the basis of 39 morphophysiological traits and 19 isozyme genes. Comparison
with Asian and African rices revealed the existence of new varietal types that do
not fit the existing classification schemes. These types are mainly lowland
cultivars grown in the high plateau region at altitudes ranging from 1,000 to
1,500 m. Based on morphophysiology, they are intermediate between indica and
tropical japonica types for most traits, although they are the tallest. Isozyme data
show a limited global gene diversity and a marked bipolar structure similar to the
classical indica-japonica structure with, however, a peculiar predominance of
allele 2 at locus Amp-1, forming multilocus types that are rare or absent in Asia.
Classical associations between some isozymes and some morphological traits
are almost nonexistent. The introduction of rices from Asia to Madagascar was
thus probably accompanied by a strong founder effect and was followed by
intensive intersubspecific recombination. Adaptation to new ecological niches
took place without pronounced disruption of subspecific complexes of coadapted
genes.
 |
| Sandy land |
The indica-japonica differentiation is the main feature of varietal diversity in Asian
cultivated rice (see Oka 1988 for a review). Such a pattern most probably arose from
multiple domestications and the associated founder effects. Post-domestication varietal
migrations were extensive, and the two types are now distributed over most Asian
regions. There remains evidence of ecological specialization, leading indica varieties
to be grown mainly under tropical lowland conditions, and japonica varieties mainly
under temperate conditions and tropical upland conditions. In some environments,
such as tropical highlands, both types are sympatric and are thus exposed to intersubspecific
introgressions. An isozymic survey of Asian traditional varieties (Glaszmann
1987, 1988) suggested that few indica-japonica intermediates exist, and that most
intermediate-like varieties are more probably consequent to the contribution of local
wild rices rather than to intervarietal recombination.
The introduction of Asian rice to other continents has enabled the study of various
aspects of rice evolution. Continental Africa has been extensively studied, and the
genetic structure of local cultivated rice is considered the result of new opportunities
for intra- and interspecific introgressions (Bezançon and Second 1984, Ghesquière
1988, Kochko 1987b). Madagascar offers a simpler situation. Because of geographic
isolation due to insularity, ancient rice introductions were limited in number. Historical
and linguistic hints (Boiteau 1977, Dez 1965, Domenichini-Ramiaramanana 1988,
Domenichini-Ramiaramanana and Domenichini 1983) suggest two main origins: from
Indonesia with the Protomalagasy people, and from the Indian subcontinent. Recent
studies showed that Madagascar exhibits a large proportion of varieties showing
particular morphophysiological and biochemical character associations, beside the
typical indica and japonica varieties (Ahmadi et al 1988, Kochko 1988, Rabary et al
1989).
In this paper, we review the peculiarities of the rices from Madagascar. We compare
them with the Asian morphological types, and with the Asian and African isozymic
types. Both sources of information lead to deductions about the evolutionary past of
cultivated rice in Madagascar that are relevant to the interpretation of cultivated rice
diversity in other parts of the world.
Materials and methods
Morphophysiological and isozymic evaluations were carried out on 179 and 182
varieties, respectively, from the Madagascar national collection maintained by the
Centre National de la Recherche Appliquée au Développement Rural (FOFIFA) and
from recent field collections by FOFIFA and the International Board for Plant Genetic
Resources; 145 varieties were common to the 2 samples.
The morphophysiological evaluation was performed at Alaotra Lake, Madagascar,
and involved 24 quantitative and 15 qualitative characters. The procedures followed
were those described by Jacquot and Arnaud (1979). The data were subjected to a factor
analysis of correspondences (FAC) (Benzecri 1973) and “nuées dynamiques” (Diday
1971) after transformation of the quantitative data into qualitative data.
The isozyme study involved 10 enzymes encoded by 19 polymorphic genes as
described by Glaszmann et al (1988), namely catalase (CAT), esterase (EST), aminopeptidase
(AMP), acid phosphatase (ACP), shikimate dehydrogenase (SDH), alcohol
dehydrogenase (ADH), isocitrate dehydrogenase (ICD), phosphogluconate dehydrogenase
(PGD), phosphoglucose isomerase (PGI), and glutamate oxaloacetate
transaminase (GOT).
