Min Yin, Showen Liu,Xi Zheng,Gung Chu, Chunmei Xu, Xiufu Zhng,Dngying Wng,, Song Chen,

aChina National Rice Research Institute,Chinese Academy of Agricultural Sciences,Hangzhou 310006,Zhejiang,China

bThe Faculty of Agriculture,Life and Environmental Sciences,Zhejiang University,Hangzhou 310029,Zhejiang,China

Keywords:

ABSTRACT New indica and japonica hybrid rice cultivars, such as the Yongyou series, provide farmers with very high yield potential.However,information on their canopy light capture and solar radiation use efficiency in the late season is limited.Field experiments were performed to compare the radiation-use parameters of four rice types:indica rice(IR),inbred japonica rice(IJR),hybrid japonica rice(HJR),and hybrid indica/japonica rice(HIJR),from 2016 to 2018 during the late season in Hangzhou, China.The grain yield, aboveground biomass, intercepted solar radiation (SI), and radiation-use efficiency (RUE) of the HIJR were on average respectively 13.4%-53.4%, 14.3%-30.6%, 7.6%-21.4%, and 8.2%-14.9% higher than those of the HJR,IJR,and IR.The leaf area index(LAI)of the HIJR was 18.2%-57.0%greater than that of the IJR and HJR at four growth stages,resulting in respectively 17.8%-38.5%and 10.7%-42.8%greater canopy light interception rates (LIR) and amount of intercepted solar radiation during the vegetative stage.The prolonged grain-filling stage also led to respectively 33.9%-52.6% and 30.5%-51.4% increases in amounts of incident and intercepted radiation for the HIJR relative to the IR during grain filling.These results indicate that the SI superiority of the HIJR was caused by canopy closure as rapid as that of the IR during the vegetative stage(greater LAI and canopy LIR during the growing season) and a grain-filling stage as long as that of the HJR.For grain-filling stage, differences in leaf Pn between HIJR, IR, and IJR were not significant,suggesting that the greater RUE of the HIJR(12.7%-52.8%higher)than that of the other rice types resulted from improved canopy architecture after flowering (FL).Principal components analysis (PCA) revealed that the superiority of the HIJR in terms of solar radiation use resulted from the greater canopy light capture capability of IR and the prolonged growth period (especially during grain filling) of japonica rice in the late growing season.

1.Introduction

Rice (Oryza sativa L.) is a major staple food worldwide [1].Faced with an increasing demand for rice,China established a“super rice” mega-project in 1996 to breed rice varieties with high yield potential [2].The use of heterosis resulting from crossing japonica with indica parents is considered a potential pathway to increasing aboveground biomass and grain yield[3,4].However, owing to poor grain filling [5]and sterility obstacles [6]that arise in the F1generation, the promotion of hybrid indica/japonica rice(HIJR)for large-scale production has been limited.Encouragingly, breeding progress has been made in recent years to overcome these problems, and highyielding HIJRs, such as the Yongyou indica/japonica hybrid series,have been made available for production[7-10].

To date, these HIJRs have been grown largely near the lower reaches of the Yangtze River,China[7],and their yields are 5%-20% greater than those of elite hybrid indica rice (IR)[7,8].HIJR such as Yongyou 12 yielded more than 13.5 t ha?1on farms over two years in this region [9].The cumulative planting area of the Yongyou japonica/indica hybrid series in the region reached 1 million ha in 2012 [11].Several morphological and physiological traits including large panicle size [12], large leaf area [10,13], prolonged leaf area duration[14],high nutrient uptake[8,13]and longer,larger deeper,and more active roots after heading [11]have been reported to underlie the yield superiority of HIJR relative to previous elite rice cultivars.However, information about the difference in solar radiation use efficiency and canopy light interception between the HIJR and other types of rice cultivar is lacking.

