Maydica 55 (2010): 33-39
DETECTION OF EPISTASIS FOR PLANT HEIGHT AND LEAF AREA PER
PLANT IN MAIZE (ZEA MAYS L.) FROM GENERATION MEANS ANALYSIS
M. Iqbal1, K. Khan2, H. Rahman3, H. Sher4,*
1
Cereal Crops Research Institute (CCRI) Pirsabak (Nowshera) NWFP, Pakistan
Department of Environmental Sciences, COMSATS Institute for Information Technology (CIIT),
Abbottabad, Hazara, NWFP, Pakistan
3 Department of Plant Breeding & Genetics NWFP Agricultural University Peshawar, Pakistan
Department of Botany & Microbiology, College of Sciences, King Saud University, Riyadh, Kingdom of Saudi Arabia
2
4
Received December 9, 2009
ABSTRACT - The role of gene action and interactions in
the inheritance of two physiologic traits in maize was examined through generations’ means analysis in a field
study during 2005 and 2006 at CCRI, Pirsabak, Nowshera,
Pakistan. A total of 20 genotypes including 4 parents, 4
F1s, 4 F2s, 4 BC1s and 4 BC2s were evaluated in a randomized complete block design with 3 replications. Data were
recorded for leaf area and plant height on plant basis. A
combined analysis of variance for generation means across
years was carried. Genetic effects were estimated by
method of weighted least square, using appropriate
weights (coefficients) for each generation mean equation.
Significant differences among generation means were indicated. Chi square values were significant for all crosses in
both traits according to a joint scaling test. Six parameters
model was applied to accommodate the digenic epistatic
interactions. Significant dominance effects were indicated
in all crosses for both traits and these effects were much
higher in magnitude than their corresponding additive effects. All crosses in the study also indicated significant additive x dominance effects for leaf area. Significant additive x additive effects were also present in some crosses
for the two traits. Complementary and duplicate gene interactions appeared operative in the inheritance of leaf
area trait. Duplicate gene interactions were seen functioning in controlling plant height in most crosses. Epistasis
played a considerable role in controlling leaf area and
plant height traits. Improvement in these two physiological
plant traits could be much slower in a selection program.
KEY WORDS: Generation means analysis; Physiological
traits; Gene action; Maize.
INTRODUCTION
The primary objective of many maize breeding
programs usually involves the evolution of maize
hybrids/varieties with high grain yield potential. In
* For correspondence (e.mail: hassan.botany@gmail.com).
the North-West Frontier Province of Pakistan, especially the mountainous areas of Malakand and Hazara divisions and the tribal areas bordering
Afghanistan, maize is the primary crop in majority
of the farming systems, the staple food of the rural
population and the main source of fodder for livestock (KHAN et al., 2003). In such situation, maize
breeders emphasize on the evolution of dual purpose maize cultivars (varieties/hybrids), having high
grain as well as fodder yield potential. Appropriate
breeding approaches are needed to improve certain
plant traits, especially leaf area and plant height, to
achieve the objective in the available time and resources.
Estimation of the types of gene action involved
in the expression of traits, the level of additive effects and the degree of dominance are very important in designing a breeding method for improving
the trait of interest. Knowledge of the way genes act
and interact will determine which breeding system
can optimize gene action more efficiently and will
help elucidate the role of breeding systems in the
evolution of crop plants (HALLAUER and MIRANDA,
1988). GAMBLE (1962) indicated that the estimates of
genetic effects can help the plant breeders to decide
the breeding procedures better suited for the improvement of the trait being analyzed. Generation
mean analysis, a biometrical method developed by
MATHER and JINKS (1982), greatly helps in the estimation of various components of genetic variance.
Breeding for improved varieties is a continuous
process and requires a thorough understanding of
the genetic mechanisms governing yield and yield
components (SALEEM et al., 2002; UNAY et al., 2004).
Several breeding procedures have been established
to increase yield, grain as well as fodder, of maize
populations and the hybrids derived from there in.
