全 文 ：作物学报 ACTA AGRONOMICA SINICA 2011, 37(2): 224234 http://www.chinacrops.org/zwxb/
ISSN 0496-3490; CODEN TSHPA9 E-mail: xbzw@chinajournal.net.cn

The study was supported by a grant from the National Science and Technology Support Program (2009BADA7B00).
*
通讯作者(Corresponding author): LI Ru-Yu, E-mail: li_ruyu@sina.com, Tel: 13606373889
Received(收稿日期): 2010-06-01; Accepted(接受日期): 2010-09-24.
DOI: 10.3724/SP.J.1006.2011.00224
Assessment of Genetic Diversity in Chinese Sorghum Landraces Using SSR
Markers as Compared with Foreign Accessions
ZHANG Han, WANG Jian-Cheng, WANG Dong-Jian, YAO Feng-Xia, XU Jin-Fang, SONG Guo-An, GUAN
Yan-An, and LI Ru-Yu*
Crop Research Institute, Shandong Academy of Agricultural Sciences, Jinan 250100, China
Abstract: The genetic variation of 184 Chinese sorghum landraces (Sorghum bicolor L.) from a broad geographic area and repre-
senting different phenotypes, and 69 representative foreign cultivated sorghum accessions (world sorghum), was assessed using 32
nuclear SSR primer pairs. Overall, lower level of genetic diversity was detected in Chinese sorghum than in world sorghum. The
allelic richness (Rs) and Nei’s allele diversity (He) for Chinese sorghum and world sorghum were 9.81 and 0.629, and 11.52 and
0.745, respectively. Fewer unique alleles were detected in Chinese sorghum than in world sorghum. Chinese sorghum had a ge-
netic diversity level lower than accessions from East Africa (He=0.732), North America (He=0.707) and South Asia (He=0.712);
and was only comparable to those from South African accessions (He=0.609). Marked differences in level of genetic variation
were revealed between Chinese sorghum landraces from 12 provinces, with Rs ranging from 3.64 to 4.88 and He from 0.517 to
0.714. Accessions from Jilin Province exhibited the highest level of genetic diversity among all regions in China, which was
comparable to the sorghum in East Africa . The results indicated a strong divergence of Chinese sorghum from world sorghum, but
a weak differentiation among Chinese sorghum both on regional and type bases. Principal component analysis (PCA) clearly
separated Chinese sorghum from world accessions but could not separate Chinese sorghum into discrete geographical or pheno-
typic groups. Analysis of molecular variance (AMOVA) indicated that 20.43% of the total genetic variation was attributable to the
difference between world and Chinese sorghum and 79.57% occurred among Chinese and world sorghum accessions. For Chinese
sorghum, partitioning the total variation revealed that genetic diversity mainly existed among accessions within regions (91.94%)
or eco-regions (94.97%) rather than among regions (8.06%) or eco-regions (5.03%). Similarly, a large portion (97.93%) of the
total variation was found within types compared to among types (2.07%). Our study supports the view that Chinese sorghum is of
African origin. Chinese sorghum may have experienced a long history of natural and human selection when largely isolated from
outside world since prehistoric time. Suggestions for sorghum breeding programs were presented in the light of these data.
Keywords: Sorghum bicolor; Genetic diversity; Microsatellites; Evolution; Origin
中国高粱地方品种遗传多样性评价及中、外高粱遗传变异水平比较
张 晗 王建成 王东建 姚凤霞 许金芳 宋国安 管延安 李汝玉*
山东省农业科学院作物研究所, 山东济南 250100
摘 要: 利用 32个高粱(Sorghum bicolor L.)核基因组多态性 SSR(simple sequence repeats)位点, 以 69份国外品种为
对照, 对 12个地区的 184份中国高粱地方品种进行了遗传多样性分析。研究结果表明, 中国高粱的遗传多样性明显
低于国外高粱。中国高粱和国外高粱的等位基因丰度(Rs)和基因多样性(He)分别为 9.81、0.629 和 11.52、0.745。中
国高粱的遗传多样性明显低于东非(He=0.732)、北美(He=0.707)和南亚(He=0.712)高粱, 与南非高粱相当(He=0.609)。
不同地区中国高粱地方品种遗传变异水平存在明显差异, 12个地区高粱种质等位基因丰度在 3.64~4.88之间, 基因多
样性值在 0.517~0.714 之间。吉林高粱地方品种遗传变异最为丰富(He=0.714), 与北美、南亚高粱相当。中国高粱与
国外高粱之间遗传分化明显, 而中国高粱地方品种地区间和类型间分化极弱。主成分分析(PCA)能够明显区分中外高
粱种质但不能将中国高粱按地区或类型分开。分子方差分析(AMOVA)表明, 中外高粱间的遗传变异占全部参试材料
遗传变异的 20.43%。中国高粱遗传变异主要存在于地区内材料间(占总变异 91.94%)或生态区内材料间(占总变异
94.97%)。在品种类型方面, 中国高粱绝大部分遗传变异存在于穗型内材料间(占总变异 97.93%)。本研究支持中国高
粱外来说的观点。
关键词: 中国高粱; 遗传多样性; 微卫星; 演化; 起源
第 2期 ZHANG Han et al.: Assessment of Genetic Diversity in Chinese Sorghum Landraces 225