According to the principles of Monteith and Moss [15],aboveground biomass can be defined as TDW=SR×LIR×RUE,where SR is incident radiation and LIR is the fraction of light intercepted by the canopy.Radiation-use efficiency(RUE)can be defined as micromoles of dry matter produced per mole of photosynthetically active photons absorbed by green canopy components [16].Thus, improved aboveground biomass is closely associated with increased incident solar radiation[15],greater canopy radiation capture [17], and enhanced leaf photosynthesis traits[16].Previous reports have documented the responses of traits associated with the superiority of yield and biomass production of HIJR compared with inbred japonica rice (IJR) and elite indica (IR) and/or hybrid japonica rice (HJR)[11,12].However, understanding of trait integration for solar radiation use and biomass production is limited.The main objectives of this study were to 1) evaluate the differences in solar radiation use (capture rate and use efficiency) among HIJR, IJR, HJR, and IR and 2) determine the profiles of solar radiation characteristics at different growth stages, including solar use efficiency, canopy light capture, and intercepted solar radiation, for HIJR,IJR,HJR, and IR.

2.Materials and methods

2.1.Experimental site

Field experiments were performed in 2016-2018 at the experimental station of the China National Rice Research Institution,Hangzhou,China(30°05′N,119°90′E,21 m altitude).The field soil was a ferric-accumulic stagnic anthrosol.The soil properties were as follows:pH,6.5;soil organic matter,18.3 g kg?1;total N,1.98 gkg?1;availableP,29.4 mgkg?1;andavailable K,35.6mgkg?1.Meteorological data were obtained from the Chinese National Meteorological Information Center: http://data.cma.cn/site/index.html.The mean temperature, daily solar radiation, and rainfall for 2016-2018 are shown in Fig.S1.

2.2.Rice cultivars and experimental design

The rice cultivars used in the current study are listed in Table S1 and their agronomic traits are described in Tables S2-S4.These cultivars were chosen based on recommendations from professional rice breeders,local rice seed distributors,and large-scale rice growers.Ecological adaptability, suitable growth period,and pest and disease resistance were favored.In 2016,the HJR cultivars were not evaluated because the growing area of commercial cultivars of hybrid japonica was limited and because seeds were difficult to acquire on the market.

Field experiments followed a randomized block design with three replications.Each plot covered at least 25 m2during each year.Pregerminated seeds were sown in seedbeds on June 23(±4) and transplanted on July 20, July 14, July 19, and July 15 in 2016, 2017, 2018, and 2019, respectively.The hill spacing was 25.0×16.7(2016-2017),26.7×13.3(2018)and 25×16(2019)cm,with 1-2 seedlings per hill for hybrid rice and 2-4 seedlings per hill for inbred rice.The total N application rate was 202.5 kg ha?1;50% was applied 1 day before transplanting (N:P2O5:K2O 15:15:15; compound fertilizer), 30% was applied 14 days after transplanting(DAT)(urea),and 20%was applied at the panicle initiation(PI)stage(urea).Approximately 100.0 kg ha?1P2O5and K2O in the form of compound fertilizer were applied as basal fertilizer one day before transplanting, and an additional 65.0 kg ha?1K2O in the form of potassium chloride was applied as a topdressing at PI.Pest and weed control and irrigation were applied as needed to avoid yield losses.

2.3.Sampling and measurements

In 2016-2018, 12 hills were sampled from each plot at the tillering stage (TS, approximately 15-20 DAT), panicle initiation (PI), flowering (FL), mid-grain filling (15 days after flowering,DAF),and physiological maturity(MS).The heading stage was defined as the date when 95% of the panicles had emerged,and the plant reached MS when 95%of the spikelets had turned yellow.

The samples were separated into green leaf blades, dead leaf blades, culms plus leaf sheaths, and panicles and then oven dried for 72 h at 75°C before determining the dry weight(DW) of the constituent organs.The aboveground biomass was calculated by summing the DWs of all organs.

Leaf area was measured with a leaf area meter (LI-3000A,LI-COR, Lincoln, NE, USA).The leaf area index (LAI) was calculated based on the area of green leaf blades,and the leaf area per tiller was calculated by multiplying the LAI by the planting density and dividing by the number of tillers.The net photosynthetic rate (Pn) of the flag leaf of representative cultivars of IR (Huanghuazhan and Tianyouhuazhan), IJR [Jia 58 and Xiushui 134(2018)or Wuyugeng 6567(2019)],and HIJR(Yongyou 538 and Yongyou 540)was measured with a portable photosynthesis system(LI-6400X,LI-COR)from 0900 to 1030 h on sunny days in 2018 and 2019.The Pnwas measured under a light intensity of 1500 μmol m?2s?1, ambient temperature, a constant CO2concentration of 380 ± 5 μmol mol?1, and a relative humidity of 75%±5%in the sample chamber.