In order to choose the best hybrid combinations, a
large number of subjectively chosen inbred lines are
34
M. IQBAL, K. KHAN, HIDAYAT-UR-RAHMAN, HASSAN SHER
crossed with each other (UNAY et al., 2004). Before
the development of high yielding maize cultivars
and/or hybrids, it is important to study the yield
components of maize and their genetic architecture.
A large number of genetic studies have been made
in the past to explore the genetic basis of yield and
yield components in maize cultivars but with very
little emphasis on the physiological traits such as
leaf area per plant and plant height which are usually considered to have substantial contribution towards increasing grain and fodder yield. AMER et al.
(2002) observed that additive genetic variance was
predominant in the inheritance of plant height, ear
height and some yield related traits while kernels
row-1 and grain yield were under the control of
non-additive genetic variance. Evaluating single
cross hybrids, SUJIPRIHATI et al. (2003) observed that
non-additive gene action played a significant role in
the inheritance of plant height, ear height and grain
yield and its related traits. KUMAR et al. (2005) reported over dominance gene action for plant height, ear
height and grain yield plant-1. The study of TABASSUM
and SALEEM (2005) revealed that the expression of
leaf area per plant, 1000-grain weight and grain
yield plant-1 was mainly governed by over dominance type of gene action while the importance of
additive gene effects was highlighted for plant
height. Similarly SINGH and ROY (2007) observed that
plant height and days to maturity were governed by
the additive type of gene effects while grain yield
plant-1 and some other characters were controlled
by non-additive gene action.
These and numerous other genetic studies have
been conducted in the past but little information is
available about gene action and interactions in the
inheritance of leaf area per plant and plant height in
crosses between contrasting groups of inbred lines
i.e., inbred lines of early maturity with short stature
and tall with late maturity. Therefore, the present
study was undertaken with the objectives 1) to elucidate the mode of inheritance of leaf area and
plant height and 2) to determine the relative magnitudes of gene effects in controlling these traits.
MATERIALS AND METHODS
Genetic material
The breeding material used in this study comprised a set of
four flint white kernel maize inbred lines, developed by manual
self pollination for 6-8 generations, having distinct genetic makeup (Table 3). Two out of these 4 inbred lines were tall in stature
and with late maturity (100-120 days) and the other two were
TABLE 1 - Development of F1 generations by crossing four maize
inbred lines.
––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––
Cross
−
−
––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––
1.
Pop.9804 X Pop.9801
2.
Pop.9804 X
FRW-4
3.
Pop.9805 X Pop.9801
4.
Pop.9805 X
FRW-4
––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––
dwarf having early maturity (90-100 days). These lines were
grown at Cereal Crops Research Institute (CCRI) Pirsabak, Nowshera, NWFP (Pakistan) located at about 1540 km North of Indian Ocean at 34°N latitude and 72°E longitude and with an altitude of 288 meters above sea level, representing a continental
climate. Six generations for each cross were developed by manual pollination procedures for crossing and selfing (RUSSELL and
HALLAUER, 1980) over two growing seasons.
Development of F1 generations
During the first growing season i.e. Spring 2005 all four
parental lines were crossed with each other to produce four F1
hybrids using the crossing pattern as shown in Table 1.
Development of F2 and back cross (BC1 & BC2) generations
Part of seed from each of the four parental inbred lines and
their four resultant F1 hybrids was planted in the field during
summer 2005 to produce F2, BC1 and BC2 generations. F2 generation of each cross was produced by selfing the F1 plants while
BC1 and BC2 generations were developed by crossing back each
F1 hybrid with its respective male and female parents in the 2nd
crop season as given in Table 2.
