[home in japanese＞home＞text]
This page is original manuscript of H.Y.Lu ed., "Proceeding of 3rd Asian Crop Science Conference, Taichun, Taiwan. 404-412, 1998".
Pedigree Analysis of Rice, Barley and Wheat
Tomohiko Yoshida
ABSTRACT
A computer program for pedigree analysis of self-pollinating crops was written in Prolog, a programming language with logical operations. For rice cultivars, coefficient of parentage to Koshihikari was computed. Significant positive correlations between the coefficient of parentage and eating quality were found among both of check cultivars and breeding lines, suggesting that the combining ability of eating quality of rice could be estimated by computing the coefficients of parentage. For malting barley, Harunanijo showed average coefficients of parentage as high as 0.561 with breeding lines. A significant positive correlation (0.421) between the coefficient of parentage and malting quality was found. For wheat, cultivars which related to Asakazekomugi had low value of flour whiteness, showing that Asakazekomugi had a poor combining ability in quality, though it was a high yield cultivar and used as a cross parent extensively. Wheat cultivars related to Kanto 107 had a high maximum viscosity value. A few ancestors constituted a great proportion of the genetic base of modern cultivars of rice, malting barley and wheat.
Key words:Barley, Coefficient of parentage, Pedigree analysis, Rice, Wheat.
INTRODUCTION
"Pedigree analysis" is to study about breeding records of crop cultivars, from which valuable informations could be obtained. But a pedigree record of recently released cultivars, which are developed after many crossings among promising parents, is very complicated and it is almost impossible to tell a simple story from the records without proper numerical analysis. In this paper, 1) Number of ancestors contained in a pedigree, 2) Contribution of ancestors to the gene pool, and 3) The relationship between the performance of cultivars and the kinship to a specific genotype are studied among newly released cultivars of rice, malting barley and wheat. Coefficients of parentage were computed to show the kinship between two genotypes. Coefficient of parentage is defined as the probability that a random gene from X is identical by descent with a random gene from Y, considering two individuals, X and Y (Kempthorne 1969).
MATERIALS AND METHODS
A computer program for calculating coefficients of parentage was written in Prolog, a programming language with logical operations for artificial intelligence and data retrieval applications (Mizuta et al. 1996). For rice and malting barley, the pedigrees of several check cultivars and breeding lines developed by Fukuoka Agricultural Research Center were studied. For wheat, check cultivars and breeding lines developed by Kyushu National Agricultural Experiment Station were studied. In addition of computing coefficients of parentage, the program enabled to count the maximum generation traced in the pedigree, the total number of ancestors and the total number of ancestors except common ones. The coefficient of parentage of a cultivar which has no ancestors is the genetical contribution by the cultivar. Contribution by ancestors to the gene pool was also estimated for these three crops. As a performance of cultivars, data of eating quality of rice, malting quality of barley and flour quality of wheat were used. Eating quality of rice was scored as a deviation from Koshihikari by a sensory evaluation test of Fukuoka Agric. Res. Center. The value of malting quality of barley was the deviation from Amaginijo in the value of malt extract measured by Tochigi Agric Exp. Station. Wheat flour quality values were deviation from Norin 61 and were kindly supplied by the Kyushu Regional Joint Wheat Quality Test. All materials were grown in Fukuoka Agric. Res. Center and the quality values were the mean of several replications.
RESULTS AND DISCUSSION
In Table 1, the maximum generation traced in the pedigree, total number of ancestors in the pedigree and total number of ancestors except common ones are shown for several cultivars of rice, malting barley and wheat. For Nipponbare of rice, the numbers are 9, 72 and 40, respectively. For Chikushi 27 of rice, the pedigree was more complicated than Nipponbare and it must be traced to the oldest ancestor up to 17 generations. It had more than 1000 ancestors in the pedigree though the number was reduced to 119 when common ones were excluded, showing that it was nearly impossible to find and trace all possible paths correctly for computing the coefficient of parentage between cultivars as Chikushi 27 with many ancestors.
In Table 2, the contribution to the gene pool by several ancestors are shown. Only a few ancestors contributed the gene pool. Five of rice, 3 of barley and 5 of wheat ancestors contribute 62.5, 71.8 and 59.4 % to the gene pool, respectively, revealing the necessity of widening the genetic background of the crops to overcome the "genetic vulnerability (Walsh 1981)".
In Fig. 1 and 2, the relationship between eating quality and coefficient of parentage to Koshihikari among several rice check cultivars and breeding lines are shown. The correlation coefficients were 0.746 and 0.583 and significant positive correlations were found. It shows that genotypes having close relationship to Koshihikari have good eating quality, or adversely, cultivars or breeding lines having no close relationship to Koshihikari have no promising as a good eating cultivar.