Results
Both morphophysiological and isozymic evaluations identified varietal types specific
to Madagascar, besides types commonly found in Asia.
68 Ahmadi et al
Morphophysiological variability
The FAC of the morphophysiological variability identified four groups, named G2,
G4, G5, and G6 (Ahmadi et al 1988) for the sake of homology with a previous study
(Jacquot and Arnaud 1979). The distribution of these groups on plane 1, 2 of the FAC
is shown in Figure 1. Axis 1 positively attracts the varieties with high tillering capacity,
thin organs, and short panicles with few branches. Axis 2 positively attracts the
varieties with high stature; intermediate tillering; and big, heavy grains with a positive
phenol reaction. The main traits of the morphological groups defined in Asia (Matsuo
1952, as reported by Angladette 1966) and in Madagascar are presented in Table 1.
The indica type in Asia, with high stature, high tillering capacity, thin organs, and
fine and light grains with a positive phenol reaction, is represented in Madagascar by
group G5, and is associated with lowland culture (with standing water) in regions
having low altitude. In Madagascar, this group could be subdivided into two subgroups,
G5A and G5B;G5A varieties display higher tiller numbers and thinner organs.
The temperate japonica type in Asia, with short stature, short panicles with little
branching, low grain shattering, and bold grains with negative phenol reaction, is
represented in Madagascar by group G2. Since these are modem varieties introduced
from Asia less than a century ago, they are not considered here.
The tropical japonica type (often referred to as javanica) in Asia, with high stature,
few tillers, long panicles with profuse branching, low grain shattering, and heavy, thick
grains with negative phenol reaction, is widely represented in Madagascar by group
G4. This type of rice has been grown under slash-and-bum culture in low-altitude forest
areas on the eastern coast for more than 1,500 yr (Labatut and Raharinarivonirina
1969).
1. Distribution of 4 groups of varieties in plane 1, 2 of the factor analysis of correspondences among 39
morphophysiological characters of rice in Madagascar (after Ahmadi et al 1988).
Traditional highland rices from intersubspecific recombination 69
Table 1. Main features of morphological groups defined in Asia (general description,
Angladette 1966) and in Madagascar (average and range, Ahmadi et al 1988). a
Varietal group
Character Temperate japonica Tropical japonica Indica Atypical
Asia Madagascar Asia Madagascar Asia Madagascar Madagascar
(G2) (G4) (G5A/G5B) (G6)
Length of
under flag
1st leaf
leaf(cm)
1st leaf
under flag
leaf(cm)
Width of 1st leaf
Tillering
Plant height (cm)
Culm diameter
Panicle length
Shattering (%)
Panicle secondary
Grain length
branches
Grain width
(L, mm)
Grain shape (L/W)
(W, mm)
100-grain
Phenol reaction
(mm)
(cm)
weight (g)
Short
Narrow
Intermediate
Short
Intermediate
Short
Little
Few
Short
Wide
Bold
Heavy
Negative
30.9
(24.6–36.2)
10.3
(8.7–12.3)
(9.5–18.1)
14.8
85
4.3
(3.6–4.8)
16.8
10.6
25.5
(74–101)
(15.2–19.5)
(4.7–24.6)
(22.1–30.5)
8.5
(7.9–9.3)
3.5
(2.8–4.1)
2.4
(2.1–2.7)
2.7
(2.0–4.9)
a Figures in parentheses are ranges.
