Analysis of the Correlation between Seed-setting Characters and Seed Investigation Time in Rice

Zhonghua SONG,Yangui ZHANG,Jinhua NING

1.Changsha Agrometeorological Experimental Station,Changsha 410125,China;2.Key Lab of Hunan Province for Meteorological Disaster Prevention and Mitigation,Changsha 410118,China

In the structural analysis of crop yield,various factors and combinations constituting biological and economic yield per unit area are determined and calculated.Seed investigation is one of the indispensable means of structural analysis of crop yield,which mainly analyzes a variety of seed-setting characters such as seed-setting rate,empty grain rate and 1 000-seed weight.Based on the investigation of seed-setting characters,the correlation between various factors that affect rice yield can be clarified[1-2].In addition,the investigation of seedsetting characters also provides reference for breeding new rice varieties[3-4]and research of stress resistance[5-7]and ecological adaptability[8-9].Seed investigation provides basic data for experimental test of rice and lays an important foundation for developing high-yield cultivation techniques of rice[10-11].Therefore,errors and deviations in rice seed investigation should be reduced as much as possible to accurately and objectively reflect the yield structure of rice,avoiding the interference of differences in seed investigation results.At present,seed investigation is commonly conducted by increasing replications,improving the representativeness of samples[12],system integration,designating the person in charge[13-14]and other methods.However,the progress of seed investigation is not considered in these methods.In fact,due to various reasons,seed investigation progresses are generally inconsistent.Sometimes seed investigation can be completed soon after sampling,while in some cases seed investigation is not completed at a very long time after sampling.Currently,little information is available on whether inconsistent progresses of seed investigation will cause differences in yield structure and whether there is a correlation between the performance of yield structure and seed investigation time.In this study,the correlation between seed investigation time and various seed-setting characters such as seed-setting rate,abortive grain rate,empty grain rate and 1 000-seed weight was analyzed,aiming at providing technical means for effectively reducing the errors or deviations and accurately reflect the seed-setting characters of rice.

Materials and Methods

Materials

With super hybrid rice cultivar Y Liangyou 1 as the experimental material,the experiment was carried out in the basic farmland protection area of Chunhua Village,Chunhua Town,Changsha County,Hunan Province.Experimental rice was sown on May 15,transplanted on June 15,and harvested on October 9.The total cultivation area was 1 067 m2.

Seed investigation

SamplingAt maturity stage,continuous 40 rice seedlings were collected to record the totaleffective spike number. Subsequently, effective spikes were cut off along the stalk and divided into three groups based on spike length:superior(group 1),moderate (group 2)and poor(group 3).The spike number in each group was counted,and all the spikes were preserved indoors under good ventilation condition.The proportion ofeach group was calculated according to the following formula:

Proportion of each group=Spike number in each group/Total effective spike number

In seed investigation,10 spike samples (intact,without grain loss)were selected from three groups with fourreplications to investigate the seed-setting rate,empty grain rate,abortive grain rate and 1 000-seed weight.Samples were not reused.

Seed investigation timeVarious seed-setting characters were investigated at 1,2,3,4,5,6,7,8,9 and 10 d post-harvest(hereinafter referred to as day 1,day 2,day 3,day 4,day 5,day 6,day 7,day 8,day 9 and day 10),respectively.Each of the ten treatments was repeated four times.In each treatment,seed investigation was conducted at 09:00-10:00.

Grain separationRice grains were separated with water-washing method to count filled grains,empty grains and abortive grains.Randomly selected spike samples were threshed until no grains remained on the spike.The collected grains were placed into the container filled with tap water,stirred fully,and placed at room temperature for about5 min;empty grains and abortive grains floating on water surface were transferred to a dish,and filled grains sinking to the bottom were transferred to another dish.In order to separate empty grains and abortive grains,two dishes were dried in an incubator at 40℃for approximately 2 h.After drying,filled grains were counted and weighed;mixed empty grains and abortive grains were placed on a frosted glass under a floodlight to count the number of empty grains and abortive grains. Subsequently, seed-setting rate,empty grain rate and abortive grain rate were calculated in accordance with the literature[15];1 000-seed weight was calculated according to the following formula:

1 000-seed weight=Total weight of filled grains/Total number of filled grains

Data processing

Statistical analysis and mapping were performedusingExcel2003 software;variance analysis,regression analysis and Duncan’s new multiple range test were performed using DPS data processing system[16].

Results and Analysis

Sample composition

Among 40 rice samples collected at maturity stage,there were 537 effective spikes in total,which were divided into three groups as previously described:148 spikes in group 1,328 spikes in group 2,and 61 spikes in group 3,accounting for 27.6%,61.1% and 11.4%,respectively.Based on the proportion of each group,among every 10 intact spike samples,the number of theoretically investigated samples in group 1,group 2 and group 3 was 2.8,6.1 and 1.1,respectively.After rounding to the nearest integer,among every 10 spike samples investigated,the number of actually investigated samples in group 1,group 2 and group 3 was 3,6 and 1,respectively(Table 1).

