Stability of QTLs for rice grain dimension and
ORIGINAL PAPER
X.Y.Wan ?J.M.Wan ?J.F.Weng ?L.Jiang J.C.Bi ?C.M.Wang ?H.Q.Zhai
Stability of QTLs for rice grain dimension and endosperm chalkiness characteristics across eight environments
Received:15October 2004/Accepted:21February 2005/Published online:5April 2005óSpringer-Verlag 2005
Abstract Rice appearance quality,including traits specifying grain dimension and endosperm chalkiness,represents a major problem in many rice-producing areas of the world.In this study,the genetic basis of six appearance quality traits of milled rice was dis-sected into quantitative trait loci (QTL)main e?ects,and the stability of these QTLs was assessed in a population of 66chromosome segment substitution lines (CSSLs)across eight environments.The CSSLs showed transgressive segregation for many of the traits,and signi?cant correlations were detected among most of the traits.Twenty-two QTLs were identi?ed on eight chromosomes,and numerous QTLs a?ecting related traits were mapped in the same regions,probably re?ecting pleiotropic e?ects.Nine QTLs,namely qGL-1,qGL-3,qGW-5,qLWR-3,qLWR-5,qPGWC-8,qPGWC-9,qACE-8,and qDEC-8,were consistently detected across the eight environments.The additive main e?ect and multiplicative interaction (AMMI)analysis showed that genotype (G)·envi-ronment (E)interaction was signi?cant for all six traits,with the ?rst three iPCA terms accounting for over 80%of the G ·E variance.Both D I values and the iPCA1-iPCA2biplots showed that the CSSLs harboring the nine QTL alleles were more stable than those carrying any of the additional 13QTL alleles,
thereby con?rming their environmental stability and pointing to their appropriateness as targets for marker-assisted selection for high-quality rice varieties.
Introduction
Rice appearance quality is determined mainly by grain length (GL),grain width (GW),length-width ratio (LWR),percentage of grains with chalkiness (PGWC),area of chalky endosperm (ACE),and degree of endo-sperm chalkiness (DEC).For improvement of milling,eating,and cooking quality,the endosperm of high-quality rice varieties should be free of chalkiness,since chalky grains have a lower density of starch granules than vitreous ones and are therefore more prone to breakage during milling (Del Rosario et al.1968).Also,since both longitudinal and transverse cracks occur easily in chalky kernels when the grain is steamed or boiled,chalkiness reduces the palatability of cooked rice (Nagato and Ebata 1959).On the other hand,rice cul-tivars with chalkiness are useful for the production of special food,an example of which is Japanese sake.Similarly,preferences for rice grain shape vary across consumer groups.Long and slender grain varieties are preferred by consumers in most Asian countries,including China,India,Pakistan,and Thailand,and in the USA,while short and bold grain cultivars are pre-ferred in Japan and Sri Lanka.Therefore,breeding for the appropriate grain shape and endosperm opacity needs to be considered in the context of market requirements.
Rice appearance quality traits are quantitatively inherited (Tan et al.2000).The identi?cation of quantitative trait loci (QTLs)for appearance quality and the elucidation of their genetic control are nec-essary for the development of marker-assisted selection (MAS)strategies aimed at improving breeding e?https://www.wendangku.net/doc/e71463409.html,ing a number of primary mapping
Communicated by D.J.Mackill
X.Y.Wan ?J.M.Wan ?J.F.Weng L.Jiang ?J.C.Bi ?C.M.Wang
National Key Laboratory for Crop Genetics and Germplasm Enhancement,
Jiangsu Plant Gene Engineering Research Center,Nanjing Agricultural University,Nanjing,210095,China
J.M.Wan (&)?H.Q.Zhai Institute of Crop Science,
Chinese Academy of Agriculture Sciences,Beijing,100081,China
E-mail:wanjm@https://www.wendangku.net/doc/e71463409.html, Tel.:+86-10-62186628Fax:+86-10-62186628
Theor Appl Genet (2005)110:1334–1346DOI 10.1007/s00122-005-1976-x
populations,Huang et al.(1997)detected12QTLs for GL,GW,and LWR,while He et al.(1999)identi?ed three QTLs a?ecting PGWC and ACE.Redon a and Mackill(1998)found that GL,GW,and LWR were controlled primarily by two loci,one each on chro-mosomes3and7,whereas Tan et al.(2000)detected a major GL QTL on chromosome3and a GW QTL on chromosome5.However,since all the above studies were conducted in a single environment,the stability of the resultant QTLs could not be evaluated.This characteristic is critical for determining the utility of a QTL for marker-assisted breeding.
In the present study,66chromosome segment substitution lines(CSSLs),derived from three back-crosses of IR24(indica)to Asominori(japonica) (Asominori/IR24//3*Asominori),were used for QTL identi?cation across eight environments.The use of CSSLs,as opposed to primary mapping populations, has distinct advantages for QTL identi?cation.Most importantly,genetic interactions between donor alleles are limited to those between genes on homozygous substituted tracts since each CSSL carries one or a few donor segments in the near-isogenic background of a recurrent genotype,thus reducing the e?ects of inter-ferences from genetic background(Howell et al.1996). Second,for the?ne mapping and positional cloning of a QTL,a secondary F2population can be derived from a further backcross between a selected CSSL and the recurrent parent(Frary et al.2000;Yano et al. 2000).Third,F2populations generated from inter-crosses between di?erent CSSLs can be used to ana-lyze epistatic interactions between pairs of QTLs(Lin et al.2000).Finally,CSSLs form the germplasm basis for the pyramiding of multiple QTLs in a plant breeding program(Li2001).
The objectives of this study were to detect stable QTLs a?ecting GL,GW,LWR,PGWC,ACE,and DEC,in order to develop MAS assays for appearance quality,and to construct secondary F2populations via backcrossing for the?ne mapping and positional cloning of stable QTLs.Materials and methods
Population development
Seventy-one recombinant inbred lines(RILs),derived from the cross Asominori·IR24,were developed by single-seed descent(Tsunematsu et al.1996).To produce a series of CSSLs in a largely Asominori genetic back-ground,19selected RILs composed of more than60% Asominori genotype were crossed and backcrossed with Asominori,without selection,until the BC3F1genera-tion.Sixty-six individuals were selected at the BC3F1 generation on the basis of a whole genome survey[116 restriction fragment length polymorphism(RFLP)loci]. These66lines,denoted L1–L66,have representation of the whole IR24genome,except for the9.8-cM region de?ned by the interval C1468-G1015on chromosome3 (Kubo et al.1999;Wan et al.2004).
Phenotypic data collection
The CSSLs and the parental varieties were grown in eight environments involving four locations(Table1). Each entry plot consisted of ten rows of ten plants each,grown in a randomized block design with two replications of each entry per environment.After drying(to about13.5%of the moisture content),the grain was stored at room temperature for3months and then milled using the Yamamoto Rice-Pal VP-30T with the following procedure:200g of puri?ed paddy rice per CSSL was put once into the miller and milled for60s(Yamamoto et al.1995).The milled rice thus obtained was used for measuring the appearance quality traits.
