Maohua DAI,Liying LIU,Changbo TANG,Dezhi LV,Zhenliang WU*

1.Dryland Farming Research Institute,Hebei Academy of Agricultural and Forestry Sciences;Hebei Provincial Key Laboratory of Crops Drought Resistance Research,Hengshui 053000,China;2.College of Agronomy and Biotechnology,China Agricultural University,Beijing 100193,China;3.Suzhou Vocational College,Suzhou 215104,China

Cotton region in mid-south area of Hebei is the main centralized producing area of cotton in Hebei and the important producing base for marketable cotton in China.Cotton is the main source of income for local farmers[1-3].Therefore,analysis on yield characters and quality characters of new cotton varieties in mid-south area of Hebei has an important practical significance.Liet al.[4]studied the correlation between lint yield and plant characters of cotton,and found that lint percentage was the most effective factor for the selection of lint yield.Heet al.[5]analyzed the correlations among agronomic characters of 66 anti-insect hybrid cottons and the results indicated that yield structure components,lint percentage and number of bolls per plant had the biggest effect on seed cotton yield and lint yield.Liet al.[6]performed path analysis and analyzed the correlations among agronomic characters of 20 cotton varieties bred in China during 1950-1990s,and found that gain in yield was mainly resulted from the increasing of lint percentage and number of bolls per plant.There are many reports about correlation analysis[7-10],principal component analysis[11-13]and cluster analysis[14-15]on the main characters of upland cotton in China and abroad.However,there is a little multianalysis on yield characters and quality characters of the newest cotton varieties in mid-south area ofHebei.Comprehensive evaluation ofyield and quality characters of the newest cotton varieties in mid-south area of Hebei was performed in the study to clear current situations and yield-increasing factors of the newest cotton varieties,and provide theoretical basis for the formulation of new breeding object and variety selection in production.

Materials and Methods

Test materials

Eleven new cotton varieties participated in the regional test of spring sowing cotton in mid-south area of Hebei were used as test materials.The cotton varieties included Jifeng 1982,Runongyan 18,Shuofengmian 1,Xinshi 71143,XP7,Xinza 16,Xinza 15,Kelong 2068,Jifengza 9,Hanwu 216 and Nongdamian 12.

Test methods

The test was performed in 11 test points of six cities including Shijiazhuang,Handan,Xingtai,Hengshui,Cangzhou and Baoding, Hebei Province during 2011-2012.Randomized block arrangement was adopted for the test and the test was repeated three times.There were four rows and the plot area was 20 m2.The management was similar with field management.

Character indexes including plant height,fruit branch number,the first fruit node,number of bolls per plant,single-boll weight,seed index,lint percentage,lint yield,upper half staple length,uniformity index,fiber strength,elongation rate and micronaire value were measured.Fiber quality index was tested by Supervision,Inspection and Test Center of Cotton Quality,Ministry of Agriculture.

Data processing

SPSS 16.0 and DPS 9.50 softwares were used to process and analyze data.

Results and Analysis

Variabilityanalysisofyield and quality characters

The average values and variation conditions of 13 characters of the tested varieties were shown in Table 1.Among yield characters,the variation coefficients of number of bolls per plant,single-boll weight,seed index and lint yield were big,and the values were 7.96%,9.83%,7.68%and 9.67%,respectively.Among fiber quality characters,the variation coefficient of micronaire value was big and the coefficient was 8.00%.The results indicated that the tested varieties had various types and characteristics.

Path analysis of yield and yield components

Path analysis of yield components on lint yield was performed (Table 2).The direct path coefficients of the four yield components from high to low ranking were lint percentage(0.762),number of bolls per plant(0.560),single-boll weight(0.530)and seed index(0.062).The results indicated that yields of the new cotton varieties in mid-south area of Hebei were codetermined by lint percentage,number of bolls per plant and single-boll weight.The indirect path coefficients of number of bolls per plant, single-boll weight,seed index and lint percentage were negative values,which indicated that single-boll weight,seed index and lint percentage would be decreased with the increasing of number of bolls per plant.Varieties with many bolls could not be blindly pursued in breeding and production whereas rational combination of single-boll weight,lint percentage and number of bolls per plant should be paid attention to.

