Twelve African cultivars of Oryza sativa were sampled to examine the biological
significance of varietal classification based on isozyme studies. Analyses of F 1
hybrid sterility and of F 2 segregations showed good correspondence between
biochemical markers and observed reproductive barriers. Genetic markers for
some vegetative and reproductive traits were identified. Such linkages could be
involved in the relationships between classifications based on various criteria.
The variability in Africa of the Asian cultivated rice species Oryza sativa has been
described and analyzed with regard to isozymic polymorphism (Ghesquière and
Second 1983; de Kochko 1987, 1988) as well as agromorphological traits (Jacquot and
Arnaud 1979, Miezan and Ghesquière 1986). These studies have shown that the genetic
diversity of O. sativa in Africa is similar to that in Asia and that it is organized in two
main groups corresponding to the indica and japonica subspecies.
The indica-japonica distinction is partly based on the existence of reproductive
barriers among Asian cultivars of O. sativa (Oka 1988). It is thus important to know
if the indica-japonica distinction, maintained in Africa by enzymatic diversity, also
reflects differences in the genetic pool, as in Asia.
Materials and methods
Observations were made at the agromorphological, fertility, and biochemical levels.
Plant material
Twelve African traditional cultivars that represent part of the isozyme variation within
the species on that continent (ORSTOM collection, Table 1) were chosen for their
enzymatic polymorphism. Three of them possess alleles that are, in Africa, characteristic
of the wild species O. longistaminata.
55
Table 1. Varieties used in crossing design.
Variety Origin Culture type Enzymatic structure reaction
Phenol
ES70-6
YS138-3
YS252-1
YS45-1
BS117
BS125
ES44
ES79
SS404
YS309
BS20
2LS102
Tanzania
Guinea
Guinea
Guinea
Guinea-Bissau
Guinea-Bissau
Tanzania
Tanzania
Senegal
Guinea
Guinea-Bissau
Mali
Phreatic
Pluvial
Phreatic, pluvial
Pluvial
Phreatic
?
Irrigated
Pluvial
Irrigated
Phreatic
Pluvial
Irrigated
Japonica
Japonica
Japonica
Japonica
Indica
Indica
Indica
Indica
Indica
lntrogressed japonica a
lntrogressed indica a
lntrogressed indica a
–
–
–
–
+
+
+
+
+
+
–
+
a See text.
F 1 fertility
A 12 × 11 crossing design was used. Variety 2LS102 was used as the female parent.
Four plants of each genotype were studied. Their panicles were bagged at heading to
avoid allopollination. Three panicles per plant were counted. Seed fertility was defined
as number of grains over total number of spikelets. The data were treated by
correspondence analysis and hierarchical ascending classification (HAC).
F 2 segregation
Numerous F 2 progenies were studied for segregation of isozyme markers by enzyme
electrophoresis as described by Second (1982) and de Kochko (1987). The terminology
used for isozyme loci and the correspondence with previous notations are given in
Pham et al (1990).
The conformity of F 2 segregation to Mendelian proportions and the sources of
distortions were tested by chi-square analysis. In segregation involving a null allele,
where heterozygous genotypes were indistinguishable, the allelic frequencies p and q
were estimated based on the frequency of the double-recessive homozygote genotype
(assuming an F 2 distribution according to p 2 :2pq:q 2 ).
Identification of markers for quantitative traits
Both agromorphological traits and genetic markers (isozyme and phenol reaction loci)
were followed in the F 2 of the cross ES70-6/SS404 (321 plants). The methods and
experimental design used to study the relationships between marker loci and quantitative
traits were as described by Pham (1990). Following Tanksley et al (1982), a
significant difference for a trait between F 2 genotypic classes of a marker locus was
interpreted as the existence of linkage between the marker locus and at least one of the
quantitative trait loci of the studied trait.
56 J.-L. Pham
Table 2. Seed fertility of F 1 hybrids and parental lines. a
Seed fertility (%) with indicated male parent
Female
ES YS YS YS BS BS ES ES SS BS YS 2LS
70-6 138-3 252-1 45-1 117 125 44 79 404 20 309 102
ES70-6
YS138-3
YS252-1
YS45-1
BS117
BS125
ES44
ES79
SS404
BS20
YS309
2LS102
87
84
67
74
0
7
85
0
43
57
33
44
83
76
85
0
9
11
34
50
30
15
82
87
88
89
0.5
41
38
0.5
60
79
43
27
70
84
84
88
43
17
21
9
74
83
28
29
12
32
67
73
87
88
90
84
89
64
18
71
16
34
70
51
75
84
44
74
81
85
12
78
39
29
73
67
80
f64
75
78
90
54
68
58
48
47
83
79
74
87
73
90
57
17
65
35
34
81
65
81
66
86
85
77
60
19
65
60
87
80
79
40
53
63
64
90
74
92
26
29
70
16
0.2
2
20
23
9
72
72
51 81
a Parental lines in boldface, Significant differences between reciprocal crosses are indicated by italics. (-) =
missing data.