The quality traits GL,GW,and LWR were measured following the method of Tan et al.(2000).GL and GW were estimated from the mean of20grains.LWR,which represents the shape of the grain,was given by the ratio GL/GW.PGWC,ACE,and DEC were evaluated
Table1The test environments in which the Asominori/IR24CSSL population was evaluated
Code Number of replications Environments
E12Nanjing Agricultural University,Nanjing,China,N31.2^A°,E118.4^A°,
May–October,2001
E22Experimental Farm of Jinhu County,Huai’an,Jiangsu,China,N32.7^A°,E119.6^A°,
May–October,2001
E32Experimental Farm of Donghai County,Lianyungang,Jiangsu,China,N35.1^A°,E118.4^A°,
June–November,2001
E42Rice Breeding Base of Lingshui County,Sanya,Hainan,China,N18.2^A°,E108.9^A°,
December,2001–May,2002
E52Nanjing Agricultural University,Nanjing,China,N31.2^A°,E118.4^A°,May–October,2002 E62Experimental Farm of Jinhu County,Huai’an,Jiangsu,China,N32.7^A°,E119.6^A°,
May-October,2002
E72Experimental Farm of Donghai County,Lianyungang,Jiangsu,China,N35.1^A°,E118.4^A°,
June–November,2002
E82Rice Breeding Base of Lingshui County,Sanya,Hainan,China,N18.2^A°,E108.9^A°,
December,2002–May,2003
1335
according to He et al.(1999)and NSPRC(1999).To separate chalky from vitreous grains,we assess100 grains per entry on a chalkiness visualizer constructed at the China National Rice Research Institute(NSPRC 1999);on the basis of these observations,we calculated PGWC.Twenty chalky grains were then selected at random,and the ratio of the area of chalkiness to the area of the whole endosperm for each grain was evalu-ated by human visual assessment on the chalkiness visualizer.The values were averaged and used as values for ACE.DEC was calculated as the product PGWC·ACE.
All six traits were measured on three replications per sample.
Linkage map construction
Genomic DNA was extracted by the CTAB method (Murray and Thompson1980).The DNA clones
Table2Summary statistics of phenotypic performance of the CSSL population and its parents for six quality traits in eight environments (SD standard deviation)
Traits a Environments Parents Population
Asominori IR24Range Mean±SD CV(%)b GL(mm)E1 5.3 5.9 4.9–5.8 5.3±0.2 3.7 E2 5.3 5.9 4.9–5.8 5.3±0.2 3.5
E3 5.3 6.0 4.9–5.8 5.4±0.2 3.5
E4 5.2 5.8 4.8–5.8 5.2±0.2 3.8
E5 5.3 5.8 4.9–5.8 5.3±0.2 3.8
E6 5.3 5.9 5.0–5.9 5.3±0.2 3.7
E7 5.3 6.0 4.9–5.8 5.4±0.2 3.6
E8 5.3 5.9 4.9–5.8 5.3±0.2 3.6
GW(mm)E1 2.7 2.4 2.3–2.8 2.6±0.1 4.3 E2 2.8 2.4 2.3–2.9 2.7±0.1 4.4
E3 2.8 2.5 2.3–3.0 2.7±0.1 4.8
E4 2.8 2.5 2.3–2.8 2.6±0.1 4.0
E5 2.8 2.5 2.3–2.8 2.6±0.1 4.2
E6 2.7 2.4 2.3–2.9 2.7±0.1 4.3
E7 2.8 2.5 2.2–3.0 2.7±0.1 5.4
E8 2.8 2.5 2.3–2.8 2.6±0.1 4.1 LWR E1 2.0 2.4 1.8–2.4 2.0±0.1 6.5 E2 1.9 2.5 1.8–2.3 2.0±0.1 6.4
E3 1.9 2.4 1.7–2.4 2.0±0.1 6.9
E4 1.9 2.3 1.8–2.4 2.0±0.1 6.4
E5 1.9 2.4 1.8–2.4 2.0±0.1 6.2
E6 1.9 2.4 1.7–2.3 2.0±0.1 6.6
E7 1.9 2.4 1.7–2.3 2.0±0.17.0
E8 2.0 2.3 1.8–2.4 2.0±0.1 6.5 PGWC(%)E125.020.0 1.3–85.928.4±22.880.4 E226.518.0 1.5–84.026.9±22.583.6
E329.018.0 1.0–91.534.3±27.078.8
E427.016.00.5–94.532.6±26.280.1
E529.021.5 3.8–92.433.0±24.775.0
E628.318.3 2.5–90.533.3±24.573.5
E728.514.8 1.5–94.834.5±26.978.1
E825.016.00.5–88.532.4±25.980.2 ACE(%)E1 4.19.2 2.2–18.77.4±3.547.1 E2 5.38.4 1.1–24.37.6±5.775.1
E3 4.58.10.5–31.37.8±6.684.4
E4 5.810.80.5–20.0 5.2±3.976.0
E5 4.38.2 1.2–28.28.2±5.364.9
E6 5.710.4 1.5–23.18.5±5.969.0
E7 4.29.1 1.2–23.8 6.7±5.074.3
E8 4.38.80.3–19.6 4.8±4.083.3 DEC(%)E1 1.0 1.80.1–15.9 2.6±3.5137.1 E2 1.4 1.50.1–20.5 2.8±4.3154.9
E3 1.3 1.50.0–20.1 3.4±4.9141.8
E4 1.6 1.70.0–17.5 2.4±3.8159.5
E5 1.3 1.80.1–16.8 3.1±3.7119.5
E6 1.6 1.90.1–20.7 3.8±5.0130.6
E7 1.2 1.40.0–21.7 3.2±4.4140.6
E8 1.1 1.40.0–17.0 2.2±3.6162.4
a GL,Grain length;GW,grain width;LWR,length-width ratio;PGWC,percentage of grains with chalkiness;ACE,area of chalky endosperm;DEC,degree of endosperm chalkiness.See Materials and methods for details of parameter calculations
b CV,Coe?cient of variation
1336
mapped by Tsunematsu et al(1996)were used as probes. DNA labeling,hybridization,and signal detection were conducted using the ECL detection system(Amersham, UK).For the whole genome survey,116RLFP loci distributed over the framework map were used in the BC3F1generation.A linkage map was constructed with 85RFLP markers evenly distributed over all12chro-mosomes using MAPMAKER/EXP3.0(Lander et al.1987). The overall map length was1,275.4cM with an average distance of15.0cM between adjacent markers,as re-ported by Kubo et al.(1999).
