Acta Genetica Sinica, August 2006, 33 (8)733745

ISSN 0379-4172

Genetic Diversity of Source Germplasm of Upland Cotton in China as Determined by SSR Marker Analysis
CHEN Guang, DU Xiong-Ming
Key Laboratory of Cotton Genetic Improvement of Agricultural Ministry, Cotton Research Institute, Chinese Academy of Agricultural Sciences, Anyang 455004,China Abstract: The genetic diversity of 43 sources of Upland cotton germplasm with different parental origins, breeding periods, and ecological growing areas in China were studied on the basis of simple sequence repeat (SSR) markers. A total of 130 gene alleles with 80% polymorphism were detected from 36 SSR primers. The number of alleles per primer ranged from two to eight with an average of 3.6. The polymorphism information content (PIC) range was 0.278-0.865, with an average of 0.62. The average genotype diversity index (H) was 1.102, the highest was 2.039 and the lowest was 0.451. The average coefficient of the genetic similarity of SSR markers among source germplasm was 0.610, ranging from 0.409 to 0.865. These indicated that the genetic diversity at the genomic level of the selected source germplasm was rich, and was representative of the diversity of the germplasms, in general. The diversity at the genome level of the base germplasm from the second and third breeding periods was decreased compared to that of the first period, indicating that the cotton genetic background in China became narrow gradually. The diversity of SSR markers among the base germplasm from early maturity cotton growing areas in the north was higher than those from the Huanghe and Yangtze growing areas. The molecular marker genetic similarity index of the domestic varieties was higher than that in the introduced varieties, which indicates that the genetic diversity in domestic cultivars was lower than that in the introduced varieties. This study gives an overview of the genetic diversity of the cotton germplasm base in China, and provides a guide for breeders to develop new cultivars efficiently. Key words: source germplasm; SSR; genetic diversity

Source germplasm for cotton (Gossypium hirsutum L.) breeding in China includes germplasms with stable genetic characters, excellent yield, adaptability, and better general combining ability, and have been used frequently as parents in many breeding programs. There are few reports on source germplasm research. Huang [1] described in the book Cotton Variety and Pedigree in China 36 source germplasms from the G. hirsutum varieties, comprising 25 foreign varieties, eight Chinese varieties and three varieties with low gossypol. Du and Liu [2] described a new method of classification of source germplasm by emphasizing the number of derived varieties in addition to their pedigrees. The authors defined the source

germplasm as being lines or varieties from which 20 or more applied varieties have been derived. According to this criterion, 47 germplasm sources of Upland cotton used in China were described, which include 16 foreign varieties, 29 Chinese varieties, and two varieties with low gossypol. Genetic diversity is the basic portion of biological diversity and is the base of biological polymorphism and species diversity. Genetic diversity and parenthood of germplasm play an important role in cotton breeding. The precise evaluation of the genetic diversity of excellent germplasm will provide a guide for choosing parents and predicting the degree of inheritance, variation, and level of heterosis, which are essential for realizing the breeding

Received: 2005-09-07; Accepted: 2005-12-19 This work was supported by the 10th Five Years Key Programs for Science and Technology Development of China (No. 2004BA525B05). Corresponding author. E-mail: ducri@public.ayptt.ha.cn; Tel: +86-372-2525 352

734

Acta Genetica Sinica

Vol.33 No.8 2006

goal. Molecular marker analysis is a modern technique, which discloses genetic differences at the DNA level in plants and is an effective tool for testing genetic diversity of germplasm in breeding programs [3, 4]. The study of cotton germplasm diversity has expanded from the phenotypic, cellular, and biochemical levels to the DNA level. Modern molecular marker techniques can illuminate the individual differences and relationships among species at the DNA level. Most cotton varieties planted in China were derived from a few sources of germplasm such as DPL, Stoneville, King, Uganda, Foster, and Trice, all of which were introduced from abroad. These varieties were the foundation of Chinese cotton breeding programs and played a decisive role in the self-breeding varieties of China. It has been indicated that the genetic base was narrow and the genetic diversity was low in Upland cotton in China. This was caused by a limited quantity of source germplasms. We will present evidence from molecular marker analysisshowing the correlation among source germplasms genetic diversity. The method using simple sequence repeat (SSR) markers is a fast and convenient molecular marker method with good reproducibility and high veracity, which can be used to evaluate cotton varieties. Genetic similarity and clustering analyses of source germplasm at the molecular level were carried out in this study to provide some data for parent selection, germplasm enhancement, and application in our cotton-breeding program. The collection, conservation, and utilization of large collections of crop germplasms have aroused peoples interest for a long time. To solve the problem of the difficulty for finding source germplasm, research of core collections of wheat (Triticum aestivum L.), rice (Oryza sativa L.), and soybean (Glycine max L.) crop germplasms was done, and it is necessary to construct a core collection for cotton in China as there are more than 7 000 accessions in the Chinese cotton collection. The study on source Upland cotton germplasm will provide a base for identifying a core collection of Chinese cotton.