Sorghum (Sorghum bicolor L. Moench) is the
world’s fifth most important grain crop and the main
food source for 500 million people in more than 30 Afri-
can nations. Great morphological variation exists within
domesticated grain sorghum landraces. Based on spikelet
morphology and inflorescence structure, Harlan and de
Wet [1]classified grain sorghum into five basic races, i.e.
bicolor, caudatum, durra, kafir and guinea, and ten inter-
mediate races, which are combinations of any two of the
basic races. Large collections of sorghum accessions
have been assembled and maintained in various national
and international genebanks. For example, the number of
sorghum accessions maintained at the International Crops
Research Institute for the Semi-Arid Tropics (ICRISAT)
exceeds 36 000[2].
China has a long history of sorghum cultivation.
Despite the widely accepted theory that sorghum was
domesticated in Africa, some scholars argued that Chi-
nese sorghum (Kaoliang) may have been domesticated in
China independent of African sorghum [3-7]. Sorghum was
the third most important cereal crop (behind rice and
wheat) in the early 20th century in China. It was culti-
vated nearly all over the country with a planting area
accounting for 16%–26% of the total area under cultiva-
tion [8]. Based on the ecological conditions and cropping
practices in different sorghum growing regions, the sor-
ghum growing area in China was divided into four major
eco-regions [9]. Sorghum had a variety of uses in Chinese
society. Since the 1970s, sorghum landraces have been
replaced gradually by hybrids. As a crop grown under
diverse ecological conditions for various uses since pre-
historic time, enormous morphological variations have
developed among Chinese sorghum landraces. Morpho-
logically, they are differentiated clearly from African
sorghum landraces [4,7-8]. In Snowden’s classification sys-
tem, Chinese sorghum was assigned to four species, i.e.
Sorghum membranaceum Chiov., S. nervosum Bess. ex
Schult., S. dochna (Forrsk.) Snowden, and S. bicolor L.
Moench [10]. Among these species, S. nervosum Bess. ex
Schult is distributed in Eastern Asia [11]. Having been
used both directly or as parents in the development of
restorer lines, Chinese sorghum landraces have played
an important role in hybrid breeding programs in
China[8,12-13]. To collect these genetic resources, three
major sorghum germplasm collection missions have been
undertaken since the 1950s [8]. Currently, China has de-
veloped one of the largest germplasm collections of cul-
tivated sorghum in the world. The National Genbank of
China (NGBC) maintains a sorghum collection of over
16 000 accessions. More than 12 000 of these accessions
are Chinese sorghum landraces [13].
Analysis of the extent and distribution of genetic
diversity in crop plants is essential for optimizing sam-
pling and breeding strategies [14]. Although great agro-
morphological variations in Chinese sorghum have long
been observed [4,7-8,15], few studies have been conducted
specifically on genetic variations in Chinese sorghum.
The genetic variation of 34 Chinese sorghum accessions
(23 landraces and 11 improved varieties) was examined
by de Oliveira et al. [16] using Inter-SSR, RAPD and
RFLP markers. They found that Chinese sorghum could
clearly be differentiated from other cultivated sorghum
races and exhibited a relatively low level of genetic di-
versity. In another study involving 12 Chinese sorghum
landraces assessed with RFLP markers, Deu et al. [17]
reported that genetic diversity of Chinese sorghum was
very restricted and Chinese sorghum was not highly dis-
tinct from other sorghum in contrast with the finding of
de Oliveira et al. [16] However, given the large amount of
Chinese sorghum landraces and their wide geographical
distribution, the sampling of Chinese sorghum in previ-
ous studies may not be adequately representative with
either small number of accessions of unknown geo-
graphical origins or accessions sampled heavily in few
regions. In addition, there have been no reports on pat-
terns of distribution of genetic diversity in Chinese sor-
ghum.
Molecular markers have proven to be powerful tools
in investigating the extent and distribution of genetic
diversity in sorghum [16-22]. Compared with other genetic
markers, SSR markers have the advantages of uniform
genome coverage, high levels of polymorphism, and
co-dominant pattern of inheritance [23]. With the deve-
lopment and mapping of large quantity of polymorphic
loci [24-27], SSR markers have been used to analyze ge-
netic diversity in subsets from global sorghum collec-
tion [2,28-29], collections from many countries [20,30-31] or
races [22].
In a previous study, we examined chloroplast DNA
variation in 185 Chinese sorghum landraces and 70 for-
eign sorghum accessions, using 14 polymorphic chloro-
plast SSR loci [32]. In this study, the genetic variation in
184 Chinese sorghum landraces was analyzed using 32
nuclear SSR loci and compared with that in 69 foreign
accessions. In addition, the distribution of genetic diver-
sity in Chinese sorghum within and among geographic
origins and types was quantified.
1 Materials and methods
1.1 Plant materials
One hundred and eighty-four Chinese sorghum
landraces were used in this study (Table 1). These acces-
sions were chosen to represent a broad sampling of phe-
notypes and regions of origin. All these accessions were
obtained from the NGBC. At present, all Chinese sor-
ghum accessions maintained at NGBC have not been
classified either according to Snowden’s [10] or Harlan
and de Wet’s [1] systems. Traditionally, Chinese sorghum
landraces were classified into four types according to
their panicle type, i.e. loose, semiloose, semicompact and
compact. Panicle types have played important roles in
sorghum landrace recognition and usage [7]. Many land-
226 作 物 学 报 第 37卷

races were named by farmers according to the compact-
ness of their panicles. Loose and semiloose types were
grown for their stalks and inflorescences as well as grains,
while semicompact and compact types were cultivated
mainly for their grains because of their high yield poten-
tial. Furthermore, semicompact and compact landraces
were also the main types that have been incorporated in
hybrid breeding programs in China [12]. In addition, 69
sorghum accessions from foreign countries, including
landraces and old varieties, were also included in this
study (Table 1). These accessions consisted of five basic
races and eight intermediate races of cultivated sorghum
based on the classification system of Harlan and de
Wet [1]. They were introduced into China in the 1980s and