Field evaluation
The material for field evaluation comprised 20 entries, generated from the above crossed and selfed combinations i.e., 4 parents, 4 F1s, 4 F2s, 4 BC1s and 4 BC2s. These 20 entries were planted in the field in triplicate, using randomized complete block
(RCB) design, for evaluation at Cereal Crops Research Institute
(CCRI) Pirsabak, Nowshera for two consecutive years (2006 and
2007) during summer growing season (July-October). The plot
size comprised two rows each for non-segregating generations
(P1, P2 and F1), 4 rows each for BC1 and BC2 generations and 8
rows each for F2 generation. Each row was 5 m long with row
spacing of 75 cm and a distance of 25 cm between plants within
row. Two seeds were planted in each hill and were later thinned
to maintain one plant hill-1. A uniform fertilizer dose of 200 kg N,
90 kg P2O5 and 90 kg K2SO4 hectare-1 was applied. Whole P2O5
(as Single Super Phosphate, SSP) and K2O (as Sulphate of Potassium) and half Nitrogen (as Urea) were applied just before planting at the time of land preparation while remaining half N was
applied as side dressing in the form of Urea, about 3 weeks after
emergence. Weeds were controlled by pre-emergence application
of Primextra Gold @ 600 ml acre-1. Insects control was carried out
through seed treatment with Confidor WS-70 before planting and
with the application of Furadon granules (3%) one month after
planting, through leaf whirl application. Hand weeding and
earthing-up operations were practiced for weed control in later
stages about four weeks after emergence. The crop was irrigated,
as and when required, till one week before maturity. Mean mini-
DETECTION OF EPISTASIS FROM GENERATION MEANS ANALYSIS
35
TABLE 2 - Development of back cross generations from four maize inbred lines.
–––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––
BC1 generations
BC2 generations
–––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––
–––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––
(Pop.9804 X Pop.9801) X Pop.9804
(Pop.9804 X Pop.9801)
X Pop.9801
(Pop.9804 X FRW-4) X Pop.9804
(Pop.9804 X FRW-4)
X FRW-4
(Pop.9805 X Pop.9801) X Pop.9805
(Pop.9805 X Pop.9801)
X Pop.9801
(Pop.9805 X FRW-4) X Pop.9805
(Pop.9805 X FRW-4)
X FRW-4
–––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––
TABLE 3 - Name, pedigree and maturity of the parental inbred lines.
–––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––
S. No.
Name
Pedigree
Maturity
–––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––
1.
Pop. 9804
CHSW X Sahard White
Late (110-120 days)
2.
Pop. 9805
Sarhad White X Azam
Late (100-110 days)
3.
Pop. 9801
Pahari X Swabi White
Early (90-95 days)
4.
FRW-4
Shaheen X Peshawer White
Early (95-100 days)
–––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––
TABLE 4 - Analysis of variance for generations combined across years.
–––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––
Source of Variation
Degrees of Freedom
Mean Squares
Expected Mean Square
–––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––
Years
y-1
=1
M1
——
Reps (years)
y(r-1)
=4
M2
——
Generations
g-1
=5
M3
δ2e+rδ2gy+ryδ2g
Generation x Year
(g-1) (y-1)
=5
M4
δ2e+rδ2gy
Pooled Error
y (g-1) (r-1)
= 20
M5
δ2e
Total
ryg-1
= 35
——
——
–––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––
mum and mean maximum temperatures ranged from 22.68°C to
35.3°C during crop growing season of 2006 (CCRI, 2006) and
14.68°C to 34.83°C in 2007 (CCRI, 2007). A total precipitation of
143.25 mm during the crop growth period of 2006 (CCRI, 2007)
and 180 mm during 2007 (CCRI, 2008) was recorded. Data on individual plant basis for the two physiological plant traits (leaf
area per plant and plant height) was recorded on 10 randomly
selected plants in each plot of P1, P2 and F1 generations, 20
plants/plot in each back cross and 30 plants/plot in each F2 generations as mentioned below:
1. Leaf area (cm2) - Length of 8th leaf from the top was measured in centimeters just after completion of pollination and
fertilization processes. At the same time width of the same
leaf was also measured at the widest point in centimeters
(TOLLENAAR et al., 2004). Leaf area was then determined using
the following formula suggested by FRANCIS et al. (1969).
Leaf Area = (Leaf Length x Leaf Width) x 0.75
Leaf area per plant was then calculated by multiplying area
2.
of a single leaf with a leaf area factor, i.e., 9.39 (PEARCE et al.,
1975).