were used in this experiment (Table 1). They were grouped using three methods. One method divided the germplasm according to their three breeding periods the first period (before 1950s), the second period (1950s-1960s), and the third period (1970s-1980s). The second method grouped germplasm on the basis of three Chinese growing areasYangtze river valley area, Huanghe river valley area, and the area in northern China. The final method divided germplasm into two groups on the basis of their origins those bred in China and those introduced from abroad. 1. 2 SSR molecular marker analysis Cotton genomic DNA was extracted on the basis of Zhangs [5] CTAB method with some modifications [6,7]. Three hundreds and ninety-eight SSR primers, which were chosen on the basis of previous studies and potential polymorphism, were used to make a primary survey among eight varieties. From this survey, 78 polymorphic primers were identified. Thirty-six primers with good amplification, clear gel bands, strong signal, and clear background were used in the PCR reactions. A 10 L volume of PCR reaction mixture contained 1.0 L 10 PCR buffer (containing 20 mmol/L Mg2+), 0.2 L dNTP (10 mmol/L), 0.3 L Taq enzyme (2 U/L), 6.2 L ddH2O, 0.65 L forward primer (5 mol/L), 0.65 L reverse primer (5 mol/L), and 1 L DNA template (50 ng/L). The PCR reaction was carried out with a PTC-100 thermocycler (MJ Research, Waltham, USA). Amplification was programmed for pre-denaturing at 94 for 3 min, 30 cycles of denaturng at 94 for 30 s, annealing at 57 for 30 s, and extension at 72 for 45 s, followed by a final extension at 72 for 7 min. PCR amplification products were separated by electrophoresis using 8% nondenaturing polyacrylamide gels and were visualized following silver staining [8]. 1. 3 Data analysis

1
1. 1

Materials and Methods

Plant materials Forty-three Upland cotton source germplasms

After observing the PCR electrophoresis results, the bands of DNA fragments were scored as present (1) or absent (0). Genetic diversity analyses were made on the basis of these scores. The statistical methods

CHEN Guang et al.: Genetic Diversity of Source Germplasm of Upland Cotton in China as Determined by SSR Marker Analysis

735

Table 1 No. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43

Source germplasm and relative information Name DPL14A DPL15 Stoneville 4 Stoneville 4B Stoneville 2B Coker100 Delfos531 Empire Foster 6 Guan nong 1 Trice Jiangsu mian 1 Jiangsu mian 2 Jiangsu mian 3 57-681 Dong ting 1 Guang ye dai zi mian DPL16 Zhong mian suo 2 Zhong mian suo 4 Zhong mian suo 3 Shan mian 4 Shan mian 5 Liao mian 3 Xu zhou 209 Xu zhou 1818 Uganda 4 Zhong mian suo 7 52-128 Gan mian 1 Shan mian 3 Chao yang mian 1 Jin mian 2 Jin mian 1 Ke ke 1543 Yi shu hong 86-1 Ji mian 1 Zhong mian suo 12 Hei shan mian 1 Zhong mian suo 10 Lambright GL-5 Micnarie 210 Ecological area America America America America America America America America America North China America Yangtze Yangtze Yangtze Yangtze Yangtze America America Huanghe Huanghe Huanghe Huanghe Huanghe North China Yangtze Yangtze Uganda Huanghe Yangtze Yangtze Huanghe North China North China North China Russia Yangtze Huanghe Huanghe Huanghe North China Huanghe America America Origin America America America America America America America America America Liaoning America Jiangsu Jiangsu Jiangsu Sichuan Hunan America America Henan Henan Henan Shanxi Shanxi Liaoning Jiangsu Jiangsu Uganda Henan Sichuan Jiangxi Shanxi Liaoning Liaoning Liaoning Russia Hubei Henan Hebei Henan Liaoning Henan America America Parent name DPL14A DPL15 Stoneville 14A Stoneville 4B Stoneville 2B Coker100 Delfos531 Empire Foster 6 King Trice DPL14A DPL14A DPL14A DPL15A DPL15A DPL15A DPL15A DPL15A DPL15A DPL15A DPL15A Stoneville 4 Stoneville 4 Stoneville 2B Stoneville 2B Uganda Uganda Delfos531 Foster 6 Foster 6 Chao Yang Mian 1 King King Ke ke 1543 Trice Stoneville 4 Stoneville 2B Uganda King King Lambright GL-5 Micnarie 210 Breeding period[2] 1 1 1 1 1 1 1 1 1 1 1 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 3 3 3 3 3 3 3