Table 1 National origin, Chinese province, eco-region and identification number of cultivated sorghum accessions included in this study
National
origin Province Eco-region Accession identification number and classification
a No. of accessions
China Anhui East 543 (SC), 559 (L), 4803 (L), 4811 (SC), 4818 (SC), 4828 (L), 4830 (SC), 4880 (L),
4893 (L), 4939 (C), 4941 (L)
11
Hebei East 394 (SL), 412 (SC), 429 (SL), 481 (SL), 482 (SL), 483 (SC), 484 (SL), 545 (SC), 560
(SL), 593 (SC), 601 (C), 602 (SL) , 628 (SC), 639 (SL), 641 (SL), 642 (SL), 643 (SL),
662 (L), 692 (SL), 693 (SC)
20
Henan East 6069 (L), 6093 ()SC, 6118 (SC), 6199 (SL), 6296 (SL), 6349 (L), 6542 (L), 6646
(SC), 6669 (L), 6685 (L), 6751 (SL)
11
Hubei East 9038 (L), 9045 (SC), 9070 (SL), 9051 (L), 9066 (SL), 9108 (L), 9111 (L), 9119 (L),
9123 (L)
9
Jiangsu East 4640 (SC), 4645 (L), 4646 (SL), 4688 (SC), 4700 (SL), 4748 (L), 4776 (SC), 7510
(L), 7518 (L)
9
Jilin Northeast 15 (SC), 267 (L), 357 (SL), 370 (L), 3670 (SL), 3739 (L), 3876 (C), 3924 (L), 3968
(C), 3975 (C)
10
Liaoning Northeast 6 (C), 19 (C), 26 (SL), 33 (SL), 39 (SL), 55 (SL), 58 (C), 82 (C), 102 (C), 105 (SC),
123 (SL), 129 (L), 144 (C), 265 (L), 279 (L), 354 (L), 7444 (L), 7445 (L), 7447 (L),
7448 (L)
20
Shandong East 404 (SL), 4982 (C), 4983 (L), 4996 (L), 5000 (C), 5001 (SL), 5008 (L), 5018 (L),
5019 (L), 5034 (L), 5046 (C), 5057 (L), 5062 (L), 5104 (SC), 5133 (L), 5142 (SC),
5153 (C), 5178 (L), 5856 (L), 7527 (L)
20
Shanxi West 504 (SL), 505 (L), 506 (SC), 507 (C), 1928 (C), 1950 (L), 1954 (SC), 1963 (L), 1964
(C), 1979 (SL), 2018 (SC), 2026 (SC), 2045 (L), 2071 (L), 2081 (SL), 2087 (L), 2091
(C), 2122 (C), 2135 (SC), 2155 (SL), 2179 (L), 2190 (L), 2260 (L), 2270 (SC)
24
Shaanxi West 9742 (L), 9755 (L), 9768 (SL), 9774 (L), 9795 (L), 9802 (C), 9809 (L), 9811 (C),
9814 (L), 9819 (L), 9826 (L), 9836 (L), 9870 (L), 9881 (L), 9883 (L), 9889 (L), 9892
(L), 9901 (L), 9906 (L), 10403 (L)
20
Sichuan Southwest 9646 (L), 11831 (SL), 11848 (SL), 11851 (SL), 11866 (L), 11888 (L), 11898 (SL),
11899 (SL), 11900 (SL), 11930 (SL), 11932 (SL), 11934 (SL), 11936 (SL), 11950
(SL), 11952 (SL), 11957 (SL), 11962 (SL), 12704 (SL), 12710 (L), 12963 (SL)
20
Yungui Southwest 10042 (L), 10053 (L), 10059 (L), 10060 (L), 10069 (L), 13084 (L), 13086 (L), 13104
(L), 13125 (L), 13130 (L)
10
Chad East Africa IS10747 (B) 1
Ethiopia East Africa IS3758 (KB), IS3798 (CB), IS11540 (C), IS12211 (C), IS18947 (DC), IS18958 (C) 6
Sudan East Africa IS2312 (D), IS2327 (DC), IS2328 (DC), IS6965 (C), IS9985 (D), IS19579 (C),
IS19596 (GC), IS20563 (C), IS25032 (GC),
9
South
Africa
South
Africa
IS2861 (K), IS3161 (KC), IS9333 (K), IS9530 (K) 4
Uganda East Africa IS2649 (G) 1
Zimbabwe South
Africa
IS2800 (C), IS2820 (DC) 2
India South Asia IS1042 (D), IS1044 (D), IS1054 (D), IS1059(DB), IS1082(D), IS2205 (DB), IS4067
(D), IS5284 (D), IS8330 (C), IS18463 (DC), IS18467 (C), IS18473 (GD), IS18484
(GC)
13
Pakistan South Asia IS8345 (D) 1
United
States
North
America
IS13 (B), IS103 (KC), IS156 (K), IS160 (K), IS172 (K), IS183 (K), IS366 (D), IS413
(DC), IS416 (DC), IS474 (DC), IS620 (C), IS627 (DC), IS633 (B), IS634 (CB), IS837
(KD), IS859 (K), IS889 (G), IS2122 (D), IS2914 (K), IS2944 (K), IS3800 (DC),
IS3977 (KD), IS3979 (D), IS10558 (K), IS10672 (DC), IS18681 (GC), IS18704 (DC),
IS22204 (GC), Broomcorn Sorghum (B)b, Tx623B (C)b
30
Mexico North
America
IS75 (C), IS511 (DC) 2
Total 253
a NGBC number (shown above without prefix) for Chinese sorghum accessions or IS number for world sorghum accessions. Letters in paren-
theses after NGBC number indicate the panicle type of Chinese sorghum landraces while those after IS number indicate the races of the accessions. C,
L, SC, and SL represent compact, loose, semi-compact and semi-loose type, respectively, whereas B, C, D, G, K, CB, DC, DB, GC, GD, KC, KB, and
KD represent bicolor, caudatum, durra, guinea, kafir, caudatum-bicolor, durra-caudatum, durra-bicolor, guinea-caudatum, guinea-durra, kafir-
caudatum, kafir-bicolor, and kafir-durra, respectively. Classification and origin of sorghum accessions follow the NGBC (http://icgr.caas.net.cn/) and
ICRISAT (http://www.icrisat.org/sorghum/Project1/pfirst.asp) database, respectively.
b Classification was made based on the system of Harlan and de Wet [1].
第 2期 ZHANG Han et al.: Assessment of Genetic Diversity in Chinese Sorghum Landraces 227


1990s and have been maintained at NGBC. All sorghum
accessions included in the study were introduced from
the short term storage collection of NGBC and had been
regenerated by selfing for unknown times. As a result,
these accessions exhibited high levels of morphological
uniformity in the field [32].
For Chinese sorghum, we combined the accessions
from Guizhou and Yunnan provinces, two neighboring
provinces on Yungui Plateau, into one sample (referred to
as “Yungui”, Table 1) because only five accessions were
sampled from each province. In addition, we combined
neighboring provinces into four eco-regions according to
tradition based on these regions’ geographical proximity,
similarity in ecological conditions and cultivation prac-
tices [9]. For foreign accessions, the following eco-regions
were adopted: East Africa, South Africa, North America,
and South Asia (Table 1).
1.2 DNA extraction
Seeds were germinated under illumination at 25℃
in a seed germinator. For each accession, green tissues
from five one-week-old seedlings were bulked to extract
DNA using the method of Doyle and Doyle [33]. The
quality and concentrations of extracted DNA were tested
by using mini-gel electrophoresis [34].
1.3 PCR amplification and electrophoresis
Thirty-two SSR primer pairs were selected to geno-
type the 253 accessions (Table 2). Primer pairs for Sb and
xtxb markers were developed by Brown, Kong, and
Bhattramakki, respectively [24-27]. Twenty-nine SSR loci
have been mapped to the linkage groups developed from
the progenies of two inbreeding lines by Menz et al.
(http://sorgblast3.tamu.edu/)[35]. The identification, re-
peat motif, linkage group and polymorphism information
content (PIC) for these SSR loci are listed in Table 2.
Polymerase Chain Reactions were set up in 25 µL
volumes containing 50 ng of template DNA, 0.2 µmol
L–1 of each primer, 100 µmol L–1 of each dNTP, 1×PCR
buffer (10 mmol L–1 of Tris-Cl, 50 mmol L–1 KCl, 1.5

Table 2 Locus identification, repeat motif, linkage group, number of alleles and PIC value of SSR primers
SSR locus Repeat motif Linkage group a Number of alleles b PIC b
Sb1-1 (AG)16 10 14 0.549
Sb4-72 (AG)16 Unknown 8 0.695
Sb5-206 (AC)13(AG)20 Unknown 19 0.230
Sb5-236 (AG)20 03 11 0.906
Sb5-256 (AG)8 01 6 0.833
Sb6-42 (AG)26 01 21 0.657
Sb-6-57 (AG)18 01 10 0.605
Sb6-84 (AG)14 02 17 0.574
Sb6-342 (AC)25 07 14 0.664
Xtxp4 (GA)23 02 24 0.803
Xtxp10 (CT)14 09 10 0.437
Xtxp12 (CT)22 04 22 0.836
Xtxp23 (CT)19 05 19 0.858
Xtxp24 (TC)21 04 24 0.914
Xtxp25 (CT)12 02 35 0.952
Xtxp26 (TG)9(AG)12 Unknown 26 0.887
Xtxp36 (GGA)7GTA(T)7+(A)7 07 4 0.397
Xtxp40 (GGA)7 07 4 0.372
Xtxp51 (TG)11 04 8 0.667
Xtxp57 (GT)21 06 18 0.804
Xtxp65 (ACC)4+(CCA)3CG(CT)8 05 8 0.544
Xtxp67 (GA)28 09 21 0.925
Xtxp96 (GA)24 02 16 0.836
Xtxp105 (TG)5+(CT)6+GTCT(GT)7 08 5 0.360
Xtxp141 (GA)23 10 18 0.840
Xtxp145 (AG)22 06 20 0.906
Xtxp177 (CT)7(GT)8 04 7 0.681
Xtxp210 (CT)10 08 12 0.837
Xtxp217 (GA) 23 10 12 0.741
Xtxp303 (GT)13 05 11 0.550
Xtxp319 (TC)17 01 2 0.857
Xtxp354 (GA)21+(AAG)3 08 14 0.736
a Linkage group designated according to Menz et al. (2002, http://sorgblast3.tamu.edu/). b Values on the entire set of 253 accessions.
228 作 物 学 报 第 37卷

