34. Genetic Correlation Analysis of Agronomic Traits in Rice

Chunhai Shi and Zongtan Shen
Agronomy Department, Zhejiang Agric. University, Hangzhou 310029, China

     Correlation analysis of agronomic traits of rice was conducted based on the additive and dominance genetic model (Zhu 1991; Zhu et al. 1993) for quantitative traits. Minque (0/1) method was used to estimate variance components. Total genetic variance (V G) can be partitioned into additive variance (VA) and dominance variance (VD) and so do covariance components (total genetic covariance (CG) = additive covariance (CA) + dominant covariance(CD))ofpairwisetraits.The correlation coefficients between trait X and Y were estimated by the formulas of phenotypic correlation genetic correlation , additive correlation, dominance correlation The Jackknife method was applied for obtaining estimates and their standard errors, and for the t-test parameters (Miller 1974).
    Incomplete diallel crosses from indica varieties of rice (Oryza sativa L.) were made using eight parents with short grain (Guangluai 4 and others) as females and five parents with long grain (Xiangzaoxian 3 and others) as males. In 1991, F1 seedlings and their parents were transplanted into the field at 15x15 cm spacing, with three replications. Seeds from 10 plants in the middle of plots were used to measure agronomic traits, including grain length (GL), grain width (GW), grain thickness (GT), ratio of grain length to width (L/W), panicle number per plant (PNP), density of grains on panicle (DGP), grains per panicle (GPP), seed fertility (SF), panicle weight (WP), grain weight (WG) and grain yield per plant (GYPP).
    Correlation analysis indicated that the phenotypic and genetic correlations were significant for most of pairwise traits studied especially for grain traits or yield traits (Table 1). The results of genetic correlation components analysis showed that additive

Table 1. Phenotypic, genetic, additive and dominance correlations of agronomic traits in crosses of indica varieties. Below diagonal: upper line - phenotypic correlation and lower line - genetic correlation; Above diagonal: upper line - additive correlation and lower line - dominance correlation
 

Traits GL GW GT L/W PNP DGP GPP SF WP WG GYPP
GL -0.86** -0.02 -0.46** -0.04 0.96** 0.59 -0.02 -0.24 0.35* -0.04 0.54** -0.20 -0.39** 0.27 0.15 0.04 -0.05 0.25* 0.09 -0.06
GW -0.79** -0.83** 0.75** 0.39 -0.97** -0.29* -0.01 -0.41 -0.17 -0.08 -0.346** 0.054 0.41* 0.39 0.13 0.16 0.48** 0.14 0.21* -0.13
GT -0.39** -0.43** 0.69** 0.73** -0.68** -0.07 -0.38* -0.04 0.35** -0.19 0.130 -0.106 0.78** -0.01 0.65* -0.07 0.89** -0.01 0.71** -0.13
L/W 0.93** 0.95** -0.95** -0.95** -0.60** -0.64** 0.11* -0.22* 0.23* -0.22 0.437** -0.455 -0.47** -0.07 -0.04 -0.16 -0.32** -0.47 -0.08 -0.27
PNP -0.04 -0.06 -0.04 -0.07 -0.10 -0.18* -0.02 -0.01 -0.84* 0.31 -0.890* 0.354 -0.56 -0.50 -0.89** 0.02 -0.41 0.35* -0.46 0.42
DGP 0.23** 0.29* -0.12 -0.15* 0.24 0.28 0.14 0.18* -0.14 -0.23 0.901** 0.903** 0.57** 0.01 0.86** 0.89** 0.50** 0.18 0.68** 0.95**
GPP 0.32** 0.37** -0.21* -0.25* 0.08 0.09 0.24* 0.28** -0.07 -0.15 0.89** 0.88** 0.41** -0.01 0.86** 0.75** 0.33** 0.14 0.65" 0.87*"
SF -0.20 -0.40 0.23* 0.45* 0.43* 0.68* -0.25* -0.48* -0.30 -0.55* 0.32* 0.57** 0.234 0.478 0.75** 0.33 0.66* 0.16 0.68* 0.09
WP 0.10 0.11 0.11 0.11 0.42* 0.47* -0.04 -0.04 -0.25** -0.33* 0.79** 0.85** 0.783** 0.817** 0.55** 0.74* 0.75** 0.25* 0.84** 0.88**
WG -0.01 -0.03 0.43** 0.45** 0.74** 0.81** -0.30** -0.32** -0.03 -0.05 0.37** 0.44** 0.245** 0.273** 0.40** 0.64* 0.55** 0.58** 0.79** 0.34*
GYPP 0.03 0.03 0.08 0.08 0.25 0.30 -0.06 -0.07 0.27** 0.23 0.62** 0.69** 0.678** 0.723** 0.35* 0.54* 0.75** 0.80** 0.42** 0.45**
    *, ** significant at 5% and 1% level, respectively.

correlations were larger than dominance correlations for most of the agronomic traits studied. Therefore, genetic correlations for pairwise traits were mainly controlled by additive effects. The additive correlations were significantly positive for some pairwise traits such as GL and L/W which showed that the plants with high L/W of grain could be indirectly obtained for selecting the plant with longer grain. Otherwise, the results which additive correlations were significantly negative among the pairwise traits such as GL and GW which showed that the indirect selection effect were also good and the variety with longer and narrower grain could be bred in rice breeding. Although significant dominance correlations were detected for the pairwise traits of GL and WG, PNP and WG, DGP and GPP, DGP and WP, DGP and GYPP, GPP and WP, GPP and GYPP, WP and WG, WP and GYPP, and WG and GYPP, significant negative dominance correlation was found only between GW and L/W. Indirect selection is applicable for traits with highly additive correlations, but the dominance correlations could be effectively used in hybrid rice breeding.

References

Miller, R. G., 1974. The jackknife, a review. Biometrika 61: 1-15. Shi, C. H. and 1. Zhu, 1993. Analysis of seed
        maternal genetic effects on milling quality characters in indica hybrid rice. RGN 10:110-111.
Zhu, J., 1992. Mixed model approaches for estimating genetic variances and covariances. Journal of
        Biomathematics 7(1): 1-11.
Zhu, J. et al., 1993. A genetic approach for anlayzing intercultivar heterosis in crops. Chinese Journal of
        Genetics 20(3): 183-191.