Data analysis
Tests of homogeneity of variance were conducted to determine whether data from multiple environments could be pooled to conduct a combined analysis of variance(ANOVA)across environments.For the com-bined analyses,the variance was partitioned into geno-type(CSSL),year,site,year·site,Rep(year·site), CSSL·year,CSSL·site,and year·CSSL·site. Phenotypic correlations among traits were estimated using the mean phenotypic values of the CSSL popula-tion,combined across the eight environments.
Following classical quantitative genetics theory (Falconer1981),the phenotypic value of a CSSL(y ihk)is described by the genetic model:
y ihk?ltG itE htGE ihte ihk;e1Twhere y ihk is the trait value for the i th CSSL(I=1,2,..., 66)of the k th?eld replication in the h th environment,l is the CSSL population mean,G i is the genotypic e?ect
(?xed)of the i th CSSL,E h$Ne0;r2
E T
is the random e?ect of
the h th environment(h=1,2,...,8),GE ih$Ne0;r2
GE T
is the
interaction e?ect between the i th CSSL and the h th envi-
ronment,and e ihk$Ne0;r2
E T
is the random residual e?ect.
The QTL parameters were estimated by composite interval mapping,a combination of interval mapping with multiple regression analysis(Zeng1994).To ob-tain empirical thresholds of the experiment,1,000per-mutations were run by randomly shu?ing the trait values(Churchill and Doerge1994).Based on the permutation test,a LOD value of 3.0was used for claiming a signi?cant main-e?ect QTL.The QTL mapping was performed based on the data in each environment using the JMP VER.3.1software package (SAS Institute1994).
The QTL e?ects were evaluated by a t-test to test the presence of signi?cant di?erences between the pheno-typic values of Asominori and those of CSSLs harboring QTL alleles derived from IR24.
The stability of a QTL was estimated using the additive main e?ects and multiplicative interaction (AMMI)model(Crossa et al.1990;Gauch and Zobel 1988)by analyzing the genotype·environment inter-action(GEI)for individual traits and thereby generating stability parameters in several dimensions.The AMMI analysis was performed using software described by Vargas and Crossa(2000).The stability parameter(D i) of the i th CSSL is calculated by:
D i?
?????????????????
X3
c?1
k c h2
ic
v u
u t
e2T
(Zhang et al.1998;Vargas et al.1999)where k c is the singular value of the c th principal components analysis (PCA)axis and h ic is the principal component values for the c th PCA axis of the i th CSSL.
The smaller the D i value,the more stable the i th CSSL is across multi-environments.When the?rst two iPCA values account for a majority of the total GEI sum of squares(SS),the AMMI biplot with the abscissa of iPCA1and the ordinate of iPCA2can also be applied to evaluate the stability of the i th CSSL.
Results
Variation of phenotypic traits
Table2shows the phenotypic variation of the CSSLs and their parents for the six quality traits across the
Table3Analysis of variance for six appearance quality traits a in the CSSL population across two years and four sites
Sources df GL GW LWR PGWC ACE DEC
MS F values MS F values MS F values MS F values MS F values MS F values Year10.25862.86**0.01711.60**0.005 3.96*2117.5566.94**101.2910.19**21.4615.95** Site30.683166.16** 1.000680.87**0.171129.97**970.0430.67**520.8052.41**56.4241.94** Year·Site30.06616.05**0.003 2.380.01611.90**697.15**22.04**63.09 6.35**24.5418.24** Rep
(Year·site)
80.10525.62**0.008 5.15**0.0129.13**336.91**10.65**30.68 3.09** 5.54 4.12** CSSL650.556135.31**0.203138.36**0.258196.00**9220.84291.50**267.0626.87**244.72181.90** CSSL·Year650.005 1.100.0010.790.0010.87122.39 3.87**18.01 1.81** 5.19 3.86** CSSL·Site1950.010 2.48**0.007 4.51**0.004 3.18**180.87 5.72**29.38 2.96**8.01 5.95** CSSL·Year*
Site
1950.0020.520.0010.950.0010.8969.90 2.21**13.96 1.41** 2.38 1.77** Error5200.0040.0010.00131.639.94 1.35
*,**Signi?cant at P￡0.05and0.01,respectively
a Abbreviations are the same as in Table2
1337
T a b l e 4Q T L s a ?e c t i n g t h e s i x q u a l i t y t r a i t s d e t e c t e d u s i n g t h e C S S L p o p u l a t i o n a c r o s s e i g h t e n v i r o n m e n t s
T r a i t a
Q T L s C h r o m o s o m e
M a r k e r i n t e r v a l
P a r a m e t e r b E 1E 2E 3E 4E 5E 6E 7E 8Q T L -C S S L s
G L q G L -11
R 210-C 955
L O D 5.34.07.45.95.65.07.16.6L 1,L 2,L 4,L 9
P V E 17.710.720.615.418.218.424.320.3A E à0.15à0.11à0.17à0.15à0.16à0.15à0.18à0.16q G L -33
R 19-C 1677
L O D 5.98.75.79.74.75.05.26.8L 16,L 17,L 18,P V E 24.551.121.649.319.535.518.542.1L 46
A E 0.290.270.240.260.270.250.240.24
q G L -22
C 601-R 3393
L O D 3.95.03.0L 13,L 14,L 61
P V E 12.414.010.6A E à0.15à0.16à0.14q G L -44X N p b 331-C 335L O D 3.16.24.84.23.8L 25,L 26,L 27
P V E 7.012.715.19.811.1A E à0.12à0.16à0.15à0.15à0.12G W
q G W -55R 3166-R 569
L O D 5.47.97.55.25.27.26.65.3L 28,L 29
P V E 23.233.930.918.729.632.627.819.2A E à0.16
à0.16
à0.19à0.11
à0.17à0.19
à0.23
à0.11q G W -11
X N p b 113-C 112
L O D 3.4L 7,L 9,L 37
P V E 10.6A E à0.13
q G W -66C 991-X N p b 12
L O D 3.9L 9,L 11,L 19,P V E 13.7L 37,L 38,L 39,A E à0.08L 61L W R
q L W R -33R 19-C 1677
L O D 3.94.24.34.13.63.73.64.0L 16,L 17,L 18,P V E 17.324.023.315.021.021.721.820.7L 46
A E 0.150.150.150.150.140.150.150.14q L W R -55
R 3166-R 569
L O D 6.19.06.36.66.96.66.15.4L 28,L 29
P V E 23.532.226.624.628.931.923.719.6A E 0.180.180.17
0.190.160.15
0.190.17q L W R -11
R 210-C 955
L O D 3.23.34.03.4L 1,L 2,L 4,L 9
P V E 11.713.418.215.3A E à0.10
à0.08
à0.12à0.10
q L W R -22X N p b 67-X N p b 132
L O D 3.2L 7,L 10,L 11,P V E 12.2L 12
A E à0.14P G W C
q P G W C -88G 1149-R 727
L O D 7.36.16.25.95.35.26.45.4L 49,L 50,L 51
P V E 21.717.527.223.719.629.827.222.9A E 25.922.033.130.226.821.833.727.9q P G W C -99
X N b p 36-X N p b 103
L O D 5.03.94.73.03.13.14.43.7L 52,L 53,L 64
P V E 12.611.719.012.611.217.819.915.8A E 32.931.422.316.833.832.7
20.919.3q P G W C -11
C 2340-C 1370
L O D 4.35.33.14.4L 5,L 6,L 7,L 8
P V E 18.518.316.318.7A E 27.129.319.727.2A C E
q A C E -88G 1149-R 727
L O D 6.28.63.214.14.54.36.512.6L 49,L 50,L 51
P V E 31.539.220.240.624.522.029.334.3A E 5.67.2
5.9
6.1
4.9
5.8
5.65.8
q A C E -22
G 1340-R 459
L O D 3.3L 7,L 10
P V E 20.7A E 3.9q A C E -99X N b p 36-X N p b 103
L O D 3.93.03.5L 52,L 53,L 64
P V E 11.114.515.7A E
4.84.0
3.9
1338
eight environments.Signi?cant di?erences were ob-served for all of the traits between the parental cultivars.The CSSL population segregated transgressively for PGWC,ACE,and DEC.The mean trait value of the CSSLs was largely consistent across the di?erent envi-ronments,except that ACE and DEC both showed lower phenotypic values in 2001,2002,and in Hainan,
indicating that the Hainan environment (N 18.2^A
°,E 108.9^A
°)tends to lessen the occurrence of chalkiness.The coe?cients of variation varied considerably among the traits,ranging from 3.7%for GL to 143.3%for DEC.For GL,GW and LWR,the ANOVA indicated highly signi?cant e?ects due to genotype (CSSL),year,site,year ·site,Rep (year ·site),and CSSL ·site for all traits,but non-signi?cant ones due to the year ·CSSL and year ·site ·CSSL interactions (Table 3).
