Triple test
cross analysis for detection of epistasis for ear characteristics in maize (Zea
mays L.)
Parvez A. Sofi, A.G.Rather and S. Venkatesh*
Division of plant Breeding &
Genetics, SKUAST-K, Shalimar, 191121, J&K, India
* Winter Maize Nursery (ICAR), Amberpet,
Hyderabad
INTRODUCTION
Maize (Zea mays L.) is one of the
most important cereals of world grown in about 70 centuries. It is the staple
food of over 200 million people in developing counties of Asia, Latin America
and Africa. Besides its importance in global agriculture, it possesses one of
the most well studied genetic systems among cereals which has motivated a rich
history of research into the genetics of various traits in maize. In fact maize
has been subjected to extensive genetic studies than any other crop [1] which
has resulted in elucidation of several basic principles that are of practical
importance to plant breeders. The general conclusion from these studies has
been that additive genetic variance is the most important in the inheritance of
quantitative traits followed by dominance whereas, epistasis is of minor
significance. However, there is growing evidence that epistasis is an important
component of genetic variance. In fact Eta-Ndu and Openshaw (1999) opined that
failure to include epistasis in the estimation of genetic components causes
bias in such estimates of expected genetic gain under selection. Therefore, the
present, investigation was undertaken to characterise the genetic architecture
of ear characters in maize by using triple test cross analysis.
The
material for present study was generated by crossing 15 white inbred lines
(Four local and 11 exotic) of maize viz., WI-9, W-6, W-7, GLET-7, GLET-27,
CML-77, CML-79, CML-111, CML-138, CML-173, CML-213, CML-214, CM:-240, CML-244
and CML-463 with three testers W3, W5 and W3 x W5. The test crosses were
generated in 2004 at winter maize nursery, Amberpet (Hyderabad). The parental
lines, testers and crosses were evaluated at two diverse locations of Kashmir
valley viz., Larnoo and Wadura in RBD with three replications at each
locations. Data was recorded for six
ear characters from 10 competitive plants from each replication and
analysed as per the procedure of Ketata et al. (1973) which is based on the
original model proposed by Kearsey and Jinks (1968) Triple test cross procedure
is an efficient genetic model and is applicable to segregating and
non-segregating populations arising from F2, Backcross or homozygous lines.
Besides, it is independent of gene frequency, linkage relationship and degree
of inbreeding. In addition to the detection of epistasis, it provides unambiguous
estimates of additive and dominance components in absence of epistasis. The
basic model for Kearsey and Jinks [4] model is :
Lijk
= M + Gij + rk + Eijk where,
Lijk
= Phenotypic
value of cross between tester Li and line j in replication K.
M = Genotypic value of the cross
between tester Li and line j.
Rk = Effect of kth
replication, and
Ekjk = Erro associated with cross ij in replication
K
The epistatic components i.e., total epistasis, [1]
type [j + l) type were tested against their environmental interaction which in
turn were tested, for significance, against their block x environment
interactions. The degree of dominance was calculated as (H/D) ½ and the
direction of dominance was determined by the correlation coefficient between
corresponding sums (L1i + L2i) and differences (L1i
- L2i).
RESULTS AND DISCUSSION
The analysis of variance presented in table-1 revealed
significant mean squares due to genotypes, lines, testers, crosses, and parents
v/s crosses indicating that substantial variability exists in the parental
lines for ear traits, and that there were significant differences between
parents and crosses. The environmental component was significant for all traits
except ears/plot whereas, G x E interaction was significant for all traits
except ear diameter, ears/plot and seed weight/ear indicating that environment
plays an important role in the expression of these traits as is expected for
quantitative traits. Similar, results
have been reported in maize for ear triats by Satyanarayan (1999) and Dodiya
and Johshi (2003)
Te analysis of variance for the detection of epistasis
(Table-2) revealed significant epistasis for all traits except ears/plot,
further establishing the fact that epistasis can not be excluded in the
estimation of genetic parameters. Non-detection of epistasis does not mean
absence of epistasis Bernardo (2002) and is primarily due to requirement of
complex mating designs, difficulty to separate epistatic variation fro additive
and dominance, large sampling errors associated with epistasis and lower
magnitude of epistasis compared to additive and dominance variation. Besides it
is also attributable to basic limitation of the Kearsey ad Jinks (1968) model
wherein the epistasis refers to loci for which the testers L1 and L2 differ and
discrepancies may occur due to genetic differences between testers and that it
is very difficult to have testers which differ at all loci for a number of
traits.