736

Acta Genetica Sinica

Vol.33 No.8 2006

and formulae used are shown below: (1) Simpson diversity index, also known as Polymorphism Information Content: PIC = 1-P i 2, where Pi represents the variation in frequency of the ith allele. (2) Shannon-weaver diversity index, also referred to as genotype diversity (H): H = -PiLnPi , where Pi indicates the variation frequency of the ith allele [9]. (3) Ne is the effective number of alleles for each locus: Ne = 1/Pi. (4) Genetic similarity coefficients (Jaccards coefficients) among varieties were calculated using the Qualitative Date program of NTSYSpc 2.1 software (Biostatistics Inc., New York, USA). (5) Clustering analyses were performed using NTSYS-pc (version 2.11a) to calculate the genetic similarity matrices, and the dendrograms were constructed by the unweighted pair-group method of arithmetic averages (UPGMA).

allele were BNL530, BNL3031, NBL1672, BNL2590, and BNL3259 having 8, 7, 6, 6, and 6 alleles, respectively. These primers possessed a strong potential for germplasm evaluation. Primers BNL3254 and TMB04 had only one polymorphic locus with less distinguishing potential among germplasms. 2. 2 Genetic similarity and clustering analyses

2. 2. 1 Source germplasm A similarity matrix was built using genetic similarity between all possible combinations of two germplasms. Genetic similarity was calculated with NTSYS 2.1 using the Jaccard coefficient according to 0-1 data from the SSR markers. The genetic similarity coefficients among source germplasm ranged from 0.409 to 0.865, with an average of 0.610. The largest similarity was 0.865 between Stoneville 4 and Stoneville 2B, which became source germplasm in China after they were introduced from Stoneville, USA. The similarity between Jiangsu mian 2 and DPL15 was 0.841. Jiangsu mian 2 had consanguinity with DPL14 and DPL15 as it was bred from a cross between DPL14 offspring and DPL15. Also, Uganda 4 had high similarity with Zhong mian suo 12 (0.816) and Zhong mian suo 7 (0.818). The common origin of these three varieties, which were derived from progenies of Uganda, had been validated by the SSR marker analysis. The lowest similarities were between Foster 6 and Jin mian 2 (0.409), Jin mian 2 and DPL14 (0.413), Jin mian 2 and DPL15 (0.423), DPL15 and Lambrigh GL-5 (0.425), and Jiangsu mian 2 and Lambrigh GL-5 (0.421). The lower similarities between these varieties indicated that they were more distant from each other. In summary, the genetic similarity between source germplasms was lower, indicating that there were differences among the source germplasms. The low genetic similarity also showed that the source germplasms selected for analysis were representative of a range of cotton germplasm. The cotton source germplasm could be divided into five groups on the basis of the average similarity coefficient (0.610) among the source germplasm (Fig. 1). The first group contained DPL14, DPL15, Uganda,

2
2. 1

Results
SSR marker multiple analyses One hundred and thirty alleles with 80% poly-

morphism among 43 source germplasms were detected. The average number of alleles for each SSR locus was 3.6 (Table 2). The effective number of alleles for each SSR locus ranged from 1.385 to 7.405, with an average of 3.066. The PIC value for the SSR loci ranged from 0.278 to 0.865, with an average of 0.62. The genotype diversity was 0.451 to 2.039, with an average of 1.102. A total of 157 unique genotypes were detected by 36 pairs of polymorphic primers from a combination of 43 source germplasms. The average number of unique genotypes for each locus was 4.36. The polymorphic loci were distributed on cotton chromosomes 3, 4, 5, 8, 9, 10, 16, 18, 20, and 23, indicating that the variations of SSR alleles were dispersed in the whole cotton genome. Among the source germplasm, two to eight SSR loci were obtained from each primer pair that was analyzed. The primers that amplified more than one

