Rice Science, 2004, 11(4): 165 170 165 http://www.ricescience.org Grain Quality and Genetic Analysis of Hybrids Derived from Different Ecological Types in Japonica Rice (Oryza sativa) LENG Yan, HONG De-lin (State Key Laboratory of Crop Genetics and Germplasm Enhancement, Nanjing Agricultural University, Nanjing 210095, China) Abstract: The performance and inheritance of 7 quality traits were studied using F 2 rice grain derived from 8 8 diallel crossing made by employing 8 parents of different ecological japonica rice types. Differences in each trait among 8 parents were not obvious, but in F 2 generation, transgressive phenomena were found in all the traits studied, indicating that the genes controlling these traits among parents were segregated. The inheritance of grain width, grain weight, chalkiness score (CS), gelatinization (GT) and gel consistency (GC) were suitable to additive-dominant model, and dominant effect contributed mainly for the 5 traits. The inheritance of grain length (GL) and amylose content (AC) did not fit into additive-dominant model, existing epistatic interactions. Dominant genes for grain width and grain weight had the efficiency of decreasing effect, and dominant genes for CS, GT and GC had the efficiency of enhancing effect. Koshihikari contained more recessive genes for gelatinization than other varieties. Zhendao 88 had more dominant genes in grain width and grain weight than other varieties. Xiushui 04 possessed more dominant genes for GL and GC, and more recessive genes for CS than other varieties. Key words: hybrid rice; grain quality; inheritance; ecological type In 1975, Yang bred the first restorer line for CMS in japonica rice (Oryza sativa L.) by introducing restorer gene from indica rice IR8 into japonica varieties through triple crosses. By utilizing the restorer line, a series of heterotic japonica hybrid rice were bred and put into use in commercialization [1]. However, because the restorer line bred by crossing indica and japonica varieties contained much genetic component of indica, the F 1 hybrids expressed the heterosis for biomass, and the decreased economic index [2], as well as the declined brown rice, milled rice and head rice rates in comparison with the inbred japonica varieties [3]. Meanwhile, owing to unceasing improvement of inbred japonica varieties, the yield potential of inbred varieties was near to that of hybrids and the grain quality was superior to hybrids. This situation gave a challenge to japonica hybrid rice breeding. Therefore, Hong et al. proposed a japonica hybrid rice breeding strategy called breeding CMS and R lines by backcrossing, screening heterotic hybrids by crossing different ecological types [2]. This means that using male sterile cytoplasm as indicator, the inbred varieties or elite lines bred in various areas were transferred to CMS and restorer lines by backcross method, then crossing parents from different ecological types for heterotic hybrids. In the process of exercising this strategy, it was found that the F 1 hybrids derived from different ecological types had shown a yield advantage on an average of 20% over the better parent, Received: 20 December 2003; Accepted: 25 March 2004 Corresponding author: HONG De-lin (nndh@jlonline.com) and the integrative agronomic traits were also ideal [4]. This paper reported the performance of quality traits of grains in F 1 plants derived from different ecological types and further the inheritance of