Signi?cant or highly signi?cant correlations were observed between GL and GW (à0.297*),GL and LWR (0.766**),GW and LWR (à0.835**),PGWC and ACE (0.691**),PGWC and DEC (0.876**),and ACE and DEC (0.885**).In addition,GW was posi-tively correlated with PGWC,ACE,and DEC,but no correlations were signi?cant between GL and any of the three chalkiness traits,suggesting that the width of rice grains is closely associated with the occurrence of chalkiness.QTL analysis
Twenty-two QTLs for the six traits were identi?ed in the eight environments and mapped to eight chromosomes with LOD values between 3.0and 23.8(Table 4,Fig.1).Of the 11GL,GW,and LWR QTLs,?ve (qGL-1,qGL-3,qGW-5,qLWR-3,and qLWR-5)were consis-tently detected across all eight environments.qGL-1and qGL-3were mapped to chromosomes 1and 3with the average percentage of phenotypic variation explained (PVE)of 18.2%and 32.8%,respectively.The IR24allele at the qGL-3locus increased GL by an average of 0.26mm,while the positive e?ect of the qGL-1locus was contributed by the Asominori allele.qGW-5was mapped to chromosome 5and accounted for 27.0%of the phenotypic variation,with the IR24allele providing a negative e?ect of 0.17mm.LWR was mostly con-trolled by loci on chromosomes 3and 5;qLWR-3and qLWR-5coincided with the major GL and GW QTLs.On average,qLWR-3and qLWR-5explained 20.6%and 26.4%of the total variation,respectively.The positive e?ects at both loci were contributed by IR24alleles.Six additional QTLs (qGL-2,qGL-4,qGW-1,qGW-6,qLWR-1,and qLWR-2)were identi?ed in one to ?ve environments,with a PVE of 7.0–18.2%.
Three PGWC,ACE,and DEC QTLs (qPGWC-8,qACE-8,and qDEC-8),were mapped in the interval G1149-R727on chromosome 8across all eight environ-ments and accounted,on average,for 23.7%,30.2%,and 37.4%of the phenotypic variation,respectively.In
D E C
q D E C -8
8
G 1149-R 727L O D 17.213.48.020.010.510.314.123.8L 49,L 50,L 51
P V E 38.848.523.640.037.834.833.242.8A E 5.96.96.76.35.5
7.07.65.8q D E C -9
9
X N b p 36-X N p b 103
L O D 4.34.66.28.14.05.28.3L 52,L 53,L 64
P V E 5.18.713.313.415.615.311.7A E 3.25.26.4
5.77.1
4.6
5.1q D E C -1a
1R 210-C 955
L O D 3.43.34.24.7L 1,L 2,L 4,L 9
P V E 5.26.412.98.4A E 1.8
2.7
2.5
2.0q D E C -1b
1
C 2340-C 1370
L O D 4.8L 5,L 6,L 7,L 8
P V E 5.8A E 1.7q D E C -22G 1340-R 459
L O D 3.33.8L 7,L 10
P V E 7.35.5A E
2.6
1.7
a
A b b r e v i a t i o n s a r e t h e s a m e a s i n T a b l e 2b P V E a n d A E r e p r e s e n t t h e p e r c e n t a g e o f p h e n o t y p i c v a r i a t i o n e x p l a i n e d a n d t h e a d d i t i v e e ?e c t o f t h e Q T L ,r e s p e c t i v e l y .T h e s e w e r e b o t h e s t i m a t e d f r o m m e a n t r a i t v a l u e s o f i n d i v i d u a l C S S L s i n t h e i n d i v i d u a l e n v i r o n m e n t s
1339
the interval XNpb36-XNpb103on chromosome 9,qPGWC-9was consistently identi?ed in all eight environments,while qACE-9and qDEC-9were only expressed in three and seven environments,respectively.The positive e?ects of these six QTLs were all contributed by IR24alleles.Five QTLs (qPGWC-1,qACE-2,qDEC-1a,qDEC-1b ,and qDEC-2)were identi?ed in one to four environments,with a PVE of 5.2–20.7%.E?ects of QTLs with high repeatability
Among the 22QTLs detected,nine (qGL-1,qGL-3,qGW-5,qLWR-3,qLWR-5,qPGWC-8,qPGWC-9,qACE-8,and qDEC-8)were detected across all eight environments (Table 4).qGL-1was located in the interval R210-C955.Four CSSLs (L1,L2,L4,and L9)carried IR24alleles in this chromosome region (Table 4).qGL-3and qLWR-3were both mapped in the interval R19-C1677.Four CSSLs (L16,L17,L18,and L46)carried IR24alleles at R19and C1677.Similarly,two or three CSSLs carried one of the qGW-5,qLWR-5,qPGWC-8,qPGWC-9,qACE-8,and qDEC-8alleles (Table 4).t-tests demonstrated signi?cant di?erences between the phenotypic values of Asominori and each CSSL carrying one of the nine QTL alleles (Table 5),indicating that the e?ects of these QTLs were signi?cant
and repeatable across the eight environments.However,the other 13QTLs were environment-speci?c,as signif-icant e?ects were detected in only one to seven environments.