mmol L–1 MgCl2, 0.1% Triton X-100, pH 8.3) and 1 U of
Taq DNA polymerase. Amplifications were performed on
a Biometra T1 Thermocycler (Biometra, Goettingen).
The PCR procedures were as follows: 2 min at 95℃,
followed by 35 cycles of 45 s at 95℃, 45 s at the appro-
priate annealing temperatures, 45 s at 72℃, and 5 min at
72℃ for final extension. PCR products were run on 6%
polyacrylamide denaturing gels (19:1acrylamide:bis, 8
mol L–1 urea ) in 1×TBE buffer (0.09 mol L–1 Tris, 0.09
mmol L–1 boric acid, 2 mmol L–1 EDTA, pH 8.0) for
1.0–1.5 h at 1 500 V in a Sequi-Gen GT (Bio-Rad, Her-
cules) apparatus with PBR322/Msp I as molecular weight
standard, then visualized with silver staining as described
by Panaud et al. [36]. Allele sizes were estimated by com-
parison against the molecular weight standard.
1.4 Data analyses
To estimate genetic diversity, we calculated the fol-
lowing statistics for each population defined by regions,
eco-regions or types (Table 1) as well as the entire sam-
ple set, using Arlequin ver 3.11 [37]: the number of poly-
morphic loci (Np), the average number of alleles per lo-
cus (A), the average observed heterozygosity (Ho), and
the average gene diversity (He) [38]. Since the number of
alleles is dependent on the sample size, allelic richness
(Rs), a measure of the number of alleles corrected for
sample size differences, was computed using FSTAT [39].
For Chinese sorghum, the average of number of rare al-
leles (allelic frequencies less than 5%, Ar) was computed.
In addition, the number of unique alleles was counted for
Chinese sorghum and foreign accessions. The PIC value
for each locus on the whole set of accessions was calcu-
lated using PowerMarker V3.25 [40].
Binary matrix was constructed by converting the
data of allele sizes into “1” (present) or “0” (absent).
Each column of the matrix corresponds to an allele. Pair-
wise Dice [41] genetic similarities (GS) between individ-
ual accessions were calculated using NTSYS-pc2.10t [42]
with the binary matrix data as input, using the formula of
Nei and Li [43]. Principal component analysis (PCA) as
implemented in MATLAB Software Release 13 (The
MathWorks Inc, Massachusetts) was performed on the
resulting genetic similarity matrixes for Chinese sorghum
landraces and for the whole set of sorghum accessions,
respectively. A scatter plot of the first and second PCA
axes was drawn to visualize the genetic relationship
among the accessions.
Analysis of molecular variance (AMOVA) as
implemented in Arlequin Ver. 3.11 [37] were performed on
the allelic data to partition total genetic variation into
components of within and among populations. The sig-
nificance of the genetic differentiation was tested with
1 000 permutations.
2 Results
2.1 SSR marker polymorphism
SSR polymorphism in 184 Chinese sorghum land-
races and 69 foreign accessions were examined using 32
pairs of nuclear SSR primers. All the SSR loci were po-
lymorphic. In total, 479 alleles were detected. The num-
ber of alleles per locus varied between 2 (xtxp319) and
35 (xtxp25), with an average of 15. The PIC values
ranged from 0.230 (sb5-206) to 0.952 (xtxp25), averag-
ing 0.726 (Table 2). Each accession could be uniquely
identified.
2.2 Genetic diversity in Chinese sorghum
Populations of Chinese sorghum landraces from 12
regions exhibited varied levels of genetic variation (Table
3). All the 32 loci were polymorphic for accessions from
Jilin, while only 26 loci were polymorphic for accessions
from Jiangsu. Across the 12 regions, the average allelic
richness varied from 3.64 to 4.88, and the mean gene
diversity ranged from 0.517 to 0.714. In terms of both the
mean allelic richness and gene diversity, the highest
value was obtained from accessions originating in Jilin
and the lowest from accessions in Jiangsu. In addition,
accessions from Jilin had the highest number of rare al-
leles in spite of the small sample size, whereas accessions
from Jiangsu had the lowest number of rare alleles. Ac-
cessions from all the regions showed low heterozygosity
with the exception of those from Hubei. The reason may
be that most of the accessions had been regenerated many
times by selfing, which significantly reduces the diversity
level within accessions. We examined the genetic diver-
sity within accessions by analyzing DNA samples from
single plant and little difference was found between
genotyping patterns of single plants (data not shown).
At the eco-regional level, all the 32 loci were poly-
morphic (Table 3). The average allelic richness and allele
diversity varied from 6.51 to 7.23 and 0.593 to 0.653,
respectively. Accessions from the Northeast had both the
highest mean values of allelic richness and gene diversity,
while the lowest values occurred for accessions from the
Southwest.
The genetic diversity in accessions of different
panicle types was also examined. All the 32 SSR loci
were polymorphic for the four types except the semi-
compact accessions, which had 31 polymorphic loci (Ta-
ble 3). The mean values of the allelic richness and gene
diversity were higher for loose and semiloose accessions
than for semi-compact and compact types. However, the
difference in the average gene diversity between the four
panicle types was small.
To compare the genetic diversity of Chinese sor-
ghum with that of accessions from other regions in the
world, we grouped all accessions into the following eco-
regions: East Africa, South Africa, South Asia, North
America and China. The allelic richness and gene diver-
sity varied from 3.250 (South Africa) to 5.093 (North
America), and from 0.609 (South Africa) to 0.732 (East
Africa), respectively. Both the allelic richness and gene
diversity for Chinese sorghum were lower than those for
any of the other eco-regions except South Africa.
第 2期 ZHANG Han et al.: Assessment of Genetic Diversity in Chinese Sorghum Landraces 229