Incident solar radiation (SR) was recorded daily using a self-recording tintype meteorological station (CR800, Techno Solutions Ltd., Beijing, China) with a pyranometer (LI-200R,400-1100 nm,LI-COR).The observed canopy light interception rate(LIRa)was measured between 0900 and 1130 h at intervals of respectively 7-10 days and 15-20 days before and after heading, using a linear photosynthetically active radiation ceptometer (AccuPAR LP-80, Decagon Devices, Pullman, WA,USA).In each plot, the light bar was placed above the canopy and 10-20 cm above the water or soil surface in succession to measure the light intensity above and within the canopy,respectively.The daily LIR values between the two measurement dates were estimated by linear interpolation.The amount of intercepted solar radiation (SI) was calculated by multiplying the daily LIR value by the SR.The average LIR for a target period was defined as the cumulative intercepted SR divided by the total incident SR during the sampling intervals.

The cumulative total DW was plotted against the cumulative intercepted radiation (PARi) for a selected growth period,and the RUE values were calculated from the slopes of the regression lines by the formula RUE=DW/∑PARi [15].

2.4.Data analysis

Tables and figures were processed with Microsoft Excel 2010(Microsoft, Redmond, WA, USA) or Origin 2020b (OriginLab,Northampton,MA,USA).All data were subjected to analysis of variance (ANOVA) using SAS 9.4 statistical software (SAS Institute Inc., Cary, NC, USA).Two-way ANOVA was used to compare the effects of rice type and year with a general linear model (GLM), for example on LIR, RUE, and cumulative SI.One-way ANOVA was performed to evaluate the effects of rice type on yield and solar radiation-use parameters.Means were subjected to Tukey’s honestly significant difference(HSD)test at the 0.05 probability level.

All parameters for the different rice cultivars were subjected to principal components analysis (PCA) to identify patterns within the rice growth and solar radiation-use parameters by dimensionality reduction.The loading graph reflects the correlations between the principal components(PCs) and the initial parameters, with greater loadings corresponding to stronger correlations.The main characteristics of the different rice types can be analyzed according to the loading graph of the PCs.PCA and cluster analysis were performed using factor analysis and data mining in R (Bell Laboratories, UoA,Auckland, New Zealand).

3.Results

3.1.Grain yield

There was significant interaction between year and rice type in yield (Fig.1).Mean grain yields were ordered as follows:HIJR ＞ HJR ＞ IR ＞ IJR, and the yield was 13.4%-53.4%significantly higher for the HIJR than for the other rice types(IR, IJR, and HJR) from 2016 to 2018, except for 2017, in which grain yield did not differ significantly between the HIJR and IR.Although the yield of IJR was the lowest, the yield variation among the cultivars showed that the IJRs could also achieve high yield, such as 8.5, 8 and 8.0 t ha?1in 2018 in Jia 58,Nangeng 46 and Xiushui 134,respectively(Table S5).

3.2.Solar radiation-use parameters of rice grown during the full season

The RUE and its related parameters, including incident radiation, intercepted radiation, light interception rate and aboveground biomass, of rice cultivars grown from 2016 to 2018 are presented in Tables 1-3.The mean RUE values were ranked in the order HIJR ＞ HJR ＞ IR ＞IJR.Although the differences in RUE among the HIJR,HJR,IJR and IR in 2017 were not statistically significant, the mean RUE of the HIJR was significantly 8.2%-14.9% greater than that of the IR, IJR, and HJR in 2016 and 2018,except for the IR in 2018.RUE values for the four rice types were variable across year and cultivar,and abnormal values for specific rice types also occurred, for example,Nangeng 9108 in 2017 for IJR,Chunyou 84 in 2017 for HIJR,and Xiushui134 in 2017 for IJR.However,these abnormal values did not affect the general trends of differences between the HJIR and the other rice types.

Values are expressed as means±SD,and within a column,the means followed by the same letters are not significantly different according to Tukey’s HSD test (P＜0.05).RUE,radiation-use efficiency; IR, IJR, HJR, and HIJR refer to indica rice,inbred japonica rice,hybrid japonica rice and hybrid indica/japonica rice,respectively.

