Program foundation: This work was supported by the Beijing Municipal Natural Science Foundation (No. 5091001), a grant from the National High Technology Research and Development Program of China (863 Program) (No. 2011AA10A106), 948 Program (No.2009-Z4), and Innovation Platform Program for Basic Research of Agricultural Breeding in Beijing (No. D08070500690801).

Wheat is one of the most important grain crops in China, plant height is an important factor affecting wheat yield, varieties with excessive plant height are susceptible to lodging under the condition of high fertilizer, which results in decreased yield; varieties with excessive plant dwarfed are susceptible to crowding of canopy leaves, and with bad ventilation and light transmission in at middle to bottom part, which affects seed-filling, thus results in abortive grain and decreased yield. Plant height is important objective traits in wheat breeding. Study on its genetic basis has important guiding significance in breeding practice, variety improvement and extension. Since the application of Nonglin 10 harboring dwarf gene in wheat breeding in the 1960s, more than 20 dwarf genes were identified (Cheng et al., 1995, Crops, (4): 36-37; Li et al., 1998). Among them, Rht1 and Rht2 from Nonglin 10 and Rht8 from red wheat were mainly used in wheat breeding, and they were located on the 4BS, 4DS, and 2DS chromosomes, respectively (Liu et al., 2003, Genetics Research in China-Abstract Book of the 7th Conference in Genetics Society of China, pp.128-129). At present, wheat production increased significantly due to breeding and popularization of the varieties harboring dwarf gene and semi-dwarf gene in the global.

Wheat plant height is usually show quantitative trait, which is controlled major dwarf genes and quantitative trait loci (QTL) affected. With the development of quantitative genetics, more and more QTLs of plant height were found. Results of previous studies showed that the major portion of 21 chromosomes in wheat related to the genetic variation of plant height (Ellis et al., 2005). Zhou et al (2004) who used the 104 RIL lines crossing between wangshuibai and Alondra, obtained four QTLs for plant height were located on the 1D, 2B, 4A and 4D chromosomes, respectively, and single QTL explained 10.3%~33.8% of the phenotypic variation. Wang et al  (2008) identified three QTLs for plant height from 218 F_2:3 lines crossing between Bainong 64 and Jingshuang 16, which were located on the 2B, 4D and 6A chromosomes, respectively. QTLs for plant height were located on the 4D and 5A chromosomes (Chu et al., 2008). Liu et al (2002) found seven QTLs for plant height, which were located on the 1B, 4B (two), 6A (two), 6D and 7A chromosomes, respectively, and single QTL explained 5.2%~50.1% of the phenotypic variation. Wang et al (2009) who analyze seven QTLs for plant height in F₂ populations, which were located on 2D, 4D, 5B and 6D chromosomes, the contribution rate was 1.6%~20%.

Previous research has shown that the materials of genetic and QTL mapping for plant height were conventional wheat varieties. In this study, a population of 234 doubled haploid derived from the cross between BS366 (photopeiod-thermo sensitive male sterile line) and Baiyu149, which were planted at two place in two years, respectively, was used to detect the QTLs for plant height based on the method of composite interval mapping (CIM). This work will provide theoretical basis for improvement of BS366 plant height, and reference value for molecular marker assisted breeding of conventional wheat plant height.

1 Results
1.1 Genetic analysis
The results showed that average plant height of BS366 and Baiyu149 were 73.8 cm and 89.5 cm under the condition of four kinds of test environmental. The difference between the two was 15.7 cm, and represent significant difference at the level of 0.01 possibility. Plant height analysis in DH population showed that significant difference between strains, and the coefficient of variation was 12%~13%. Distributions of plant height (Figure 1) and skewness (Table 1) from DH population indicated that plant height have shown a continuous distribution and two-dimensional transgressive segregation in four kinds of environmental conditions, which was conform to normal distribution. This study implied that plant height was controlled by multiple quantitative trait genes, and consistent with the requirements of QTL analysis.

Figure 1 Distribution of plant height in DH population

Table 1 Statistical analysis of plant height in DH population

1.2 Variance analysis
Variance Analysis for plant height is performed by SPSS 16.0 software, results showed that location, year, genotype, location and genotype interaction, and year and genotype interaction represent significant difference (Table 2).

Table 2 Variance analysis for plant height in DH population

1.3 QTL analysis
Based on the method of composite interval mapping (CIM), 25 QTLs of plant height, which were located on chromosomes 1A, 1B, 2A, 2B, 2D, 3B (two), 3D (two), 4B, 5A (two), 6A, 6B (three), 6D, 7A (three), 7B (four) and 7D were detected in four kinds of environmental conditions, the contribution rate was 2.11%～9.64%, the additive effect was between 3.35 and 16.19.

From 2007 to 2008, a total of twenty-two QTLs have been detected in Beijing, in which one QTL was detected in both years. It is located on chromosome 6BL, with interval was from barc14 to gwm58, which is closely linked to gwm58. Genetic distance was 1.93 cM, and the contribution rate was 2.11%~9.64%, and the additive effect was between 5.67 and 9.64; A total of twenty-two QTLs have been detected in Fu’nan. In 2007, a total of fourteen QTLs for plant height have been detected in Beijing and Fu’nan, in which seven QTLs were detected in both years. They are located on chromosomes 1B, 2A, 3B (two), 3D, and 6B (two), respectively, and explained 2.12%~9.19% of the phenotypic variation; In 2008, a total of twelve QTLs for plant height have been detected in Beijing and Fu’nan, in which five QTLs were detected in both years. They are located on chromosomes 2BS, 6BL, 7BL (two), 7DS, respectively, and the contribution rate of single QTL was 2.87%~3.68% (Table 3).