CHEN Guang et al.: Genetic Diversity of Source Germplasm of Upland Cotton in China as Determined by SSR Marker Analysis

737

Table 2

Allelic variation of 36 SSR loci among 43 source germplasms No. of No. of polymorphic alleles alleles 8 4 4 2 3 3 2 4 6 4 3 6 3 2 7 2 2 6 5 4 5 3 3 2 3 2 4 3 2 2 3 3 4 3 5 3 3.6 8 3 3 2 2 2 2 4 5 3 2 4 2 2 6 1 2 5 4 3 4 1 2 2 2 2 3 2 2 2 3 2 4 1 5 2 2.889 Polymorphism (%) 1 0.75 0.75 1 0.667 0.667 1 1 0.833 0.75 0.667 0.667 0.667 1 0.857 0.5 1 0.833 0.8 0.75 0.8 0.333 0.667 1 0.667 1 0.75 0.667 1 1 1 0.667 1 0.333 1 0.667 No. of genotypes 12 5 4 3 3 3 3 11 6 5 3 4 4 3 7 2 3 7 6 5 5 2 3 3 3 3 5 3 3 3 4 3 6 2 7 3 4.361 Chromosome 4 5 20 923 923 10 10 9 16 10 9 923 18 10 3 23 18 26 25 20 5 22 PIC 0.865 0.696 0.673 0.397 0.614 0.635 0.489 0.722 0.800 0.682 0.622 0.784 0.590 0.310 0.834 0.341 0.436 0.780 0.718 0.737 0.778 0.665 0.602 0.477 0.551 0.496 0.681 0.590 0.499 0.278 0.593 0.605 0.708 0.651 0.779 0.625 0.620 H 2.039 1.266 1.227 0.586 1.008 1.054 0.682 1.327 1.680 1.240 1.027 1.607 0.961 0.488 1.867 0.525 0.628 1.588 1.358 1.360 1.558 1.096 0.989 0.670 0.873 0.689 1.229 0.962 0.692 0.451 0.967 1.000 1.305 1.074 1.548 1.037 1.102 Ne 7.405 3.285 3.056 1.658 2.588 2.742 1.957 3.600 4.990 3.141 2.647 4.634 2.439 1.449 6.019 1.518 1.774 4.536 3.550 3.804 4.512 2.986 2.513 1.911 2.229 1.982 3.139 2.441 1.996 1.385 2.458 2.535 3.422 2.868 4.523 2.669 3.066

Primer BNL530 BNL830 BNL1053 BNL1231 BNL1317 BNL1414 BNL1421 BNL1495 BNL1672 BNL1694 BNL2449 BNL2590 BNL2634 BNL2960 BNL3031 BNL3254 BNL3255 BNL3259 BNL3383 BNL3442 BNL3474 BNL3482 BNL3806 BNL3948 BNL3976 BNL4030 JESPR65 JESPR101 JESPR114 JESPR152 TMG10 TMK19 TMP02 TMB04 TME12 TMH08 Average

Note: PIC = polymorphic information content, H = Shannon-weaver diversity index, Ne = effective number of alleles.

738

Acta Genetica Sinica

Vol.33 No.8 2006

and the source germplasms derived from them. The second group was mainly composed of Stoneville and their derived source germplasms. Lambright GL-5 and Micnarie 210 with very low gossypol content also belonged to this group. The third group included three source germplasms: Foster 6, Trice, and Ji mian 1. Jian mian 2 and Delfos531 formed the fourth and fifth group, respectively. 2. 2. 2 The source germplasm from different breeding periods There were 11 source germplasms (Table 3) from an early breeding period, most of which were intro-

duced from the USA except Guan nong 1. According to the similarity matrix, Stoneville 2B and Stoneville 4 had the greatest similarity (0.865), whereas DPL15 and Stoneville 4 had the least similarity (0.446). The average similarity coefficient of these source germplasms was 0.587. Guan nong 1 was derived from King of USA, and was an early maturity variety bred in 1930 in China. It was popular in the early maturity cotton growing area of northern China, including the Liaoning Province, and had become a source germplasm of the early period. More than 100 varieties were derived from Guan nong 1. According to the