Compared with the HJR, IJR, and IR, the HIJR had significantly greater aboveground biomass (14.3%-30.6%, except for HJR in 2017) and intercepted solar radiation (7.6%-21.4%), while the IJR had the lowest value among the rice types.The total incident solar radiation for the HIJR reached 2098 MJ m?2on average, which was 2.8% and 7.7% greater than that for the IJR and IR, respectively, but there was no difference between the HIJR and HJR.The mean light interception rate was highest for the HIJR (61.1%), slightly higher than that for the IR, but 6.5% and 15.0% greater(P＜0.05) than that for the HJR and IJR, respectively(Tables 1-3).

3.3.Plant radiation use and growth traits at different growth stages

There was significant interaction effect of rice type and year for the RUE during the vegetative stage (RUETS-PI) and reproductive stage (RUEPI-FL) (Fig.2).The RUETS-PIof the HIJR was significantly 21.5%, 15.9% and ?19.1% greater than that of the IR, IJR and HJR in 2018, but did not differ significantly from that of the IR,IJR and HJR in 2016 and 2017.With respect to RUE at PI-FL (RUEPI-FL), the HIJR presented higher values than did the IR in 2016,but lower than that of the IR in 2017,with the values not differing significantly in 2018.For RUE at the grain-filling stage, the RUEFL-MSof the HIJR showed 12.7%-52.8% greater values than did the IR, IJR and HJR in 2016-2018, except for Chunyou 84 in 2017 and Chunyou 927 in 2018 (Table S5).

Fig.1– Grain yields of four rice types grown during the late season in 2016–2018.Data are mean and sd.Means followed by different letters are significantly different according to Tukey’s HSD test(P＜0.05).IR, indica rice;IJR,inbred japonica rice;HJR,hybrid japonica rice;HIJR,hybrid indica/japonica rice.

Table 1 RUE and its related parameters for four rice types grown during the late season in 2016.–

Table 2–RUE and related parameters for four rice types grown during the late season in 2017.

Table 3–RUE and related parameters for four rice types grown during the late season in 2018.

Fig.2–Radiation use efficiency(RUE)at three growth stages of four rice types grown during the late season in 2016–2018.A–C show radiation use efficiencies in 2016–2018,respectively.Values are mean and sd.Means followed by different letters at the same growth stage are significantly different according to Tukey’s HSD test (P＜0.05).TS,tillering stage,defined as 20 days after transplanting;PI,date of panicle initiation;FL,flowering stage;MS,maturity stage;IR,indica rice;IJR,inbred japonica rice;HJR,hybrid japonica rice;HIJR,hybrid indica/japonica rice.

The amount of canopy-intercepted solar radiation of the HIJR was similar to that of the IR but was significantly greater than that of the IJR and HJR at the TP-PI stage in 2016-2018,except in 2018, in which no difference between the HIJR and HJR was detected.At the PI-FL stage, the amount of canopyintercepted solar radiation of the IJR was significantly lower than that of the IR and HIJR from 2016 to 2018,except for HIJR in 2016.In contrast to the stages before FL,the solar radiation interception by the IR at the grain-filling stage was 18.2%-26.9%, 22.2%-23.7%, and 23.4%-33.9% lower than that by the IJR, HJR, and HIJR, respectively (Fig.3A-C and Table S6).The average LIR of the HIJR did not differ drastically from that of the IR but was significantly greater than that of the IJR and HJR at all stages from 2016 to 2018,except at the FL-MS stage(Fig.3D,Table S6).

The differences in LAI between the rice types at TS, PI, FL,and 15 DAF were similar in 2016 and 2017 (Fig.4).The LAI of the HIJR did not differ drastically from that of the IR but was significantly greater than those of the IJR and HJR, except the FL and 15 DAF stages in 2017, in which there was significant difference between HIJR and IR; In 2018, the LAI of the HIJR was not significantly different from those of the IR and HJR at TS, the IR at PI, and the HJR at 15 DAF (Table S6).In contrast,crop growth rates from tillering to PI (CGRTS-PI) of the HIJR were similar to those of the IR and HJR and were significantly greater than those of the IJR (Fig.5A and Table S5).The tiller numbers at the tillering stage (TNTS) for the IR were 25.9%-48.5%, 18.9%-38.7%, and 23.7%-35.9% greater than those for the IJR,HJR,and HIJR,respectively(Fig.5B and Table S5).