F 1 hybrids
No genotype effect was detected in F 1 seed from crosses, germination rate (71%), or
losses after sowing (<3%).
Table 2 shows the seed fertility of F 1 hybrids and parental lines, which covered a
uniform range (0–90%). The sample of varieties presents various situations, showing
that a diversity of relationships between genotypes corresponds to the diversity of the
genotypes.
Differences occurred between reciprocal crosses for about one-third of the observed
combinations (Table 2).
Classification of varieties by F 1 fertility
The data were subjected to correspondence analysis after joining Table 2 and its
transposition (although the information given by the crosses with 2LS102 is thus lost,
this genotype is less discriminating). The varieties were then classified by HAC using
as variables the factorial coordinates on axes 1 to 4 (Fig. 1). The plane defined by the
first two axes of the correspondence analysis in Table 2 allows a quick visualization
of the results (Fig. 2).
Two main clusters appear: 1) the japonica (as classed by their enzymatic genotype)
varieties ES70-6, YS138-3, YS45-1, YS252-1, and YS309 and the sole indica BS20;
and 2) the indica varieties SS404, ES44, ES79, BS117, BS125, and 2LS102. For all
varieties, classification into either of these two groups is the same for both male and
female parents.
Genetic diversity in rice ( Oryza sativa L.) in Africa 57
– –
–
–
–
–
–
–
–
–
–
–
–
–
–
1. Hierarchical ascendant classification of 12 cultivars from fertility of their F 1 hybrids. Roman type =
cultivar as male parent; italics = cultivar as female parent.
The japonica cluster + BS20 includes three unequal groups:
• The first consists of typical japonica varieties, which generate fertile withinjaponica
hybrids and sterile hybrids when crossed with indicas. They are ES70-
6 and YS138-3 (male and female), and YS45-1 and YS252-1 (male).
• BS20 (male and female) and YS252-1 (female) are close, and YS45-1 (female)
belongs to this group. These varieties could be called wide compatible (Ikehashi
and Haraki 1984), but only BS20 was wide compatible in both reciprocal crosses.
• YS309 is alone. This genotype generates mostly sterile or semisterile hybrids
and could be called a narrow compatibility variety.
The differences between male and female behavior play a clear part in the
classification of indica varieties (Fig. 1,2). BS125, ES44, SS404, and BS117 form a
group that is notably homogeneous in male behavior but is more scattered in female
behavior, where two subgroups appear. The first female subgroup consists of varieties
(ES79, BS125, and BS117) showing extreme reactions with japonicas, especially with
ES70-6. The second consists of ES44, SS404, and 2LS102, which generate more
intermediate reactions. These results suggest the existence of cytoplasmic variability
in indica cultivars.
Usefulness of biochemical markers
Is there any correspondence between the information from F 1 analyses and that from
biochemical markers?
58 J.-L. Pham
2. Twelve cultivars plotted in the plane defied by axes 1 and 2 of correspondence analysis of F 1 fertility table.
Heavy frame = female parent, light frame = male parent, = japonica cultivar, = indica cultivar.
Enzymatic polymorphism. The classification into two main groups based on F 1
fertility corresponds to that based upon enzymatic criteria. The only exception is BS20,
which is grouped with japonica varieties, though enzymatically it is classed as an
indica.
Furthermore, the enzymatic classification of japonica varieties (de Kochko 1988)
separates YS45-1 and YS252-1 from YS138-3 and ES70-6. This separation is also
apparent in fertility analysis. This correspondence is all the more interesting because
YS45-1 and YS252-1 have “hybrid” genotypes (Second 1982) resulting from reciprocal
introgressions between indica and japonica “ancestral” genotypes. These intermediate
genotypes present a trend toward wide compatibility (as the female parent only).
Our observations agree with those of Clément and Poisson (1986), who describe the
wide compatibility of the japonica varieties of the G3 group (Jacquot and Arnaud
1979); however, reciprocal crosses were not studied by Clément and Poisson (1986).
