1. Is Yunnan a rice diversity center in isozyme variation?

Yin-Li SUN, Hong-Wei CAI and Xiang-Kun WANG

Agronomy Department, Beijing Agricultural University, Ma-Lian-Wa, Beijing, 100094 China

To compare diversity and the degree of Indica-Japonica differentiation, a total of 463 rice cultivars (O. sativa) were examined for isozymes, which included 115 from Yunnan, 126 from other provinces of China, 83 from Southeast Asia (Vietnam, Laos, Cambodia, Thailand and Indonesia), 83 from South Asia (India and Bangladesh, excluding Aus types) and 56 cultivars of the Aus type. The isozymes studied were 8 polymorphic loci, i.e., Amp-1, Amp-2, Acp-1, Acp-2, Cat-1, Est-2, Est-10, and Mal-1. The PAGE (polyacrylamide gel electrophoresis) method was used for isozyme analysis.

From the results, average gene diversity [H=1/nSIG(1-SIG{i}x2\i\i\), Nei 1975],"genotype frequency" (no. of genotypes/no. of cultivars; F), and squared correlation coefficient between isozyme loci (R2; Sano and Morishima 1992) were computed in each regional group, separately for the Indica and Japonica types.

The Indica and Japonica types were classified on the basis of frequencies of isozyme alleles at 6 loci: Acp-1, Acp-2, Amp-2, Cat-1, Est-10 and Mal-1, as follows: For each allele at each locus, discriminant parameter was computed as D`i`=F`i`/(F`i`+F`j`), where F`i` and F`j` are the frequencies of a given allele in the Indica and Japonica groups of control cultivars, respectively. Then, the average of D`i` values for the six loci of each accession representing its allelic constitution was computed for quantifying isozyme variations. This computation showed that the Indica and Japonica types were distinguished clearly, and that the cultivars from different regions contained the both types except for the Aus group which appeared to be wholy Indicas.

The results obtained for average gene diversity (H) and "genotype frequency" (F) are given in Table 1. The table shows that the Indica type generally had higher H and F values than the Japonica type, although the difference varies among the regional groups. The varieties from Southeast Asia had high H and F values in both the Indica and Japonica groups, followed by the South Asian varieties. Yunnan varieties, particularly of the Indica type, had higher H and F values than the varieties from other provinces of China.

The Aus group which was all Indicas had relatively low values of H and F, suggesting that the varieties are not very variable in isozyme. They are early varieties which are planted in spring and mature in late summer, while other varieties from South Asia excluding the Aus type contain so-called Aman (early summer to late fall) and Boro (winter) rice.

Table 1.  Average gene diversity (H) and genotype frequency (F; genotype no.
/cultivar no.)  estimated for each regional variety group divided into Indica
(1) and Japonica (J) types
_______________________________________________________________________________
          Yunnan      China excluding      Southeast       South Asia
                         Yunnan              Asia    __________________________
Item                                                    Aus    Exc.   Aus
          I    J         I    J            I      J      I     I       J
_______________________________________________________________________________
No. of    34   81        55   71           35     48     56    61      23
varieties
  H       .246 .141      .166 .158         .301   .353   .147  .339    .310
  F       .618 .284      .364 .324         .857   .750   .250  .754    .652
_______________________________________________________________________________
Further, the association pattern of isozyme alleles was examined by the squared correlation coefficient of allelic frequencies (Sano and Morishima 1992). Almost all loci combinations showed high correlation coefficients exceeding the significance level, indicating non-random association of alleles. However, Est-2 had relatively low correlations, some being insignificant, and the Southeast Asian group generally showed relatively low correlations (Table 2).

The nonrandomness of allele association observed indicates the trend of rice varieties to be differentiated into the Indica and Japonica types. If the correlations were computed within each type, they would be quite low as was observed for diagnostic characters (Oka 1988, p. 143). The results of the present study


Table 2.  Squared correlation coefficient (R2) for nonrandomness of allelic
association between isozyme loci
_______________________________________________________________________________
Locus                  Amp-2  Cat-1  Mal-1        Acp-1  Acp-2  Est-2  Est-10

Chromosomal
location                 8      6      1            12     12     6      1
_______________________________________________________________________________
Yunnan province (115)                R2=.429

Cat-1                  .45**

Mal-1                  .58**  .49**

Acp-1                  .66**  .44**  .46**

Acp-2                  .84**  .55**  .55**        .80**

Est-2                  .02    .01    .01          .03    .04*

Est-10                 .58**  .36**  .41**        .66**  .61**  .09*
_______________________________________________________________________________
China   (126)                        R2=.504

Cat-1                  .61**

Mal-1                  .38**  .57**

Acp-1                  .64**  .84**  .49**

Acp-2                  .51**  .79**  .48**        .87**

Est-2                  .39**  .50**  .25**        .63**  .51**

Est-10                 .53**  .70**  .45**        .76**  .67**  .66**
_______________________________________________________________________________
Southeast Asia (83)                R2=.157

Cat-1                  .33**

Mal-1                  .43**  .39**

Acp-1                  .08*   .08*   .18**

Acp-2                  .07*   .07*   .20**        .47**

Est-2                  .04    .02    .20**        .02    .03

Est-10                 .08*   .03**  .37**        .13**  .13**  .15**
_______________________________________________________________________________
Aus rice (56)                     R2=.390

Cat-1                  .43**

Mal-1                  .32**  .32**

Acp-1                  .43**  .43**  .32*

Acp-2                  .65**  .65**  .51**        .65**

Est-2                  .10**  .10**  .32**        .10*   .15**

Est-10                 .18**  .18**  .89**        .18**  .37*   .18**
_______________________________________________________________________________
South Asia (except Aus rice)   (83)      R2=.256

Cat-1                  .05 *
Mal-1                  .30**  .17**
Acp-1                  .34**  .10**  .22**
Acp-2                  .21**  .09**  .28**        .46**
Est-2                  .19**  .16**  .41**        .12**  .20**
Est-10                 .36**  .15**  .40**        .16**  .31**  .49**
_______________________________________________________________________________
An unknown locus of Amp was investigated in addition to the 7 loci, and the results were included in computing R2.

suggest that Yunnan rice varieties are not particularly diverse in isozymes even if they are highly diverse in key characters.

References

Nei, M. 1975. Molecular Population Genetics and Evolution. North-Holland Publ., Amsterdam. 288 pp.

Oka, H. I., 1988. Origin of Cultivated Rice. Elsevier/Japan Sci. Soc. Press, Amsterdam/Tokyo. 254 pp.

Sano, R. and H. Morishima, 1992. Indica-Japonica differentiation of rice cultivars viewed from variations in key characters and isozymes, with special reference to landraces from the Himalayan hilly areas. Theor. Appl. Genet. 84. 266-274.