Differences in seed-setting characters

According to different investigation time,seed-setting rate in various treatments was 77.5%,77.6%,77.2%,70.8%,70.1%,68.4%,67.9%,69.3%,68.2% and 65.5% ,respectively.Specifically,seed-setting rate reached the highest(77.6% )at 2 d post-harvest and the lowest(65.6% )at 10 d post-harvest.The average seed-setting rate was 71.3% ;rangeR=12.0% ;variation coefficientCV=6.3%.Analysis of variance showed that:among treatments,F=15.218 0,significance level reached 0.000 0,indicating extremely significant differences in seedsetting rate;among groups,F=1.994 0,significance level reached 0.138 6,indicating insignificantdifferences in seed-setting rate(Table 2).

According to different investigation time,abortive grain rate in various treatments was 11.5%,11.8%,13.0%,19.1%,18.9%,18.8%,24.0%,22.2%,22.3% and 23.1%,respectively,with an average of 18.5%.Specifically,abortive grain rate reached the highest(24.0% )at 7 d post-harvest and the lowest(11.5% )at 1 d post-harvest;rangeR=12.5% ;variation coefficientCV=2.6%.Analysis of variance showed that:among treatments,F=21.205 0,significance level reached 0.000 1,indicating extremely significant differences in abortive grain rate;among blocks,F=2.045,significance level reached 0.131 2,indicating in-significant differences in abortive grain rate(Table 2).

Table 1 Sample composition and grouping

Table 2 Variance analysis of seed-setting characters at different seed investigation time

According to different investigation time,empty grain rate in various treatments was 11.0%,10.6%,9.7%,10.1%,11.0%,12.9%,8.2%,8.5%,9.4% and 11.4%,respectively,with an average of 10.3%.Specifically,empty grain rate reached the highest(12.9% )at 6 d post-harvest and the lowest(8.2% )at 7 d post-harvest;rangeR=4.7% ;variation coefficientCV=13.6%.Analysis of variance showed that:among treatments,F=6.596,significance level reached 0.000 1,indicating extremely significant differences in empty grain rate;among blocks,F=2.924, significance level reached 0.051 9,indicating insignificant differences in empty grain rate(Table 2).

According to different investigation time,1 000-seed weight in various treatments was 25.58,25.58,25.47,25.15,24.89,24.92,24.84,24.49,24.39 and 24.28 g,respectively,with an average of 24.60 g,Specifically,1 000-seed weight reached the maximum(25.58 g)at 1,2 d post-harvest and the minimum (24.28 g)at 10 d post-harvest;rangeR=1.3 g;variation coefficientCV=1.9%.Analysis of variance showed that:among treatments,F=3.246,significance level reached 0.008 5,indicating extremely significant differences in 1 000-seed weight;among blocks,F=0.877,significance level reached 0.465 1,indicating insignificant differences in 1 000-seed weight(Table 2).

Changes of seed-setting characters with seed investigation time

The correlation between seedsetting characters and seed investigation time was analyzed by regression analysis to study changes of seed-setting characters with the investigation time(Table 3).According to the results of regression analysis,the correlation coefficient of seed-setting rate and seed investigation time was-0.913 2,P=0.000 2,indicating an extremely significantnegative correlation between seed-setting rate and seed investigation time,and the seed-setting rate showed a decreasing trend with the postponement of seed investigation;the correlation coefficient of abortive grain rate and seed investigation time was 0.918 6,P=0.000 2,indicating an extremely significant positive correlation between abortive grain rate and seed investigation time,and the abortive grain rate showed an upward trend with the postponement of seed investigation;the correlation coefficient of empty grain rate and seed investigation time was-0.181 7,P=0.615 4,indicating no significant correlation between empty grain rate and seed investigation time;the correlation coefficient of 1 000-seed weight and seed investigation time was-0.982 6,P=0.000 1,indicating an extremely significantnegative correlation between 1 000-seed weight and seed investigation time,and the 1 000-seed weight showed a downward trend with the postponement of seed investigation.

Duncan’s new multiple range test of seed-setting characters

According to the results of Duncan’s new multiple range test of seedsetting rate at different seed investiga-tion time,seed-setting rate at day 1,day 2 and day 3 exhibited no significant differences,which were extremely remarkably higher than that in other treatments.Seed-setting rate at day 4 and day 5 exhibited no significant differences,which both showed no significant differences compared with that at day 6,day 7,day 8 and day 9.However,seed-setting rate at day 4 was extremely significantly higher than that at day 10;seed-setting rate at day 5 was significantly higher than that at day 10.Moreover,seed-setting rate at day 6,day 7,day 8 and day 9 exhibited no significant differences,which also showed no significant differences compared with that at day 10(Table 4).