Fig.3–Canopy-intercepted solar radiation and average light interception(LIR)at three growth stages of four rice types grown during the late season in 2016–2018.A–C show the canopy-intercepted solar radiation in 2016–2018;D shows the average canopy light interception across 2016–2018.Values are mean and sd.Means followed by different letters at the same growth stage are significantly different according to Tukey’s HSD test(P＜0.05).TP,transplanting;PI,panicle initiation;FL, flowering stage;MS, maturity stage;IR,indica rice;IJR,inbred japonica rice;HJR,hybrid japonica rice;HIJR,hybrid indica/japonica rice.

The differences in Pnbetween rice cultivars across the two years are shown in Fig.6.Although the differences in Pnat PI,FL (2018) and 35 DAF (2018) stage were not significant, the Pnwas 6.1%-28.4%significantly greater for the HIJR(Yongyou 538 and Yongyou 540) than for the IR and IJR at FL in 2019, and 16.2% higher for the HIJR than for the IR at 35 DAF in 2018.However, no difference in Pnwas detected between the IR(Huanghuazhan) and HIJR (Yongyou 540), which showed significantly lower values than did the other cultivars at the late grain-filling stage(22 DAF) in 2019.

Fig.4–Leaf area index(LAI)at four growth stages of four rice types grown during the late season in 2016–2018.A–C show the leaf area index in 2016–2018,respectively.Values are mean and SD.Means followed by different letters at the same growth stage are significantly different according to Tukey’s HSD test (P＜0.05).TS,tillering stage,defined as 20 days aftertransplanting(DAT);PI,date of panicle initiation;FL, flowering stage;15DAF,15 days after flowering;MS,maturity stage;IR,indica rice;IJR,inbred japonica rice;HJR,hybrid japonica rice;HIJR,hybrid indica/japonica rice.

3.4.PCA of solar radiation-use parameters

Fig.7B shows the loading scores of PCs 1 (Dim1)and 2 (Dim2)based on the analysis of rice cultivars of the four rice types grown in the late season in 2016-2018.Dim1 and Dim2 explained respectively 32.3% and 14.9% of the total variation.The cumulative percentage explained by Dim1 and Dim2 was nearly 50% (Fig.7C).According to the coordinates of the variables (Fig.7A), the parameters involved in canopy light capture(the mean LIR during the different growth stages;the SI of the vegetative stage;LAI at PI,FL,and 15 DAF;etc.)were clustered along Dim1.The traits associated with Dim1 of the HIJR were similar to those of the IR but differed sharply from those of the IJR and HJR.Parameters describing phenology,such as the SR and SI of the grain-filling period, fell into the same group (Dim2).The HIJR was consistent with the IJR and HJR in terms of Dim2 but was significantly different from the IR.These results showed that the capacity of the HIJR to use solar radiation resulted from the high light capture capability of the IR (32.3% of the total variation) and the high SR and SI during the grain-filling stage of the IJR and HJR(14.9%of total variation).

Fig.5–Crop growth rate(CGR,A),tiller number(TN,B),leaf area per tiller(C)and specific leaf area(SLA,D)during the tillering stage of four rice types grown during the late season in 2016–2018.Values are mean and sd.Means followed by different letters in the same year are significantly different according to Tukey’s HSD test(P＜0.05).TS,tillering stage,defined as 20 days after transplanting;PI,date of panicle initiation;TNTS,tiller number per hill at tillering stage;CGRTS-PI,crop growth rate during the tillering stage;IR, indica rice;IJR,inbred japonica rice;HJR,hybrid japonica rice;HIJR,hybrid indica/japonica rice.

3.5.Contribution to total biomass of solar radiation-use parameters at three growth stages

The Pearson’s correlations between solar radiation-use parameters and total biomass during the target stage are shown in Fig.S2.Total biomass was positively correlated with the SI at the vegetative stage(r=0.47,P＜0.01)and grain-filling stage(r = 0.42, P＜0.05), and RUE at the vegetative stage (r = 0.35,P＜0.05)and grain filling stage(r=0.54,P＜0.001).Thus,RUE at the grain-filling stage contributes the most to the biomass difference among cultivar variation.

4.Discussion

HIJR shows greater heterosis for biomass production than do other elite rice cultivars.In accord with the results of previous studies [4,5,8], HIJR achieved 13.4%-53.4% higher grain yield than the IR, IJR and HJR from 2016 to 2018 in the late season(Fig.1).In terms of yield components, the HIJR had higher spikelets per panicle and greater sink size as compared to IR,IJR and HJR(Table S2),which was in agreement with previous study [8,11].Aboveground biomass was 14.3%-30.6% greater for the HIJR than for the others in 2016-2018 (Tables 1-3),indicating the superiority of the newly released HIJR in the late season.

