1-Plant breeding department , Faculty of Agriculture . Shahrekord University, Shahrekord, Iran.
Email ah_rezaei@yahoo.com or rezaei@agr.sku.ac.ir
2-Agronomy department , Faculty of Agriculture . Shahrekord University , Shahrekord, Iran.
Email v_roohi@yahoo.com
A complete set of diallel crosses among 10 corn inbred lines were used to determine the genetic control of yield and some of its components, at Research Farm, College of Agriculture, Shahrekord University, Iran, in 2002. A triple lattice design was used. Significant differences (P<0.01) were detected among genotypes for all the 10 traits studied. Genotypic variances were partitioned into additive and non-additive genetic components based on the method of Hayman. Both of these components were significant (P<0.05) for all traits. The highest and lowest degrees of dominance were observed for ear height and number of kernel rows, respectively. Broad-sense heritability estimates ranged from 0.91 for cob percentage to 0.68 for number of kernel rows. Narrow-sense heritability estimates ranged from 0.73 for days to tassel emergence to 0.06 for grain yield. The parental lines with the most dominance or recessive allelic frequencies were determined based on their distribution along the regression line of array covariances on array variances.
In a set of complete diallel crosses genotypic variances were partitioned to additive and non-additive components based on the Hayman method. Most of these components were significant for all traits.
Key Words
Corn, Diallel crosses, Heritability, Variance components, Heterosis
Genetic improvement of corn yield depends on knowledge of the size and nature of gene action and genetic control of related traits such as yield components. Also the choice of the most efficient breeding strategy depends on this information. Whereas dominance gene action would favor the production of hybrids, additive gene action indicates standard selection procedures such as mass selection would be effective in bringing about changes to the selected character. Different genetic analyses have been used to obtain information about genetic control of quantitative traits. Diallel cross analysis has been used most extensively (1,2,7). According to Hayman’s analysis, significant mean squares due to a component indicates presence of additive genetic variation in the materials under investigation. Significant b component indicates the presence of dominance component of variation. A significant b1 component shows directional dominance whereas a significant b2 component implies asymmetry in gene distribution and b3 indicates the presence of the residual non-additive variation which is identical to Griffing’s specific combining ability. The c and d components indicate variation due to general reciprocal differences and specific reciprocal differences, respectively, and in the absence of any cytoplasmic or maternal effects both provide estimate of environmental component of variation (6).
A complete set of diallel crosses among ten Iranian corn inbred lines were evaluated in a triple-lattice design at Research Farm, Shahrekord College of Agriculture, Shahrekod University, Shahrekord, Iran in 2002. A three-row plot, 2.5m long and 70cm wide, with a within-row spacing of 20cm was used. Fertilizer treatments were 200 kg/ha of ammonium phosphate applied prior to planting plus an additional 200 kg/ha N topdressed at thinning and tasselling. Ten traits including grain yield (kg/ha at 14% moisture), yield components, and others traits including days to tasselling, and plant and ear height were determined. Analysis of variance was performed based on a triple-lattice (no relative efficiency was found over randomized complete block design for all traits studied). Hayman (3,4) and Jinks and Hayman (5) analyses were performed to estimate genetic parameters. Also reciprocal effects were partitioned to maternal and non-maternal effects. Array variance and covariance, parental and progeny means, and the variance of progeny means were used to estimate different genetic parameters such as D (additive variance), H1 and H2 (dominant variance), F (covariance between additive and dominance effects), and broad and narrow sense heritability estimates.
The components of variance according to Hayman’s method are given in Table 1. Significant (P<0.01) differences were observed among genotypes for all the studied traits. Hayman analysis of variance indicated both additive (a) and dominance (b) effects to be highly significant. All traits showed directional dominance (b1) and dominance effects common to the progeny of a particular parent (b2). All traits except number of leaves and number of kernel rows showed dominance effects specific to particular crosses (b3). There were significant differences between maternal effects (c) for grain yield and number of kernel rows. Also, significant differences were detected for non-maternal reciprocal effects (d) for grain yield, length of ear and number of tassel branches. Additive components of genetic variances (D) were greater than the two dominance components (H1 and H2) for all traits, except grain yield, ear height, and ear length (Table 2). The highest (1.53) and the lowest (0.5) degrees of dominance were observed for ear length and number of kernel rows, respectively. The regression coefficients ranged between 0.61 for ear length to 0.98 for number of kernel rows and number of tassel branches. The highest and lowest values of average heterosis were observed for grain yield and number of leaves per plant. Heterosis was negative for cob percentage and days to tassel emergence (Table 2). The sign of F indicated that dominant alleles were more frequent than recessive alleles for most traits. Broad-sense heritability estimates ranged from 0.68 for number of grain rows to 0.91 for cob percentage. Narrow-sense heritability estimates were also widely-varying ranging from 0.73 for days to tasselling to 0.06 for grain yield (Table 2). Figures 1 to 4 presents the distribution of parents along the regression line of Wr on Vr, and the sign and amount of intercepts for four traits. A zero, positive, or negative intercept indicate complete, partial and over dominance gene action, respectively. Finally, the parents closer to the origin point posses more dominant alleles and those further from the origin contain more recessive alleles for the respective traits.