On the other hand, no correspondence is found in indica varieties between
enzymatic and fertility classification. This lower efficiency of enzymatic analysis
confirms the possible importance of cytoplasmic variability for classifying these
cultivars.
Cytoplasmic variability. Polymorphism of chloroplast DNA (restriction fragment
length polymorphism) of the parental lines (except ES70-6) was studied by Z.H. Shang
(pers. comm.) at ORSTOM, Montpellier, following the method described by Dally and
Second (1989). Using the restriction endonuclease Eco RI, the indica varieties showed
Genetic diversity in rice ( Oryza sativa L.) in Africa 59
three restriction patterns, while all the japonicas showed a common pattern. This
difference in variability cannot be discussed because of the small size of the sample.
Although there is no overlapping between groups of indica varieties obtained from
fertility analysis and the observed restriction patterns, these results show the convergence
of different methods to demonstrate the cytoplasmic diversity of indica varieties.
This conclusion is supported by an analysis of agromorphological traits (data not
shown), which shows numerous differences between reciprocal crosses. Cytoplasmic
variability must be considered in germplasm management and in breeding programs,
especially those using multi-origin populations.
F 2 progenies
Thirty-two F 2 progenies were studied for marker loci segregation, including isozyme
loci and phenol reaction locus.
Analysis of marker loci segregation
Table 3 shows the conformity of F 2 segregations to Mendelian proportions for each
progeny and chromosome. All the abnormal segregations are analyzed in Table 4.
Crosses within subspecies. Two progenies of japonica/japonica crosses and three of
indica/indica crosses were examined. Five chromosomes were marked. All segregations
were normal.
Intersubspecific crosses. Of 15 F 2 progenies studied, 8 showed at least 1 distorted
segregation. Among the seven marked chromosomes, six carried loci that are subject
to distortion. The most common chromosomes with abnormal segregations were
chromosome 6 (loci Est-2 and Pgi-2 ) and chromosome 12 ( Sdh-1 and Acp-1 ). For the
other loci, distortions seemed to be relevant to particular cases. Most of the abnormal
segregations (Table 4) showed unequal allelic frequencies. Random assortment of
gametes was generally observed. Thus, as noted by Pham et al (1990) for other
progenies, deviations from Mendelian ratios result from gametic selection rather than
from zygotic selection.
All distortions did not have the same range (Table 4). The F 2 progenies BS125/
ES70-6, ES70-6/ES79, and YS138-3/ES79 presented extreme distortions at locus Est-
2 (chromosome 6), since only one recessive homozygote plant was observed instead
of the 70 theoretical plants. In the F 2 of ES70-6/SS404, the ratio of allelic frequencies
was nearly 1:3. Similar distortions were observed at locus Cat-1 (chromosome 6) in the
F 2 of BS125/ES70-6 and on chromosome 12 (locus Sdh-1 and Acp-1 ) in the F 2 of ES70-
6/SS404. Other distortions presented an allelic frequency ratio of about 2:3. All allelic
excesses were in favor of the indica allele.
Crosses involving BS20 and YS309. Of 13 F 2 progenies observed, 6 showed at least
one skewed segregation. Seven of the nine marked chromosomes carried loci subject
to distortions. Chromosome 6 was the most susceptible, but the allelic frequency did
not exceed 60:40.
60 J.-L. Pham
Table 3. Conformity of marker loci segregations for each F 2 progeny and marked chromosome. a
Cross
Chromosome and locus (loci)
12
Acp-1
Sdh-1
1
Got-1
Est-5
2
Amp-1
3
Pgi-1 Ph
4 6
Pgi-2
Est-2
6
Cat-1
7 11
Est-9 Adh-1
Pgd-1
Est-1
?
Got-3
?
Chromosomes
(no.) with
distorted loci
Marked
chromosomes
(no.)