Stability of QTL e?ects
The AMMI analysis demonstrated signi?cant GEI for all six traits (Table 6),indicating that QTL-by-environ-ment interactions (QEI)were signi?cant for each trait studied.It was therefore necessary to analyze the sta-bility of the CSSLs harboring the detected QTL alleles and to assess the stability of individual QTLs across the eight environments.The ?rst three iPCA terms of each trait accounted for over 80%of the GEI SS,indicating that the stability of the CSSLs can be precisely evaluated with the D i values calculated from the ?rst three iPCA values (Table 6).For GL QTLs,the average D i value of four CSSLs (L1,L2,L4,and L9)harboring the qGL-1allele was 0.13,and those of the CSSLs carrying the qGL-2,qGL-3and qGL-4alleles were 0.24,0.13and 0.31,respectively (Table 7).Using the abscissa of iPCA1and the ordinate of iPCA2in the AMMI biplot,we observed that eight CSSLs (L1,L2,L4,L9,L16,L17,L18,and L46)harboring the qGL-1and qGL-3alleles were closer to the origin,while six CSSLs (L13,
L14,
Fig.1Map locations of
identi?ed QTLs a?ecting grain length (GL ),grain width (GW ),length-width ratio (LWR ),percentage of grains with chalkiness (PGWC ),area of chalky endosperm (ACE ),and degree of endosperm chalkiness (DEC )of Asominori ·IR24CSSL population,detected in the eight environments.8Es represents QTLs detected in eight environments,such as in E1,E2,E3,E4,E5,E6,E7,and E8.The same terminology is used for all of the following QTLs;i.e.,1Es is one environment;2Es ,two environments,etc
1340
T a b l e 5t -t e s t f o r t h e d i ?e r e n c e s o f p h e n o t y p i c v a l u e s b e t w e e n A s o m i n o r i a n d t h e C S S L c a r r y i n g a n y o n e o f t h e Q T L a l l e l e s w i t h h i g h r e p e a t a b i l i t y i n e i g h t e n v i r o n m e n t s
T a r g e t Q T L s a n d C S S L s E 1a
E 2
E 3
E 4E 5E 6E 7E 8
M e a n s b
P
M e a n s
P
M e a n s P M e a n s P M e a n s P M e a n s P M e a n s P M e a n s P
q G L -1
A s o m i n o r i A s o 5.35.35.35.25.35.35.35.3L 15.00.034*c
5.00.023*5.10.004**5.00.025*5.00.006**5.10.017*5.10.003**5.10.010**L 25.00.031*d
5.10.033*5.10.005**5.00.017*5.00.006**5.00.012*5.00.003**5.00.012*L 45.10.027*e
5.10.038*5.10.003**5.00.005**5.10.007**5.00.027*5.10.003**5.10.008**L 95.00.01**f 5.00.039*
5.00.001**4.90.011*4.90.003**5.00.058*
4.90.002**
5.00.006**
q G L -3
A s o m i n o r i 5.35.35.35.25.35.35.35.3L 165.80.007**5.80.004**5.80.006**5.80.004**5.80.002**5.90.004**5.80.011*5.80.006**L 175.80.006**5.80.004**5.80.001**5.80.001**5.80.003**5.80.008**5.80.014*5.80.006**L 185.70.025*5.70.007**5.70.008**5.60.002**5.60.010**5.70.013*5.70.016*5.70.010**L 465.80.015*
5.80.012*
5.70.001**5.80.002**5.80.009**5.80.008**
5.80.015*
5.80.002**
q G W -5
A s o m i n o r i 2.72.82.82.82.82.72.82.8L 282.40.001**2.30.002**2.40.001**2.30.022*2.30.003**2.40.009**2.40.013*2.30.004**L 292.30.000**
2.40.005**
2.40.002**2.30.006**2.30.003**2.40.001**
2.40.001**
2.40.010**
q L W R -3
A s o m i n o r i 2.01.91.91.91.91.91.92.0L 162.30.003**2.30.007**2.20.016*2.30.006**2.30.023*2.30.002**2.20.008**2.20.008**L 172.30.003**2.20.008**2.20.000**2.30.006**2.30.009**2.20.004**2.20.017*2.30.003**L 182.30.015*2.20.008**2.30.003**2.30.020*2.20.009**2.20.003**2.30.003**2.30.014*L 462.30.005**
2.30.008**
2.30.007**2.30.008**2.30.032*
2.30.011*
2.30.001**
2.30.017*
q L W R -5
A s o m i n o r i 2.01.91.91.91.91.91.92.0L 282.30.007**2.30.005**2.40.010**2.30.014*2.30.015*2.30.007**2.30.001**2.30.002**L 292.40.001**
2.30.006**
2.30.014*2.40.007**2.40.011*
2.30.007**
2.30.005**
2.40.002**
q P G W C -8
A s o m i n o r i 25.026.529.027.029.028.328.525.0L 4985.00.003**84.00.011*90.00.003**87.00.003**89.50.002**90.50.011*91.00.000**88.50.002**L 5084.30.006**84.00.003**83.00.004**81.00.001**82.60.002**83.30.004**85.00.005**80.50.001**L 5183.50.005**
75.00.001**
91.50.002**94.50.000**
90.80.004**
90.50.008**
94.80.002**
87.30.014*
q P G W C -9
A s o m i n o r i 25.026.529.027.029.028.328.525.0L 5277.80.005**71.50.000**84.00.004**86.50.000**84.10.005**79.30.023*87.50.004**83.80.003**L 5376.70.003**76.50.003**78.50.006**74.50.003**77.40.001**79.00.003**76.80.005**77.30.004**L 6485.90.003**
81.50.000**
89.50.003**
94.00.000**
92.40.005**
88.80.015*
92.30.002**
87.00.014*
q A C E -8
A s o m i n o r i 4.15.34.55.84.35.74.24.3L 4918.70.026*24.30.015*22.40.017*20.00.047*18.90.042*23.10.016*23.80.006**19.20.004**L 5018.50.056*18.60.005**21.50.004**18.70.008**18.90.042*18.10.041*20.40.005**19.60.039*L 5114.40.016*
15.60.012*
15.30.013*
14.60.025*
16.80.022*
17.10.020*
14.20.011*
14.80.002**
q D E C -8
A s o m i n o r i 1.01.41.31.61.21.61.21.1L 4916.00.022*20.50.029*20.10.008**17.50.041*16.80.019*20.70.001**21.70.006**17.00.001**L 5016.80.036*15.60.005**17.90.003**15.10.002**15.70.041*15.00.013*17.40.012*15.80.031*L 5112.10.011*
11.70.006**
14.00.005**
13.80.011*
15.30.023*
15.60.028*
13.40.002**
12.9
0.008**
*,**S i g n i ?c a n t l e v e l s o f t -t e s t s a t P ￡0.05a n d 0.01,r e s p e c t i v e l y a
E 1–E 8,T h e e i g h t e n v i r o n m e n t s a s i n T a b l e 1b M e a n r e p r e s e n t s t h e t r a i t a v e r a g e v a l u e s o f t w o r e p l i c a t e s i n t h e ?e l d e x p e r i m e n t c ,d ,e ,f V a l u e s d e n o t e t h e s i g n i ?c a n t d i ?e r e n c e b e t w e e n A s o m i n o r i a n d C S S L 1,C S S L 2,C S S L 4,o r C S S L 9i n t h e E 1(2000,N a n j i n g )a t P ￡0.05,P ￡0.05,P ￡0.05,a n d P ￡0.01,c o r r e s p o n d i n g l y .T h e s a m e p a r a m e t e r s a r e t r u e f o r e a c h o f t h e Q T L s i n t h e t a b l e
1341
L61,L25,L26,and L27)carrying the qGL-2and qGL-4 alleles were far from this point(Fig.2A).These results showed that the stability of qGL-1and qGL-3was higher than that of qGL-2and qGL-4across the eight https://www.wendangku.net/doc/e71463409.html,ing the same inference,qGW-5was the most stable among the three GW QTLs(Table7, Fig.2B),and the stability of qLWR-3and qLWR-5was higher than that of qLWR-1and qLWR-2(Table7, Fig.2C).
