2002; Xi et al. 2016). In the progress of grain filling, the starch content was increased(Liang et al. 1994; He et al. 2003) while the protein content was decreased (Lin et al. 2010). This could be caused by the reduced photosynthetic production because of limited leaf area in BH treatment (Wang et al. 2014a) or because the weakened light intensity in AH treatment inhibited the carbon metabolism while the application of more N improved the N metabolism, which resulted in low amylose and high protein content of rice. Similar results have been also obtained when the balance between carbon and N metabolism was disturbed by stressed conditions (Ahmed et al. 2015 ).

Therefore, according to the combined effect of shading time and N levels on the grain filling and quality in japonica super rice, whether high N application is needed or not under shading conditions depends on the requirement of rice growers. In shaded conditions, obtaining high yield and improved milling quality of rice requires an appropriate increase in N fertilizer. However, to achieve good eating and cooking quality of rice, limited application of N fertilizer would be recommended.

5. Conclusion

Compared with 150N NS, the grain yield of 150N BH was decreased by 21.07-26.07% because of reduced panicle size and weight, while yield of 150N AH was decreased by 9.46-10.60% mainly due to the reduced grain weight per panicle. Shading combined with 300N increased the number of secondary branches, grain number, and weight per panicle, and consequently improved the yield of rice. In superior grains, compared with 150N NS, the grain weight of 150N BH was decreased by 4.18-5.91% because of shortened T99of 1.6 to 1.7 days, while the grain weight was reduced by 13.39-13.92% in 150N AH due to the slow filling rate of grain. Shading combined with 300N decreased the grain weight due to lower GRmean, particularly in the treatment of 300N AH. Under 150N, the milling and appearance qualities as well as the eating and cooking quality were all decreased with shading. Higher N under shading conditions could improve the milling quality and decrease the rate of chalky kernel of rice, while the eating and cooking quality decreased with increased chalky area and chalkiness.

Acknowledgements

We are grateful for grants from the National Key Technology R&D Program of China (2016YFD0300503), the Key Research Program of Jiangsu Province, China (BE2016344),the earmarked fund for China Agriculture Research System(CARS-01-27), the National Nature Science Foundation of China (31701350), the Program for Scientific Elitists of Yangzhou University, China, and a project funded by the Priority Academic Program Development of Jiangsu Higher Education Institutions, China.