Long 46.3
(39.4–51.9)
Wide
Low
Tall
Thick
Long
Intermediate
Many
Inter-
Wide
mediate
Big
Heavy
Negative
16.0
(10.7–18.1)
(5.5–16.6)
8.7
120
(97–137)
5.3
(3.8–6.5)
(18.8–29.3)
24.4
7.9
(1.7–17.9)
44.0
(28.0–66.4)
(7.1–11.3)
9.4
3.6
2.6
(2.2–4.4)
3.0
(3.1–4.2)
(2.1–3.6)
Long
Narrow
High
Tall
Intermediate
Intermediate
Much
Intermediate
Long
Thin
Fine
Light
Positive
(23.0–40.1)
32.8
10.6
(8.0–13.8)
14.6
10.9–23.2)
111
(73–135)
4.3
(3.3–5.6)
20.7
17.2–23.2)
17.4
(4.5-32.8)
31.5
21.5–39.5)
10.0
(7.8-11.8)
(2.5–3.4)
2.9
3.4
(2.7–4.2)
2.7
(2.2–3.5)
+
(29.8–57.7)
39.1
(9.7–17.3)
12.4
(8.2–19.7)
11.3
(104–145)
122
5.0
(3.3–6.0)
(17.7–26.6)
22.1
12.5
(5.1–32.0)
(23.3–50.4)
35.9
9.6
(7.6–12.1)
3.3
(3.0–3.8)
2.8
(2.0–3.8)
3.0
+/-
(2.7–3.8)
The group specific for Madagascar, namely group G6, encompasses varieties grown
under lowland conditions, having the highest stature, and whose other characters are
intermediate between those of the tropical japonica group (G4) and the indica group
(G5). Group G6 is numerically important in Madagascar and has no equivalent in Asia.
Thus, the morphophysiological variability of rice is particular in Madagascar:
beside varietal groups G2, G4, and G5, which fully correspond to the temperate
japonica, tropical japonica, and indica Asian types, respectively, an additional group,
G6, stands out with atypical character combinations. Its members, particularly those
with the vernacular designations “Rojo” and “Latsika,” are grown mostly at altitudes
higher than 1,000 m and have a high level of cold tolerance (Rasolofo et al 1986).
Isozymic variability
The FAC of the isozymic variability identified two main clusters close to the indica and
japonica groups (groups I and VI of Glaszmann 1987; Fig. 2). Two representatives of
70 Ahmadi et al
–
2. Distribution of
isozyme data at 15
from Madagascar,
varieties from Madagascar in plane 1, 2 of the factor analysis of correspondences of
loci. Delimitations of the 6 varietal groups found in Asia are shown. Among the varieties
129 cluster in the shaded area close to group I, 48 cluster in the shaded area close to group
VI, 2 fall into group V, and 3 are intermediate (from Rabary et al 1989).
group V, otherwise characteristic of the Indian subcontinent, were also found. The
main features of the genetic diversity in the main enzymatic groups in Asia (Glaszmann
1987, 1988), in Africa (Kochko 1987a), and in Madagascar (Rabary et al 1989) are
given in Table 2.
Among the 19 loci surveyed, all polymorphic in Asia and Africa, only 13 exhibit
some variation in Madagascar, and they display 30 alleles vs 46 in Asia and 33 in
Africa. The missing alleles are those rare in Asia. Thus, for most polymorphic loci, the
allele most frequent in Asia further gained importance in Africa and Madagascar. This
tendency is reversed in Madagascar for several loci. The reversal is particularly
pronounced for Amp-1, where allele 2, rare in Asia (13%) and in Africa (10%), has
become frequent (61%).
When the indica and japonica groups are compared with their Asian counterparts,
the loss of alleles is less pronounced than when the overall allele frequencies are
considered, because the alleles specific for the minor Asian groups are no longer taken
into account. However, some of the reversals noted above become stronger.
For the indica group, Madagascar is characterized by a reversal of allele frequencies
at loci Amp-1, Est-2, and Sdh-1 as compared with Africa, and at those loci and Pgi-2
as compared with Asia. The most significant peculiarity is the high frequency of allele
2 at locus Amp-1, which is very rare in Asia and much less frequent in Africa.