Fig.6–Leafnetphotosyntheticrateatthreegrowthstagesofsixricecultivarsgrownduringthelateseasonin2018(A)and 2019(B).Valuesaremeanandsd.Meansfollowedbydifferentlettersatthesamegrowthstagearesignificantlydifferent accordingtoTukey’sHSDtest(P＜0.05).IR,indicarice;IJR,inbredjaponicarice;HJR,hybridjaponicarice;HIJR,hybridindica/japonicarice.

Biomass accumulation is the product of canopyintercepted solar radiation and RUE [16,18], and greater canopy interception of solar radiation can be achieved by increasing canopy LAI, prolonging the growth period, or both[19].In the present study, the HIJR and IR had larger average LAI (Fig.4) at four growth stages, and greater maximum LIRa(Fig.S3) than did the IJR and HJR before FL.The consistently greater LAI before FL suggested a faster canopy closure of the HIJR than of the IJR and HJR, which increased the average canopy LIR and slightly resulted in a greater canopy SI.In terms of these dynamic canopy traits,it seemed that the HIJR inherited the fast canopy establishment of elite indica cultivars [20].However, canopy architecture dynamics can be affected by tillering capacity, leaf expansion, leaf erectness,and other traits [21].Notably, the canopy establishment process in IR seems to be even faster than in HIJR, especially during the vegetative stage (Fig.4 and Fig.S3), but the differences in architecture traits were cultivar- and yeardependent.The mean LAI and tiller number were not significantly different or slightly lower for the HIJR than for the IR at the TS,while the average leaf area per tiller and SLA did not differ dramatically between HIJR and IR(Figs.4 and 5).These results indicate that both HIJR and IR have superiority in canopy-intercepted solar radiation over the other rice types during the vegetative stage.However,study interpreting such superiority based on canopy architecture traits is still lacking.

The degrees of radiation interception during grain filling in the HIJR and HJR did not drastically differ but were significantly greater than those of the IJR and IR.These differences were attributed mainly to the prolonged growth period of the first two rice types [7].The HIJR also showed a longer growth period than the IR and IJR, especially a longer grain-filling stage (Tables S3 and S4), a finding in agreement with the findings of previous studies for a single season [7].This difference would increase the amount of incident solar radiation for the HIJR compared with the IR.The higher LAI at FL and 15 DAF stages also led to a larger leaf area duration in the HIJR than in the IJR and HJR during grain filling(Fig.4),leading to a slight and significant increase in radiation interception during grain filling by the HIJR compared with the japonica rice(IJR and HJR)and IR(Fig.3).The longer growth period during the grain filling stage was generally attributed to the low temperature tolerance of japonica rice [22].In the present study, the daily low temperatures in the late grainfilling period (20 days before maturity) were 4.5 °C lower for the HIJR than for the IR(Fig.S1).However,stay green traits in HIJR, such as high LAI, leaf chlorophyll content (SPAD value)and Pnvalues during the late grain-filling stage, were found[11]in HIJR varieties in contrast to the other types,suggesting that leaf senescence occurred more slowly in HIJR than in IR,a finding inconsistent with ours(Figs.3 and 4).Thus,the greater biomass heterosis of HIJR rice than of the other rice types may also have resulted from its advantage with respect to both the growth period and prolonged greenness of the canopy during grain filling.