Table 3 Regression analysis of seed-setting characters and seed investigation time(x)

Table 4 Duncan’s multiple comparisons of seed-setting characters at different seed investigation time

Table 5 Correlation analysis among seed-setting rate,abortive grain rate and empty grain rate

According to the results of Duncan’snew multiple range test of abortive grain rate at different seed investigation time,abortive grain rate at day 7 exhibited no significant differences compared with that at day 8,day 9 and day 10,but it was extremely significantly higher than that in other treatments.Abortive grain rate at day 8,day 9 and day 10 exhibited no significant differences,which were significantly higher than that at day 4,day 5 and day 6,and extremely significantly higher than that at day 1,day 2 and day 3.Abortive grain rate at day 4,day 5 and day 6 exhibited no significant differences,which were extremely significantly higher than that at day 1,day 2 and day 3.In addition,there were no significant differences among abortive grain rate at day 1,day 2 and day 3(Table 4).

According to the results of Duncan’s new multiple range test of 1 000-seed weight at different seed investigation time,1 000-seed weight at day 1,day 2 and day 3 exhibited no significant differences,which showed no significant differences compared with that at day 4,day 5,day 6 and day 7 but were extremely significantly higher than that at day 8,day 9 and day 10.Moreover,there were no significant differences among abortive grain rate at day 4,day 5,day 6,day 7,day 8,day 9 and day 10(Table 4).

Although empty grain rate at different seed investigation time exhibited a significance level of 0.000 1,there was no significant correlation between empty grain rate and seed investigation time.Therefore,Duncan’s new multiple range test of empty grain rate was not performed.

Correlation analysis among seedsetting rate,abortive grain rate and empty grain rate

The correlation analysis among seed-setting rate,abortive grain rate and empty grain rate was conducted to analyze the mutual influence and mutual restriction among these three indicators.According to the results,the correlation coefficient of seedsetting rate and abortive grain rate wasr=-0.956 2,indicating an extremely significantnegative correlation between seed-setting rate and abortive grain rate;the correlation coefficient of seed-setting rate and empty grain rate wasr=0.021 9,indicating no significant correlation between seed-setting rate and empty grain rate;the correlation coefficient of abortive grain rate and empty grain rate wasr=-0.313 4,indicating no significant correlation between abortive grain rate and empty grain rate(Table 5).

Conclusion and Discussion

The results indicated that seedsetting rate,abortive grain rate,empty grain rate and 1 000-seed weight of rice varied at different seed investigation time.With the postponement of seed investigation,seed-setting rate and 1 000-seed weight showed downward trends, abortive grain rate showed an upward trend,while empty grain rate exhibited a significant correlation with seed investigation time.

In general,seed-setting rate is restricted by abortive grain rate and empty grain rate.Seed-setting rate is inversely proportionally related to empty grain rate and abortive grain rate.In the present study,seed-setting rate exhibited an extremely significant negative correlation with abortive grain rate but had no remarkable correlation with empty grain rate,indicating that the reduction in seed-setting rate with the postponement of seed investigation was mainly due to the increase of abortive grains,which was related to the formation mechanism of abortive grains and empty grains.The formation of empty grains is resulted from abnormal pollen development or fertilization;the formation of abortive grains is mainly caused by insufficient filling.The analysis of yield structure was carried out after rice harvest,and the proportion of empty grains would not change with seed investigation time,resulting in the insignificant correlation between seed-setting rate and empty grain rate.However,after the harvest,grain respiration continued,and some grains between filled and abortive forms might become abortive grains due to respiratory consumption,which increased abortive grains and reduced filled grains, thereby enhancing abortive grain rate and declining seedsetting rate.Furthermore,the increase of abortive grain rate might also be due to the factor that some grains between filled and abortive forms became abortive grains after dehydration.

In this study,differences in rice yield structure were caused by different seed investigation time but not related to other internal factors affecting yield structure such as rice variety,climatic and ecological condition,soil,and cultivation technology.In order to avoid differences in yield structure caused by seed investigation time,improve the accuracy of results,provide reliable data sources for relevant research,and reduce errors and deviations in the conclusion,the analysis of yield structure should be conducted at a specific time.Currently,there are no rigid requirements for seed investigation time in yield structure analysis of rice and other crops.Tanget al.[16]suggested that yield structure analysis should be completed within one month after the harvest.According to the results,seed-setting rate,abortive grain rate and 1 000-seed weight exhibited no significant differences within the first three days after the harvest;since day 4,seed-setting rate and 1 000-seed weightdeclined remarkably,while abortive grain rate increased significantly,which to some extent indicated that changes of yield structure caused by investigation time should be concerned in yield structure analysis of rice and other crops to improve the accuracy of relevant research.In conclusion,the analysis of seed-setting rate,abortive grain rate,empty grain rate,and 1 000-seed weight should be completed within the first three days after the harvest.