Fig. 1

Dendrogram of 43 source germplasms based on SSR similarity coefficient

Table 3

SSR genetic similarity coefficient of source germplasms from different breeding periods Breeding period Similarity coefficient 0.587 0.630 0.630 0.610 Range 0.446-0.865 0.445-0.818 0.525-0.782 0.409-0.865 No. of samples 11 25 7 43

First period (before the 1950s) Second period (1950s-1960s) Third period (1970s-1980s) Overall

CHEN Guang et al.: Genetic Diversity of Source Germplasm of Upland Cotton in China as Determined by SSR Marker Analysis

739

Fig. 2

Dendrogram of first period source germplasm originating before the 1950s based on SSR similarity coefficient

similarity coefficient, Guan nong 1 has a close relationship with DPL cotton. The early breeding period source germplasms were classified into three groups by the average similarity coefficient (0.578) (Fig. 2). The first group was DPL type including DPL15, DPL14, Guan nong 1, Coker 100, and empire cotton. The second group was Stoneville type containing Stoneville 2B, Stoneville 4, Stoneville 4B, Foster 6, and Trice. Delfos531 itself became the third group. Tracking the breeding history of these three groups, the ancestors of Delfos, DPL, and Trice had some relationship with Foster cotton. Foster with some specific characters was bred by the Foster farm, in USA, in 1904. This common ancestor shows that the genetic background of cotton was very narrow. The second period source germplasm included 25 germplasms, 22 of which were derived from nine of the first-period source germplasms. The sources of the other three germplasms were Ke ke 1543, Zhong mian suo 7, and Uganda 4. The average similarity coefficient of these 25 source germplasms was 0.630. The similarity between Uganda 3 and Zhong mian

suo 7 was the greatest (0.818). The lowest similarity was between Jin mian 2 and Jiangsu mian 2 (0.445), which were derived from Guan nong 1 and DPL14, respectively. The second period source germplasms were divided into three groups on the basis of the average similarity coefficient (0.63) (Fig. 3). The first group included varieties derived from DPL, Uganda 4, Zhong mian suo 7, 52-128, and Guan nong 1. The second group contained Foster offspring (Gan mian 1 and Shan mian 3), a DPL15-derived line (57-681), and a Stoneville-derived line (Liao mian 3). The third group contained four source germplasms, Xu zhou 1818, Yi shu hong, Chao yang 1, and Jin mian 2. The third-period source germplasm included seven varieties, and their average similarity coefficient was 0.630. The two closest basic germplasms were Ji mian 1 and Zhong mian suo 12 with a similarity coefficient of 0.728. The two farthest basic germplasms were Zhong mian suo 12 and Lambright GL-5 (0.525). Zhong mian suo 12 was developed from the cross of Uganda 4 and Ji mian 1. Zhong mian suo 12 and Ji mian 1 were classified into one

740

Acta Genetica Sinica

Vol.33 No.8 2006

Fig. 3 Dendrogram of second period source germplasm originating between 1951 and 1970 based on SSR similarity coefficient

group by SSR cluster analysis, indicating that the derived variety had a close relationship with its parents at the DNA level. The third-period source germplasms were classified into two groups by average similarity coefficient (0.63) (Fig. 4). The first group was Lambright GL-5 and Micnarie 210 with lower gossypol content and 86-1 with resistance to the Fusarium wilt. The second group contained Ji mian 1, Zhong mian suo 12, and two early maturity cottons, Hei shan mian 1 and Zhong mian suo 10, which were derived from Guan nong 1. The comparison of SSR genetic similarity coefficients of different breeding periods is shown in Table 3. The average similarity coefficient of the first-period source germplasm was the lowest (0.587) with a broad range (0.446-0.865), indicating that this germplasm had many differences at the genome level. The similarity coefficients of the second- and the third-period source germplasms were greater than that of the first period, and the range was also smaller. This indicates that the similarity in the second- and the third-period source germplasms increased. The main

reason was that most of the source germplasms of the second and the third periods were derived from those of the first period source germplasms. Moreover, some genetic differences among the varieties were lost in the process of selection of high-quality and high-yield traits in the breeding program, and this resulted in the genetic base for breeding becoming narrower. 2. 2. 3 Source germplasm from different cotton growing areas From the similarity coefficient of source germplasms

from different cotton areas (Table 4), we found the following results. The similarity among the introduced American varieties was the lowest, which indicated significant genetic differences among the American varieties. The similarity among the source germplasm from the Huanghe River valley cotton area was the highest. The similarity among the source germplasm of Yangtze River valley area was also high, as their average similarity coefficient reached 0.610. These coefficients showed that the variation of source germplasms in the two main cotton-growing areas of China declined compared with that of the germplasms introduced originally. However, the