1-Baker, R. J. 1978. Issues in diallel analysis. Crop Sci., 18 : 533 - 536.
2-Griffing, B. 1956. Concept of general and specific combining ability in relation to diallel crossing system. Aust. J. Biol. Sci. ,9 : 463 - 493.
3-Hayman, B. I. 1954a. The analysis of variance of diallel tables. Biometrics, 10 : 235 - 244.
4-Hayman, B. I. 1954b. The theory and analysis of diallel crosses. Genetics, 39 : 789 - 809.
5-Jinks, J. L. and Hayman, B. I. 1953. A survey of the genetical basis of heterosis in a variety of diallel crosses. Heredity, 9 : 223 - 238.
6-Walterrs, D. E. and Gal, J. S. 1977. A note on the Hayman analysis of variance for a full diallel table. Heredity. 38 : 401- 407.
7-Wright, A. J. 1985. Diallel designs, analysis and reference population. Heredity, 54 : 307 -311.
Table 1. ANOVA according to Hayman’s method for grain yield and others traits
sources |
df |
Grain yield |
Percentage of cob |
Number of leaves |
Plant height |
Ear height |
Number of days to tasseling |
Length of leaves |
Number of kernel row |
Length of ear |
Number of tassel branches |
rep |
2 |
9.4** |
31.23** |
10.24** |
391.8** |
118.8 ns |
67.4** |
6.4 ns |
8.84** |
2.73 ns |
49.2** |
a |
9 |
2.5** |
52.55** |
54.55** |
8082** |
1259.** |
546.6** |
750.2** |
42.3** |
41.23** |
426** |
b |
45 |
5.0** |
119.3** |
2.9 ** |
1191** |
377.3** |
29.3 ** |
79.7** |
3.46 * |
9.23** |
41.8** |
b1 |
1 |
90** |
1072** |
21.87** |
21297** |
4625.** |
71.1** |
1528** |
10.45** |
143.7** |
440** |
b2 |
9 |
6.8** |
413.1** |
3.43 ns |
1368** |
280.8** |
68.9** |
58.3** |
5.14 ** |
14.3** |
32.4 ns |
b3 |
35 |
2.1 ** |
16.6** |
2.22 ns |
571.4 ** |
280.7** |
17.9 * |
43.8 * |
2.82 ns |
4.07 * |
32.8 * |
c |
9 |
4.4** |
7.4ns |
1.13 ns |
242.6 ns |
125.7 ns |
13 ns |
30.4 ns |
3.93 ** |
2.9 ns |
19.6 ns |
d |
36 |
1.8 ** |
7.8 ns |
1.53 ns |
248.4 ns |
131.7 ns |
11.2 ns |
30.8 ns |
2.39 ns |
6** |
36.35 ** |
error |
198 |
0.59 |
5.9 |
1.45 |
240.6 |
70.1 |
11.1 |
27.5 |
2.2 |
2.2 |
19.15 |
total |
299 |
*,**, ns : significant at 5% and 1% probability level and not significant respectively
Table 2. Genetic variance components and related statistics in parental inbred lines of corn
parameter |
Grain yield |
Cob % |
Number of leaves |
Plant height |
Ear height |
Number of days to tasseling |
Length of leaves |
Number of kernel row |
Length of ear |
Number of tassel branches |
D |
2.4 |
126 |
2.57 |
1067 |
111 |
45.5 |
2.57 |
6.6 |
5.65 |
23.5 |
H1 |
4.6 |
184 |
1.5 |
935 |
261 |
27.5 |
1.5 |
1.6 |
7.94 |
18.6 |
H2 |
2.94 |
75 |
0.97 |
635 |
205 |
12.2 |
0.97 |
0.86 |
4.7 |
15.2 |
Sqr(H1/D) |
1.34 |
1.2 |
0.76 |
0.94 |
1.53 |
0.78 |
0.76 |
0.50 |
1.2 |
0.89 |
h2b |
0.80 |
0.91 |
0.81 |
0.84 |
0.79 |
0.85 |
0.81 |
0.68 |
0.77 |
0.73 |
h2n |
0.06 |
0.69 |
0.71 |
0.52 |
0.34 |
0.73 |
0.71 |
0.58 |
0.40 |
0.57 |
b |
0.80 |
0.94 |
0.73 |
0.89 |
0.61 |
0.90 |
0.73 |
0.98 |
0.62 |
0.98 |
F |
3.9 |
232 |
-0.44 |
845 |
87.7 |
25.2 |
6.3 |
4.7 |
6.3 |
-0.06 |
Figure 1: Distribution of parents along regression line for grain yield
Figure 2: Distribution of parents along regression line for number of tassel branches
Figure 3: Distribution of parents along regression line for cob percentage
Figure 4: Distribution of parents along regression line for plant height