This relationship also shows that the promising crossings can be predicted by computing the coefficients of parentage of lines which will be developed by candidate crossings. If the lines have high value of the coefficient, it is expected to have good eating quality. The coefficients of parentage for candidate crosses among 10 breeding lines were computed and shown in Table 3. Crosses with * had value more than 0.5. Some lines had high combining ability and their offsprings generally had high values. On the other hand, some crossings result into low values, showing that we had better not include these cross combinations.
For malting barley, Harunanijo, is Koshihikari in malting barley. Harunanijo is a good malting quality cultivar and used extensively as a cross parent. A significant positive correlations (0.421) between the coefficient of parentage with Harunanijo and malting quality was found among local cultivars released during a few decades (Fig.3). It shows that we can estimate good quality candidate crosses by computing the coefficient of parentage as in the case of good eating rice.
For wheat, it seems that there is no leading variety used extensively as a cross parent, having high performance as a parent. Table 4 shows that cultivars related to Asakazekomugi had low value of flour whiteness, showing that Asakazekomugi had a poor combining ability in quality though it was a high yield cultivar and used as cross parent often. Cultivars which related to Kanto 107 had high maximum viscosity values. Cultivars related Siroganekomugi had high protein content, showing that Kanto 107 and Siroganekomugi might have good combining ability to flour quality.
In conclusion, with the help of a programming language with logical operations, several valuable informations were obtained by analyzing the pedigree of modern day's 3 grain crop cultivars. Rice cultivars related to Koshihikari have good eating quality. Barley cultivars related to Harunanijo have high malting quality. Cross combinations with high performance can be partially estimated in advance by computing the coefficients of parentage. It will give a way to the planning of more reasonable and theoretical breeding strategy for higher yield and quality and we can decide cross combination in advance with the help of a computer, not by the conventional instinct and experience.
REFERENCES
Kempthorne O. 1969. An Introduction to Genetic Statistics. Iowa State Univ. Press, U.S.A.
Mizuta,K., A.Sasaki and T.Yoshida 1996. Prolog computer program for evaluation of
coefficients of parentage and its application to the analysis of pedigree of malting
barley cultivars. Agricultural Information Research. 5:19-28. (In Japanese with English summary)
Walsh,J. 1981. Genetic vulnerability down on the farm. Science 214:161-164.
Table 1. Maximum generation traced in the pedigree, total
number of ancestors and number of ancestors except common ones.
| --------- | ---------- | ---------- | ---------- |
| Crop | Maximum | Number of | ancestors |
| | generation | ----- | ---- |
| Cultivar | in the | Total | Except |
| | pedigree | | common ones |
| --------- | ---------- | ---------- | ---------- |
| Rice |
| Nipponbare | 9 | 72 | 40 |
| Chikushi 19 | 13 | 560 | 86 |
| Chikushi 20 | 13 | 724 | 97 |
| Chikushi 27 | 17 | 1238 | 119 |
| Barley |
| Kyushu Nijo 1 | 5 | 26 | 15 |
| Nishino Gold | 8 | 64 | 24 |
| Asaka Gold | 7 | 76 | 37 |
| Kyushu Nijo 12 | 10 | 196 | 47 |
| Wheat |
| Saikai 160 | 8 | 78 | 42 |
| Saikai 172 | 9 | 84 | 40 |
| Saikai 177 | 9 | 112 | 45 |
| Saikai 179 | 9 | 138 | 66 |
| --------- | ---------- | ---------- | ---------- |
Table 2. Ancestors contributing to the gene pool.
| --------- | ---------- | ------------- | ------------- |
| Crop | Ancestor | Mean | Accumulated |
| No. | | coefficient | contribution |
| | of parentage | to gene pool(%) |
| --------- | ---------- | ------------- | ------------- |
| Rice |
| 1 | Aikoku | 0.163 | 16.3 |
| 2 | Asahi | 0.134 | 29.7 |
| 3 | Shinriki | 0.122 | 41.9 |
| 4 | Joshu | 0.110 | 52.9 |
| 5 | Ooba | 0.096 | 62.5 |
| Barley |
| 1 | Golden Melon | 0.417 | 41.7 |
| 2 | Chevallier | 0.159 | 57.6 |
| 3 | Sapporo 7 | 0.142 | 71.8 |
| Wheat
|
| 1 | Chunaga | 0.359 | 35.9 |
| 2 | Eshima | 0.081 | 44.0 |
| 3 | Hirakikomugi | 0.061 | 50.2 |
| 4 | Hayakomugi | 0.049 | 54.9 |
| 5 | Shinriki | 0.045 | 59.4 |
| --------- | ---------- | ------------- | ------------- |
Rice; for 17 cultivars developed in Fukuoka Agr.Exp.Stn.