Plant height (cm) - Height of each plant was measured as the
distance from ground level to the auricle of the flag leaf
(GUZMAN and LAMKEY, 2000) with the help of a measuring rod
on 10 randomly selected plants for P1, P2 and F1 generations,
20 plants each for BC1 and BC2 generations and 30 plants for
F2 generation.
Statistical analysis
Combined analysis of variance (GOMEZ and GOMEZ, 1983)
was carried out, using MSTAT-C (User’s guide to MSTAT-C, 1991)
computer program, according to an ANOVA format (Table 4) to
detect significant differences among generations for the two
plant parameters.
Generation means analysis was applied, on parameters
which showed significant differences among generations, to determine the mode of inheritance and the magnitude of gene ac-
36
M. IQBAL, K. KHAN, HIDAYAT-UR-RAHMAN, HASSAN SHER
TABLE 5 - Coefficients α and β utilized for the construction of different models in generation means analysis.
–––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––
Generation
Genetic effects
M
A
D
Aa
Ad
Dd
–––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––
1
1
-0.5
1
-1
0.25
P1
P2
1
-1
-0.5
1
1
0.25
F1
1
0
0.5
0
0
0.25
F2
1
0
0
0
0
0
BC1
1
0.5
0
0.25
0
0
BC2
1
-0.5
0
0.25
0
0
–––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––
TABLE 6 - Combined analysis of variance for generation means showing source of variation, degrees of freedom and mean squares for leaf
area per plant and plant height.
–––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––
Mean Squares
Source of
Plant Height (cm)
DF
Leaf Area (m2)
Variation
Cross 1
Cross 2
Cross 3
Cross 4
Cross 1
Cross 2
Cross 3
Cross 4
–––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––
Years
1
1669367*
146713ns
1939903*
327932*
244.93*
876.16**
1.25ns
48.07ns
Reps(year)
4
139820
148211
141431
5249
24.64
6.07
11.84
12.35
Generations
5
809943
371711**
631830**
250818**
2160.64**
1072.86**
2255.18**
1920.15**
Gen x Year
5
106828*
211971**
170369ns
140851**
40.92ns
148.26**
131.68**
345.57**
Pooled Error
20
28214
40573
87084
20641
24.67
5.67
6.35
11.75
C.V (%)
11.79
12.28
20.62
8.78
3.38
1.77
1.73
2.49
–––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––
** = Significant at 1 % probability level; * = Significant at 5 % probability level; ns = Non significant.
tion for the traits under investigation. This analysis was accomplished in several steps:
1) Generation mean for each trait, showing significant differences among generations in the combined analysis of variance, was calculated from the two years data by using Microsoft Excel computer program.
2) For a given trait, each generation mean was expressed in
terms of its genetic effects, using the following equation suggested by HAYMAN (1958).
–
G = m + αa + βd + α2aa + 2αβad + β2dd
–
Where G = observed mean for generation; m = mean effect;
a = average additive effects; d = average dominance effects;
aa = average interactions between additive effects; ad = average interactions between additive and dominance effects; dd
= average interactions between dominance effects, α and β
are the coefficients of a and d which are listed in Table 5.
Because the various generation means are not known with
equal precision (HAYMAN, 1958), weights were calculated for
each generation; the appropriate weights being the reciprocals of the squared standard errors of each mean (MATHER
and JINKS, 1971).
4) With the set of equations obtained in step 2 and considering
the weights associated with each generation, genetic effects
were estimated by the method of weighted least squares
(multiple linear regressions) by utilizing matrix algebra as described by ROWE and ALEXANDER (1980). Briefly, these three
matrices were defined as: W (weights matrix), X (matrix of
coefficients of genetic effects) and Y (column vector of observed generation means). The estimates of genetic effects
were derived from the column vector β, defined as: β =
(X2WX)-1(X2WY).