Path analysis of yield and quality characters

Path analysis ofeach quality character on lint yield was performed(Table 3).The direct path coefficients of the five quality characters on lint yield from high to low ranking were uniformity index (1.530),fiber strength(0.254),micronaire value(-0.525),elongation rate (-0.636)and upper half staple length (-1.283).The direct path coefficients of uniformity index and fiber strength on lint yield werepositive values,which indicated that improvements of the two quality indexes were achieved together with yield increasing.The direct path coefficient of upper half staple length on lint yield was a negative value but it played a positive role on lint yield through uniformity index,fiber strength and elongation rate indirectly.The direct path coefficient of elongation rate on lint yield was a negative value but it had a positive role on lint yield based on upper half staple length and micronaire value.The direct path coefficient of micronaire value on lint yield was also a negative value,which indicated that fiber would not be shortened and the quality would not be decreased with the increasing of yield.

Table 1 Variation conditions of main characters of the testing varieties

Table 3 Path analysis of fiber quality characters on lint yield

Table 4 Principal component analysis of yield and quality charactes

Principalcomponentanalysisof yield,yield characters and quality characters

Due to the complexrelations among yield characters,quality characters and yield,correlation coefficient matrix of ten yield and quality characters was used forprincipal component analysis to objectively and comprehensively evaluate the tested varieties[16-17].The results were shown in Table 4.Accumulated variance contribution rate of the first three principal components reached 87.263%,which basically represented the most informations of the ten main characters of the tested cotton varieties.The first three principalcomponents were shown as follows:

Among them,z1-z10represented number of bolls per plant,single-boll weight,seed index,lint percentage,upper half staple length,uniformity index,fiber strength,elongation rate,micronaire value and lint yield,respectively.

Table 4 showed that the contribution rate of the first principal component(Prin1)was 53.640%.The score coefficients of the ten characters in the first principal component(Prin1)were positive values except that of number of bolls per plant and elongation rate,which illustrated that single-boll weight,seed index,lint percentage,upper half staple length,uniformity index,fiber strength and micronaire value increased together with lint yield that these characters could be used for the comprehensive selection of yield and quality.The contribution rate of the second principal component(Prin2)was 20.097%.The score coefficients of number of bolls per plant and lint percentage in the second principal component were positive values,which could be used as factors for yield selection.The contribution rate of the third principal component(Prin3)was 13.525%.The score coefficients of upper half staple length and fiber strength in the third principal component were the highest,which could be used for yield selection.

Clustering analysis

Square sum of deviations was used on Chi-square clustering for hierarchical clustering and the dendro-gram was constructed(Fig.1),and the average values of the main characters of each group were calculated(Table 5).Fig.1 and Table 5 showed that the 11 cotton varieties could be clustered into three groups.The first group contained three varieties,which was the representative group of high yield and good quality with the highest lint percentage and single-boll weight.The yield of the first group was the highest and the upper half staple length,uniformity index and fiber strength were high.The second group contained seven varieties,which was the representative group of middle yield and middle quality with middle number of bolls per plant,single-boll weight and lint percentage.The upper half staple length and fiber strength of fiber quality index of the second group were high.The third group only included one variety,which was the representative group of low yield and low quality with low single-boll weight and lint percentage.The yield of the third group was the lowest and the upper half staple length and fiber strength of fiber quality index were the lowest.

Table 5 Average values of the ten characters of cotton varieties in different clustering groups

Discussions

Discussion of variability analysis of yield and quality characters

The study results showed that among the 13 characters of the tested varieties,the variance coefficient of single-boll weight was the biggest.The variances of number of bolls per plant and seed index among yield characters were obvious.The variance coefficient of lint percentage was small and the value was 3.37%.Among the five quality characters,the variance coefficients of upper half staple length,uniformity index,fiber strength and elongation rate were small.The results illustrated that there were certain degree of variances among yield characters of different varieties while the quality characters had a high similarity.Fiber quality was hardly to be significantly improved according to the hybridization between varieties in midsouth area of Hebei.In future,quality breeding should be mainly focused on distant hybridization to create new germplasms[18-22]and on research,developmentand application ofelite genes[23-26].

Path analysis and clustering analysis on multiple characters of cotton varieties

The results ofpath analysis showed that number of bolls per plant,single-boll weight,seed index and lint percentage had a certain negative correlation.The increasing of number of bolls per plant would result in the decreasing of single-boll weight,seed index,lint percentage and yield.The results of clustering analysis also indicated that the characteristics of highyield variety were the highest singleboll weight and lint percentage.The research was performed in mid-south area of Hebei and rainfall in the area mainly distributed in June to August.The tested varieties were the newest cotton varieties in mid-south area of Hebei.Currently,to increase number of bolls per plant is very hard.In addition,rainfall affects boll setting percentage in flower and boll stage.Therefore,yield increasing should be mainly focused on the improvement of singleboll weight and lint percentage.

