Testcross Performance and Heterosis of Selected Quality Protein Maize (QPM) (Zea mays L.) Inbred Lines
One hundred thirty four genotypes were evaluated applying simple lattice design during 2007 cropping season. In this study BHQ542 maize variety and other checks were used at two locations. The study objective was to determine performance of the test crosses, heterosis and heterotic patterns of the inbred lines and testers and to identify promising crosses for future advanced trial. All agronomic and physiological data were recorded and analyzed. There was significant differences among genotypes for most of the traits with few exceptions. Based on the mean value of grain yield, sixteen crosses at Awassa and only four crosses at Melkassa were superior to QPM hybrid (BHQ542) but no crosses out yielded this check in the combined data. However, five lines crossed with T1 and seven lines crossed with T2 gave superior yield than the second best common check BH540 in the combined data. In this study crosses with T2 were superior to that of crosses with T1. Heterosis value varied highly for most of the traits studied at individual location and across locations. The highest heterosis over mid and better parent for grain yield was 189.47% (L36xT2) and 100.8% (L36xT2) at Awassa and 172.22% (L18xT2) and 62.65% (L35xT2) at Melkassa, respectively. Among the twenty maximum heterosis of grain yield only L5, L10, L15, L16, L18, L23, L28, L35, L36, L37 and L41 were common for both locations.
Received: December 22, 2013;
Accepted: February 03, 2014;
Published: April 26, 2014
Maize is one of the most important cereals after wheat and rice and is widely
distributed in the world. Maize originated in Central America and introduced
to West Africa in the early 1500s by the Portuguese traders (Dowswell
et al., 1996) and it was introduced to Ethiopia (Huffnagel,
1961) during the 1600-1700s (Worku et al., 2001b).
Its production covers 1.3 million ha in the country.
In Ethiopia, maize grows from low lands to high lands (Mulatu
et al., 1993). Mulatu et al. (1993)
noted that maize research in Ethiopia was started half a century ago. Since,
the start of maize research more that 35 maize varieties have been released
through research for different agro-ecologies of the country (Pixley,
2003). However from these, the share of quality protein maize was 7, 283
ha (Krivanek et al., 2007). The quality aspect
(specially the quality protein) of this crop has got attention just few years
ago, even on the world level (CIMMYT, 2000).
Quality protein maize was discovered by Purdue University scientists in 1961
(Krivanek et al., 2007). After repeated research
on quality protein to improve the associated defects Hussain
et al. (2006), many breeding programs worldwide began converting
normal maize endosperm populations and inbred lines to o2 versions through a
direct backcross approach (Prasanna et al., 2001).
After conversion quality protein maize tasted like normal maize and yielded
as much and more the normal maize while containing nearly twice the quantity
of the essential amino acids (CIMMYT, 2000). Similarly,
Tulu et al. (2001) reported that the quality
protein maize hybrids have the potential to out yield their normal maize counterparts
and disprove the common belief that quality protein maize is low yielding than
Taking into account the importance of quality protein maize, Ethiopian Agricultural
Research Institute started research on quality protein maize just few years
ago and released the first QPM maize (BHQ542) in 2001 (Worku
et al., 2001a). This research project was initiated with the following
objects (1) Determine performances of the top crosses, (2) Estimate heterosis
and determine heterotic pattern of the 44 quality protein maize inbred lines
and to identify promising crosses for further advanced trials.
MATERIALS AND METHODS
Site description: The research was conducted during the 2007 rainy season
at two locations at Hawassa and Melkassa. The Hawassa site is located at 7°N
and 38°31E. latitude and
longitude near Hawassa city; the capitals of South Nations Nationalities People
Regional State, Ethiopia and lies at an altitude of 1700 m above sea level with
the average annual rainfall of 1100 mm and with the average annual maximum and
minimum temperatures of 27 and 12°C, respectively. The second site, Melkassa
is located in the central Rift Valley of Ethiopia, 117 km from Addis Ababa southeast
direction. It is situated at about 8°24'N latitude and 39°21'E longitude
at an elevation of 1550 m above sea level. The area receives annual rain fall
of 763 mm of which 70% comes during June to September and has an average maximum
and minimum temperature of 28.4 and 14°C, respectively.