Barley; for 51 malting cultivars developed in Fukuoka Agr.Exp.Stn.
Wheat; for 16 cultivars developed in Kyushu Natl. Agr.Exp.Stn.
Table 3. Coefficients of parentage between Koshihikari and expected lines which could be obtained by dialle crosses among 10 rice cultivars.
| ----- | ----- | ----- | ----- | ----- | ----- | ----- | ----- | ----- | ----- | ----- |
| Tikushi | 1 | 8 | 12 | 13 | 14 | 15 | 18 | 19 | 20 | 23 |
| ----- | ----- | ----- | ----- | ----- | ----- | ----- | ----- | ----- | ----- | ----- |
| 8 | 0.702* |
| 12 | 0.702* | 0.637* |
| 13 | 0.582* | 0.517* | 0.517* |
| 14 | 0.571* | 0.506* | 0.506* | 0.386 |
| 15 | 0.650* | 0.585* | 0.585* | 0.465 | 0.453 |
| 18 | 0.569* | 0.504* | 0.504* | 0.385 | 0.373 | 0.452 |
| 19 | 0.586* | 0.521* | 0.521* | 0.401 | 0.390 | 0.469 | 0.388 |
| 20 | 0.566* | 0.501* | 0.501* | 0.381 | 0.370 | 0.449 | 0.369 | 0.385 |
| 23 | 0.603* | 0.538* | 0.538* | 0.419 | 0.407 | 0.486 | 0.406 | 0.423 | 0.403 |
| 24 | 0.734* | 0.669* | 0.669* | 0.550* | 0.538* | 0.617* | 0.537* | 0.553* | 0.534* | 0.571* |
| ----- | ----- | ----- | ----- | ----- | ----- | ----- | ----- | ----- | ----- | ----- |
*; The value is larger than 0.5.
Table 4. Correlations between wheat quality and coefficient of parentage with the cultivar in the table among 10 cultivars.
| -------------- | --------- | -------- | ---------- | --------- |
| Cultivar | Protein | Flour | Flour | Maximum |
| | content | yield | whiteness | viscosity |
| -------------- | --------- | -------- | ---------- | --------- |
| Kanto 107 | -0.402 | 0.456 | 0.575 | 0.893** |
| Asakazekomugi | 0.184 | -0.375 | -0.662* | -0.838** |
| Shiroganekomugi | 0.709* | -0.451 | -0.179 | -0.527 |
| -------------- | --------- | -------- | ---------- | --------- |
*,**; Significant at 5,1% level, respectively.
start:-(point(Z)->retract(point(Z));true),assert(point(0)),d_clear,
write(hinsyu_1),read(X),nl,write(hinsyu_2),read(Y),kinkou(X,Y).
kinkou(X,Y):-not(o_srch(X,Y,1,[[],[X]])),not(o_srch(Y,X,1,[[],[Y]])),
not(p_srch(X,Y,1,[X])),nl,write('Coefficient of Relationship = '),
point(P),write(P),nl.
p_srch(X,Y,P,L):-(r(X,V,_);r(X,_,V)),float(*,P,'0.5',A),append(L,V,L1),
not(o_srch(V,Y,A,[L1,[V]])),not(p_srch(V,Y,A,L1)),fail.
o_srch(V,Y,P,[L1,L2]):-(r(Y,V,_);r(Y,_,V)),float(*,P,'0.5',A),append(L2,Y,L3),
retract(point(N)),float(+,N,A,A1),assert(point(A1)),!,fail.
o_srch(V,Y,P,[L1,L2]):-(r(C,V,_);r(C,_,V)),not(member(C,L1)),
float(*,P,'0.5',A),append(L2,C,L3),not(o_srch(C,Y,A,[L1,L3])),fail.
not(X):-X,!,fail.
not(_).
append([],X,[X]).
append([A|X],Y,[A|Z]):-append(X,Y,Z).
member(A,[A|L]).
member(A,[_|L]):-member(A,L).
r('cult_1','parent_1','parent_2'). /*database of cross parents*/
．
．
Fig. 1. Program in Prolog for calculating coefficient of parentage.
