Crop Science
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Published online 30 July 2007
Published in Crop Sci 47:1384-1392 (2007)
© 2007 Crop Science Society of America
677 S. Segoe Rd., Madison, WI 53711 USA
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF) Free
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Wang, B.
Right arrow Articles by Zhang, T.
Right arrow Search for Related Content
PubMed
Right arrow Articles by Wang, B.
Right arrow Articles by Zhang, T.
Agricola
Right arrow Articles by Wang, B.
Right arrow Articles by Zhang, T.
Related Collections
Right arrow Cotton
Right arrow Crop Genetics

CROP BREEDING & GENETICS

QTL Analysis and Epistasis Effects Dissection of Fiber Qualities in an Elite Cotton Hybrid Grown in Second Generation

Baohua Wang, Yaoting Wu, Wangzhen Guo, Xiefei Zhu, Naitai Huang and Tianzhen Zhang*

National Key Lab. of Crop Genetics and Germplasm Enhancement, Cotton Research Institute, Nanjing Agricultural Univ., Nanjing 210095, ChinaM

* Corresponding author (cotton{at}njau.edu.cn).


   ABSTRACT
TOP
 NOTES
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
The purpose of the research presented here was to explore the genetic basis of cotton (Gossypium hirsutum L.) fiber quality traits through quantitative trait locus (QTL) analysis and epistasis effects dissection, and further discuss the mechanism of heterosis. An immortalized F2 population was developed by intercrossing Xiangzamian 2–derived recombinant inbred lines (RILs) (XZM2). Fiber quality traits were investigated in F1 and F2 generations of hybrid XZM2, its two parents, and the immortalized F2 population in multiple environments in China. The low level of heterosis in XZM2 and in the immortalized F2 population suggested a lack of dominant and dominant x dominant interaction. In general, the low correlations of genotypic heterozygosity with trait performance and midparent heterosis showed that heterozygosity was not always advantageous for performance, and they excluded overdominance as a major genetic basis of heterosis. A total of 50 QTLs for fiber quality were identified by single-locus QTL analysis. Although partial dominance and overdominance were detected, additive genetic variance was predominant. Common QTLs were detected both in the homozygous RILs and in the heterozygous immortalized F2 populations. Additionally, single-locus heterotic effects and epistasis effects at the two-locus level were detected. Our results indicated that additive gene action was the primary mechanism responsible for genetic variability in fiber quality traits. Additionally, we found that single-locus heterotic effects and epistasis effects contributed to heterosis of fiber quality traits in XZM2.

Abbreviations: AA, additive-by-additive effects • AD, additive-by-dominance effects • DA, dominance-by-additive effects • DD, dominance-by-dominance effects • FB, fiber yellowness • FE, fiber elongation • FL, fiber length • FMAT, fiber maturity • FMIC, micronaire • FR, fiber reflectance • FS, fiber strength • FSFI, short fiber index • FUR, fiber uniformity ratio • PV, phenotypic variation • QTL, quantitative trait locus • RIL, recombinant inbred line


   INTRODUCTION
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
MOLECULAR MARKER analysis methods have demonstrated (Stuber et al., 1992; Xiao et al., 1995) the viability of the two classical hypothesis regarding heterosis, namely, dominance (Bruce, 1910) and overdominance (Shull, 1908; East, 1936). Epistasis, another heterosis hypothesis (Stuber et al., 1973, 1992), has recently been suggested to play an important role in heterosis of Shanyou 63, an elite rice hybrid (Oryza sativa vs Shanyou 63) (Yu et al., 1997; Hua et al., 2003). Recent research has also suggested epistasis and dominance to be the primary genetic basis of heterosis in rice (Li et al., 2001; Luo et al., 2001). Genetic mechanisms of heterosis are complex and can differ, depending on the crop or various crosses within the same crop.