Table 3 The QTL identified of plant height in DH population

In Beijing and Fu’nan, based on QTL analysis for plant height in both years, there is small difference in detected QTLs under the conditions of the different places and the same years, and there is great difference in detected QTLs under the conditions of the different years and the same places (Figure 2). QTL of located chromosome 7A has been detected in Beijing.

Figure 2 The position of QTLs for plant height in wheat

2 Discussion
The results of this research indicated that all detected QTLs for 2007 year explained 49.82% and 31.44% of the phenotypic variation in Beijing and Fu’nan, respectively; all detected QTLs for 2008 year explained 29.09% and 33.33% of the phenotypic variation in Beijing and Fu’nan, respectively; this implied that parts QTLs for plant height has not detected. Four major reasons for above results: firstly, constructed linkage map of wheat was not complete, and molecular was not tightness; secondly, the height-controlling genes in their parents are partially identical, lacking of polymorphism; thirdly, a part of minor QTLs have not detected due to testing error and personal error; fourthly, mathematical model and statistical method for QTL mapping affected efficiency of QTL detected.

QTLs of plant height with different researcher detected are different, Repeatability of these QTLs are unstable in different combinations and environment due to a large number of gene loci and their alleles. Meanwhile, Different genetic materials with different genes and genetic background will lead to quite detected QTLs, but the major QTL is stable. In addition, the group size also affects the efficiency of QTL detection. The effects of molecular marker’s segregation distortion are reduced in larger groups, with the error of the phenotypic data reduced, which thus leads to the increase of the accuracy of QTL localization. While the small groups can reveal certain genetic characteristics, but test results were affected due to the emergence of molecular data deviation. Therefore, traits like plant height are subject to change in environment, we should select larger group to eliminate differences of the local environment, which thus leads to the phenotype error and reduce to the effects of molecular marker’s segregation distortion. But a group too large undoubtedly increases the workload, the number of groups range from 150 to 300 per plant (lines) (Zhang, 2002, Chinese Science Bulletin, 51(9): 2223-2231).

In Beijing and Fu’nan, QTL for plant height on chromosome 6B was detected in both years, the result indicated this QTL was subject to little effect in environment. At present, other researchers have not yet found the relevant reports of the QTL, which may be a new QTL loci controlling plant height. In this study, we detected QTL for plant height on chromosome 6A, with interval was from barc3 to barc107. Results of the study are in line with the works by Wang et al (2008) and Liu et al (2009), QTLs were detected by them near Xgwm570, which is closed to the barc3. we also detected a QTL in barc163-gwm495 interval on the 4B chromosome, this result is in line with the works by Liu et al (2009), while Liu detected a QTL in Xgwm107.2-Xgwm113 interval, this result suggested that the same QTL has been detected on the 4B chromosome. Stability and credibility of detected QTL for genetic improvement of plant height and marker-assisted breeding for dwarf and semi-dwarf wheat.

Detection of different mapping population leads to the discovery of more QTLs, and comprehensive grasp of gene location in the chromosomes is the basis of genetic research for traits. The QTL detection in different groups may be relatively stable loci formed in the long-term process of evolution, and value in use of molecular marker-assisted selection.

3 Materials and Methods
3.1 Experimental materials
In this study, we obtained F1 generation by crossing sterile line BS366 and Baiyu149, which were used to be as maternal and paternal materials, respectively, and provided by Beijing Research Center for Hybrid Wheat, Beijing Academy of Agriculture and Forestry Sciences. 234 doubled haploids are derived from F1 generation through anther culture and chromosome doubling, and then formed a mapping population. After years of observation and identification, BS366 has a stable photoperiod and temperature sensitive genie male sterile characteristic.

3.2 Field experiment
In 2007-2008 and 2008-2009 year, the parents and 234 DH lines were grown in Haidian Experimental Station of Beijing and Fu’nan Experimental Station of Anhui, Random block design, two replications, one row for each strain, 1.5 m length, 40 grains in each row, and conventional management strategies in grower fields. Plant height from the ground to top excluding awns is counted by centimeters, and average height of effective stem is marked plant height for the row.

3.3 Construction of linkage map
In this research, 1428 pairs of primers derived from published sequences of Röder et al (1995), Smulders et al (1997) and Gupta et al (2002), were synthesized by SBS Genetech Co., Ltd.

Linkage analysis of marker results by using QGA_CN software, which was provided Professor Zhu Jun in Zhejiang University. According to the linkage distance between markers and high-density microsatellite map of wheat published by Somers et al (2004), we drew a linkage map by using Mapdraw v2.1 software. 130 pairs of primers were marked on the chromosome 21 of wheat, the total length of linkage map was 1711.52 cM, the average genetic distance was 13.17 cM between markers.