The average D i value of four CSSLs(L5,L6,L7,and L8)harboring the qPGWC-1allele was3.90,and those of the CSSLs carrying the qPGWC-8and qPGWC-9 alleles were1.44and1.21,respectively(Table7).The four CSSLs(L5,L6,L7,and L8)were far from the origin in the iPCA1-iPCA2biplot(Fig.2D).Therefore, qPGWC-8and qPGWC-9were more stable than qPGWC-1across the eight environments.Similarly,the e?ects of qACE-8and qDEC-8were the most stable of the eight ACE and DEC QTLs(Table7,Fig.2E,F).
In summary,the nine QTLs(qGL-1,qGL-3,qGW-5, qLWR-3,qLWR-5,qPGWC-8,qPGWC-9,qACE-8,and qDEC-8)were relatively stable across the eight envi-ronments based on the presence and direction of sig-ni?cant QTL e?ects,the consistency of QTL detection, and the AMMI analysis of CSSLs harboring consistent QTL alleles.
Clustering of QTLs
The QTLs for related traits detected in at least three environments were frequently identi?ed in the same genome regions(Table4,Fig.1).The GL,LWR and DEC,QTLs were clustered in the interval de?ned by R210and C955on chromosome1.The IR24allele de-creased GL and LWR but increased DEC.Similarly,common GL and LWR QTLs were detected in the interval R19-C1677on chromosome3;in both cases,the IR24allele had a positive e?ect on the traits.In addi-tion,the region de?ned by markers R3166and R569on chromosome5harbored GW and LWR QTLs,al-though these act in opposite directions.Additionally,the PGWC,ACE,and DEC QTLs could be simultaneously mapped in the interval G1149-R727and XNpb36-XNpb103on chromosomes8and9,respectively.The IR24alleles at both loci had the positive e?ects on the three traits.
Discussion
Of the22QTLs identi?ed in this study,nine(41%)were relatively stable across the eight environments(Table4), unlike the outcome reported in other studies[29.7%, 1.5%,0%,and0%in Li et al.(2003),Hittalmani et al. (2003),Teulat et al.(2003),and Campbell et al.(2003), respectively].The high percentage of stable QTLs found in the present study may be due to one or a combination of the following factors.(1)The stable QTLs were responsible for major e?ects and were associated with high LOD scores(average 6.8)and PVE(average 25.7%).As suggested by Tanksley(1993)and Zhuang et al.(1997),QTLs with major e?ects are more likely to be stable across multiple environments.(2)The herita-bility of GL,GW,and LWR all exceeded95%,while those of PGWC,ACE,and DEC were over83%(Lin et al.2001;Xing et al.2001).Highly heritable traits tend to be more repeatable and stable across multiple envi-ronments(Paterson et al.1991).(3)Each CSSL carries a small number of IR24segments in a largely Asominori background,and thus genetic interactions between IR24 alleles are minimized.
Table6Analysis of genotype-by-environment interaction using the AMMI model for six appearance quality traits among the CSSLs harboring22QTL alleles detected in eight environments
Traits Sources Error Environment Genotype Environment·Genotype iPCA1iPCA2iPCA3 GL df127715105211917 SS0.350.18**32.25**0.43*0.18**0.10*0.07
SS%40.922.916.8 GW df9571177171513 SS0.080.11** 2.77**0.42**0.20**0.10**0.05**
SS%46.924.111.4 LWR df127715105211917 SS0.10.07**9.29**0.35**0.14**0.08**0.07**
SS%4022.820.5 PGWC df9571177171513 SS3116.123834.73**115207.79**5774.64**3639.57**1337.52**421.02
SS%6323.27.2 ACE df797963151311 SS524.01212.57**4180.39**775.73**353.03**203.93**119.6
SS%45.526.315.4 DEC df135716112222018 SS328.51247.33**8070.35**974.74**374.73**305.91**151.51** SS%38.431.415.5
*,**Signi?cant levels of t-tests at P￡0.05and0.01,respectively a Abbreviations are the same as in Table2b SS%,the percentage of SS(sum of squares)of iPCA1-3to that of enivronment·genotype
1342
Of the stable QTLs,qGL-1,qPGWC-9,qACE-8,and qDEC-8are reported here for the?rst time,while the remaining?ve QTLs(qGL-3,qGW-5,qLWR-3,qLWR-5,and qPGWC-8)are located in the vicinity of QTLs a?ecting these traits detected in the other mapping populations(Huang et al.1997;Redon a and Mackill 1998;He et al.1999;Tan et al.2000).These?ve QTLs each accounted for a signi?cant proportion of the overall phenotypic variation,with an average PVE of 42.0%,44.0%,24.6%,31.8%,and22.8%,respectively across various studies.This is taken to mean that the e?ects of these?ve QTLs are more stable not just across environments,but also across varied genetic back-grounds.Nine CSSLs(L16,L17,L18,L46,L28,L29,
Table7The stability parameters of all the CSSLs carrying22QTL alleles detected across eight environments
Traits QTLs D i values of parents and the target CSSLs harboring any of the22QTL allele
GL Parents Aso IR24
D i0.170.09
qGL-1CSSL L1L2L4L9
D i0.110.100.140.15
qGL-3CSSL L16L17L18L46
D i0.130.160.050.20
qGL-2CSSL L13L14L61
D i0.280.320.14
qGL-4CSSL L25L26L27
D i0.170.440.33
GW Parents Aso IR24
D i0.170.25
qGW-5CSSL L28L29
D i0.130.08
qGW-1CSSL L7L9L37
D i0.210.140.26
qGW-6CSSL L9L11L19L38
D i0.140.230.390.20
CSSL L39L61
D i0.400.19
LWR Parents Aso IR24
D i0.120.27
qLWR-3CSSL L16L17L18L46