Previous research has suggested that variable genetic effects are the basis of fiber quality heterosis in cotton (Gossypium hirsutum L.). Meredith and Bridge (1972) found that additive effects predominated in fiber strength, elongation, and fineness, while both additive and dominance effects were involved in fiber length. Al-Rawi and Kohel (1970) noted that the magnitude of the additive genetic component was larger than the dominance component for all fiber properties except fineness, and that all fiber properties were within the partial dominance range, except for fiber fineness, that showed overdominance. Additional study has indicated that epistasis and dominance might be equally important for heterotic effects (Al-Rawi and Kohel, 1969).

Cotton, a major source of fiber, is important in the global economy. However, its use is limited by the lack of combinations with high heterosis and the high cost of producing F1 seeds. Interestingly, recent research has reported that F2 heterosis exists in cotton (Weaver, 1984; Meredith, 1990; Tang et al., 1993a, 1993b; Zhang and Pan, 1999; Wu et al., 2004). Although the F1 and F2 hybrids had significant heterosis in cotton yield, heterosis of fiber properties was comparatively low; otherwise the deviation would lead to poor uniformity of fiber quality. If the fiber quality traits of the parents segregated little, then the performance in the F2 hybrid would be nearly equal to that in the F1. Al-Rawi and Kohel (1969) found there were no effects noted in any fiber trait resulting from inbreeding depression. One of the major challenges in cotton research is to elucidate the genetic mechanisms of cotton heterosis, so as to improve the efficiency of hybrid selection.

Fiber quality traits of cotton are inherited in a complex manner and tend to vary with the environment. Epistasis has been suggested to be the foundation of these complex traits (Cheverud and Rountman, 1995; Rieseberg et al., 1996). As well, genotype-by-environment interaction is another important factor regulating inheritance of complex traits in cotton (Saranga et al., 2001; Paterson et al., 2003). Immortalized mapping populations could help resolve problems of low heritability and high experimental error in the identification of quantitative traits. Intercrosses between recombinant inbred lines (RILs) produce a permanent population, or the immortalized F2 population (Hua et al., 2002, 2003), and are extremely useful in study of heterosis and quantitative trait locus (QTL) analysis. The objectives of this study were to detect QTLs associated with fiber quality traits in an immortalized F2 population and determine the role of heterosis in the expression of these traits.


   MATERIALS AND METHODS
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
 DISCUSSION
 REFERENCES
 
Plant Materials
A population of 180 RILs constructed by single-seed descent from a cross between two upland cotton varieties, Zhongmiansuo-12 (ZMS12) and 8891 (B.H. Wang et al., 2006), was intercrossed following a design for developing an immortalized F2 population. The two varieties were parents of Xiangzamian-2 (XZM2), the most widely cultivated cotton hybrid planted both in F1 and F2 populations before transgenic Bt hybrids were extensively released in China; XZM2 was released in the Hunan province in 1997 (Li et al., 1997) and in the Yangtze River Valley cotton-growing region in 2001. Crosses were made between RILs chosen by random permutations of the 180 RILs. In each round of permutation, the 180 RILs were randomly divided into two groups and were paired up at random, without replacement, to provide parents for 90 crosses. Each of the RILs was used only once in each round of pairing and crossing. This procedure was repeated twice, resulting in a population of 180 crosses. Seeds for field trials were obtained from 171 crosses.

Field Trials
A total of six generations, namely, P1 (ZMS12), P2 (8891), F1, BC1, BC2, and F2, were planted in 1999 and 2000 at the Lingqing Experimental Station, Cotton Research Center, Shandong Academy of Agriculture Sciences in the Yellow River Valley cotton-growing region and at the Jiangpu Experimental Station, Nanjing Agricultural University in the Yangtze River Valley cotton-growing region. The current study did not include data from Lingqing for the year 2000 due to heavy damage caused by bollworms. Yield and yield-component traits were examined in 60 P1, P2, and F1 plants, 180 BC1 and BC2 plants, and 300 F2 (unpublished data, 2006) plants, while fiber quality traits were investigated in P1, P2, F1, and F2 plants. Fiber samples were tested by the High-Volume Instrument (HVI900; USTER Company, Switzerland) at the Supervision Inspection and Testing Center of Cotton Quality, Ministry of Agriculture, China. Fiber quality trait evaluation included fiber length (FL, mm), fiber strength (FS, cN/tex), micronaire (FMIC), which is a measure of the maturity and/or the fineness of fiber, fiber elongation (FE), and fiber uniformity ratio (FUR).