Little information on RUE differences between HIJR and other rice types in China is available.In the present study,the HIJR showed 5.6%-10.8% higher mean RUE than did the IR,HJR, and IJR during the growing season, with wide cultivar variation (Tables 1-3).In most crop species, variation in RUE results primarily from CO2assimilation rate and quantum efficiency[16],However,the effects of photosynthesis manipulation can be diminished when scaled to the canopy owing to complexities associated with canopy light interception[21,23].Thus, in addition to maximum leaf Pn, the geometry of light interception may alter RUE [16,24].Duncan et al.[25]attributed a greater canopy RUE mainly to an improved Pnwhen the LAI was less than 3 to 4, but canopies with more erect leaves had a greater RUE when the LAI exceeded this range.In the present study, the plant aboveground biomass was mainly positively correlated with the SI rather than RUE during the TS-FL(Fig.S2),indicating that the Pnof the HIJR(at the TS,the LAI was 0.9-1.6) did not differ significantly from those of the other rice types during the vegetative stage.But the HIJR had a single tiller leaf area comparable to that of the IR, but greater leaf area per tiller than the IJR from 2016 to 2018 (Fig.5), in partial agreement with the finding [26]that the HIJR showed the larger mean length and width of upper fifth leaves than IR and IJR.In a previous study of leaf angle [27], HIJR showed a smaller leaf basic angle and dropping angle than did IR and IJR at heading stage, indicating that the HIJR had a greater RUE because of its canopy architecture with erect leaves.In the present study,the greater RUE of HIJR compared with those of the other rice types was confirmed at the grain-filling stage(Fig.2).Given that differences in leaf Pnbetween the HIJR, IR,and IJR were not significant at mid-grain filling in both 2018 and 2019(Fig.6),the greater RUE of the HIJR than those of the other rice types may have resulted from improved canopy architecture rather than an increased leaf Pnafter FL.

Fig.7–Coordinates of the parameters(A),loading scores of principal components 1 and 2(B),and eigenvalues(C)from principal components analysis of the solar radiation-use parameters of rice cultivars in the late growing seasons of 2016–2018.RUE,solar radiation-use efficiency;tot,the full growing season;SI,canopy solar radiation interception;SR,incident solar radiation;LIR,canopy light interception rate;VEG,vegetative period;REP,reproductive period;15DAF,15 days after flowering;GF,grainfilling period;CGR,crop growth rate;LAI,leaf area index;TS,tillering stage;PI,panicle initiation;FL,flowering stage;DAF,days after flowering;SLA,specific leaf area.

The heterosis of subspecies hybrids has been highlighted to illustrate the yield and biomass superiority of HIJR compared with other rice types for years[3], and many yieldrelated traits have been investigated, including panicle size,leaf/root morphophysiology, nutrient absorption, and phenology[11,13,28].However,full understanding of the source of these traits in HIJR is lacking.PCA showed that the HIJR grouped with the IR along Dim1 (canopy light capture capability) but with the IJR/HJR along Dim2 (prolonged grain filling).IR is more sensitive to temperature and has greater tillering and leaf expansion under high temperature; japonica rice prefers a cooler environment and shows fewer tillers and a slower leaf growth rate, but is highly tolerant to low temperature during grain filling [22].Accordingly, in view of the parameters describing solar radiation use, we hypothesize that the superiority of HIJR with respect to solar radiation use results from the integration of the high canopy light capture capability of IR and the prolonged growth period(especially during grain filling) of japonica rice in the late growing season.

5.Conclusions

Compared with the other rice types, the HIJR showed greater potential for grain yield and biomass production during the late growing season.This potential is attributed to 1) its increased canopy radiation interception throughout the growing season;specifically,the fast canopy closure before FL and the large leaf area duration due to the prolonged grain-filling period and resistance to cold temperature;2)its increased solar RUE,which may have been conferred by optimized canopy architecture with erect leaves during the grain-filling stage.The superiority of the HIJR in solar radiation use resulted from the greater canopy light capture capability of IR and the prolonged growth period(especially during grain filling)of japonica rice in the late growing season.

CRediT authorship contribution statement

Min Yin performed the experiment and co-wrote the MS;Shaowen Liu performed the experiment and data analysis;Xi Zheng assisted in data analysis and revised the MS; Guang Chu assisted with the field experiment and revised the MS;Chunmei Xu assisted in Pnevaluation and revised the MS;Xiufu Zhang revised the MS; Dangying Wang designed the experiment and revised the MS; Song Chen designed the experiment, performed data analysis,and co-wrote the MS.

Declaration of competing interest

Authors declare that there are no conflicts of interest.

Acknowledgments

This research was supported in part by grants from the National Key Research and Development Program of China(2016YFD0300108, 2016YFD0300208-02), the National Natural Science Foundation of China(31671638),the China Agriculture Research System (CARS-01-04A), and Central Public Interest Scientific Institution Basal Research Fund(2017RG004-1).

Appendix A.Supplementary data

Supplementary data for this article can be found online at https://doi.org/10.1016/j.cj.2020.06.010.