The Design And Analysis Methods Of Field Experiment In Wheat Founder Parent And Derivative Progeny

Founder parent is a base of improved variety in wheat breeding.A new and good wheat variety cannot be bred or developed without the integrated elite characters of Founder parent.Not only the descendant of Founder Parent were selected wheat varieties with large area as materials in most areas of china.But also It derives a lot of breeding parents who have a widely application.But because of experimental material are made up of founder parents and their progeny. And founder parent and its derivative progeny have a relative affinity.Field trial and the corresponding analytical methods for materials of founder parent and its derivative progeny with basing on the Large quantity and Affinitiessimilarity which are worth disscussing and researching. This study is consider that the 254 materials of founder parents and their progeny from Yangling dwarf wheat variety trials in 2007-2008.The scheme of design(the design of hierarchical structure) for the number of founder parent and its derivative progeny was determined in prophase.The strain effects and correlation of the material are being solved by using hierarchical mixed model,And significant differences in traits between founder parents and their progeny in different generations was illustrated using methods of Multiple Comparison,Linear Grouped Contrast,Breeding value estimates,Multivariate Statistics analysis and Comprehensive Evaluation.The evolutionary tendency and contribution between founder parents and their progeny has been analyzed more further.Through analyzing the plot yield data which would be partly simulated from YangLing in 2007-2008 of ShanXi Province,and partly measured from TaiAn in 2007-2008 of ShanDong Province.The more relative efficiency and stable models were selected by simulation.Finally,We set R language environment as Development platform,and set R-Commander as Design Template.We constructed a software analytical module.It convenient to analyze and manipulate in the correlation of the material and large number of reginal trials and in the future for experimentalist and research worker.1. Based on the characteristic affinities similarity in founder parent and its derivative progeny of treatment design. According to the ideas of hierarchical design.“intrablock similarity”and“Relative balance on each derivative generation”should be basic principle during conducted incomplete block design with hierarchical treatment. Using Two-direction main area block Arrange in field, Latin square Arrange in intrablock treatment of main area and Setting center CK,random CK,side CK in main area during Environmental design.2. The S.E.means on the estimated treatment effects for 14 traits of general linear model and hierarchical mixed model value were 187.58 and 120.89,respectively,the S.E.means on the estimated treatment effects of hierarchical mixed model value was lowered 35.55% than general linear model value. It is observed that hierarchical mixed model on grouping block within repeat could effectively reduce the errors on the estimated treatment effects.Suggestions are useing hierarchical mixed model to estimated treatment effects when it has a Large number of treatments and Data are missing and imbalancing.Through ICC(Interclass Correlation Coefficient),We find that Four traits:plot yield,tillernumber,leaf length,plant height,which has higher ICC values.It shows that individual have higher Similarity within class in 4 traits. according to grouping results by two methods on multiple comparisons.the letter was overlap in Tukey grouping test. It is delicate, Intensive to be Compared on small samplesize.the letter wasn’t overlap in Scott-Knott grouping test.It is succinct,easy to be explained on large samplesize. It is thus suggested that Using the scott-knott method can conduct better the grouping results on large samplesize.3. Base on 3 founder parents:Funo,Bima4,Nanda2419 and its derivative progeny as research object.then their 12 traits were identified in YangLing of ShanXi province in 2008. In view of the means for 12 traits from founder parent and its first filial generations. Founder parent of Funo can transfer genetic background which have high cumulative level on alleles from 7 traits:plot yield, tillernumber, the ability of stripe rust-resistant, grain number per spike, grain weight per spike, leaf length, leaf width to Its first filial generations. But the cumulative level on alleles from 5 traits: heading stage, spikelet number per spike, 1000 grains weight, ear length, plant height are improved due to the influence of Other parents.Founder parent of Bima4 can transfer genetic background which have high cumulative level on alleles from 9 traits:plot yield, tillernumber, the ability of stripe rust-resistant, spikelet number per spike, ear length,grain number per spike, grain weight per spike, leaf length, leaf width to Its second filial generations.But the cumulative level on alleles from 3 traits: heading stage, 1000 grains weight, plant height are improved due to the influence of Other parents. Founder parent of Nanda2419 can transfer genetic background which have high cumulative level on alleles from 4 traits:plot yield, tillernumber, spikelet number per spike, leaf width to Its first filial generations.But the cumulative level on alleles from 8 traits: heading stage, the ability of stripe rust-resistant, ear length,grain number per spike, grain weight per spike, leaf length,plant height are improved due to the influence of Other parents. By analysing the 9 traits .It has remained basically steady on 3 founder parents:Funo,Bima4,Nanda2419 and their first filial generations,second filial generations,third filial generations, fourth filial generations, fifth filial generations, It shows that the Founder parent can Steady transfer good genetic background to its derivative progeny.In view of the breeding value for 12 traits from 3 founder parent:Funo,Bima4,Nanda2419 and its derivative progeny.Founder parent of Funo can transfer favorable