The parental lines, F1, the 180 RILs, and the 171 immortalized F2 crosses were planted at Jiangpu Experimental Station in 2004 and 2005. A randomized complete block design with two replications was used in the field trials. Each plot consisted of four rows: two rows of the hybrid and one row for each of the respective parents; fifteen 30-d-old seedlings were transplanted in each row, with a plant-to-plant distance of 30 cm and a row-to-row distance of 80 cm. The five plants in the middle of each row were tagged for scoring and harvesting seed cotton. Fiber samples were tested in the Cotton Quality Testing Center of Henan province (HVI Spectrum, Zhengzhou, China). Evaluation of fiber quality traits included FL, FS, FMIC, FE, FUR, fiber maturity (FMAT), fiber yellowness (FB), fiber reflectance (FR), and short fiber index (FSFI).

Molecular Markers and Linkage Maps
Molecular marker data for RIL populations have been previously described (B.H. Wang et al., 2006). A total of 4106 single-sequence repeat primer pairs and 384 amplified fragment length polymorphism primer combinations were used to screen parents, resulting in 127 and 18 polymorphic loci, respectively. Additionally, two of the 1040 random amplified polymorphic DNA primers (OPAJ15 and OPAK4), one of 30 sequence-related amplified polymorphism (Li and Quiros, 2001) primer pairs (em1*me4), and a dominant yellow anther gene P1 from 8891 were included as RIL population screening markers. The total number of polymorphic loci was 149. The genotype for each cross in the immortalized F2 population was deduced from RIL data. Linkage maps were constructed using Mapmaker 3.0 (Lander et al., 1987; Lincoln et al., 1992). Assignments of linkage groups to subgenomes or chromosomes were based on backboned linkage maps (Han et al., 2006; K. Wang et al., 2006).

Data Analysis
Both single-locus QTLs and epistatic QTLs were mapped for the immortalized F2 population. Single-locus QTLs were analyzed by composite interval mapping (Zeng, 1994) using QTL Cartographer 2.0 (Basten et al., 2001). A significant QTL was defined as having a likelihood ratio (LR) threshold of 13.8 (equal to a logarithms of odds [LOD] score of 3.0), which restricted the occurrence of Type I statistical errors (false positives) to less than 5% in the large cotton genome (Jiang et al., 1998). To detect a suggestive QTL, an LR value between 9.2 and 13.8 (equal to a LOD score of 2.0–3.0) was used (Lander and Kruglyak, 1995). Confidence intervals (90–95%) associated with QTL locations were set as the map interval, corresponding to 1 LOD decline on either side of the peak. ANOVA showed that genotype-by-environment interactions were not significant for fiber quality traits, consequently, QTL mapping was conducted with data sets both from a single environment (separate analysis) and from data that incorporated the means of two environments (joint analysis). QTL nomenclature that is typically used in rice was adopted (McCouch et al., 1997). A QTL Network 2.0 program, based on a mixed linear model (Yang et al., 2005, Yang and Zhu 2005), was used to determine epistatic QTL conditioning fiber quality traits, with a permutation test of 1000 times at a significance level of p = 0.005.


   RESULTS
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
DISCUSSION
 REFERENCES
 
Trait Performance and Midparent Heterosis in F1, F2 of XZM2, and the Immortalized F2 Population
Previous research has demonstrated that XZM2 is a cotton hybrid with high heterosis in yield (unpublished data, 2006). In the current study, fiber properties had comparatively lower heterosis (Tables 1 and 2), while performance of most fiber quality traits in the F1 and F2 of XZM2 fell between that of the two parents. All traits expressed transgressive segregation in both directions in the immortalized F2 population (Table 2). Interestingly, some crosses of the immortalized F2 population displayed even higher trait performance and midparent heterosis than that noted for XZM2.