Conclusions

Mid-south area of Hebei belongs to the warm temperate zone with abundant light and warm resources,which is one of the most suitable areas for cotton planting[27-29].Path analysis,principal component analysis,clustering analysis,etc.were used in the research.The results indicated that in cotton area ofmid-south area of Hebei,varieties with many bolls could not be blindly pursued in cotton production whereas single-boll weight,lint percentage and number of bolls per plant should be rationally combined.

[1]HUANG ZK(黃滋康).China Cotton Varieties and the Pedigrees (中國棉花品種及其系譜)[M].Bejing(北京):Chinese Agricultural Press (中國農(nóng)業(yè)出版社),2007.

[2]CUI RM(崔瑞敏).Cotton Varieties in Hebei Province(河北省棉花品種志)[M].Shijiazhuang(石家莊):Hebei Science and Technology Publishing House (河北科學技術出版社),2013.

[3]WANG HQ(王恒銓).Hebei Cotton(河北棉花)[M].Shijiazhuang(石家莊):Hebei Science and TechnologyPublishing House(河北科學技術出版社),1992.

[4]LI Z(李胄),LIU YL(劉有良),LIU DP(劉黨培).Study on correlativity between lint yield and plant character in upland cotton(陸地棉皮棉產(chǎn)量與植株性狀相關性研究)[J].Acta Agriculturae Boreali-Occidentalis Sinica (西北農(nóng)業(yè)學報),1997,6(3):44-49.

[5]HE TJ(何團結).(抗蟲雜交棉農(nóng)藝性狀相關性分析)[J].(安徽農(nóng)業(yè)科學),2003,31(6):929-930,943.

[6]LI YS(李永山),TANG BH(唐秉海),ZHANG K(張凱),et al.Studies on yield components,fiber qualities and pedigree analysis of varieties released in different years(不同年代棉花品種產(chǎn)量構成、纖維品質及其系譜分析)[J].Cotton Science (棉花學報),2001,13(1):16-19.

[7]GU SP(顧雙平),CHANG XY(常曉陽).Correlation analysis of characters of cotton fiber quality(棉花纖維品質性狀的相關剖析)[J].Acta Agriculturae Jiangxi(江西農(nóng)業(yè)學報),2002,14(2):24-28.

[8]LIU YX(劉英欣),HAN XM(韓祥銘),SONG XL(宋憲亮),et al.Genetical correlation and differences of the main economic charactersin upland cotton(陸地棉主要經(jīng)濟性狀的遺傳相關及遺傳差異)[J].Cotton Science(棉花學報),2001,13(2):95-99.

[9]AZHAR FM,NAVEED M,ALI A.Correlation analysis of seed cotton yield with fiber characteristics inGossypium hirsutumL[J].International Journal of A-griculture and Biology,2004,6(4):656-658.

[10]IQBAL M,KHAN RSA,SADIQ A,et al.Correlation and path coefficient analysis for earliness and yield traits in cotton(G.hirsutumL.)[J].Asian Journal ofPlant Sciences,2006,5(2):341-344.

[11]QI ZJ(齊子杰),LIU GP(劉光萍).Major factor and path analysis of quantitative characters of new cotton varieties of Henan(棉花新品種數(shù)量性狀的主成分分析與通徑分析)[J].Journal of Henan Institute of Science and Technology(河南科技學院學報),2009,37(4):7-10.

[12]CHEN RJ(陳榮江),WANG WF(王文峰),ZHU MZ(朱明哲).Canonical relation and principal components analysis of new varieties of cotton yield and quality characteristics(棉花新品種產(chǎn)量性狀與品質性狀的典型相關與主成分分析)[J].System Sciences and Comprehensive Studies in Agriculture (農(nóng)業(yè)系統(tǒng)科學與綜合研究),2007,23(2):187-189.

[13]WANG PZ(王沛政),YAN S(煙杉),GAO X(高翔).Genetical correlation and principal component analysis on quantity character in upland cotton with the drought/salt resistance and dwarf characters(抗旱、耐鹽堿、矮稈陸地棉數(shù)量性狀的遺傳相關及主成分分析)[J].Cotton Science (棉花學報),2002,14(3):154-157.