genetic effect from 5 traits: the ability of stripe rust-resistant, high grain number per spike, high grain weight per spike, leaf length, leaf width to Its derivative progeny.And stripe rust-resistant, leaf length can transfer more favorable genetic effect to Its derivative progeny.breeding value which have unfavourable genetic effect is small on spikelet number per spike.It shows that Founder parent of Funo can transfer smaller genetic load which have unfavourable genetic effect to Its second filial generations. Founder parent of Bima4 can transfer most favorable and additive genetic effect from 1 traits: tillernumber to Its derivative progeny.But Founder parent of Bima4 can transfer genetic load which have unfavourable genetic effect to Its derivative progeny from 9 traits: low plot yield, Feeling stripe rust, less spikelet number per spike,short ear length,less grain number per spike, light grain weight per spike,short leaf, narrow leaf,high plant height.Founder parent of Nanda2419 can transfer most favorable and beneficiated genetic effect from 4 traits: more spikelet number per spike,high 1000 grains weight,long ear length, heavy grain weight per spike to Its derivative progeny.But Founder parent of Nanda2419 can transfer genetic load which have unfavourable genetic effect to Its derivative progeny from 4 traits: less tillernumber, Feeling stripe rust, short leaf, high plant height.4. The 17 grouping materials were clustered by using Euclidean distance and group-average method.It suggested that the materials could be divided into 8 groups: GroupⅠis Bima4.GroupⅡis Funo.GroupⅢis inter parents of Bima4.GroupⅣis Nanda2419. GroupⅤare first filial generations, second filial generations, third filial generations, fourth filial generations, fifth filial generations of Funo. GroupⅥare first filial generations, second filial generations, third filial generations, fourth filial generations, fifth filial generations of Nanda2419. GroupⅦis fifth filial generations of Funo. GroupⅧare second filial generations, third filial generations, fourth filial generations of Bima4 respectively.through a Correspondence Analysis, announcing to the characteristic of each groups: The results maximum present Similarity and dissimilarity between Founder parents and its derivative progeny.Through the factor loading was estimated by principal component analysis. The 8 factors will be named:The factor of grain number and weight per spike; The factor of 1000 grains weight; The factor of growth period ; The factor of plot yield; The factor of leaf Width; The factor of spikelet number per spike; The factor of plant height; The factor of leaf length;We find that the factor of grain number per spike and grain weight whose contribution of variance accounted to 14%.This factor play a decisive role in yield-building.Through the comprehensive evaluation and analysis by improved Radar chart.We can get the comprehensive score for each sample. The top 5 generations followed by:the fifth filial generations of Nanda2419,the fourth filial generations of Nanda2419,the third filial generations of Funo,the fourth filial generations of Funo,the second filial generations of Nanda2419. The lowest comprehensive score is Bima4. In view of the Comprehensive scores for genotype value and EBVs from3 founder parent and its derivative progeny.Founder parent of Nanda2419 can transfer more favorable and beneficiated genetic effect to Its derivative progeny. Founder parent of Bima4 can transfer more unfavorable and reductive genetic effect to Its derivative progeny.Comprehensive transmission ability for 3 founder parents and its derivative progeny:The best is Bima4,The next is Nanda2419,The weakest is Funo.5. Through the effectiveness analysis of the Primitive linear model which based on the 100 times of simulation data for plot yield from Yangling dwarf wheat variety trials in 2007-2008.The results show that Over 50 percent of F value which was calculated by the 100 times of simulation data for Three fixed effects: Repetition, Treatment,Block within Repetition is greater than F value which was calculated by raw data. We think original probability value from three fixed effects is quite acceptable. It is show that the result of ANOVA is effectiveIt is available to produced simulation data by the Primitive linear model, and It can be analyzed.Through analysis the data from 2 locis, it has been confirmed that the simplified generalized grid design analysis (SGL) is better control the error from field experiment,and It Can be more precise estimateeffects of treatments by increasing the number of Repetition.Test efficiency level in (SGL)model on 2 replications as well as Random complete block designs(RB)model on 3 replications. It is proved that (SGL)model on 3 replications can highly increase test efficiency in large number of experiment.6. This research based on R language environment platform.And taken R-Commander as the design of the template,established a module of the mixed models.Thismethod could be used in split plot design analysis and Nested design analysis in Design of Experiment.Experimentalist and Research worker can easily choose the maximum likelihood estimation(MLE)and the restricted estimation maximum likelihood (REML) in the parameter menu.And It can estimated the effects for each levels in grouping factors of different hierarchies in Fixed effects and random effects parameters.an increase of estimating adjustment mean modules.It can easily adjust means in each levels of treatment factors for Analysis of covariance in the future,and Marginal Means were simply estimated for assuming the balanced situation in unbalanced data.an increase of Multiple comparison analysis and Linear grouping contrasts modules,and It has increased some methods on the previous(for example:Added some methods of Multiple comparison analysis for P value adjustment and Scheffe、Tukey Linear grouping contrasts for arbitrary contrasts).It enrich the methods of Multiple comparison analysis and Linear grouping contrasts which were selected by Experimentalist and Research worker.