II.50 Relation between yield and protein content of barley.

Tomohiko Yoshida. Kyushu Agr. Exp. Sta., Fukuoka, 833 Japan.

In this study, genetic correlations between yield (grain weight per plant) and protein content of barley were computed using F₁, F₂ and backcross generations. Plants were harvested individually (excluding borders) and grain weight and protein content were measured on a per plant basis. Grain weight and protein content of the parents are given in Table 1. The number of replication and the average number of plants harvested per replication are presented in Table 2. Mean values of grain weight and protein content of F₁, F2 and backcross generations (B₁, B₂) are given in Table 3.

Table 1. Parental values

Table 2. Materials used

Table 3. Mean values of F1, F2, and backcross generations

Variances and covariances of yield and protein content were computed for each replication and averaged. Covariance between two variables P and Q was computed as:

COV (P,Q) = COV (P_G,Q_G) + COV (P_E,Q_E)
where, COV (P_G,Q_G) is total genetic covariance computed as:
COV (P_G,Q_G) = COV (P_G,Q_G)_A, + COV (P_G,Q_G)_N
where, COV (P_G,Q_G)_A is additive genetic covariance, COV (P_G,Q_G)_N is nonadditive genetic covariance and COV (P_E,Q_E) is environmental covariance.

Like Mather's D,H method, COV (P_G,Q_G) is estimated by
COV (F₂) - 1/3 (COV (P₁) + COV (P₂) + COV (F₁))
and COV (P_G,Q_G)_A is estimated by
2COV (F₂) - COV (B₁) - COV (B₂).

In Table 4, covariances are listed.

From the values mentioned above, broad and narrow sense heritability values (h²_B and h²_A, respectively) were computed for each trait. Also, phenotypic correlation (rp), total genetic correlation (r_T) and additive genetic correlation (r_A) were computed. The values are presented in Table 5. Both broad and narrow sense estimates of heritability of yield was relatively high for all crosses. In addition, broad sense heritability of protein content was high for all crosses but the narrow sense values were not consistent (Table 5). In the F2 population, the phenotypic correlation between yield and protein content was negative for all crosses. Also total genetic correlation was negative. Thus it appears that the negative correlation between yield and protein content is genetic. The additive genetic correlation was not always negative. In the cross Saikai-10 x Hiproly, r_A = -0.569 but in the cross Saikai-l x Hiproly, r_A = 0 370. In the cross Saikai-12 x Hiproly, r_A could not be estimated but the additive genetic covariance was positive (4.1473). Non-additive genetic covariance was negative for all crosses.

Table 4. Covariance values
Table 5. Heritability and correlation coefficient

The results of this study indicate that a phenotypic correlation between yield (grain weight per plant) and grain protein content of barley was negative. The negative phenotypic correlation was genetic. And the negative genetic correlation was due primarily to non-additive genetic interaction.

BGN 6 toc
BGN Main Index