Table 3 Genetic variation within populations of Chinese sorghum landraces defined by geographical regions, eco-regions or panicle typesa
Region/eco-region
/panicle type N Np A (range) Ar (range) Rs Ho He
Regions Anhui 11 30 4.16 (1–9) 0.84 (0–4) 4.02 0.03 0.593
Hebei 20 31 5.25 (1–12) 1.50 (0–7) 4.37 0.01 0.577
Henan 11 31 4.52 (1–9) 1.25 (0–5) 4.35 0.01 0.621
Hubei 9 29 3.84 (1–9) 0.78 (0–3) 3.76 0.11 0.543
Jiangsu 9 26 3.66 (1–9) 0.72 (0–4) 3.64 0.01 0.517
Jilin 10 32 5.00 (2–8) 1.91 (0–5) 4.88 0.04 0.714
Liaoning 20 32 5.25 (2–13) 1.56 (0–9) 4.39 0.02 0.596
Shaanxi 20 31 5.63 (1–12) 1.78 (0–7) 4.66 0.02 0.607
Shandong 20 31 5.03 (1–14) 1.41 (0–10) 4.15 0.02 0.539
Shanxi 24 31 5.34 (1–12) 1.63 (0–7) 4.25 0.03 0.566
Sichuan 20 31 5.16 (1–14) 1.50 (0–9) 4.32 0.06 0.563
Yunhui 10 29 4.25 (1–8) 1.13 (0–5) 4.21 0.04 0.598
Eco-regions Northeast 30 32 7.34 (2–18) 3.13 (0–13) 7.23 0.02 0.653
Central 80 32 8.22 (2–21) 3.72 (0–14) 6.87 0.03 0.599
West 44 32 7.09 (2–15) 2.72 (0–9) 7.19 0.02 0.599
Southwest 30 32 6.47 (2–18) 2.22 (0–12) 6.51 0.05 0.593
Panicle types Loose 86 32 9.22 (2–17) 4.72 (1–15) 7.36 0.03 0.630
Semiloose 49 32 7.75 (2–20) 3.41 (0–14) 7.46 0.04 0.623
Semicompact 27 31 6.25 (1–15) 2.19 (0–8) 6.22 0.04 0.603
Compact 22 32 6.22 (2.15) 2.25 (0–7) 6.25 0.02 0.609
a N: sample size; Np: number of polymorphic loci; A: average number of alleles per locus; Ar: average number of rare alleles (allelic frequencies
less than 5%) per locus; Rs: allelic richness; Ho: average observed heterozygosity; He: average gene diversity (Nei, 1987).

We combined the accessions from foreign countries
into one group (world sorghum), in the light of the well
established fact that cultivated sorghum in these regions
is of African origin [11,45-46,22], and compared the genetic
diversity of Chinese sorghum with that in world sorghum.
Both the allelic richness and gene diversity for Chinese
sorghum were lower than those for world sorghum. In
addition, there were 129 unique alleles for 69 world sor-
ghum accessions, while 99 alleles for 184 Chinese sor-
ghum accessions. Chinese sorghum landraces exhibited
much lower genetic variation than foreign accessions as a
whole.
Principal component analyses were conducted on
the Dice genetic similarity matrixes for Chinese sorghum
landraces and all accessions. The first two principal
components explained 40.72% of the total variation. For
Chinese sorghum, scatter plot of the first two principal
components was drawn (Fig. 1). Accessions from 12 re-
gions intermixed and no clear groups formed on the basis
of region. However, accessions from Yungui, Sichuan,
Hubei and Shaanxi formed a relatively separate group.
Similarly, principal component analysis could not sepa-
rate accessions of different panicle types into distinct
groups (Fig. 2).
For all the Chinese and foreign accessions, the first
and second principal components accounted for 29.36 %
and 6.29% of the total variation, respectively (Fig. 3).
Chinese sorghum landraces and foreign accessions were

Table 4 Genetic variation within populations defined by eco-regions in the world, Chinese sorghum and world sorghum
Population N Np A (range) Au Rs He
Eco-regions East Africa 17 31 6.59 (2–13) 5.085 0.732
South Africa 6 28 3.54 (1–6) 3.250 0.609
South Asia 14 32 6.00 (2–13) 4.891 0.712
North America 32 32 8.13 (3–14) 5.093 0.707
China 184 32 11.06 (2–25) 4.563 0.629
Chinese sorghum 184 32 11.06 (2–25) 99 9.815a 0.629
World sorghum 69 32 11.38 (4–28) 129 11.521a 0.745
N: sample size; Np: number of polymorphic loci; A: average number of alleles per locus; Au: total number of unique alleles; Rs: allelic richness;
He: average gene diversity. a Value calculated for Chinese and world sorghum only.
230 作 物 学 报 第 37卷



Fig. 1 PCA plot indicating the genetic relationships between 184 Chinese sorghum landraces
Accessions are labeled according to their geographical origins.



Fig. 2 PCA plot indicating the genetic relationships between 184 Chinese sorghum landraces
Accessions are labeled according to their panicle types.

clearly separated from each other except for several ac-
cessions mixing with foreign accessions, indicating the
two germplasm collections are genetically differentiated.
AMOVA analysis of the total genetic variation
among the 184 Chinese sorghum landraces and all the
Chinese and foreign accessions showed all variance
components to be significant (P<0.001) (Table 5). Ge-
netic variation in Chinese landraces mainly existed
among accessions within regions, which accounted for
91.94% of total variation, while among regions variation
accounted for the remaining 8.06%. At the eco-regional
level, among and within eco-regions genetic variations
made up 5.03% and 94.97% of the total variation, re-
spectively. In terms of panicle types, most genetic varia-
tion was distributed within accessions of panicle types,
accounting for 97.93% of the total variation, while varia-
tion among panicle types made up 2.07% of the total
variation. For the entire set of Chinese and foreign ac-
cessions, AMOVA analysis indicated that 20.43% of total
variation was attributed to the difference between Chi-
nese and foreign sorghum, and 79.57% was among ac-
cessions within China and foreign countries.
3 Discussion
3.1 Genetic diversity in Chinese sorghum landraces
Our data showed that Chinese sorghum landraces
had lower level of genetic diversity than world sorghum.
It was reported that the average similarity index for Chi-
第 2期 ZHANG Han et al.: Assessment of Genetic Diversity in Chinese Sorghum Landraces 231




Fig. 3 PCA plot indicating the genetic relationships between 253 cultivated sorghum accessions
Accessions are labeled according to their geographical origins.