D i0.110.150.160.15
qLWR-5CSSL L28L29
D i0.110.14
qLWR-1CSSL L1L2L4L9
D i0.230.170.120.31
qLWR-2CSSL L7L10L11L12
D i0.150.300.340.20 PGWC Parents Aso IR24
D i 1.77 2.11
qPGWC-8CSSL L49L50L51
D i 1.22 2.140.95
qPGWC-9CSSL L52L53L64
D i0.79 1.860.98
qPGWC-1CSSL L5L6L7L8
D i 2.51 4.12 4.42 4.53 AC
E Parents Aso IR24
D i0.95 1.22
qACE-8CSSL L49L50L51
D i 1.11 1.020.53
qACE-9CSSL L52L53L64
D i 2.37 2.67 1.97
qACE-2CSSL L7L10
D i 2.63 1.63
DEC Parents Aso IR24
D i0.640.74
qDEC-8CSSL L49L50L51
D i0.800.600.65
qDEC-9CSSL L52L53L64
D i 2.19 2.02 2.18
qDEC-1a CSSL L5L6L7L8
D i 2.780.94 1.320.98
qDEC-1b CSSL L1L2L4L9
D i0.94 1.31 1.660.74
qDEC-2CSSL L7L10
D i 1.32 1.23
1343
Fig.2The AMMI biplots for six appearance quality traits drawn using the phenotypic values of target CSSLs carrying any of the 22QTL alleles detected in eight environments.AMMI biplots:A GL,B GW,C LWR,D PGWC,E ACE,F DEC.A detailed description of E1–E8is give in Table 1
1344
L49,L50,and L51),all harboring one of these?ve QTL alleles,have been backcrossed to Asominori,and nine secondary F2populations are currently being exploited for the?ne mapping and positional cloning of these QTLs.
Appearance quality traits are quantitatively inher-ited.It is thus di?cult for breeders to e?ciently improve rice grain appearance using conventional methods due to a lack of discrete phenotypic segregation in rice progeny.Moreover,before maturity,the phenotypic identi?cation of appearance traits can not be carried out during conventional breeding.Therefore,it is particu-larly helpful for enhancing the e?ciency of selection and shortening the course of breeding to use markers closely linked to these above?ve QTLs to screen target geno-types directly for related traits in early generations.
Classical quantitative genetics assumes that trait correlations are the result of either pleiotropic e?ects or the tight linkage of genes.As numerous QTLs a?ecting related traits were mapped to similar genomic regions (Fig.1),pleiotropy was the most probable genetic basis of the high trait correlations,which ranged from à0.835**between GW and LWR to0.885**between ACE and DEC.Where pleiotropy is implicated,a coincidence of location and direction of genetic e?ect is expected for positively related traits.The results ob-tained here are in good agreement with this expectation. The detection of chromosomal segments harboring clusters of QTLs and the directions of the e?ects of these intervals were only slightly a?ected by environmental factors(Table4,Fig.1).
Three stable QTLs(qPGWC-8,qACE-8,and qDEC-8)were mapped in the interval G1149-R727on chro-mosome8(Fig.1).Interestingly,three stable QTLs (qAC-8,qTD-8,and qIVOE-8)a?ecting,respectively, the amylose content,tenderness,and palatability of cooked rice have also been detected in the same region (Wan et al.2004).In addition,Jiang(2002)and Fujita et al.(1999)mapped the soluble starch synthase III (SSSIII)gene and isoamylase(ISA)gene in the V115-R1813and C10122S-G1149intervals on chromosome8, respectively.Since these two regions overlap the G1149-R727interval in high-density maps(Causse et al.1994; McCouch et al.2002),these QTL clusters appear to be associated with the synthesis and structure of amylose and amylopectin controlled by the SSSIII and ISA genes (Smith et al.1997;Nakamura2002).The elucidation of the molecular e?ects of starch properties on the deter-mination of rice quality will require?ne mapping and positional cloning of the QTL clusters using secondary F2populations between target CSSL and Asominori. Acknowledgements We are extremely grateful to Prof.A.Yoshim-ura,Kyushu University,Japan,for kindly providing the CSSL population and genotype data.We thank https://www.wendangku.net/doc/e71463409.html, for linguistic correction.This research is supported by the grants from the National High Technology Research and Development Program of China(No.2003AA222131; 2003AA207020),the National Natural Science Foundation of China(No.30270811).References
Campbell BT,Baenzigar PS,Gill KS,Eskridge KM,Budak H, Erayman M,Dweikat I,Yen Y(2003)Identi?cation of QTLs and environmental interactions associated with agronomic traits on chromosome3A of wheat.Crop Sci43:1493–1505 Causse MA,Fulton TM,Cho YG,Ahn SN,Wu KS,Xiao JH,Yu ZH,Ronald PC,Harrington SE,Second G,McCouch SR, Tanksley SD(1994)Saturated molecular map of the rice gen-ome based on an interspeci?c backcross population.Genetics 138:1251–1274
Churchill GA,Doerge RW(1994)Empirical threshold values for quantitative trait mapping.Genetics138:963–971
Crossa J,Gauch HG,Zobel RW(1990)Additive main e?ects and multiplicative interaction analysis of two international maize cultivar trials.Crop Sci30:493–500
Del Rosario AR,Briones VP,Vidal AJ,Juliano BO(1968)Com-position and endosperm structure of developing and mature rice kernel.Cereal Chem45:225–235
Falconer DS(1981)Introduction to quantitative genetics,2nd edn.