quality traits. Materials MATERIALS AND METHODS Rough rice grains harvested from plants of 8 parents and 56 F 1 hybrids obtained from 8 8 diallel crossing were used. The parents were Koshihikari, Akenohoshi, 3726, Liuqianxin, Zhendao 88, Wuyujing 3, Bing 8979 and Xiushui 04. Koshihikari and Akenohoshi were bred in Japan and belonged to medium season rice with earlymaturity ecological type in Nanjing. Varieties 3726 and Zhendao 88 were developed in Siyang Cotton Basic Seed Farm and Zhenjiang Institute of Agricultural Sciences in Jiangsu Province, respectively and came of medium season rice with medium-maturity ecological type in Nanjing. Liuqianxin, Wuyujing 3 and Bing 8979, which were bred in Jiangsu Academy of Agricultural Sciences, Wujin Rice and Wheat Basic Seed Farm, Jiangsu Province, and Jiaxing Institute of Agricultural Sciences in Zhejiang Province respectively, belonged to medium season rice with late-maturity ecological type in Nanjing. Xiushui 04 was bred in Zhejiang Province and belonged to late-maturity ecological type in Nanjing. Methods After drying under the sun and storing for some time at room, rough rice was processed to brown
166 Rice Science, Vol. 11, No. 4, 2004 rice and milled rice using a machine KETT made in Japan. Grain length and grain width of head rice was measured using vernier caliper with the precision of 0.02 mm. Grain weight was measured using LA114 type analytical balance (0.0001 g). For each trait of every material, 30 grains were measured in two replicates. Chalkiness score and gelatinization (alkali spread value, ASV) were measured and graded according to the method as published by the Ministry of Agriculture, Husbandry, and Fishery, P. R. China [5]. Gel consistency was measured and classified according to Tang et al [6]. Amylose content was determined according to Li et al [7]. Data were analyzed with Excel software. Although the length and width of glume were conditioned by the maternal genotype, the length and width of head rice were associated with the extent of starch filling. Therefore, all the grain quality traits studied in this paper were considered as 3N endosperm traits, and the genetic parameters were analyzed according to the diallel model for 3N endosperm traits proposed by Mo [8]. RESULTS Performance of grain quality traits of the eight parents From Table 1 it is evident that the measurement values in the 7 quality traits did not show great variations among the 8 parents. Range of grain length, grain width and grain weight was 4.52 to 5.13 mm, 2.36 to 2.95 mm and 1.57 to 2.30 mg, respectively. Range of chalkiness score was 0.30% to 3.33%, belonging to grade 1 to 2. Range of GT-ASV was 4.00 to 6.93 (grade), belonging to medium or low gelatinization. Range of gel consistency was 31.2 to 78.3 mm, belonging to soft or medium grade. Amylose content varied from 17.0% to 19.9%, falling to low amylose content. The results indicated that the performance differences of the 7 quality traits were not obvious among the parents, which were the popular varieties in commercialization. Averaged performance of F 2 grain quality in the 56 combinations It can be noticed from Table 2 that the range of the quality traits in F 2 was wider than that in parents. Range of grain length was 4.63 to 5.40 mm, with 21 combinations having grains longer than long-grain parent and 7 combinations having grains shorter than short-grain parent. Range of grain width was 2.45 to 3.14 mm, with 16 combinations