5) Standard error associated to each estimate of a genetic effect
was obtained as the diagonal elements of the solution equation SE(β) = 兹莥莥莥莥莥
(XX)-1σ2, where σ2 was the error variance, estimated by the mean square of the two years average
(HOLTHAUS et al., 1996; FRANK and HALLAUER, 1997).
6) Significance of each genetic effect estimate was evaluated as
described by SNEDECOR and COCHRAN (1989), utilizing a t-test.
7) In order to test the adequacy of the model, chi-square (χ2)
tests were performed as outlined by ROWE and ALEXANDER
(1980).
Steps 3 to 7 were accomplished utilizing the Microsoft Excel
computer program.
The joint scaling test (CAVALLI, 1952) was used to detect epistasis for both traits measured. In the presence of epistasis, additive (a), dominance (d) effects and non-allelic interaction components (aa, ad, and dd) of generation means were estimated to explain the inheritance of various traits, using HAYMAN (1958, 1960)
and MATHER and JINKS (1971) models. A three parameter model
also known as additive-dominance model was used to explain
genetic variability for those traits which showed non significant
values for chi-square (χ2). By observing a significant χ2 value, a
six parameter model was applied to accommodate the digenic
epistatic interactions.
DETECTION OF EPISTASIS FROM GENERATION MEANS ANALYSIS
37
TABLE 7 - Estimates of genetic effects for two physiological traits in 4 maize crosses evaluated at CCRI combined over 2 years (summer, 2006
and summer, 2007).
–––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––
Type of
non-allelic
Parameter
Cross
m
a
d
aa
ad
dd
χ2
interaction
–––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––
Leaf area per plant Pop.9804 X FRW-4
3614.52* 149.97ns 4398.61* 2406.38* -567.40*
-004.23*
309.65** Duplicate
Pop.9804 X Pop.9801 4223.76*
318.21*
1896.45*
437.43*
-230.57*
104.59ns
47.32** ——
Pop.9805 X FRW-4
4445.03* -154.28ns 1754.71* -435.12ns -901.50* 1871.32* 109.19** Complementary
Pop.9805 X Pop.9801 4824.51*
351.45*
1402.92* -482.70ns -227.19*
514.53*
8.34* Complementary
Pop.9804 X FRW-4
142.86*
-1.63ns
78.11*
30.46*
9.28*
-18.01*
52.88** Duplicate
Pop.9804 X Pop.9801 134.64*
-3.33*
52.36*
17.26*
-2.23ns
-23.09*
9.28* Duplicate
Pop.9805 X FRW-4
146.69*
-5.08*
56.70*
9.21ns
7.73*
1.23ns
34.85** ——
Pop.9805 X Pop.9801 140.42*
14.50*
56.92*
12.17*
17.50*
-24.30*
115.03** Duplicate
–––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––
m = mean; a = additive; d = dominance; aa = additive x additive; ad = additive x dominance; dd = dominance x dominance.
** = Significant at 1% probability level; * = Significant at 5% probability level; ns = Non significant.
Plant Height
RESULTS
The combined analysis of variance indicated significant differences among generations for leaf area
per plant and plant height (Table 6). All chi- square
(χ2) values were significant for all 4 crosses in both
traits according to joint scaling test (CAVALLI, 1952),
thereby, indicating the inadequacy of three parameters model to explain the genetic variability for
these traits (Table 7). Therefore, a six parameters
model was applied to accommodate epistasis for
these traits.
Estimates of genetic effects from generation
mean analysis according to a 6 parameter model
and the chi-square (χ2) values for leaf area per
plant in 4 maize crosses are presented in Table 7.
Estimates of genetic effects for dominance (d) and
additive x dominance (ad) were significant for all
crosses where as significant additive (a) effects were
found only in 2 of the 4 crosses (Pop.9804 x
Pop.9801 and Pop.9805 x Pop.9801). Moreover, significant dominance x dominance (dd) interaction effects were seen in 3 crosses and the additive x additive (aa) interaction effects of significant nature
were observed in two of the 4 crosses examined.