The portioning of epistasis and its fixable [I] and
non-fixable [j + l} revealed significance of both components for all traits
except ear length, ears/plot and seed weight/ear for which additive interaction
was non-significant and ears/plot for which (j + l) type was non-significant.
Epistasis as well as its components interacted significantly with environment
for most of the traits. The comparative analysis revealed that non-fixable [j +
l] component of epistasis was greater than its corresponding fixable component
for all traits except ear height, ear diameter and kernel rows/ear, where
reverse was the case. The preponderance of non-additive epistasis indicates that hybrid breeding can be
employed to exploit this component. However, both components, i.e., [I] and [j
+ l] can be exploited in intra as well as inter population improvement. Similar
results in maize have been reported by Wolf and Hallauer (1997) and Leon et
al. (2005)
The estimates of genetic components of variance
significance of both additive and dominance components for all traits except
ear diameter for which (H) component was non-significant. Additive component
was greater than its corresponding dominance components for all traits. Both
additive and dominance components interacted with the environment for kernel
rows/ear and seed weight/ear whereas, for other traits they were relatively
stable across environments.
The degree of dominance was in the range of partial
dominance whereas, the direction of dominance (F) was non-significant
indicating ambidirectional nature of dominance. Similar, results in maize have
been reported by Gautam (2003) and Bhatnagar et al. (2004) The preponderance of
additive component would mean that simple selection schemes would be successful
for bringing about improvement in the traits studied but existence of
substantial epistasis implicates that the selection process would have to be
delayed as non fruitful results would
be obtained in immediate
generations. Recurrent selection procedure may be more useful in such
situations to exploit both additive and dominance components by increasing the
frequency of favourable alleles while maintaining genetic variation in breeding
population Doerksen et al
(2003)
CONCLUSION
The
presence of epistasis for almost all ear characters indicates that the
estimates of components of variation would be biased to an unknown extent if
they are estimated by genetic models assuming absence of epistasis. Regardless
of the type of epistasis, the bias tends to be greater in additive component
than that of dominance component [13] which causes over estimation of narrow
sense heritability and consequently the predicted genetic gain would have an
additional bias proportional to that of heritability. It would thus, be logical
to search for epistasis rather than attributing it to left over variance after
additive and dominance are accounted for.
REFERENCES
1.
Hallauer,
A.R. and Miranda, J.B. 1988. Quantitative Genetics in Maize Breeding. Lowa
State College Press, Ames. 468 pp.
2.
Eta-Ndu, T
and openshaw, S.J. 1999. Epistasis for grain yield in two G2 populations in
maize. Crop Sci. 39(2) : 346-352.
3.
Ketata, H.,
Smith, E., Edwards, L. and McNew,R. 1976. Detection of epistatic, additive and
dominance variation in winter wheat. Crop Sci. 16 : 1-4.
4.
Kearsey, M.J.
and Jinks, J.L. 1968. Ageneral method of detecting additive, dominance and
episstatic variation for metric traits. I. Theory Heredity 23 : 403-409.
5.
Satyanarayan,
E. 1996. Combining ability for quantitative traits of early single crosses
maize hybrids under rainfed conditions. Madras J. of Agri. 30 : 204-208.
6.
Dodiya, N.
and Joshi, V. 2003. Heterosis and combining ability for quality and yield in
early maturing single crosses of maize. Crop Res. 26(1) : 114-118.
7.
Bernardo, R.
2002. breeding for quantitative traits in plants. Stemma Press, Woodburg. P.
141.
8.
Wolf, D.P and
Hallauer, A.R. 1997. Triple test cross analysis to detect epistasis in maize.
Crop Sci. 37 : 763-770.
9.
Leon, N.D.,
Goors, J. and Kaeppler, S. 2005. Genetic control of prolificacy and related
traits in goldern glow maize population II. Genetypic analysis. Crop Sci. 45 :
1370-1378.
10.
Gautam, A.S.
2003. Combining ability studies for grain yield and other traits in inbred
lines of maize. Crop Res. 26(3) : 482-485.
11.
Bhatnagar,
S., Betran, F. and Rooney, L. 2004. Combining ability of quality protein maize
inbreds. Crop Sci. 44 : 1997-2005.
12.
Doerksen, T., Kannenberg, L. and Lee, E.A. 2003. Effect of recurrent selection on combining ability in
maize breeding populations. Crop Sci. 43 : 1652-1658.
13.
Viana, J.M.