London New York
Frary An,Nesbitt TC,Frary Am,Grandillo S,Knaap EVD,Cong B,Liu JP,Meller J,Elber R,Alpert KB,Tanksley SD(2000) fw2.2:A quantitative trait locus key to the evolution of Tomato fruit size.Science289:85–88
Fujita N,Kubo A,Francisco PB,Nakakita M,Harada K,Minaka N,Nakamura Y(1999)Puri?cation,characterization,and cDNA structure of isoamylase from developing endosperm of rice.Planta208:283–293
Gauch HG,Zobel RW(1988)Predictive and postdictive success of statistical analysis of yield trails.Theor Appl Genet76:1–10 He P,Li SG,Qian Q,Ma YQ,Li JZ,Wang WM,Chen Y,Zhu LH (1999)Genetic analysis of rice grain quality.Theor Appl Genet 98:502–508
Hittalmani S,Huang N,Courtois B,Venuprasad R,Shashidhar HE,Zhuang JY,Zheng KL,Liu GF,Wang GC,Sidhu JS, Srivantaneeyahul S,Singh VP,Bagali PG,Prasanna HC, Mclaren G,Khush GS(2003)Identi?cation of QTL for growth-yield and grain yield-related traits in rice across nine locations of Asia.Theor Appl Genet107:679–690
Howell PM,Marshall DF,Lydiate DJ(1996)Towards developing intervarietal substitution lines in Brassica napus using marker-assisted selection.Genome39:348–358
Huang N,Parco A,Mew T,Magpantay G,McCouch S,Guider-doni E,Xu JC,Subudhi P,Angeles ER,Khush GS(1997) RFLP mapping of isozymes,RAPD and QTLs for grain shape, brown planthopper resistance in a doubled haploid rice popu-lation.Mol Breed3:105–113
Jiang HW(2002)The cloning and characterization of the gene for starch synthesis in rice endosperms and studies on the molecular physiological e?ects of high temperature on rice grain quality forming.PhD thesis,Zhejiang University,Hangzhou China Kubo T,Nakamura K,Yoshimura A(1999)Development of a series of Indica chromosome segment substitution lines in Japonica background of rice.Rice Genet Newsl16:104–106 Lander ES,Green P,Abrahamson J,Barlow A,Daly MJ,Lincoln SE,Newburg L(1987)MAPMAKER:An interactive computer package for constructing primary genetic linkage maps of experimental and natural populations.Genomics1:174–181
Li ZK(2001)QTL mapping in rice:a few critical considerations.Rice genetics.In:Khush GS,Brar DS,Hrady B(eds)Proc4th Int Rice Genet Symp.IRRI,Los Banos,Philippines,pp153–171
Li ZK,Yu SB,La?tte HR,Huang L,Courtois B,Hittalmani S, Vijayakumar CHM,Liu GF,Wang GC,Shashidhar HE, Zhuang JY,Zheng KL,Singh VP,Sidhu JS,Srivantaneeyakul S,Khush GS(2003)QTL·environment interactions in rice.I.
Heading date and plant height.Theor Appl Genet108:141–153 Lin HX,Yamamoto T,Sasaki T,Yano M(2000)Characterization and detection of epistatic interactions of3QTLs,Hd-1,Hd-2 and Hd-3,controlling heading date of rice using nearly isogenic lines.Theor Appl Genet101:1021–1028
1345
Lin JR,Wu MG,Shi CH(2001)Analysis on genetic e?ects of appearance quality traits in japonica hybrid rice.Chin J Rice Sci 15:93–96
McCouch SR,Teytelman L,Xu YB(2002)Development and mapping of2240new SSR markers for rice(Oryza sativa L.).
DNA Res9:199–207
Murray MG,Thompson WF(1980)Rapid isolation of high molecular weight plant DNA.Nucleic Acids Res8:4321–4325 Nagato K,Ebata M(1959)Studies on white–core rice kernel II.On the physical properties of the kernel.Proc Crop Sci Soc Jpn 28:46–50
Nakamura Y(2002)Towards a better understanding of the meta-bolic system for amylopectin biosynthesis in plants:rice endo-sperm as a model tissue.Plant Cell Physiol43:718–725 NSPRC(National Standard of People Republic of China)(1999) High quality paddy,GB/T17891-1999,Standards Press of China Paterson AH,Damon S,Hewitt JD,Zamir D,Rabinowitch HD, Lincoln SE,Lander ES,Tanksley SD(1991)Mendelian factors underlying quantitative traits in tomato:comparison across species,generations,and environments.Genetics127:181–197 Redon a ED,Mackill DJ(1998)Quantitative trait locus analysis for rice panicle and grain characteristics.Theor Appl Genet 96:957–963
SAS Institute(1994)JMP statistics and graphics guide:version3: SAS Institute,Cary,N.C.
Smith AM,Denyer K,Martin C(1997)The synthesis of the starch granule.Annu Rev Plant Physiol.Plant Mol Biol48:67–87 Tan YF,Xing YZ,Li JX,Yu SB,Xu CG,Zhang QF(2000)Ge-netic bases of appearance quality of rice grains in Shanyou63, an elite rice hybrid.Theor Appl Genet101:823–829 Tanksley SD(1993)Mapping ploygenes.Annu Rev Genet27:205–233
Teulat B,Zoumarou-Wallis N,Rotter B,Salem MB,Bahri H,This D(2003)QTL for relative water content in?eld-grown barley and their stability across Mediterranean environments.Theor Appl Genet108:181–188Tsunematsu H,Yoshimura A,Harushima Y,Nagamura Y,Ku-rata N,Yano M,Sasaki T,Iwata N(1996)RFLP framework map using recombinant inbred lines in rice.Breed Sci46:279–284
Vargas MH,Crossa J(2000)The AMMI analysis and graphing the biplot.Universidad Auto noma Chapingo,Biometrics and Sta-tistics Unit,CIMMYT,Mexico.https://www.wendangku.net/doc/e71463409.html,/ biometrics
Vargas M,Crossa J,van Eeuwijk FA,Ram?rez ME,Sayre K (1999)Using partial least squares regression,factorial regres-sion,and AMMI models for interpreting genotype·environ-ment interaction.Crop Sci39:955–967
Wan XY,Wan JM,Su CC,Wang CM,Shen WB,Li JM,Wang HL,Jiang L,Liu SJ,Chen LM,Yasui H,Yoshimura A(2004) QTL detection for eating quality of cooked rice in a population of chromosome segment substitution lines.Theor Appl Genet 110:71–79
Xing YZ,Tan YF,Xu CG,Hua JP,Sun XL(2001)Mapping quantitative trait loci for grain appearance traits of rice using a recombinant inbred line population.Acta Bot Sin43:840–845 Yamamoto R,Horisue N,Ikeda R(1995)Rice breeding manual.
Miscellaneous Publication of the National Agriculture Re-search Centre in Japan.No.30,October
Yano M,Katayose Y,Ashikari M,Yamanouchi U,Monna Lisa, Fuse T,Baba T,Yamamoto K,Umehara Y,Nagamura Y, Sasaki T(2000)Hd-1,a major photoperiod sensitivity quanti-tative trait locus in rice,is closely related to the Arabidopsis ?owering time gene CONSTANS.Plant Cell12:2473–2483 Zeng ZB(1994)Precision mapping of quantitative trait loci.
Genetics136:1457–1468
Zhang Z,Lu C,Xiang ZH(1998)Analysis of variety stability based on the AMMI model in silkworm.Sci Agric Sin31:62–68 Zhuang JY,Lin HX,Lu J,Qian HR,Hittamani S,Huang N, Zheng KL(1997)Analysis of QTL·environment interaction for yield components and plant height in rice.Theor Appl Genet95:799–808
1346