having grains wider than wide-grain parent and 7 combinations having grains narrower than narrow-grain parent. Range of grain weight was 1.64 to 2.54 mg, with 27 combinations having grain weight heavier than heavy-grain parent, and 8 combinations having grain weight lighter than lightgrain parent. Range of chalkiness score (CS) was 0.03% to 8.17%, falling into grade 1 to 3, with 23 combinations having CS lower than low CS parent and 13 combinations having CS higher than high CS parent. Range of alkali spread value (ASV), standing for gelatinization, was 2.83 to 6.93, belonging to high, medium and low grade, with 10 combinations having grain ASV higher than high ASV parent and 14 combinations having ASV lower than the low ASV parent. Range of gel consistency (GC) was 2.59 to 8.75 cm, belonging to soft and medium grades, with 11 combinations having GC softer than the soft GC parent, and 14 combinations having GC harder than the hard GC parent. Range of amylose content (AC) was 12.6% to 18.3%, belonging to low AC grade, with 50 combinations having AC lower than low AC parent. The transgressive phenomena in F 2 indicated that the genes controlling the traits among parents were segregated. It is necessary to conduct genetic analysis further. Table 1. Averaged performance of grain quality traits in the eight parents. Parent Grain length Grain width Grain weight (mg) Chalkiness score (ASV, grade) Gel consistency (cm) Koshihikari 4.70 2.78 1.92 1.47 5.20 7.60 17.4 Akenohoshi 4.52 2.36 1.57 0.30 4.00 7.83 18.3 3726 5.01 2.70 1.97 1.77 6.93 5.71 17.6 Liuqianxin 5.17 2.89 2.15 0.80 6.47 3.91 17.0 Zhendao 88 4.91 2.74 2.16 1.87 4.67 4.65 17.8 Wuyujing 3 4.77 2.92 2.30 2.93 5.40 3.12 18.6 Bing 8979 5.13 2.95 2.27 1.67 5.40 3.80 19.9 Xiushui 04 4.75 2.91 1.99 3.33 6.11 5.31 17.6 ASV, Alkali spread value. Amylose content
LENG Yan, et al. Grain Quality and Genetic Analysis of Hybrids Derived from Different Ecological Types in japonica Rice 167 Table 2. Mean of the quality characters of F 2 rice grain in 56 hybrids. Cross Grain length Grain width Grain weight (mg) Chalkiness score (ASV, grade) Gel consistency (cm) Amylose content 1 2 4.94 2.76 2.12 1.33 3.10 5.65 15.1 1 3 4.65 2.54 1.80 1.03 6.33 4.84 13.3 1 4 4.56 2.80 1.71 1.93 6.30 4.83 14.0 1 5 4.78 2.75 2.03 0.53 4.45 5.33 16.2 1 6 4.63 2.80 1.97 1.00 7.57 5.75 15.5 1 7 4.88 2.82 2.10 1.00 6.17 8.40 16.8 1 8 4.93 2.89 2.26 1.10 2.83 5.52 17.4 2 1 4.95 2.75 2.03 8.17 3.53 8.71 13.8 2 3 5.14 2.66 2.02 0.23 3.53 7.33 13.0 2 4 4.95 2.64 1.85 0.55 4.00 8.09 17.4 2 5 4.52 2.60 1.82 3.17 4.53 8.21 16.0 2 6 4.57 2.45 1.61 1.53 4.17 8.75 17.0 2 7 4.69 2.60 1.82 1.00 4.60 7.06 16.7 2 8 5.04 2.89 2.21 1.23 4.23 6.32 15.4 3 1 4.85 2.61 1.82 0.93 6.60 5.30 14.5 3 2 5.15 2.75 2.11 0.37 3.73 5.37 13.9 3 4 5.14 2.85 2.23 1.23 6.50 5.59 16.4 3 5 5.40 2.85 2.31 2.63 5.97 5.23 14.0 3 6 4.93 2.96 2.34 2.00 5.73 5.78 16.1 3 7 5.05 2.78 2.08 2.40 6.03 5.44 13.4 3 8 4.86 2.82 2.02 3.57 6.93 4.16 17.2 4 1 4.84 2.95 2.27 1.77 5.57 5.01 15.0 4 2 4.59 2.61 1.67 0.30 3.00 5.94 15.0 4 3 5.36 2.82 2.25 2.37 6.03 4.42 16.3 4 5 4.67 2.59 1.74 0.03 5.73 4.55 15.4 4 6 4.87 2.93 2.30 2.73 6.07 4.31 15.4 4 7 5.25 2.87 2.28 1.37 6.00 4.93 14.4 4 8 5.12 2.80 2.02 3.70 6.63 4.49 14.2 5 1 4.93 2.81 2.10 1.50 4.67 4.64 15.4 5 2 4.92 2.75 2.10 1.60 4.47 4.60 16.6 5 3 5.24 2.82 2.30 1.23 5.77 4.21 17.7 5 4 5.07 2.76 2.16 0.47 5.67 4.22 15.5 5 6 4.81 2.75 2.13 1.97 3.73 4.86 17.1 5 7 4.86 2.64 1.98 0.30 5.17 4.27 12.7 5 8 4.73 