Traditional highland rices from intersubspecific recombination 71
Table 2. Genetic diversity of the main enzymatic groups in Asia (Glaszmann 1988), in Africa (including Madagascar, Kochko 1987b),
and in Madagascar (Rabary et al 1989); allele frequencies; and diversity indices (H) (Nei 1975). a
Allele frequency (%)
Locus Allele Whole sample Indica group Japonica group
Asia Africa Mada- Asia Africa
Madagascar Madagascar
Asia Africa
(n= (n= gascar (n=900) (n=359) Indica I I* (n= (n= Japonica J J*
1688) 688) (n=181) (n=129) (n=32) (n=97) 451) 329) (n=47) (n=41) (n=6)
Cat-1
Sdh-1
Pgi-1
Pgi-2
Est-1
Est-2
2
1
3
H
1
3
2
4
H
2
1
H
2
1
3
4
H
0
1
H
0
2
1
H
83
17
100 99
1
100
96
4
tr
0.08
99
1
81
19
0.31
100
0
98
2
0.04
100
71
29
0.41
37
62
1
1
0.48
49
51
0.50
60
29
8
3
0.55
9
91
0.17
36
42
22
0.65
97
3
0.05
13
87
97
3
0.05
50
50
0.50
100
0
97
3
0.06
88
12
0.21
19
31
50
0.62
65
35
93
7
100
tr 100
0
100
0
100
0
100
0
100
0
100
0
0.28
30
67
3
0.46
39
61
0.48
92
7
tr
tr
0.15
32
68
0.44
36
38
26
0.66
0
58
40
2
0.02
82
18
0
7
93
0 0.46 0.13
43
54
2
0.52
1
87
13
0.23
47
53
100 100
0.22
66
34
0.44
98
1
1
0.04
12
88
0.21
64
22
12
0.52
0.50
76
24
0.36
84
14
1
1
0.27
0.30
90
10
0.18
98
2
0.13
87
13
0.23
98
2
0.04
6
94
0.11
0.02
tr
100
tr
100
0 0
98
2
0.04
100
100
0
100
0.04
93
7
0.13
59
24
17
0.57
0
25
75
0.38
76
24
0
29
71
0.41
74
24
2
0.39
0.50
3
97
0.06
13
47
40
0.60
0
67
33
0.44
53
47
0
25
75
0.38
75
23
2
0.36
100
0
21
30
49
0.63
72
22
6
0.43 0.37 0.50
continued
– – – – – – – – – – – –
– – – –
– – – – – – – –
– – – – – – – – –
– –
–
–
– –
– –
–
– –– – – – –– –– –– –– –– – – – – – – – –
– –
– – ––
Table 2 continued.
Allele frequency (%)
Locus Allele Whole sample Indica group Japonica group
Asia Africa Mada- Asia Africa
(n= (n= gascar (n=900) (n=359) Indica I I* (n= (n= Japonica J J*
1688) 688) (n= 181) (n=129) (n=32) (n=97) 451) 329) (n=47) (n=41) (n=6)
Madagascar Madagascar
Asia Africa
Est-5 0 tr
1 99
3
2
97
1
H 0.02 0.06
Est-9 1 40
2 60
43
H 0.48 0.49
57
Cat-1 1 71 83
Amp-1 1 78 88
2 13 10
3 4 2
4 5
5 tr
H 0.37 0.22
Amp-2 1 39 P
2 61 P
3 tr
4 tr
H 0.48 –
100 100
tr
94
6
0.11
43
57
0.29
100
78
19
3
100
0
100 100 100
0 0 0
100
0
100
0
100
0
100
0
continued
100 100 100
0
62
38
0.47
97
39
61
tr
64
34
0.45
100
93
tr
6
1
0.13
1
99
0
81
19
0.23
97
100
0
87
13
0.28
100
100
0
tr
100
tr
4
90
62
38
0.47
99
25
75
100
0
65
100
tr
tr
P
100
0
81
88
12
0.21
100
0
100
0
93
100
0
100
0
10
tr
0.47
30
70
0.35
P
P
0.38
2
98
0.04
0
100
0
3
97
0.18
99
tr
tr
tr
0.42 0.02 0 0.06 0.02
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
– –
– –
– –
–
–
–
–
–
Table 2 continued.