Japonica/japonica
YS252-1/YS45-1
ES70-6/YS45-1
BS125/SS404
lndica/indica
SS404/BS125
ES79/BS117
Indica/japonica or
japonica/indica
BS117/YS138-3
BS125/ES70-6
BS125/YS45-1
ES70-6/ES79
ES70-6/SS404
ES79/YS45-1
SS404/ES70-6
SS404/YS45-1
SS404/YS252-1
YS138-3/ES79
YS252-1/ES79
YS252-1/SS404
YS45-1/ES44
YS45-1/ES79
YS45-1/SS404
0
+
0
0
0
0
BS117/BS20
With introgressed varieties
BS20/SS404
BS20/YS309
BS20/YS45-1
ES70-6/BS20
ES70-6/YS309
SS404/BS20
YS138-3/BS20
YS309/BS125
YS138-3/YS309
YS309/SS404
YS45-1NS309
0
0
0
+
0
0
0
0
0
0
0
0
0
0
+
+
0
0
+
0
0
0
+
0
0
0
0
+
0
+
+
0
+
+
0
0
0
0
0
0
+
0
+
+
0
0
0
0
0
+
+
+
0
0
+
+
0
0
0
0
0
0
0
0
+
0
0
0
0
0
0
0
0
0
0
0
0
+
0
+
0
0
0
0
+
0
+
+
+
0
0
0
0
0
0
0
+
+
+
0
0
0
0
0
0
0
0
0
0
0
0
0
0
+
0
+
+
0
+
0
0
0
0
0
0
0
0
2
0
4
3
0
2
2
0
0
1
0
0
1
1
0
2
4
1
0
3
0
2
0
2
0
0
2
2
5
4
5
4
5
3
5
6
2
3
2
2
5
2
3
3
3
3
2
6
7
4
5
7
2
5
2
2
2
3
a 0 = conformity, + = distortion.
Genetic diversity in rice ( Oryza sativa L.) in Africa 61
Table 4. Distorted F 2 segregations.
Allelic c 2 test a
F 2 plants frequency
Chromosome Cross (P 1 /P 2 ) (no.) Locus Homogeneity F 2 distribution
P 1 P 2 of allelic (p 2 :2pq:q 2 )
frequency
1
1
3
4
4
6
6
6
6
6
6
6
6
6
6
6
6
6
6
6
7
11
11
12
12
12
12
12
12
12
12
12
?
?
?
?
?
ES70-6/ES79 120
BS20/YS45-1 258
YS45-1/ES44 52
BS20/YS309 79
ES70-6/SS404 281
BS125/ES70-6 48
ES70-6/ES79 115
BS125/ES70-6 65
BS20/SS404 100
ES70-6/ES79 120
ES70-6/SS404 398
ES70-6/YS309 243
SS404-ES70-6 60
SS404/YS252-1 80
YS138-3/ES79 96
YS309/BS125 80
BS20/YS309 80
ES70-6/SS404 398
SS404/ES70-6 97
YS138-3/BS20 100
ES70-6/ES79 117
BS20/YS309 76
ES70-6/YS309 309
ES70-6/SS404 394
BS125ES70-6 64
BS20/SS404 100
ES70-6/SS404 387
SS404/ES70-6 99
SS404/YS252-1 80
YS309/SS404 59
YS309/BS125 80
YS45-1/SS404 52
BS20/YS309 79
BS125/ES70-6 60
ES70-6/ES79 46
ES70-6/YS309 243
YS138-3/BS20 60
Est5
Got-1
Pgi-1
Ph
Ph
Cat-1
Cat-1
Est-2
Est-2
Est-2
Est-2
Est-2
Est-2
Est-2
Est-2
Est-2
Pgi-2
Pgi-2
Pgi-2
Pgi-2
Est-8
Adh-1
Pgd-1
Acp-1
Sdh-1
Sdh-1
Sdh- 1
Sdh- 1
Sdh- 1
Sdh- 1
Sdh- 1
Sdh- 1
Got-3
Pox-3
Pox-4
Pox-4
Est- 1
.42
.56
.38
.61
.44
.75
.40
1 .00
.40
.00
.24
.44
.78
.59
.10
.47
.34
.40
.58
.52
.40
.38
.62
.36
.61
.39
.36
.65
.69
.19
.30
.48
.58
.59
.33
.48
.61
.58
.44
.62
.39
.56
.25
.60
.00
.60
1 .00
.76
.56
.22
.41
.90
.53
.66
.60
.42
.48
.60
.63
.38
.64
.39
.61
.64
.35
.31
.81
.70
.52
.42
.41
.67
.52
.39
6.52*
2.77 ns
0.93 ns
6.83**
24.00***
8.42**
8.00**
4.90*
0.63 ns
15.63***
32.16***
4.64*
0.32 ns
9.85**
9.50**
37.39***
62.54***
6.1 3*
9.68**
64.83***
16.99***
22.50***
43.93***
25.60***
0.15 ns
5.94*
4.03*
11.13**
1.00 ns
0.59 ns
0.10 ns
2.78 ns
0.41 ns
6.27*
1.48 ns
2.47ns
3.25 ns
5.86*
0.92 ns
1.28 ns
0.07 ns
0.38 ns
0.02 ns
5.92*
0.73 ns
0.36 ns
0.18 ns
3.46 ns
2.90 ns
4.99*
0.1 6 ns
2.58 ns
0.01 ns
6.1 3*
a ns = nonsignificant; significance at the 5% (*), 1% (**), and 0.1% (***) levels.