View this table:
[in this window]
[in a new window]

  		Table 1. Trait performance and midparent heterosis of fiber quality traits in F1 and F2 of XZM2.

 

View this table:
[in this window]
[in a new window]

  		Table 2. Trait performance and midparent heterosis of fiber quality traits in the immortalized F2 population, XZM2 and its two parents measured on means of 2 yr.

 
Correlations between fiber quality traits measured as the means of two environments in the immortalized F2 population were also calculated (Table 3). The results indicated that a number of fiber quality traits were significantly associated with each other (Table 3). Our data suggested that while FL was significantly correlated with FR, it was negatively correlated with FMIC, FMAT, FE, and FB, consistent with the results of the RIL population (B.H. Wang et al., 2006). Fiber strength was significantly correlated with FUR and FMAT, while it was negatively correlated with FSFI and FE. The highest positive correlation was detected between FMIC and FMAT, two traits that are both measured in micronaire units (r = 0.925), while the highest negative correlation was between FUR and FSFI (r = –0.759).


View this table:
[in this window]
[in a new window]

  		Table 3. Correlation analyses between fiber quality traits measured on means of two environments in the immortalized F2 population.

 
Dissection of Single-Locus Heterotic Effects Based on QTL Analysis
Marker genotypes of 149 loci in the immortalized F2 population were determined from the RIL population. A molecular linkage map was constructed spanning 1025.1 cM, which was longer than the 865.2-cM map constructed from the RIL population (B.H. Wang et al., 2006). Given the deficiency in RIL marker data, greater deficiency for the deduced marker data in the immortalized F2 population was expected, potentially resulting in larger distances in the linkage map. The RIL (B.H. Wang et al., 2006) and immortalized F2 population maps were highly consistent, with the exception of a few linkage groups. Additionally, the RIL-based linkage map was used to tag QTLs for fiber quality traits. The results of the detection of single-locus QTLs in 2004 and 2005 are shown in Table 4. Additionally, 50 QTLs for fiber quality traits were identified in the two environments using composite interval mapping, of which 21 were detected by joint analysis. Suggestive QTLs detected at least twice, either in two environments, or both in joint analysis and separate analysis, as well as significant QTLs, are discussed briefly in the following sections.


View this table:
[in this window]
[in a new window]

  		Table 4. Quantitative trait loci (QTLs) for all traits in the two environments detected by QTL Cartographer 2.0 independently (separate analysis).

 
Fiber Length
In total, two significant QTLs, qFL-A10-1 and qFL-D2-1, were detected both by joint and separate analysis, explaining 10.46 and 21.33% of the phenotypic variation (PV), respectively. Notably, they were also detected in the RIL population (B.H. Wang et al., 2006). The suggestive QTL qFL-D6-1 was detected in 2005 and explained 11.11% of PV. It was also detected in joint analysis, and alleles from ZMS12 were in the direction of increasing FL.

Fiber Strength
We identified two suggestive QTLs, qFS-LG08-1 and qFS-D4-1, in separate analysis in 2004, as well as in joint analysis, explaining 5.20% and 6.89% of PV, respectively. The suggestive QTL qFS-D2-1, explaining 7.51% of PV, was detected in 2004 and was also identified in the RIL population. Alleles from ZMS12 were in the direction of increasing FS for all three QTLs.

Micronaire
The suggestive QTL qFMIC-LG08-1, explaining 6.28% of PV, was detected in 2004 and in joint analysis, and alleles from ZMS12 were in the direction of increasing FMIC. The suggestive QTL qFMIC-A10-1 and the significant QTL qFMIC-D2-1 were detected both in separate and in joint analysis, explaining 6.17 and 9.10% of PV, respectively. Additionally, both QTLs were detected in the RIL population. The significant QTL qFMIC-A11-1 was detected both in 2005 and in joint analysis, explaining 14.72% of PV, and alleles from 8891 were in the direction of increasing FMIC. The significant QTL qFMIC-D2-2 was detected in 2005 in separate analysis, explaining 7.68% of PV, and alleles from ZMS12 were in the direction of increasing FMIC.