Table 5 Summary of results of AMOVA (values given are %
variation)
Variance components
Source of
variation Groups Among
groups
Among acces-
sions within
groups
China
Eco-region 4 5.03 94.97
Region 12 8.06 91.94
Panicle type 4 2.07 97.93
China vs. World 2 20.43 79.57

nese sorghum accessions was lower than that for acces-
sions from other regions or races in a study by de Olive-
ira et al. [16] using Inter-SSR, RAPD, and RFLP markers.
Deu et al.[17] found that the genetic diversity of Chinese
sorghum landraces were very restricted. Our study con-
firmed these authors’ findings with a broader and more
balanced sampling of Chinese sorghum using nuclear
SSR markers. However, the gene diversity for our Chi-
nese sorghum accessions (0.629) was larger than that
(0.10) observed by Deu et al. [17]. This could be explained
by the differences in both the Chinese sorghum acces-
sions and the markers used in the two studies. The sam-
pling of Chinese sorghum accessions in our study is more
balanced, representing both the geographical regions and
phenotypes of Chinese sorghum. On the other hand, SSR
markers are more polymorphic than RAPD and RFLP
markers and could reveal more genetic variation [23]. Al-
though the overall genetic variation in Chinese sorghum
landraces was lower than that in accessions from East
Africa, South Asia and North America, it was comparable
to that in South Africa accessions. Genetic diversity in
South Africa accessions was also found lower than that in
other regions [14, 46].
Although the overall genetic diversity level in Chi-
nese sorghum was low, our study revealed a marked re-
gional difference in genetic variation level among Chi-
nese sorghum landraces. Accessions from Jilin Province
exhibited the highest genetic diversity level among all 12
regions, a level only slightly lower than that detected in
East African accessions, but comparable to that found in
accessions from South Asia or North America. In a pre-
vious study, we detected four chloroplast haplotypes
among the same 10 Jilin accessions used in this study, the
highest haplotype number of all regions [32] .This con-
gruence agrees with the observation that genetic variation
level in chloroplast genome is consistent with that in nu-
clear genome [47].
3.2 Differentiation of Chinese sorghum landraces
among regions of origin or types
Both the AMOVA and PCA analyses revealed a
weak differentiation among landraces from different re-
gions or eco-regions (Fig.1, Table 5). Similar results were
reported in previous studies [14,20,28,30,48-49]. The weak dif-
ferentiation among populations of sorghum landraces
from different regions or eco-regions could be explained
as follows: (1) cultivated sorghum has a relatively short
history, thus accessions in different regions or eco-
regions have insufficient time to fully differentiate; (2)
human migration and seed trade between regions or eco-
regions have weakened regional differences. However,
the regional differentiation revealed in our study was
weaker than reported in some other studies. In a study of
Eritrean sorghum landraces, it was found that 23% of the
total variation was attributable to variation among popu-
lations while 77% was from accessions among popula-
tions [30]. The possible reason for this discrepancy might
be that East Africa which Eritrea belongs to is the major
center of diversity and the region where sorghum was
232 作 物 学 报 第 37卷

first domesticated [18,50-51]. As a result, sorghum landraces
in that region might have differentiated more sufficiently.
In contrast, Chinese sorghum is of foreign origin with a
shorter cultivation history. Moreover, China was a highly
developed agricultural society where seed exchange
among regions might have occurred more frequently.
Our results also showed a weak differentiation
among populations of panicle types. (Fig. 2, Table 5). In
a previous study, Menkir et al.[14] reported 86% of the
total variation existed between accessions and 14% be-
tween races. Cross pollination among sorghum landraces
of different types grown in proximity may have been the
major factor for weak differentiation among panicle
types.
3.3 Origin and evolution of Chinese sorghum
The origin of Chinese sorghum has been an issue of
contention. One view is that Chinese sorghum is of Afri-
can origin, which was introduced to China en sea or land
routes (African origin) [4,52]. This view is based mainly on
the fact that there have been no findings of annual wild
sorghum in China. Another view, mainly supported by
some Chinese scholars, holds that Chinese sorghum was
independently domesticated in China (domestic origin).
They argued that Chinese sorghum possesses distinct
morphological traits and extensive morphological varia-
tion. Furthermore, archeological findings suggest that
China might have started sorghum cultivation as far back
in history as 3000 B.C. [3-6], a time even earlier than that
speculated for the start of sorghum domestication in Af-
rica by some authors [11,44].
Chinese sorghum exhibited low genetic diversity
level compared with sorghum from other races or re-
gions [16-17]. These findings would be expected if Chinese
sorghum had been introduced from overseas. Chinese
sorghum accessions also exhibited a significantly lower
chloroplast genetic diversity than foreign sorghum [32].
Our data based on nuclear SSR markers are consistent
with the previous studies. In this study, Chinese sorghum
landraces exhibited lower genetic diversity in terms of
their mean number of alleles per locus, mean gene diver-
sity and the number of unique alleles than world sorghum.
It is noteworthy that a relative high number of unique
alleles was detected in our study compared with that de-
tected by de Oliveira et al. [16]. This is possibly due to the
fact that our sample of Chinese sorghum accessions was
much larger (184 vs. 34) and more representative. Fur-
thermore, SSR loci were more polymorphic than RFLP
or RAPD loci [23].
However, our research indicated a clear genetic dif-
ferentiation of Chinese sorghum from foreign accessions
(Fig. 3, Table 5). Similar results have been observed in
previous studies [16-17]. This differentiation is consistent
with the morphological divergence of Chinese sorghum
from foreign sorghum. Compared to foreign sorghum,
Chinese sorghum landraces are insensitive to daylength
and adaptive to cold to warm temperate climate condi-
tions; their plants have weak tillering, dry pith and white
leaf midribs; most of them are characterized by sessile
spikelets with glabrous, crustaceous to papery, and more
or less striately nerved glumes; the semi-compact and
compact panicle types have more primary and secondary
branches [4,7]. Chinese sorghum landraces also show an
extensive range of agro-morphological variations, rang-
ing from primitive types, such as “wild sorghum” (land-
races similar to primitive bicolor race but with strongly
hairy glumes and shattering spikelets at maturity),
through intermediate types (showing varying degrees of
evolution in a number of characteristics, such as panicle
compactness, glume covering, glume texture, varying
degrees of membranaceous part in relation to the whole
glume, to highly evolved types (such as compact type
that features little glume covering, easy threshability, and
high number of and large grains). Starting from the very
primitive types, through the varied intermediate types to
the highly evolved ones, there appears to have been a
clear “trail” of selections on a number of agro-morpho-
logical traits in Chinese sorghum landraces.
Many authors speculated Chinese sorghum were in-
troduced from India [11,45,53], and postulated that Chinese
sorghum evolved from bicolor race sorghum. de Oliveira
et al. [16] observed that Chinese sorghum was genetically
most closely related to race bicolor and suggested that a
subset of African lines (probably bicolor lines) gave rise
to Chinese sorghum. Cultivated sorghum was first do-
mesticated in the northeast quadrant of Africa [18,50-51].
Early race bicolor is thought to be a set of primitive and
heterogeneous forms originating from wild sorghum [54]
and relates to the early domesticates prior to other race
differentiation [19]. Given that Chinese sorghum contains
large number of primitive bicolor-soghums-like landraces,
we speculate that the early introductions to China might
be primitive bicolor race sorghum. Both the nuclear and
chloroplast genome diversity revealed in this and our
earlier study [32] indicate the early introductions might
have had a quite broad genetic base. More or less sepa-
rated from the outside world, these early introductions
have experienced intense human selection for different
purposes at diversified habitats during a long period in
history. In time, Chinese sorghum have gained their dis-
tinct morphology. Many authors speculated that culti-
vated sorghum may have been introduced into China
through different routes at different times in history [11].
These latter introductions may have lost their typical
morphology due to introgression with local varieties.
3.4 Implication for the exploitation of Chinese
sorghum landraces in breeding programs
A successful pattern for sorghum hybrids develop-
ment in China is using male sterile lines of foreign origin
and restorer lines of Chinese sorghum genetical back-
ground as parents [12-13]. With few exceptions, the restorer
lines are semicompact or compact panicle types. In the
1960s and 1970s, a small number of semicompact and
第 2期 ZHANG Han et al.: Assessment of Genetic Diversity in Chinese Sorghum Landraces 233