2005. Dominance, Epistasis, heritabilities and expected genetic grains.
Genetics and Mol. Biol. 28(1) : 67-74.
|
Source of variation |
d.f. |
Ear height (cm) |
Ear length (cm) |
Ear diameter (cm) |
Ears/plot |
Kernel rows/ear |
Seed weight/ear (g) |
|
Environments |
1 |
2177.58** |
251.66** |
10.61** |
1.337 |
2.19** |
535.81** |
|
Genotypes |
62 |
920.28** |
21.23** |
0.80** |
41.22** |
3.77** |
39.63* |
|
Parents (Lines) |
14 |
273.08** |
21.36** |
0.44** |
14.15 |
0.99* |
92.59** |
|
Parents (testers) |
2 |
175.50** |
51.99** |
2.00** |
44.34** |
1.40** |
49.78* |
|
Crosses |
44 |
739.60** |
17.53** |
0.55** |
53.16** |
2.45** |
125.79* |
|
Parents v/s crosses |
1 |
1577.25** |
402.79* |
17.94** |
1039.18** |
110.41** |
1064.74** |
|
G x E interaction |
62 |
227.30** |
4.34** |
0.19 |
9.23 |
1.14* |
10.82 |
|
Error |
258 |
39.20 |
1.54 |
0.16 |
9.55 |
0.50 |
19.99 |
Table 2: Analysis
of variance for detection of epistasis for and ear characteristics in maize
|
Source of variation |
d.f. |
Ear height (cm) |
Ear length (cm) |
Ear diameter (cm) |
Ears/plot |
Kernel rows/ear |
Seed weight/ear (g) |
|
Epistasis (L1i + L2i-2L3i) |
15 |
451.36** |
55.68** |
2.07 |
146.48 |
4.53** |
95.59** |
|
[i] type epistasis |
1 |
493.455** |
24.52 |
2.58* |
10.21 |
5.53** |
0.66 |
|
[j +l] type epistasis |
14 |
448.367** |
57.90** |
2.03* |
152.13 |
4.45** |
102.38** |
|
Epistasis x blocks |
30 |
56.28 |
7.04 |
0.16 |
44.42 |
1.15 |
37.59 |
|
[I] type x blocks |
2 |
38.93 |
3.97 |
0.91 |
13.73 |
0.03 |
52.85 |
|
[j + l} type x blocks |
28 |
57.52 |
7.25 |
0.17 |
42.13 |
1.23 |
36.50 |
|
Epistasis x environments |
15 |
127.42** |
18.09** |
0.93** |
86.12** |
1.13** |
29.14** |
|
[i] type x environments |
1 |
103.18* |
18.64** |
1.12** |
13.31 |
2.10 |
4.35 |
|
[j + l] type x environments |
14 |
129.13** |
18.05** |
0.91** |
91.32** |
1.42** |
30.91** |
|
[i] type x blocks x environments |
2 |
13.96 |
3.19 |
0.06 |
6.91 |
3.71 |
1.34 |
|
[j + l) type x blocks x environments |
28 |
46.80 |
7.25 |
0.17 |
15.14 |
0.12 |
10.48 |
|
Epistasis x blocks x environments |
30 |
44.61 |
6.98 |
0.16 |
14.37 |
0.24 |
9.87 |
Table 3 : Estimates
of additive (D), dominance (H), additive x environment (G2D), dominance x
environment (G2H), degree and direction of dominance for ear characteristics in
maize
|
Additive component (D) |
1181.81**
± 428.47 |
13.03**
± 4.95 |
0.37*
± 0.14 |
0.20** ± 0.01 |
2.38** ± 0.11 |
81.33**
± 3.69 |
|
Additive x environment (G2D) |
36.32
± 24.50 |
1.06
± 0.71 |
± 0.08 ±0.01 |
0.14** ± 0.01 |
0.23** ±0.02 |
4.26** ±0.22 |
|
Dominance component (H) |
80.36* ± 37.05 |
5.40*
± 2.41 |
0.10 ±0.05 |
0.13** ± 0101 |
0.67** ± 0.03 |
24.02** ±1.48 |
|
Dominance x environment (G2H) |
24.40 ± 16.36 |
2.58**
± 0.97 |
0.06
± 0.04 |
0.01 ± 0.01 |
016** ± 0.01 |
15.70**
± 0.75 |
|
Degree of dominance (H/D) V2 |
0.26 |
0.64 |
0.52 |
0.80 |
0.53 |
0.54 |
|
Direction of dominance (F) |
38.21 |
1.38 |
0.01 |
0.48 |
-4.16 |
-1.78 |