62 J.-L. Pham
The case of chromosome 6
The region of chromosome 6 marked by loci Pgi-2 and Est-2 is apparently very
susceptible to distortions. Several researchers have noted distortions on loci wx and C,
which are located on the same chromosome (see Nakagahra et al 1974). Figure 3 shows
the intervarietal relationships of these loci. Some results were added involving parent
108 used by Oka (1958) as an indica tester. The results support the classification
obtained from studying F 1 hybrids. Japonicas ES70-6 and YS138-3 are opposed to
indica varieties. The wide compatibility of YS252-1, YS45-1, and BS20 was confirmed,
although a light reciprocal effect was observed with SS404. The uniqueness of
YS309 was also confirmed.
The rate of distortion in the F 2 was not proportional to F 1 sterility. One explanation
is that the genetic systems involved in hybrid sterility are not all located near loci Est-
2 and Pgi-2, and that other mechanisms are involved in distortions (Nakagahra et al
1974).
Relationships between agromorphological traits and genetic markers
We will discuss here some of the results obtained from studying the japonica/indica F 2
progeny of ES70-6/SS404 (Pham 1990).
Table 5 presents the significant effects obtained for three loci that were used for
classifying cultivated rice: Pgi-2 (Second 1982), Acp-1 (Inouye and Hagiwara 1980,
Shahi et al 1969) and Ph (Oka 1958). Our detection of quantitative trait loci indicates
that these loci have different allelic states in ES70-6 and SS404. Polymorphism among
quantitative trait loci seems therefore to correspond to enzymatic polymorphism.
Assuming that these results could be extended to other varieties, the linkage of these
markers with traits that were often considered useful in classifying cultivated rice
(tillering, shape of the grain, size of the flag leaf, length of the panicle) could partially
3. Intervarietal relationships of loci Est-2 and Pgi-2 on chromosome 3. Broken lines indicate normal F 2
segregations; solid lines correspond to skewed F 2 segregations. Arrows show parent whose allele is in excess.
Genetic diversity in rice ( Oryza sativa L.) in Africa 63
explain the correspondence between the classifications obtained using enzymatic and
morphological criteria.
Locus Pgi-2 appears to be a marker common to morphological and reproductive
traits, since linkage was also demonstrated with fertility. This result agrees with the
genetic map of rice, which locates numerous genetic sterility systems on chromosome
6. The existence of reproductive barriers contributing to the isolation of indica and
japonica subspecies could therefore be correlated with the preservation of a morphological
identity.
Conclusion
The genetic structure of the sample of African varieties may be clearly revealed by
studying the diversity of biochemical markers. The indica-japonica distinction revealed
by isozyme studies corresponds to a distinction based on reproductive barriers
like hybrid sterility and abnormal transmission of genetic information in F 2 progeny.
Although our study was limited by the small number of varieties, the results favor using
biochemical markers for evaluating rice germplasm collections, since the resulting
classifications have biological significance.
Table 5. F 2 progeny of ES70-6/SS404. Tests for difference between genotypes for 3 marker
loci with respect to some quantitative traits. a
Locus
Trait ACp-1 Pgi-2 Ph
ES/ ES/ SS/ Test ES/ ES/ SS/ Test ES/ ?/ Test
ES SS SS ES SS SS ES SS
No. of tillers at 50 d
after sowing
Heading date (no. of
days from sowing)
Plant height (cm)
Length-width ratio
of flag leaf
No. of primary
branches of panicle
Length-width ratio
of grain
Seed fertility (%)
107 100 104
24.8 26.1 27.2
12.2 11.3 1.7
2.76 2.88 2.93
ns
ns
**
**
**
ns
4.72 5.70 6.11
99.0 92.8 94.5
110 102 102
11.6 11.3 11.9
31.2 24.4 34.0
***
**
ns
ns
**
26.0
12.7
26.5
11.2
ns
ns
ns
***
ns
ns
a 321 plants were studied. ES/ES = homozygous for ES70-6 allele, SS/SS = homozygous for SS404 allele, ES/
SS = heterozygous, ?/SS = ES/SS and SS/SS are indistinguishable. ns = nonsignificant, * = significant at the
5% level, ** = 1%, *** = 0.1%.
No comments:
Post a Comment