compact type landraces with short statures were used in
breeding programs for development of restorer lines.
Many later restorer lines were derived from crosses in-
volving these early restorer lines [12-13]. This practice
leads to a narrowing genetic base for restorer breeding
materials [12-13]. Our study indicates that genetic variation
mainly exists within regions and diversity levels in ac-
cessions from different regions are varied with accessions
from Jilin Province showing highest genetic diversity.
Therefore, accessions from this region should receive
more attention in genetic diversity assessment and ex-
ploitation in breeding programs. Moreover, our study
showed that genetic variation in Chinese sorghum existed
mainly within accessions of different panicle types. The
genetic diversity in semicompact and compact accessions
is considerable compared to loose and semiloose types.
The genetic diversity in semicompact and compact land-
races may not have been fully used, given the importance
of semicompact and compact type landraces in breeding
programs. Emphasis should be given to the assessment of
genetic diversity in Chinese sorghum landraces of these
two panicle types. Since the number of landraces is so
huge, a core collection should be constructed to facilitate
their use in breeding programs. However, some studies
indicated that differentiation of agro-morphological
characters is higher than genetic differentiation between
regions [49]. Therefore, both morphological and genetic
diversity should be taken into account in exploitation of
genetic variation.
4 Conclusion
Lower level of genetic diversity was detected in
Chinese sorghum than in world sorghum. Chinese sor-
ghum had a lower genetic diversity level than accessions
from East Africa (He=0.732), North America (He=0.707)
and South Asia (He=0.712); and was only comparable to
those from South African accessions (He=0.609). Marked
differences in level of genetic variation were revealed
between Chinese sorghum landraces from 12 provinces,
with the highest level of genetic diversity from Jilin. The
results indicated a strong divergence of Chinese sorghum
from world sorghum, but a weak differentiation among
Chinese sorghum both on regional and type bases. Our
study supports the view that Chinese sorghum is of Afri-
can origin. And data from this study have implications
for sorghum breeding programs in China.

Acknowledgments: We would like to thank Mr. Lu Ping,
National Genebank of China (NGBC) for providing all
the plant materials for this study. We also thank Ms. Liu
Yongjie and Ms. Li Qun, Crop Research Institute, Shan-
dong Academy of Agricultural Sciences (SAAS), for
their technical assistance.
References
[1] Harlan J R, de Wet J M J. A simplified classification of cultivated
sorghum. Crop Sci, 1972, 12: 172–176
[2] Grenier C, Deu M, Kresovich S, Bramel-Cox P J, Hamon P.
Assessment of genetic diversity in three subsets constituted from
the ICRISAT sorghum collection using random vs. non-random
sampling procedures: B. Using molecular markers. Theor Appl
Genet, 2000, 101: 197–202
[3] Li F( 璠李 ), Li J-Y(李敬仪), Lu Y(卢晔), Bai P(白品), Cheng
H-F(程华芳). New discovery of the ancient remains of agricul-
ture in the Neolithic site of Dong-Hui hill at Min-Luo county in
Gansu. Agric Archaeol (农业考古), 1989, (1): 56–65 (in Chese
with English abstract)
[4] Wang F-D(王富德), Liao J-L(廖嘉玲). Origin and domestication
of Chinese sorghum. Liaoning Agric Sci (辽宁农业科学), 1981,
(4): 23–26 (in Chinese)
[5] Wei S(卫斯). Origin of cultivated sorghum in China. Agric Hist
China (中国农史), 1984, (2): 45–50 (in Chinese)
[6] Li F( 璠李 ). A brief discussion on the origin and development of
the cultivated plants in China. Agric Archaeol (农业考古), 1993,
(1): 49–55(in Chinese)
[7] Wang F-D(王富德), Liao J-L(廖嘉玲). Classification of Chinese
cultivated sorghum. Liaoning Agric Sci (辽宁农业科学), 1981,
(2): 18–22(in Chinese)
[8] Lu Q, Dahlberg J A. Chinese sorghum genetic resources. Econ
Bot, 2001, 55: 401–425
[9] Qiao K-D(乔魁多), Wei Z-S(魏振山). Agro-Morphological De-
scriptions of Chinese Sorghum (中国高粱品种志). Beijing:
Agricultural Press, 1980 (in Chinese)
[10] Snowden J D. The cultivated races of Sorghum. London: Allard
and Son, 1936
[11] de Wet J M J, Huckabay J P. The origin of Sorghum bicolor: II.
Distribution and domestication. Evol Int J Org Evol, 1967, 21:
787–802
[12] Wang F-D(王富德), Lu Q-S(卢庆善). An analysis of the pedi-
grees of the main hybrid sorghum in China. Acta Agron Sin (作物
学报), 1985, 11: 9–14 (in Chinese with English abstract)
[13] Li Y, Li C. Genetic contribution of Chinese landraces to the de-
velopment of sorghum hybrids. Euphytica, 1998, 102: 47–55
[14] Menkir A, Goldsbrough P, Ejeta G. RAPD based assessment of
genetic diversity in cultivated races of sorghum. Crop Sci, 1997,
37: 564–569
[15] Li Y, Li C. Phenotypic diversity of sorghum landraces in China.
International Conference on Genetic Improvement of Sorghum
and Pearl Millet, 1997. pp 659–668
[16] de Oliveira A C, Richter T, Bennetzen J L. Regional and racial
specificities in sorghum germplasm assessed with DNA markers.
Genome, 1996, 39: 579–587
[17] Deu M, Rattunde F, Chantereau J. A global view of genetic di-
versity in cultivated sorghum using a core collection. Genome,
2006, 49: 168–280
[18] Aldrich P R, Doebley J. Restriction fragment variation in the nu-
clear and chloroplast genomes of cultivated and wild Sorghum
bicolor. Theor Appl Genet, 1992, 85: 293–302
[19] Deu M, Gonzalez-de-Leon D, Glaszmann J C, Degremont I,
234 作 物 学 报 第 37卷