One of the 4 crosses (Pop.9804 x FRW.4) having
significant positive estimate of dominance (d) effects was found to have significant negative estimate for dominance x dominance (dd) as well.
However, two other crosses (Pop.9805 x FRW.4 and
Pop.9805 x Pop.9801) were found to have significant and positive estimates both for dominance (d)
and dominance x dominance (dd) type of interaction. The estimates of dominant gene (d) effects for
leaf area per plant were considerably higher in
magnitude compared to additive (a) gene effects in
all crosses involved in the investigation.
The observed values of chi-square (χ2) and the
estimates of genetic effects for plant height in a 6
parameter model for 4 maize crosses are given in
Table 5. Significant and positive estimates for dominance (d) gene effects were found in all crosses
where as significant estimates for additive (a) gene
effects were observed in 3 out of 4 crosses involved
in the study. Moreover, the relative magnitudes of
dominance (d) gene effects were much higher compared to those of additive (a) gene effects for this
trait. Two of the 4 crosses possessed significant negative estimates for additive (a) effects where as one
cross (Pop.9805 x Pop.9801) had significant positive
additive (a) effect. Each of the 3 interaction effects,
namely additive x additive (aa), additive x dominance (ad) and dominance x dominance (dd), were
found significant for 3 of the each 4 crosses involved. Except for Pop.9805 x FRW.4 where dominance x dominance (dd) interaction effects were
not significant, all other crosses were found to have
significant negative effects for dominance x dominance (dd) and with their corresponding dominance
(d) effects as positive and significant.
38
M. IQBAL, K. KHAN, HIDAYAT-UR-RAHMAN, HASSAN SHER
DISCUSSION
The presence of significant estimates for dominance (d) effects and additive x dominance (ad) effects in all crosses and that of dominance x dominance (dd) effects in ¾ of the crosses show that
dominance or dominance type of epistatic gene effects could have largely been involved in the control of leaf area trait in the present study. The role
of dominance for this trait is also evident from the
existence of significant dominance effects which
were much higher in magnitude compared to additive and additive x additive type of effects. However, the presence of significant estimates for additive
(a) and additive x additive (aa) gene effects in 2 of
the 4 crosses indicates that some additive or additive type of gene action could also be operative in
the inheritance of this trait. Complementary type of
non-allelic interaction was evidenced in two of the
4 crosses (Pop.9805 x FRW.4 and Pop.9805 x
Pop.9801) where estimates both for dominance (d)
and dominance x dominance (dd) effects were significant and positive. Moreover, duplicate type of
epistasis was also detected for one the crosses
(Pop.9804 x FRW.4) in which estimates for dominance (d) and dominance x dominance (dd) effects
had significant values with opposite signs. Results
of the present investigation regarding gene effects
for leaf area are in consonance with the findings of
TABBASSUM and SALEEM (2005) who reported that
over dominance gene effects were important in
governing leaf area in maize in their study on diallel
crosses among 8 promising inbred lines of maize. In
a study on the nature and magnitude of gene action, using 6 generations of maize (P1, P2, F1, F2,
BC1 and BC2), RAVIKANT et al. (2006) reported that
other traits, beside grain yield, were governed by
epistatic gene effects.
The higher magnitude of significant dominance
estimates for all crosses implicate a much larger role
of dominance compared to that of additive gene effects in the inheritance of plant height in these crosses. These results are in agreement with those reported by MOUSA (2004) who observed preponderance
of non-additive gene action in the inheritance of
plant height in his study on evaluation of 54 combinations from 27 newly developed maize hybrids. Inter-allelic interactions were also seen operative in
controlling plant height since estimates for additive x
additive (aa), additive x dominance (ad) and dominance x dominance (dd) effects were found significant in most of the crosses investigated in our study.