Allele frequency (%)
Locus Allele Whole sample Indica group Japonica group
Madagascar Madagascar
Asia Africa Mada- Asia Africa Asia Africa
(n= (n= gascar (n=900) (n=359) Indica I I* (n= (n= Japonica J J*
1688) 688) (n = 181) (n=129) (n=32) (n=97) 451) 329) (n=47) (n=41) (n=6)
51
47
2
67 62
33 38
0.44 0.47
Amp-3
Acp-1
Adh-1
Pgd-7
All loci
P
P
P
1
71
28
P
P
0
1
2
3
4
5
6
H
1
2
3
H
0
1
2
3
H
1
2
3
H
4
48
43
1
3
tr
1
0.58
62
67
1
0.48
95
tr
4
1
0.09
67
6
27
0.47
0.62
P
P
73 100
27
63
37
76
24
73
27
tr
0.39
96
4
0.08
99
1
0.02
68
32
0.47
91
9
0
17
83
0.36
98
2
0.39
1
99
0.41
71
29
0.52
98
2
99
1
0.02
P
59
17
24
0.56
0.30
52
48
0.50
P
P
2
100 98 100
0.42 0.28
99
0.04 0.16 0.04 0.02
tr
88
12
tr
0.21
96
0 0.04 0
99 100 99 P 100 100 100
1
0.02
76
21
3
0.42
0.29
1
0.02
61
15
24
0.55
0.28
1
0.02
63
37
0.47
0.1 5
0
100
0 0
85 83
15 17
0.26 0.28
0.05 0.12
64 70
9
27
0.51
0.38
84
16
4
0.08
0.1 0
30
0.42
0.16
0
0.03
0.44
0.23
0.27
0.20
a tr = traces, 0.5 > tr > 0; P = present, >0, but undetermined. * = presence of allele 2 at locus Am p-1.
– – – – – – – – – –
– – – – – – – – – – –
– – – – – – – – – – – –
– – – – – – – – – – –
– – – – – – – – – – – –
– – –
– – –
– – – – – – – – – – – –
– – – – – – – – – – –
– – – – – – – – – –
– – – – – – –
– – – –
– – – – –
– – – –
For the japonica group, Madagascar differs from Africa by the reversal of allele
frequencies at locus Est-1, and from Asia by the reversal at locus Cat-1. For this group,
too, the frequency of allele 2 at locus Amp-1 is significant, whereas this allele is absent
from the japonica group in Asia and Africa.
On the basis of the allele at locus Amp-1, Rabary et al (1989) tentatively distinguished
subgroups I and I* in the indica group, and J and J* in the japonica group,
* indicating the presence of allele 2. The subgroups I and J display only minor
differences from their African and Asian counterparts. The same is true for J*, besides
the difference at locus Amp-1. For I*, more significant differences account for all the
peculiarities of the indica group in Madagascar.
Thus, as morphophysiological variability does, isozyme variability of rice in
Madagascar identifies specific types beside the usual indica and japonica types
observed in Asia.
Congruence between the two classification schemes
The distribution of the 144 varieties analyzed for both morphology and isozymes on
plane 1, 2 of the FAC of the morphological data is shown in Figure 3. The isozyme
groups and the culture type—upland vs lowland—are distinguished, and two classes
of elevation are considered.
Enzymatic group I contains all lowland varieties, distributed mostly in morphological
group G5, but a few falling into group G6. All are grown at elevations lower than
1,000 m.
Varieties of enzymatic group I* are scattered in all morphological groups at all
altitudes with, however, higher concentration in group G6 and at elevations above
1,000m. Among the 78 I* varieties, 3 are grown under upland conditions.
Varieties of enzymatic groups J and J* cluster around morphological groups G4 and
G6. In G4, most are upland varieties (tropical japonica) grown at low elevations,
whereas in G6, all are lowland varieties belonging to the vernacular families “Vary
Lava” and “Latsika.” The former are well known for their very long and wide grains;
the latter are grown only above 1,500 m.