2.83 2.00 7.03 6.17 5.08 17.3 6 1 4.73 2.87 2.02 1.73 5.00 6.88 13.3 6 2 4.94 2.74 2.16 0.17 6.10 4.71 13.0 6 3 4.90 2.95 2.33 2.80 6.20 3.61 15.6 6 4 5.30 3.08 2.48 3.71 5.07 2.93 14.4 6 5 4.88 2.74 2.12 1.03 5.33 2.59 13.9 6 7 4.82 2.92 2.26 2.07 6.37 3.22 17.5 6 8 4.82 3.14 2.50 1.57 6.67 3.01 17.0 7 1 5.00 2.88 2.20 0.90 5.00 4.18 12.6 7 2 4.94 2.82 2.13 0.87 4.83 4.60 14.9 7 3 5.17 2.88 2.35 2.70 5.42 4.63 14.6 7 4 5.19 2.84 2.31 1.63 5.00 4.51 17.2 7 5 5.01 2.77 2.26 1.97 3.77 4.26 18.7 7 6 5.02 2.91 2.29 0.63 5.07 3.93 17.8 7 8 4.93 2.94 2.41 0.47 6.67 4.55 18.3 8 1 4.92 2.88 2.19 0.40 5.13 4.39 15.1 8 2 5.20 2.81 2.19 1.13 4.77 4.97 16.2 8 3 4.99 2.94 2.44 0.70 6.30 4.79 15.4 8 4 4.81 2.93 2.28 2.97 4.80 4.32 16.4 8 5 5.13 2.93 2.24 2.90 6.00 4.74 16.9 8 6 4.96 3.09 2.54 1.03 6.67 4.74 17.4 8 7 4.98 3.03 2.46 0.87 6.50 4.81 17.4 1, Koshihikari; 2, Akenohoshi; 3, 3726; 4, Liuqianxin; 5, Zhendao 88; 6, Wuyujing 3; 7, Bing 8979; 8, Xiushui 04. The same as in Table 4.
168 Rice Science, Vol. 11, No. 4, 2004 Table 3. ANOVA on grain length, width, weight, chalkness score, gelatinization, gel consistency, amylose content. ANOVA df Grain length Grain width Grain weight Chalkiness score Gel consistency Amylose content MS F MS F MS F MS F MS F MS F MS F Between replications 1 0.02 0.03 0.01 0.17 0.06 0.36 0.07 Between combinations 63 5.26 26.77 ** 2.51 13.07 ** 60.50 28.81 ** 325.50 7.00 ** 148.20 32.93 ** 242.49 25.77 ** 4.60 6.94 ** Error 63 0.20 0.21 0.21 46.52 4.54 9.41 60.07 ** Significant at 1% level. Genetic analyses Analysis of variance The results of analysis of variance showed the significant differences (α=0.01) existed in all the 7 grain quality traits between combinations (Table 3), indicating there was real genetic variation among them. The genetic analysis could be conducted for the 7 quality traits. Genetic analysis on 3N endosperm traits Embryo and endosperm of seed borne on F 1 plant represent F 2 generation. The variance (Vr) of parent array and covariance (Wr) between parents and progeny obtained from the analysis of Vr-Wr for the 3N endosperm traits have been shown in Table 4. Calculated from Table 4, the regression coefficients of Wr to Vr for the 3N endosperm traits are listed in Table 5. It can be observed from Table 5 that, for the 5 traits of grain width, grain weight, chalkiness score, gelatinization (ASV) and gel consistency, the regression coefficients were significantly different from 0 but not from 1, indicating that the gene action fitted into additive-dominant model. For grain length, the regression coefficient was significantly different from both 0 and 1. For amylose content, the regression coefficient was not remarkably different from 0 but significantly different from 1. These results indicated that the gene action for grain length and amylose content was not suitable well to the additive-dominant model. For the 5 traits fitted into additive-dominant model, genetic parameters were estimated according to the equations deduced by Mo [8] and listed in Table 6. Table 6 demonstrated that the dominant effects were larger than additive effects for the traits of grain width, grain weight, chalkiness score, gelatinization and gel consistency. Relative coefficients between Wr+Vr and means of parents were positive for the grain width and grain weight, implying that dominant alleles for grain width and grain weight had negative effect on Table 4. Variance of parent array and covariance between parent