CHEN Guang et al.: Genetic Diversity of Source Germplasm of Upland Cotton in China as Determined by SSR Marker Analysis

741

Fig. 4 cient

Dendrogram of third period source germplasm originating between 1971 and 1990 based on SSR similarity coeffi-

similarity among the source germplasm in the north early maturity cotton area was lower than that in the Huang-he and Yangtze River areas. The breeders usually focused on using the varieties with high quality, high yield, and stronger adaptability, and this likely made the genetic background among the varieties in the Huang-he and Yangtze River areas narrower. 2. 2. 4 The domestic and introduced source germplasm Many of the Chinese breeding source germplasms had been based on the introduction, selection, and domestication of germplasms from other countries. As a result, the average genetic similarity coef ficient of domestic source germplasm (0.624) was higher than
Table 4 SSR genetic similarity coefficient of source germplasm from different cotton growing areas Cotton-growing areas Similarity coefficient No. of samples Yangtze River areas Huanghe areas Northern China American Overall 0.610 0.651 0.586 0.576 0.610 10 11 6 14 43

that of the introduced germplasm (0.585, Table 5). This indicated that it would be difficult for the genetic base of domestic germplasms to exceed foreign germplasms, which could be the result of the limited genetic resources used in China. It was necessary to analyze the introduced source germplasm as the source germplasms in China were all derived from introduced varieties. There were four groups of foreign source germplasms measured with the average similarity coefficient of source germplasm. The first group was composed of DPL type, Empire, Coker100, and Uganda cotton. The second group was composed of Stoneville type, Foster, and Trice. The third group was composed of those with lower gossypol content and Ke ke 1543 introduced from Russia. The Delfos cotton formed a group by itself (Fig. 5).
Table 5 SSR genetic similarity coefficient of domestic and introduced germplasm Source germplasm Similarity coefficient Domestic Introduced 0.624 0.585 No. of samples 27 16

742

Acta Genetica Sinica

Vol.33 No.8 2006

Fig. 5

Dendrogram of introduced source germplasm based on SSR similarity coefficient

Discussion

The genetic relationships among source germplasms of Upland cotton in China have been analyzed on the basis of SSR molecular markers. The main results are summarized below: 1. The average genetic similarity coefficient of source germplasms was 0.610 and ranged from 0.409 to 0.865. This suggested that the analyzed source germplasms possessed vast diversity, a large extent of variation, and were a general representative of the diversity of source germplasms. 2. The average genetic diversity coefficients of the SSR markers for the source Upland germplasms of the first, second, and third breeding periods were 0.587, 0.630, and 0.630, respectively. This implied that the genomic difference among the modern source germplasm had gradually decreased compared to that of the early source germplasm. This was probably a result of the narrow breeding base as the breeders used a limited number of varieties with high quality and high yield. 3. The similarity coefficients of the SSR markers

among the source germplasms in different Chinese cotton growing areas were different. The diversity of the source germplasm in the northern early maturity cotton area was largely preserved. However, the diversity declined distinctly in the Huanghe and Yangtze River areas compared to the diversity of the introduced source germplasm. The varieties with high quality, high yield, and stronger adaptability were preserved, whereas those with poor adaptability were eliminated through selection, and this lessened the genetic difference among the varieties in Huanghe and Yangtze River areas. 4. The SSR similarity coefficient of domestic source germplasm (0.624) was higher than that of the introduced germplasm (0.585). This suggests that the introduced source germplasm had adapted to the environment and climate in China, and many varieties were derived from them. However, the limited genetic diversity of domestic source germplasm could never exceed the diversity of the introduced germplasms. China is not a native cotton growing area; therefore, its cotton breeding and production was based on

CHEN Guang et al.: Genetic Diversity of Source Germplasm of Upland Cotton in China as Determined by SSR Marker Analysis