• About 75% of the morphologically atypical varieties (G6) are also atypical
• A higher frequency of atypical varieties is observed above 1,000 m.
• The cold-tolerant lowland japonica varieties do not morphologically resemble the
• Some upland varieties possess both the typical morphology of tropical japonica
These are the main noteworthy outputs of this comparison:
regarding isozymes.
temperate Asian japonica varieties.
(G4) and basically indica isozymes (I*).
Discussion
The genetic diversity of rice in Madagascar is characterized by the presence both of
varietal groups similar to the indica and japonica groups from Asia and Africa and of
Traditional highland rices from intersubspecific recombination 75
3. Varieties from Madagascar scattered in plane 1, 2 of the factor analysis of correspondences among 39 morphophysiological
characters (see Figure 1) and differentiated according to enzymatic group and culture type
for 2 classes of elevation. = I, lowland; = I*, lowland; = I*, upland; = J, lowland; = J, upland,
= J*, lowland; =group V, lowland; * = intermediate, lowland (Rabary et al 1989).
groups specific to the island, where morphophysiological and isozymic variations are
not fully congruent.
Comparison of isozyme allele frequencies among Asia, Africa, and Madagascar
reveals that the introduction of rice to Africa and Madagascar was accompanied by the
loss of minor Asian alleles. But in Madagascar, a few exceptions are noted, such as
allele 2 at locus Amp-1, which is restricted to the Indian subcontinent in Asia and is
predominant in Madagascar.
76 Ahmadi et al
Thus, rice introduction in Madagascar caused genetic drift linked to founder effects.
Allele Amp-1 2 in Asia is found mainly among varieties that differ from indica and
japonica types in several other specific alleles; in a survey of 1,688 Asian varieties
(Glaszmann 1987,1988), only 2 indicas displayed this allele: one from southern India
and one from Sri Lanka. In Madagascar, this allele is found in an array of genotypes,
all very rare or absent in Asia. Its introgression from local wild rices can be excluded,
for only Oryza longistaminata is present on the island, and it does not possess this allele
(Ghesquière 1988, Second 1985). Therefore, these genotypes must have arisen from
local genetic mixing within O. sativa. Moreover, since this allele is found in both indica
and japonica isozymic backgrounds, mixing has involved both intrasubspecific and
intersubspecific recombinations. The higher frequency of japonica-prone alleles
Est-2 0 and Sdh-1 2 in group I * than in group I confirms that I * varieties probably arose
from intersubspecific recombination. The situation is different in Africa, where the frequency
of Amp-1 2 is higher than in Asia only among indicas, with a low frequency of
alleles Est-2 0 and Sdh-1 2 .
Evidence drawn from the isozyme data alone is reinforced by the morphological
evidence of many intermediate forms. The environmental conditions of high plateaus
in Madagascar certainly exerted particular selection pressures that favored certain
recombinant forms, as shown by the high frequency of these forms at high elevations
and their cold tolerance (Rasolofo et al 1986).
The data are conclusive, thanks to the peculiarities of Madagascar, where varietal
migrations and gene introgression from wild rices did not complicate the pattern of
variation. Some firm conclusions are relevant to other, less favorable situations. In
particular, it is striking that intersubspecific recombination has left so few intermediates
as far as isozymes are concerned. This is to be compared with the tendency of
parental gene combinations to increase among the progenies of indica/japonica crosses
(see Oka 1988 for a review). Those genes that kept their initial assortment are certainly
tightly linked to components of coadapted gene complexes involved in the maintenance
of indica-japonica differentiation. These are Pgi-1 on chromosome 3, Cat-1 on
chromosome 6, Est-9 on chromosome 7, Amp-2 on chromosome 8, and Acp-1 on chromosome
12 (Wu et al 1988), using the new chromosome nomenclature. Therefore,
other regions in Asia or Africa where these genes form multilocus associations should
not be considered without indica-japonica gene introgression.
No comments:
Post a Comment