Epistasis in the form of duplicate gene interactions
appeared functioning in plant height expression in 3
(Pop.9804 x FRW.4, Pop.9804 x Pop.9801 and
Pop.9805 x pop.9801) of the 4 crosses as these
crosses had positive and significant estimates for
dominance (d) effects with their corresponding significant and negative estimates for dominance x
dominance (dd). The inheritance of plant height in
one cross (Pop.9805 x FRW.4) seemed to have been
controlled mainly by additive (a), dominance (d)
and additive x dominance (ad) gene actions in this
study. Similar findings have also been reported in
other such studies on the inheritance of plant height
trait in maize crop. KASSEM et al. (1978a,b) in their
study on maize populations have concluded that additive x additive (aa) and dominance x dominance
(dd) epistatic effects played a major role in the inheritance of plant and ear height. SUJIPRIHATI et al.
(2003) observed that non-additive gene action
played a significant role in the inheritance of plant
height beside other traits while evaluating single
cross hybrids from a 12 x 12 diallel crosses among
maize inbred lines while TABBASSUM and SALEEM
(2005) and SINGH and ROY (2007) in their studies
have pointed out the importance of additive gene effects in controlling the plant height trait in maize.
The preponderance of dominance and dominance type of epistatic gene effects for leaf area and
plant height traits in the present study reveals that
the expression of these traits was largely controlled
by many genes having small effects and also dominant in their action. The presence of inter-allelic
gene action for controlling leaf area and plant
height in our four crosses between inbred lines with
contrasting maturity and plant height was also evident. Results of the present investigation amply suggest that the parental inbred lines used in our crosses could be a useful source of favorable dominant
genes and/or gene combinations for the development of maize hybrids with increased vigor for leaf
area and plant height. However, improvement of
these traits (gain from selection) in a breeding program would be slow due to presence of lower magnitudes of additive effects compared dominance
and dominance type of epistatic effects.
CONCLUSIONS
In light of the findings of the present study, it
can be inferred that 1) epistasis plays a considerable
role in controlling leaf area per plant and plant
DETECTION OF EPISTASIS FROM GENERATION MEANS ANALYSIS
39
height traits in crosses between inbred lines with
contrasting plant heights and maturities and 2) dominance and dominance type of interaction effects
are more important than additive or additive x additive interaction effects in the inheritance of plant
height and leaf area traits. Hence, gain from selection in these two physiological plant traits could be
much slower in a selection program. Further studies
are suggested to confirm these results by using inbred lines with different genetic backgrounds.
KHAN K., M. IQBAL, Z. SHAH, B. AHMAD, A. AZIM, H. SHER, 2003
Grain and stover yield of corn with varying times of plant
density reduction. Pak. J. Biol. Sci. 6: 1641-1643.
REFERENCES
KUMAR R., M. SINGH, M.S. NARWAL, S. SHARMA, 2005 Gene effects
for grain yield and its attributes in maize (Zea mays L.). Nat.
J. Pl. Imp. 7: 105-107.
AMER E.A., A.A. EL-SHENAWY, H.E. MOSA, 2002 Evaluation of some
new inbred lines of maize for combining ability. Annals Agri.
Sci. Moshtohor 40: 791-802.
CAVALLI L.L., 1952 An analysis of linkage in quantitative inheritance. pp. 135-144. In: E.C.R. Reeveand, C.H. Waddington
(Eds.), Quantitative Inheritance HMSO London.
CCRI, 2007 Annual technical report (2006-07) unpublished. Cereal Crops Research Institute (CCRI), Pirsabak, Nowshera,
NWFP, Pakistan.
CCRI, 2008 Annual technical report (2007-08) unpublished. Cereal Crops Research Institute (CCRI), Pirsabak, Nowshera,
NWFP, Pakistan.
FRANCIS C.A., J.N. RUTGER, A.F.E. PALMER, 1969 A rapid method
for plant leaf area estimation in maize. Crop Sci. 9: 537-539.
FRANK T.E., A.R. HALLAUER, 1997 Generation means analysis of
the twin-ear trait I maize. J. Heredity 88: 469-474.
GAMBLE E.E., 1962 Gene effects in corn (Zea mays L.) I. Separation and relative importance of gene effects for yield. Canadian J. Plant Sci. 42: 339-348.