Chantereau J, Lanaud C, Hamon P. RFLP diversity in cultivated
sorghum in relation to racial differentiation. Theor Appl Genet,
1994, 88: 838–844
[20] Djè Y, Forcioli D, Ater M, Lefèbvre C, Vekemans X. Assessing
population genetic structure of sorghum landraces from North-
western Morocco using allozyme and microsatellite markers.
Theor Appl Genet, 2000, 99: 157–163
[21] Menz M A, Klein R R, Unruh N C, Rooney W R, Klein P E,
Mullet J E. Genetic diversity of public inbreds of sorghum deter-
mined by mapped AFLP and SSR markers. Crop Sci, 2004, 44:
1236–1244
[22] Folkertsma R H, Frederick H, Rattunde W, Chandra S, Raju G S,
Hash C T. The pattern of genetic diversity of Guinea-race Sor-
ghum bicolor (L.) Moench landraces as revealed with SSR mark-
ers. Theor Appl Genet, 2005, 111: 399–409
[23] Powell W, Machray G C, Provan J. Polymorphism revealed by
single sequence repeats. Trends Plant Sci, 1996, 1: 215–222
[24] Brown S M, Hopkins M S, Mitchell S E, Senior M L, Wang T Y,
Duncan R R, Gonzalez-candelas F, Kresovich S. Multiple meth-
ods for the identification of polymorphic simple sequence repeats
(SSRs) in sorghum [Sorghum bicolor (L.) Moench]. Theor Appl
Genet, 1996, 93: 190–198
[25] Kong L, Dong J, Hart G E. Characteristics, linkage-map positions,
and allelic differentiation of Sorghum bicolor (L.) Moench DNA
simple sequence repeats (SSRs). Theor Appl Genet, 2000, 101:
438–448
[26] Bhattramakki D, Dong J, Chhabra A K, Hart G E. An integrated
SSR and RFLP linkage map of Sorghum bicolor (L.) Moench.
Genome, 2000, 43: 988–1002
[27] Schloss S J, Mitchell S E, White G M, Kukatla R, Bowers J E,
Paterson A H, Kresovich S. Characterization of RFLP probe se-
quences for gene discovery and SSR development in Sorghum
bicolor (L.) Moench. Theor Appl Genet, 2002, 105: 912–920
[28] Djè Y, Heuertz M, Lefèbvre C, Vekemans X. Assessment of ge-
netic diversity within and among germplasm accessions in culti-
vated sorghum using microsatellite markers. Theor Appl Genet,
2000, 100: 918–925
[29] Casa A M, Mitchell S E, Hamblin M T, Sun H, Bowers J E,
Paterson A H, Aquadro C F, Krosovich S .Diversity and selection
in sorghum: simultaneous analysis using simple sequence repeats.
Theor Appl Genet, 2005, 111: 23–30
[30] Ghebru B, Schmidt R J, Bennetzen J L. Genetic diversity of Eri-
trean sorghum landraces assessed with simple sequence repeat
(SSR) markers. Theor Appl Genet, 2002, 105: 229–236
[31] Abu Assar A H, Uptmoor R, Abdelmula A A, Salih M, Ordon F,
Friedt W. Genetic variation in sorghum germplasm from Sudan,
ICRISAT and USA assessed by simple sequence repeats (SSRs).
Crop Sci, 2005, 45: 1636–1644
[32] Li R, Zhang H, Zhou X, Guan Y, Yao F, Song G, Wang J, Zhang
C. Genetic diversity in Chinese sorghum landraces revealed by
chloroplast simple sequence repeats. Genetic Resources Crop
Evolution, 2010, 57: 1–15
[33] Doyle J J, Doyle J L. A rapid DNA isolation procedure for small
quantities of fresh leaf tissue. Phytochem Bull, 1987, 19: 11–15
[34] Sambrook J, Fritsch E F, Maniatis T. Molecular cloning. New
York: Cold Spring Harbor Laboratory Press, 1989
[35] Menz M A, Klein R R, Mullet J E, Obert J A, Unruh N C, Klein P
E. A high-density genetic map of Sorghum bicolor (L.) Moench
based on 2926 AFLP, RFLP and SSR markers. Plant Mol Biol,
2002, 48: 483–492
[36] Panaud O, Chen X, McCouch S R. Development of microsatellite
markers and characterization of simple sequence length poly-
morphism (SSLP) in rice (Oryza sativa L.). Mol Gen Genet, 1996,
252: 597–607
[37] Excoffier, Laval L G, Schneider S. Arlequin Ver. 3.0: An inte-
grated software package for population genetics data analysis.
Evolutionary Bioinformatics Online, 2005, 1: 47–50
[38] Nei M. Molecular Evolutionary Genetics. New York: Columbia
University Press, 1987
[39] Goudet J. FSTAT, a program to estimate and test gene diversities
and fixation indices (version 2.9.3). Available from http://www.
unil.ch/izea/softwares/ fstat.html, 2001
[40] Liu K, Muse S V. PowerMarker: Intergrated analysis environ-
ment for genetic marker data. Bioinformatics, 2005, 21: 2128–
2129
[41] Dice L R. Measures of the amount of ecological associations be-
tween species. Ecology, 1945, 26: 297–302
[42] Rohlf FJ. NTSYS-pc version 2.10 t. Applied Biostatistics. SSPS,
Chicago, 2001
[43] Nei M, Li M H. Mathematical model for studying genetic varia-
tion in terms of restriction endonuclease. Proc Natl Acad Sci USA,
1979, 76: 5269–5273
[44] de Wet J M J, Harlan J R. The origin and domestication of Sor-
ghum bicolor. Econ Bot, 1971, 25: 128–131
[45] de Wet J M J. Systematic and evolution of Sorghum sect.
Sorghum (Gramineae). Am J Bot, 1978, 65: 477–484
[46] Aldrich P R, Doebley J, Scherts K F, Stec A. Patterns of al-
lozyme variation in cultivated and wild Sorghum bicolor. Theor
Appl Genet, 1992, 85: 451–460
[47] Garris A J, Tai T H, Coburn J, Kresovich S, McCouch S. Genetic
structure and diversity in Oryza sativa L. Genetics, 2005, 169:
1631–1638
[48] Ayana A, Bryngelsson T, Bekele E. Genetic variation of Ethio-
pian and Eritrean sorghum (Sorghum bicolor ssp. verticiliflorum
(L.) Monech) germplasm assessed by random amplified poly-
morphic DNA (RAPD). Genet Resour Crop Evol, 2000, 47:
471–482
[49] Djè Y, Ater M, Lefèbvre C, Vekemans X. Pattern of morphologi-
cal and allozyme variation in sorghum landraces of Northwest
Morocco. Genet Resour Crop Evol, 1998, 45: 541–548
[50] Doggett H. Sorghum, 2nd edn. London: Longman, 1988
[51] Harlan J R, Stemler A B L. Races of Sorghum in Africa. In:
Harlan J R, de Wet J M J, Stemler A B L, eds. Origins of African
Plant Domestication. The Hague: Mouton Press, 1976
[52] Chen W-M(陈文明). Agricultural Archaeology in China (农业考
古). Beijing: Cultural Relics Publishing House, 2002 (in Chinese)
[53] Mann J A, Kimber C T, Miller F R. The origin and early
cultivation of sorghum in Africa. Texas Agricultural & Mechani-
cal Station Bulletin No. 1454, Texas Agricultural & Mechanical
University, College Station, Tex., USA. 1983. pp 1–21
[54] Harlan J R. The Living Fields, Our Agricultural Heritage. Cam-
bridge: Cambridge University Press, 1995