[14]SU JJ(宿俊吉),DENG FJ(鄧福軍),CHEN H(陳紅),et al.Analysis of correlation,cluster and variation on the main agronomic characters of upland cotton cultivars of lines(陸地棉主要農(nóng)藝性狀的變異性、聚類和相關性分析)[J].Xinjiang Agricultural Sciences(新疆農(nóng)業(yè)科學),2011,48(8):1386-1391.

[15]WANG PZ(王沛政),CHEN AM(陳愛民),ABU LET(阿布來提),et al.Principal component and clustering analysis of quantitative characters of upland cotton with different origin(不同來源陸地棉數(shù)量性狀主成分及聚類分析)[J].Acta Agriculturae Boreali-Occidentalis Sinica(西北農(nóng)業(yè)學報),1999,8(4):50-52.

[16]CHEN RJ(陳榮江),ZHU MZ(朱明哲),SUN CF(孫長法).Comprehensive evaluation and clustering analysis of yield and quality characters of new cotton varieties(棉花新品種產(chǎn)量品質性狀的綜合評價及聚類分析)[J].Acta Agriculturae Boreali-Occidentalis Sinica(西北農(nóng)業(yè)學報,2007,16(4):264-268.

[17]TANG QY(唐啟義).DPS Data Processing System(DPS數(shù)據(jù)處理系統(tǒng))[M].Beijing(北京):Science Press(科學出版社),2010.

[18]HUA JP(華金平),ZHAN XJ(詹先進).Discussion of study direction of cotton breeding in Hubei(再論湖北棉花育種的研究方向)[J].China Cotton(中國棉花),2001,28(10):6-8.

[19]YU SX(喻樹迅),FAN SL(范術麗).The evolutions and prospect of cotton genetics and breeding in Chian(我國棉花遺傳育種進展與展望 [J].Cotton Science(棉花學報),2003,15(2):120-124.

[20]PAN JJ(潘家駒),ZHANG SZ(張樹榛),LIU JL(劉記麟),et al.Pandect of Crop Breeding(作物育種學總論)[M].Bejing(北京):Chinese Agricultural Press(中國農(nóng)業(yè)出版社),1994.

[21]GUO XM(郭香墨),FAN SL(范術麗),WANG HM(王紅梅),et al.Achivements of technical innovation about cotton genetics and breeding in China(我國棉花育種技術的創(chuàng)新與成就)[J].Cotton Science(棉花學報),2007,19(5):323-330.

[22]JOANNE CD.Ecology and resistance management forBacillus thuringiensistransgenic plants[J].Biocontrol Science and Technology,1994,4(4):563-571.

[23]QIN YH(秦永華),QIAO ZX(喬志新),LIU JY(劉進元).Application of transgenic technology in cotton breeding(轉基因技術在棉花育種上的應用)[J].Cotton Science(棉花學報),2007,19(6):482-488.

[24]WANG RH(汪若海),LI XL(李秀蘭).Present situation and proposals for the Appl-resistant cotton in China(我國轉基因抗蟲棉應用現(xiàn)狀及建議)[J].Biotechnology Information (生物技術通報),2000(5):1-6.

[25]PERLAK FJ,FUCHS R L,DEAN DA,et al.Modification of the coding sequence enhances plant expression of insect control protein genes[J].Proc Natl Acad Sci USA,1991,88(8):3324-3328.

[26]MCGAUGHEY WH.Insect resistance to the biological insecticide Bacillus thuringiensis[J].Science,1985,229(4709):193-195.

[27]ZHAI LS(翟禮生).Research on Integrated Physical Regionalization and Building System in Town,Village,Region,Province of China(中國省域村鎮(zhèn)建筑綜合自然區(qū)劃與建筑體系研究)[M].Beijing(北京):Geological Publishing House(地質出版社),2008.

[28]CHENG SL(程樹林).Hebei Agroclimate and Compartment(河北省農(nóng)業(yè)氣候及其區(qū)劃)[M].Beijing (北京):China Meteorological Press(氣象出版社),1988.

[29]SHI ZL(石振禮),WANG ZF(王整風).Currentsituation and countermeasures of the development of cotton industrialization in Hebei(關于河北棉花產(chǎn)業(yè)化發(fā)展的現(xiàn)狀與對策)[J].China’s Cotton and Linen Economic(中國棉麻經(jīng)濟),2004(5):23-25.