743

the introduced germplasms. Source germplasm is the foundation of cotton breeding, the basis of derived varieties, and plays an important role in the cotton production in China. Therefore, it is necessary to study the source germplasm. The genetic diversity of 43 source germplasms of Upland cotton, classified according to the quantity of derived varieties and their pedigrees, have been assessed in this study. The large differences among the source germplasms covered 95% of the diversity of varieties bred in China and proved that the source germplasms selected were a representative sample of Chinese cotton [2]. The source germplasms contributed considerably to breeding programs in China, although their types and amounts were limited. This increased the similarity among their derived varieties, and made the genetic base to narrower. This is explained by the following experimental results: the genetic diversity of the source germplasm of the second and the third breeding period was lower than that of the first period; the diversity of domestic source germplasm declined compared to that of the introduced germplasm; and the diversity among the derived varieties decreased compared to those of their source parents. The main reasons were that the source germplasms of the latter two breeding periods were mostly derived from the first period and the domestic source germplasms were mostly derived from introduced germplasm sources. Other research[10-14] on cottons genetic diversity using RAPD, ISSR, and SSR markers also showed that the genetic base of cotton breeding was narrow in China. For instance, Bie et al.[10] clustered 30 varieties from three main cotton-growing areas in China into three groups and showed that their genetic base was narrow. Their diversity also had some relationship with their pedigrees on the basis of RAPD marker and phenotype analyses of the varieties. Xu et al.[15] reported that the genetic diversity of the varieties bred by CRICAAS from the end of the 1970s to the middle of the 1990s was higher than that of the varieties from the Hebei Province. The 19 varieties bred in the Hebei Province also had a much narrower base. Xu et al.[16] analyzed the genetic diversity of the varieties resistant to Fusarium wilt using RAPD

markers. This research pointed out that the genetic diversity of Upland cotton was lower in China, but it was vital for introducing Fusarium wilt resistance genes from G. arboreum, G. barbadense, and other species into Chinese G. hirsutum. Wu et al.[17] studied 36 domestic and introduced cultivars of Upland cotton with SSR markers and morphological characters and showed that there were some genetic differences among these materials, but the genetic diversity of these cotton cultivars was low. This again suggested that the genetic base of Upland cotton was narrow. When Xu et al.[18] compared Upland cotton in Yangtze River and Huang-he River areas by RAPD marker analysis, the genetic diversity of the varieties in both areas were similar. This further supported the view that common germplasm sources and same breeding targets, similar breeding methods, and policies could be the most important factors to influence the genetic diversity of these two cotton-growing areas in China. Molecular markers can show genetic diversity at the genome level, whereas the phenotypes express the interactions of genes and environments. The majority of domestic source germplasm and their derived varieties have been bred in China by selecting and domesticating introduced varieties. The domestic varieties have higher yield and improved quality, but the selection pressure and limited source germplasm has narrowed the genetic base in domestic cultivars. The development of a core collection, which identifies a small number of unique germplasms to represent the genetic diversity present in a larger collection of germplasm, is a very important research area. On the basis of this study, it is obvious that source germplasms are the most diverse part of the cotton collection in China. This research of source germplasms by SSR marker analyses will provide important methods and technologies for the construction of a Chinese cotton core collection. It will also provide data for variety improvement, and updating and enhancing the diversity of germplasms. Furthermore, both source germplasms and their derived varieties are studied systematically to understand the genetic base of cotton breeding more clearly in our program.