Fiber Uniformity Ratio
The significant QTL qFUR-D9-1, explaining 10.08% of PV, was detected both in 2004 and in joint analysis. The other significant QTL, qFUR-A5-1, was detected in 2005 and explained 10.48% of PV, and alleles from ZMS12 were in the direction of increasing FUR. The suggestive QTL qFUR-A9-1 was detected in 2005, explaining 5.93% of PV, and was also detected in joint analysis, with alleles from 8891 in the direction of increasing FUR.

Short Fiber Index
The suggestive QTL qFSFI-A10-1 was detected in 2004 and 2005, and alleles from 8891 were in the direction of increasing FSFI. The suggestive QTL qFSFI-LG04-1 explaining 13.74% of PV was detected in 2005 in separate analysis and in joint analysis, and alleles from ZMS12 were in the direction of increasing FSFI. Two other suggestive QTLs, qFSFI-LG08-1 and qFSFI-A11-1, were detected both in separate and in joint analysis, and alleles from 8891 were in the direction of increasing FSFI.

Fiber Elongation
The suggestive QTL qFE-LG08-1 was detected in 2004 and 2005 and alleles from 8891 were in the direction of increasing FE. The significant QTL qFE-D4-1 detected in 2004 explained 10.13% of PV, and alleles from 8891 were in the direction of increasing FE. The suggestive QTL qFE-D6-2 was detected in 2005 and was also detected in the RIL population; alleles from ZMS12 were in the direction of increasing FE.

Fiber Maturity
The QTL qFMAT-LG08-1 was detected in 2004 and 2005 in separate analysis and also simultaneously identified in joint analysis. Alleles from ZMS12 were in the direction of increasing FMAT. QTL qFMAT-A11-1 was detected in both years in separate analysis, and it was also detected in joint analysis; alleles from 8891 were in the direction of increasing FMAT. The significant QTL qFMAT-D12-1 was identified in 2004 and in joint analysis, explaining 12.44% of PV. The suggestive QTL qFMAT-D2-1 was detected in 2005 and joint analysis, as well as in the RIL population. The significant QTL qFMAT-D2-2 was detected in 2005 in separate and joint analysis.

Reflectance
The significant QTL qFR-A11-1 was detected in 2004 in separate analysis, explaining 13.24% of PV, and alleles from ZMS12 were in the direction of increasing FR. The other significant QTL, qFR-A5-2, was detected in 2005, explaining 8.67% of PV, and alleles from 8891 were in the direction of increasing FR.

Yellowness
The significant QTL qFB-A5-1 was detected in 2004 in separate analysis, explaining 20.18% of PV. It was also detected in the RIL population, and alleles from 8891 were in the direction of increasing FB. The second significant QTL, qFB-A9-1, was detected in 2004. The third significant QTL, qFB-LG09-1 on LG09, was detected in 2005 in separate analysis, and alleles from ZMS12 were in the direction of increasing FB.

Altogether 50 QTLs were detected in 2 yr, with 21 simultaneously detected in joint analysis. Additionally, 9 QTLs were detected in the RIL population. Single-marker analysis was performed on QTL flanking markers, as well as comparisons of trait performance between different genotypes. The data suggested that in most cases, trait performance of the heterozygote was between that of the two homozygotes (data not shown). Heterozygosity at 149 marker loci was calculated for 171 crosses of the immortalized F2 population. Correlations of heterozygosity of the marker genotypes with trait performance and midparent heterosis were not significant for any trait, with the exception of FMIC in 2005 (data not shown).

Detection of Epistasis Effects for Fiber Quality Traits
Epistatic QTLs for fiber quality traits were analyzed with QTL Network 2.0 (Yang et al., 2005). Digenic effects were categorized as additive-by-additive effects (AA), additive-by-dominance effects (AD), dominance-by-additive effects (DA), and dominance-by-dominance effects (DD) (Table 5).


View this table:
[in this window]
[in a new window]

  		Table 5. Epistatic quantitative trait loci (QTLs) detected by QTL Network 2.0 for fiber quality traits using associated data of two environments.