and offspring. Parent Grain length Grain width Grain weight Chalkiness score Gel consistency Amylose content array Vr Wr Vr Wr Vr Wr Vr Wr Vr Wr Vr Wr Vr Wr 1 0.0136-0.0048 0.0099 0.0101 0.0154 0.0019 1.7179 1.4537 1.0749 0.7350 0.8178 0.7089 1.4680 0.2479 2 0.0438 0.0113 0.0205 0.0206 0.0403 0.0160 2.5814 0.2184 0.2848-0.1364 0.5247 0.7227 1.0500-0.1623 3 0.0379 0.0235 0.0142 0.0122 0.0282 0.0216 0.7283 0.7421 1.0130 0.7186 0.3295 0.7453 1.3610-0.6686 4 0.0552 0.0404 0.0146 0.0182 0.0439 0.0378 1.6460 2.1441 0.7824 0.6392 1.0582 1.4572 0.3020-0.0060 5 0.0177 0.0163 0.0057 0.0048 0.0097 0.0073 2.0600 2.5900 0.4359 0.5505 0.6398 1.1702 0.5980-0.3000 6 0.0014 0.0160 0.0226 0.0237 0.0375 0.0284 0.7530-0.1775 0.8523 0.7383 1.8316 2.1060 1.2300 2.3800 7 0.0178 0.0283 0.0102 0.0156 0.0196 0.0210 0.3429-0.4431 0.4207 0.4251 0.8870 1.3462 2.0100 0.3650 8 0.0103-0.0037 0.0055 0.0081 0.0271 0.0171 5.6419 3.8810 1.0197 0.5551 0.2788 0.7157 0.9090 0.5500 Mean 0.0247 0.0300 0.0129 0.0142 0.0278 0.0189 1.9340 0.9376 0.7355 0.5282 0.7959 0.1211 1.1160 0.3278 Table 5. Regression parameters of Wr to Vr and t test of regression coefficient for the seven 3N traits. Trait b s b H 0 : β=0, t value H 0 : β=1, t value Grain length 0.4842 0.0311 15.9582 16.1961 Grain width 0.9539 0.1560 5.9768 0.2888 Grain weight 0.7236 0.1915 3.7786 1.4433 Chalkiness score 0.7401 0.2915 2.5389 0.8916 0.7135 0.2436 2.9291 1.1760 Gel consistency 0.9260 0.1717 5.3927 0.4310 Amylose content 0.2929 0.1977 1.4815 3.5766 *, ** significant at 5% and 1% levels, respectively.
LENG Yan, et al. Grain Quality and Genetic Analysis of Hybrids Derived from Different Ecological Types in japonica Rice 169 Table 6. Estimates of genetic parameters for grain width, grain weight, chalkiness score, gelatinization and gel consistency. Genetic parameter Grain width Grain weight Chalkiness score Gel consistency D 0.0166 0.0247 0.4439 0.3984 1.1190 H 0.5154 1.4603 94.5400 22.3296 4.9749 H 0.8783 1.9264 135.8480 23.3286 15.5968 F 0.0481 0.0953 4.6747-0.8299 1.5211 r 0.5966 0.8672-0.7339-0.1153-0.5609 Table 7. Analysis of variance of Wr (i) +Vr (i) and Wr (i) -Vr (i) for grain length and amylose content. Trait Source df Wr (i) +Vr (i) Wr (i) -Vr (i) MS F MS F Grain length Between common parents 7 0.0017572 60.0 ** 0.0003686 10.2 ** Within common parents 8 0.0000293 0.0000360 Amylose content Between common parents 7 0.000112 3.2 NS 0.0000883 6.1 * Within common parents 8 0.000035 0.0000144 *, ** Significant at 5% and 1% levels, respectively; NS Not significant. efficiency. Relative coefficients between Wr+Vr and means of parents were negative for the 3 traits of chalkiness score, gelatinization and gel consistency, indicating that the dominant alleles for chalkiness score, gelatinization and gel consistency acted as efficiency enhancing genes. The F value for gelatinization was less than 0, suggesting that recessive alleles frequency carried by the parents was higher than dominant alleles. For the other 4 traits, the F values were larger than 0, suggesting that dominant alleles frequency carried by the parents was higher than that of recessive alleles. From Fig. 1 it can be viewed that for the trait of grain width, parents 5 (Zhendao 88) and 8 (Xiushui 04) carried more dominant alleles, and parents 6(Wuyujing 3) and 2 (Akenohoshi) bore more recessive alleles (Fig. 1- A). For the trait of grain weight, parents 1 (Koshihikari) and 5 (Zhendao 88) carried more dominant alleles, and parents 4 (Liuqianxin) and 6 (Wuyujing 3) had more recessive alleles (Fig. 1-B). For the trait of chalkiness score, parents 3 (3726), 6 (Wuyujing 3) and 7(Bing 8979) carried more dominant alleles, and parents 8 (Xiushui 04) possessed more recessive alleles (Fig. 1-C). For the trait of gelatinization, parents 2 (Akenohoshi) carried more dominant alleles, and parents 1 (Koshihikari) and 3 (3726) carried more recessive alleles (Fig. 1-D). For the trait of gel consistency, parents 8 (Xiushui 04), 3 (3726) and 2 (Akenohoshi) carried more dominant alleles, while parent 6 (Wuyujing 3) had more recessive alleles (Fig. 1-E). For the two traits not fitting into additive-dominant model, both dominant and epistatic effects were found in grain length, and only epistatic effects were found in amylose content according to the variance analysis of Wr (i) +Vr (i) and Wr (i) -Vr (i) (Table 7). Fig. 1. Regression of Wr to Vr for the five 3N traits fitted to additivedominant model.
170 Rice Science, Vol. 11, No. 4, 2004 DISCUSSION In this study, we found that the genes conditioning the traits of grain quality were scattered among the parents with the transgressive phenomena appearing in F 2 generation, though the appearance difference was not obvious in the parents used. This suggested that it is possible to breed F 1 hybrid with high yield potential and superior grain quality by crossing different ecological type of japonica rice. Considering Table 6 and Fig. 1, we can expect that if Zhendao 88 or Xiushui 04 is used as one of the parents, the progeny will have narrow grains; and the progeny will have heavier grains if Koshihikari or Zhendao 88 is used as one of the parents. Similarly, if 3726 or Wuyujing 3 and Bing 8979 is utilized as one of the parents, the progeny will have grains with low chalkiness score; if Akenohoshi is used as one of the parents, the progeny will have grains with low gelatinization ; and the progeny will have grains with long gel consistency if 3726 or Xiushui 04 is employed as one of the parents. ACKNOWLEDGMENTS We would like to express our sincere thanks to Miss XIAO Fangjing and Mr. RONG Hua for the measurement of partial laboratory data. REFERENCES 1 Yang Z Y. Progresses in the breeding of japonica hybrid rice. Hybrid Rice, 1994, (3 4): 46 49. (in Chinese) 2 Hong D L, Pan E F, Chen C Q. Comparative studies on harvest index between hybrids and pure lines in japonica rice (Oryza sativa L.). J Nanjing Agric Univ, 1998, 21(4): 12 18. (in Chinese with English abstract) 3 Hong D L. Fertility restoration ability and offspring economic characters of BT type iso-cytoplasmic restorers in japonica rice (Oryza sativa L.). Jiangsu Agric Sci, 1998, (5): 2 7. (in Chinese) 4 Hong D L, Yang K Q, Pan E F. Heterosis of F 1 s derived from different ecological types and combining ability of their parents in japonica rice (Oryza sativa L.). Chinese J Rice Sci, 2002, 16(3): 216 220. (in Chinese with English abstract) 5 The Standard issued by the Ministry of Agriculture, Husbandry and Fishery of Republic of China. Measurement for Rice Grain. NY147 88. (in Chinese) 6 Tang S X, Khush G S. Modified single grain analysis for gel consistency. Chinese J Rice Sci, 1990, 4(2): 55. (in Chinese) 7 Li R, Huang C W. Single-kernel measurement for amylose content of rice. Guangdong Agric Sci, 1988, (5): 7 9. (in Chinese) 8 Mo H D. Genetic analysis for endosperm traits in diallel designs. J Jiangsu Agric Coll, 1988, 9(3): 1 10. (in Chinese with English abstract)