744

Acta Genetica Sinica

Vol.33 No.8 2006

Acknowledgments: We thank Prof. Liu Guo-Qiang and deputy Prof. Sun Junling, who provided valuable suggestions for the paper , and helped review this article. We also acknowledge Dr. Lori Hinze, USDA-ARS, College Station, TX 77845, for a critical review of this article. References:
[1] Huang Z K, Ji D P, Sun S K, Wang R H, Zhou S H, eds. Cotton Variety and Pedigree in China. Beijing: Chinese Agricultural Press, 1996. [2] Du X M, Liu G Q, Chen G. Basic germplasm for cotton breeding in China. Journal of Plant Genetic Resources, 2004, 5(1) : 69-74 (in Chinese with an English abstract). [3] Jia J Z. Molecular germplasm diagnostics and molecular marker assisted breeding. Scientia Agricultura Sinica, 1996, 29(4) : 1-10 (in Chinese with an English abstract). [4] Xie J, Cai Z, Liu X H, Li F H, Cao H L, Luan Y C. Application of biotechnology on evaluation of genetic diversity of germplasm. CROPS, 1998(Supplement), 71-76 (in Chinese). [5] Zhang J, Stewart J, Mac D. Economical and rapid method for extracting cotton genomic DNA. J Cotton Sci, 2000, 4 : 193-201. [6] Shen F F, Yu Y J, Lu F Z, Yin C Y. Isolation of nuclear DNA from cotton and its RAPD analysis. Acta Gossypii Sinica, 1996, 8(5) : 246-249 (in Chinese with an English abstract). [7] Song G L, Cui R X, Wang K B, Guo L P, Li S H, Wang C Y, Zhang X D. A rapid improved CTAB method for extraction of cotton genomic DNA. Acta Gossypii Sinica, 1998, 10(5) : 273-275 (in Chinese with an English abstract). [8] Zhang J, Wu Y T, Guo W Z, Zhang T Z. Fast screening of microsatellite markers in cotton with PAGE/silver staining. Acta Gossypii Sinica, 2000, 12(5) : 267-269 (in Chinese with an English abstract). [9] Dong Y C, Cao Y S, Zhang X Y, Liu S C, Wang L F, You G X, Pang B S, Li L H, Jia J Z. Establishment of candidate core collections in Chinese common wheat germplasm. Journal of Plant Genetic Resources, 2003, 4(1) :

l-8 (in Chinese with an English abstract). [10] Bie S, Kong F L, Zhou Y Y, Zhang G M, Zhang Q Y, Wang X G. Genetic diversity analysis of representative elite cotton varieties in three main cotton regions in China by RAPD and its relation with agronomic characteristics. Scientia Agricultura Sinica, 2001, 34(6) : 597-603 (in Chinese with an English abstract). [11] Guo W Z, Zhang T Z, Pan J J, Wang X Y. A preliminary study on genetic diversity of Upland cotton cultivars in China. Acta Gossypii Sinica, 1997, 9(5) : 242-247 (in Chinese with an English abstract). [12] Liu W X, Kong F L, Guo Z L, Zhang Q Y, Peng H R, Fu X Q, Yang F X. An analysis about genetic basis of cotton cultivars in China since 1949 with molecular markers. Acta Genetica Sinica, 2003, 30 (6) : 560-570 (in Chinese with an English abstract). [13] Liu W X. Studies on genetic improvement in cotton planted in China since 1949 [Dissertation]. China Agricultural University, 2004 (in Chinese). [14] Ma X, Du X M, Sun J L. SSR fingerprinting analysis on 18 colored cotton lines. Journal of Plant Genetic Resources, 2003, 4(4) : 305-310 (in Chinese with an English abstract). [15] Xu Q H, Zhang X L, Feng C D, Nie Y C. Genetic diversity analysis on cultivars (G. hirsutum L.) developed by Hebei Province and CCRI by RAPD markers. Acta Gossypii Sinica, 2001, 13(4) : 238-242 (in Chinese with an English abstract). [16] Xu Q H, Zhang X L, Nie Y C, Femg C D. Genetic diversity evaluation of cultivars (G. hirsumtum L.) resistant to Fusarium wilt by RAPD markers. Scientia Agricultura Sinica, 2002, 35(3) : 272-276 (in Chinese with an English abstract). [17] Wu Y T, Zhang T Z, Yin J M. Genetic diversity detected by DNA markers and phenotypes in Upland cotton. Acta Genetica Sinica, 2001, 28(11) : 1040-1050 (in Chinese with an English abstract). [18] Xu Q H, Zhang X L, Nie Y C. Genetic diversity evaluation of cultivars (G. hirsumtum L.) from the Changjiang River Valley and Yellow River Valley by RAPD markers. Acta Genetica Sinica, 2001, 28(7) : 683-690 (in Chinese with an English abstract).

CHEN Guang et al.: Genetic Diversity of Source Germplasm of Upland Cotton in China as Determined by SSR Marker Analysis

745

SSR

/ 455004 398 BNLJESPR TMB SSR 43 SSR 8%, PIC Shannon-weaver H NTSYSpc2.1 Jaccard UPGMA 3 45891016182023 36 130 80%, 28 3.6 PIC 0.2780.865 0.62H 0.451 2.039 1.102 SSR 0.610 0.4090.865 SSR 0.5870.6300.630 SSR SSR SSR 0.6240.585 SSR 1978- E-mailcgcjca637@sohu.com