GOMEZ K.W., A.A. GOMEZ, 1983 Statistical procedures for agricultural research. 2nd ed. London: John Wiley and Sons.
KASSEM E.S., E.A. HASSABALLA, M.A. EL-MARSHIDY, M.A. KHALIFA,
1978a Relative importance of effects in the inheritance of: I.
plant height, ear height and flowering of maize plants. Egypt
J. Agron. 3: 203-212.
KASSEM E.S., M.A. EL-MARSHIDY, E.A. HASSABALLA, M.A. KHALIFA,
1978b Relative importance of effects in the inheritance of:
II. Yield and yield components in maize. Egypt J. Agron. 3:
213-225.
MATHER K., J.L. JINKS, 1971 Introduction to biometrical genetics.
Cornell University Press. Ithaca, NY. pp. 382.
MATHER K., J.L. JINKS, 1982 Biometrical Genetics. 3rd ed. Chapman and Hall Ltd, London. UK.
MOUSA H.E., 2004 Comparison between two types of testers for
evaluating new white inbred lines of maize. Annals Agri. Sci.
Moshtohor 42: 475-487.
PEARCE R.B., J.J. MOCK, T.B. BAILY, 1975 Rapid method of estimating leaf area per plant in maize. Crop Sci. 15: 691-694.
RAVIKANT R. PRASAD, CHANDRAKANT, 2006 Gene effects for metric
traits in quality protein maize (QPM) (Zea mays L.). Crop Improvement 33: 94-101.
ROWE K.E., W.L. ALEXANDER, 1980 Coputations for estimating the
genetic parameters in joint-scaling test. Crop Sci. 20: 109-110.
RUSSELL W.A., A.R. HALLAUER, 1980 Corn. pp. 299-312. In: W.R.
Fehr, H.H. Hadley (Eds.), Hybridization of crop plants. Am.
Soc. Agron. Crop Sci., Madison.
SALEEM M., K. SHAHZD, M. JAVID, A. AHMAD, 2002 Genetic analysis
for various quantitative traits in maize (Zea mays L.) inbred
lines. Intl. J. Agric. Biol. 4: 379-382.
SINGH P.K., A.K. ROY, 2007 Diallel analysis of inbred lines in
maize (Zee mays L.). Intl. J. Agri. Sci. 3: 213-216.
GUZMAN P.S., K.R. LAMKEY, 2000 Effective population size and
genetic variability in the BS11 maize population. Crop Sci.
40: 338-346.
SNEDECOR G.W., W.G. COHCRAN, 1989 Statistical methods. Eight
edition. Iowa State University Press. Ames, IA.
HALLAUER A.R., J.B. MIRANDA, 1988 Quantitative genetics in maize
breeding. 2nd ed. Iowa State Univ. Press. Ames. IA, USA.
SUJIPRIHATI S., G.B. SALEH, E.S. ALI, 2003 Heritability, performance and correlation studies on single cross hybrids of tropical maize. Asian J. Plant Sci. 2: 51-57.
HAYMAN B.I., 1958 The separation of epistatic from additive and
dominance variation in generation means. Heredity 12: 371390.
HAYMAN B.I., 1960 The separation of epistatic from additive and
dominance variation in generation means. II. Genetics 31:
331-146.
HOLTHAUS J.F., J.B. HOLLAND, P.J. WHITE, K.J. FREY, 1996 Inheritance of beta glucon content of oat grain. Crop Sci. 36: 567572.
TABASSUM M.I., M. SALEEM, 2005 Genetic trend of maize (Zee
mays L.) under normal and water stress conditions. Pak. J.
Boil. Sci. 8: 571-580.
TOLLENAAR M., A. AHMADZADEH, E.A. LEE, 2004 Physiological basis
of heterosis for grain yield in maize. Crop Sci. 44: 2086-2094.
UNAY A., H. BASAL, C. KNONAK, 2004 Inheritance of grain yield in
a half-diallel maize population. Turk J. Agric. 28: 239-244.
User’s guide to MSTAT-C. 1991. Michigan State University, USA.