 
Fiber Length
A locus at the interval NAU460–NAU1298 on D12 interacted with a locus at the interval E11M4b–BNL1423 on A9, with AA explaining 6.13%, AD 2.83%, DA 23.36%, and DD 12.05% of PV.

Fiber Strength
A locus at the interval NAU997–CIR332 on A3 interacted with a locus at the interval NAU1426–NAU1200 on A5, with AA explaining 5.19%, AD 13.09%, and DA 1.39% of PV.

Fiber Uniformity Ratio
A locus at the interval NAU3654–NAU1043 on LG04 interacted with a locus at the interval E4M12–NAU1426 on A5, with AD explaining 7.46%, DA 5.30%, and DD 1.51% of PV.

Short Fiber Index
A locus at the interval NAU1043–NAU3810 on LG04 interacted with a locus at the interval NAU3948–OPAJ15, with AA explaining 13.37%, AD 2.62%, DA 8.72%, and DD 28.37% of PV. A locus at the interval CIR407–E22M8 on D6 interacted with a locus at the interval E9M14–P1 on A5, with AA explaining 2.47%, DA 8.23%, and DD 5.03% of PV.

Fiber Elongation
A locus at the interval NAU3377a–NAU3848 on LG09 interacted with two loci on D9. Total effects of AD explained 7.77%, DA 9.96%, and DD 19.99% of PV.

Fiber Maturity
Two pairs of interactions were detected for fiber maturity. A locus at the interval E17M5–E11M4b on A9 interacted with a locus at the interval NAU4073–E4M9 on D12, with AA explaining 20.08%, AD 14.07%, and DD 3.20% of PV. A locus at the interval NAU797–NAU1269 on D5 interacted with a locus at the interval CIR071–NAU2700 on D6, with AA explaining 10.76% and AD 5.81% of PV.


   DISCUSSION
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
REFERENCES
 
The strategy of exploiting F2 is determined by the selection of parents. Meredith and Brown (1998) suggested that three points should be emphasized in producing high-yielding F2 populations, including the choice of at least one parent that is well adapted to the targeted region; the region of origin of the second parent is not relevant. Additionally, if fiber quality is a breeding objective, at least one parent has to have above average fiber quality, as well as be a good yield combiner. Dever and Gannaway (1992) argued that the variability associated with F1 and F2 hybrids was a function of original parental variability, that is, when original parental variability was low and the fiber properties of both parents were similar, small fiber variability in the F2 hybrid could be achieved. Later, Tang et al. (1993a) suggested the evaluation of F2 heterosis should be based on the performance of the F2 hybrids in the environments of interest.

In this current study, an immortalized F2 population was constructed by randomly intercrossing of RILs for the purpose of exploring the genetic basis of heterosis for fiber quality traits. Midparent heterosis in F1 and F2 of XZM2 and the immortalized F2 population were calculated for fiber quality traits (Tables 1 and 2). Our data suggested F2 heterosis was often lower than that in F1, and the coefficient of variance in F2 was near to that of the high-value parent or higher than those in the two parents (Table 1), suggesting there existed larger variability in F2 compared with both parents and F1. The results indicated that to utilize F2 in cotton production, parents with elite fiber qualities should be used. Moreover, there should be little difference between the parents in fiber-related traits. The range of average midparent heterosis in the immortalized F2 population was from –1.61% for FE to 1.67% for FS (Table 2). The low level of heterosis indicated that a major portion of the genetic effects in the immortalized F2 population was additive. The generally low correlations between genotypic heterozygosity with trait performance and midparent heterosis suggested heterozygosity was not always advantageous for performance, and also excluded overdominance as major genetic basis of heterosis. Midparent heterosis did not increase much in the F1 hybrid of XZM2, with the highest value of 1.87% noted in FL. The elite cotton hybrid XZM2 had high heterosis in yield, although heterosis of fiber quality traits showed less heterosis when compared with yield and yield components, and F1 performance of most fiber quality traits fell between those of the two parents (Table 2). This could be explained by previous studies suggesting that genetic control of fiber quality traits was largely additive (Tang et al., 1993b), with little nonadditive gene action (Meredith, 1990).