| kosihikari･ | ･norin22･･･ | ･norin8････ | ･aikoku (1) |
| | ･ | ･ | ･asahi |
| | ･ | ･norin6････ | ･joshu |
| | ･ | | ･kiryoyosi |
| | ･norin1････ | ･oba |
| | (2) | ･rikuu132･･ | ･aikoku |
| | | | ･kamenoo |
|
| norin1････ | ･oba |
| (2) | ･rikuu132･･ | ･aikoku (1) |
| | | ･kamenoo |
* >>> kosihikari norin22 norin8 aikoku
** aikoku rikuu132 norin1 <<< 0.03125 (1)
* >>> kosihikari norin1
** norin1 <<< 0.5 (2)
Coefficient of parentage = 0.53125
Fig. 2. Computing the coefficient of parentage between Koshihikari and Norin 1.
| asakazek･･ | ･hiyokuk･･･ | ･junreik･･･ | ･n52･･･････ | ･chunaga | (1,2) |
| | ･ | ･ | ･ | ･esimasinri | ･esima |
| | ･ | ･ | ･ | (3) | ･sinriki |
| | ･ | ･ | ･n26･･･････ | ･chunaga | (4,5) |
| | ･ | ･ | | ･saitama29 |
| | ･ | ･saikai95･･ | ･n36･･･････ | ･chunaga | (6,7) |
| | ･ | | ･ | ･esimasinri | ･esima |
| | ･ | | ･ | (8) | ･sinriki |
| | ･ | | ･n61･･･････ | ･fukuokakom | ･fukuokakom | ･hizakiri |
| | ･ | | | ･ | ･ | ･akabozu |
| | ･ | | | ･ | ･esima (9) |
| | ･ | | | ･chunaga | (10,11) |
| | ･siroganek･ | ･sirasagik･ | ･chunaga |
| (12) | ･ | ･n59･･･････ | ･n7････････ | ･goshu13 |
| | ･ | | ･ | ･sirochabo |
| | | ･ | | ･n5････････ | ･yushoki |
| | | ･ | | | ･hirosimasi |
| | | ･saikai104･ | ･hirakik |
| | | ･n20･･･････ | ･chunaga |
| | | | | ･esimasinri | ･esima |
| | | | | | ･sinriki |
| siroganek･ | ･sirasagik･ | ･chunaga | (1,4,6,10) |
| (12) | ･ | ･n59･･･････ | ･n7････････ | ･goshu13 |
| | ･ | | ･ | ･sirochabo |
| | ･ | | ･n5････････ | ･yushoki |
| | ･ | | | ･hirosimasi |
| | ･saikai104･ | ･hirakik |
| | | ･n20･･･････ | ･chunaga | (2,5,7,11) |
| | | | ･esimasinri | ･esima (9) |
| | | | (3,8) | ･sinriki |
* >>> asakazek hiyokuk junreik n52 chunaga
** chunaga sirasagik siroganek <<< 0.015625 (1)
* >>> asakazek hiyokuk junreik n52 chunaga
** chunaga n20 saikai104 siroganek <<< 0.0078125 (2)
* >>> asakazek hiyokuk junreik n52 esimasinriki
** esimasinriki n20 saikai104 siroganek <<< 0.0078125 (3)
* >>> asakazek hiyokuk junreik n26 chunaga
** chunaga sirasagik siroganek <<< 0.015625 (4)
* >>> asakazek hiyokuk junreik n26 chunaga
** chunaga n20 saikai104 siroganek <<< 0.0078125 (5)
* >>> asakazek hiyokuk saikai95 n36 chunaga
** chunaga sirasagik siroganek <<< 0.015625 (6)
* >>> asakazek hiyokuk saikai95 n36 chunaga
** chunaga n20 saikai104 siroganek <<< 0.0078125 (7)
* >>> asakazek hiyokuk saikai95 n36 esimasinriki
** esimasinriki n20 saikai104 siroganek <<< 0.0078125 (8)
* >>> asakazek hiyokuk saikai95 n61 fukuokakomugi18 esima
** esima esimasinriki n20 saikai104 siroganek <<<0.001953125 (9)
* >>> asakazek hiyokuk saikai95 n61 chunaga
** chunaga sirasagik siroganek <<< 0.015625 (10)
* >>> asakazek hiyokuk saikai95 n61 chunaga
** chunaga n20 saikai104 siroganek <<< 0.0078125 (11)
* >>> asakazek siroganek
** siroganek <<< 0.5 (12)
Coefficient of parentage = 0.611328125
Fig. 3. Computing the coefficient of parentage between
Asakazekomugi and Siroganekomugi
(End of text)