[1]GAO LY(高良艳),ZHOU HF(周鸿飞).Relationship between yield component factors and yield in rice(水稻产量构成因素与产量的分析)[J].Liaoning Agricultural Sciences(辽宁农业科学),2007(1):26-28.

[2]LIN HX(林洪鑫),PAN XH(潘晓华),SHI QH(石庆华),et al.Effects of row-spacing on yield of super-high-yielding early and late rice(行株距配置对超高产早晚稻产量的影响)[J].Chinese Journal of Rice Science(中国水稻科学),2011,25(1):79-85.

[3]XU J(徐津),TAN XL(谭学林),WANG B(王梆),et al.Breeding and application of new Dian-typejaponicahybrid rice combination Dianza 46(滇型杂交粳稻滇杂46的选育与应用)[J].Seed(种子),2012,31(4):109-112.

[4]LI ZY(李铮友),CHEN LJ(陈丽娟),ZHANG SH(张树华),et al.Breeding and utilization of Dian-type hybrid rice Yuza 29(籼粳杂交水稻榆杂29的选育与应用研究)[J].Hybrid Rice(杂交水稻),1996(5):9-11.

[5]LIU LL(刘玲玲),PENG XL(彭显龙),LIU YY(刘元英),et al.Effects of potassium on rice blast and yield under different nitrogen management in cold area of China(不同氮肥管理条件下钾对寒地水稻抗病性和产量的影响)[J].Journal of Integrative Agriculture(中国农业科学),2008,41(8):2258-2262.

[6]GUO MS(郭萌生),ZHANG HL(张红林),XIE Y(谢勇),et al.Influence of temperature conditions on setting percentage of middle-late hybrid rice(温度条件对杂交中晚稻结实率的影响)[J].Chinese Journal of Agrometeorology(中国农业气象),2008,29(3):304-307.

[7]LI SL(黎世龄),CHEN YF(陈云风),LUO XP(罗筱平),et al.Analysis on identificational temperature conditions in flowering stage of rice in Yiehun(宜春地区水稻抽穗扬花期耐热性鉴定温度条件分析)[J].Seed(种子),2012,31(4):78-80.

[8]DONG ZQ(董占强),XIAO XZ(肖秀珠),YAO KM(姚克敏).Meteorological indexes of critical stages and climatic adaptation in Jiangsu in two-line hybrid rice 65396(Peiai 64S/E32)(两系法杂交稻65396关键期的气象指标及其在江苏的气候适应性)[J].Chinese Journal of A-grometeorology(中国农业气象),2002,23(3):5-11.

[9]YOU ZM(尤志明),JIANG ZW(姜照伟),CHENG XH(程雪华),et al.High-yield conditions and climatic/ecological adaptability of hybrid-rice Ⅱyou131(杂交稻Ⅱ优131高产结构与气候生态适应性研究)[J].Fujian Journal of Agricultural Sciences(福建农业学报),2008,23(3):281-287.

[10]CHEN YM(陈益明).Yield structure and super high-yield supporting technologies of super hybrid rice(超级稻的产量结构与超高产配套技术)[J].Acta Agriculturae Jiangxi(江西农业学报),2008,20(2):20-22.

[11]ZHOU J(周静),MA GH(马国辉).Advances in super high-yield cultivation of super hybrid rice(超级杂交稻超高产栽培研究进展)[J].Crop Research(作物研究),2006(5):422-425.

[12]NY/T 1300-2007,Technical Procedures for Rice Variety Trials(农作物品种区域试验技术规范-水稻)[S].

[13]CHEN MG(陈铭官).Statistics of grain number per panicle in rice seed investigation(在水稻考种中穗实粒数统计方法的研究)[J].Fujian Science and Technology of Rice and Wheat(福建稻麦科技),2000,19(1):12-13.

[14]WU HS(吴洪山),HONG ZC(洪中川),LI YH(李玉环).A brief introduction to the issues in improvement of regional trial of hybrid rice varieties(浅谈提高杂交水稻品种区试质量问题)[J].Bulletin of Agricultural Science and Technology(农业科技通讯),2008(8):121-122.

[15]China Meteorological Administration(中国气象局).Specifications for Agricultural Meteorological Observation(农业气象观测规范)[M].Beijing:China Meteorological Press(北京:气象出版社),1993:32-42.

[16]TANG QY(唐启义),FENG MG(冯明光).Practical Statistical Analysis and Data Processing System(DPS)(实用统计分析及其DPS数据处理系统)[M].Beijing:Science Press(北京:科学出版社),2002:33-92.