In single-locus QTL analysis (Table 4), a total of 50 QTLs for fiber quality traits were detected in separate analysis. Although some partial dominance and overdominance were detected, additive genetic variance was predominant. Additionally, single-marker analysis of QTL flanking markers demonstrated that trait performance of the heterozygote did not significantly deviate from the average of two homozygotes (data not shown). These results suggested that the major portion of genetic variability for fiber quality traits resulted from additive gene action. Furthermore, 9 of the 50 QTLs were simultaneously detected in the RIL population in which QTL additive effects were the major genetic basis of fiber quality traits (B.H. Wang et al., 2006). These common QTLs detected both in the homozygous RIL and the heterozygous immortalized F2 populations further reinforced the idea of additive gene action.

Epistasis effects at the two-locus level were also detected with QTL analysis identifying AA, AD, DA, and DD epistatic effects (Table 5). Previous research has shown inconsistent results for gene effects involving epistasis in explaining heterosis. Miller and Marani (1963) reported that small but significant F2 deviations from linear regression for lint yield, boll weight, FS, and earliness in the generation means suggested the presence of epistasis for those traits. Marani (1968) found more dominance and epistasis in the G. barbadense crosses, and FL showed small dominance effects in both G. barbadense and G. hirsutum crosses and some epistasis for G. barbadense. Baker and Verhalen (1975) reported that individual F2 means for all traits were very similar to the means of their midparent and F1 performance, while all estimates of F2 deviation were small and statistically nonsignificant, suggesting that additive effects and dominance were more important than epistasis effects. Interestingly, Meredith and Bridge (1972) indicated that the detection of epistasis is dependent on the particular genetic material involved, the traits being studied, and the particular genetic model used. In the current study, significant epistasis effects were detected for fiber quality traits including FL, FS, FUR, FSFI, FE, and FMAT using an immortalized F2 population, while no epistasis was detected for FMIC, FB, and FR. Dominance and epistasis effects controlling fiber qualities were smaller than those controlling yield traits (unpublished data, 2006). Moreover, fiber quality traits were not highly segregated, and the performance in theF2 hybrid was nearly equal to that of F1 (Table 1), leading to the feasible exploitation in F2 of XZM2. These results confirmed that although nonadditive effects were detected, additive effects play a major role as the genetic basis of fiber qualities in XZM2. Our results indicated that the major portion of genetic variability for fiber quality traits was additive gene action, while, to a certain degree, both single-locus heterotic and epistasis effects contributed to heterosis of fiber quality traits in XZM2.


   ACKNOWLEDGMENTS
 
This work was supported by grants from the State Key Basic Research and Development Plan of China (2006CB101708), National Natural Foundation for Outstanding Youth (30025029), High-Tech Program 863 (2002AA207006), the Changjiang Scholars and Innovative Research Team in University, and the Teaching and Research Award Program for Outstanding Young Teachers in Higher Education Institutions of MOE, China. The authors would like to thank Prof. Dr. Jun Zhu from the Zhejiang University, and Prof. Dr. Jinping Hua from the China Agricultural University for their help in data analysis.


   NOTES
TOP
 NOTES
ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
All rights reserved. No part of this periodical may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying, recording, or any information storage and retrieval system, without permission in writing from the publisher. Permission for printing and for reprinting the material contained herein has been obtained by the publisher.

Received for publication October 10, 2006.


   REFERENCES
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 





This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF) Free
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Wang, B.
Right arrow Articles by Zhang, T.
Right arrow Search for Related Content
PubMed
Right arrow Articles by Wang, B.
Right arrow Articles by Zhang, T.
Agricola
Right arrow Articles by Wang, B.
Right arrow Articles by Zhang, T.
Related Collections
Right arrow Cotton
Right arrow Crop Genetics

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
The SCI Journals Agronomy Journal Vadose Zone Journal
Journal of Natural Resources
and Life Sciences Education		 Soil Science Society of America Journal
Journal of Plant Registrations Journal of
Environmental Quality		 The Plant Genome