Experimental materials: Initially 45 inbred lines and their 90 crosses
and checks were included in the experiment. The experiment whose results are
reported in this paper consists of 44 quality protein maize inbred lines, two
testers (the open pollinated variety Obatampa and the quality protein maize
single cross hybrid CML144xCML159), their 88 top-crosses and checks in one experiment
(Set I). The less vigorous lines were tested in a separate experiment (Set II)
adjacent to the main trial (Set I) to avoid shading and resource competition
effects of the crosses, testers and checks. Different checks including an already
released quality protein maize variety from their respective location were included
in the experiment. The checks were A-511, Melkassa 1, 2, 3, 4, 5, Katumani,
Acv3, Acv6, BHQ542, Gibe and BH540 at Awassa and pool 15-c7 Melkassa 2, BHQ542,
BH540 and Gibe at Melkassa. Table of Pedigree of parents and testers was not
presented in this manuscript because of large volume.
The experiment was planted in two Sets as Set I and Set II. Set I experiment
consisted of 100 genotypes (90 crosses and 10 checks). Set II experiment consisted
of 49 genotypes (45 parents, 2 testers and 2 checks). At Awassa, the trial was
planted on May 3, 2007 and at Melkassa it was planted on June 19, 2007. Each
treatment was tested on a two- row plot. Each plot consists of two rows, each
4 m length and 1.5 m width. A spacing of 75 cm between rows and 25 cm between
plants within a row was used. Plot area was, therefore, 6 m2. During
planting, two seeds per hill were sown and then thinned to recommended plants
per plot at knee height.
Two types of fertilizers, DAP and urea, were used at the rate of 100 kg ha-1
DAP at planting and 100 kg ha-1 of urea in split application of 25%
at planting and 75% at 37 days after emergence at Awassa and Melkassa. All cultural
practices were used as recommended for each location. Border rows were planted
at the two ends of each block to avoid border effects.
Data collection and analysis: The data collected were days to emergence,
days to tasseling, days to silking, days to maturity, stand count at harvest,
grain yield (kg plot-1), total biomass (kg plot-1), disease
and lodging % (1-5), 1000 seed weight, plant height (cm), ear height (cm), No.
of ears per plant, No. of seeds per plant, No. of tassel branching, weight of
grain per plant, ear length, ear diameter, No. of kernel rows per ear, No. of
kernels per ear and umber of kernels per row. All data collected were subjected
to analysis using SAS software at individual location and over locations but
in this study those showed significant were reported. Those traits which manifested
significant differences between the genotypes at either or both locations, subjected
to LxT analysis of variance using SAS software; the elaborations given by Singh
and Chaudhary (1977) were followed. The LxT analysis for data combined over
location are presented in Table 2. One degree of freedom contrasts
were used to compare the performance of crosses and lines, lines and testers
and to compare the crosses of T1 with the crosses of T2. Locations were considered
as fixed while genotypes were considered as random effect.
During analysis, the Better Parent Heterosis (BPH) and Mid Parent Heterosis
(MPH) in percent were calculated for the characters that showed significant
difference between genotypes as suggested by Falconer (1994).
Initially 45 quality protein maize lines with their 90 crosses and local checks
were used in the experiment following simple lattice design. The data of line
15 was missing from both replications at Melkassa. Therefore, the line 15 and
its crosses were discarded from all analysis and the remaining 88 top crosses
that resulted from 44 lines crossed with two testers (Obatampa and CML144/CML159)
were analyzed and presented in this study. The analysis was conducted for 20
traits of all genotypes used in the experiment (Table 1).
Crosses were compared with checks, hybrids or open pollinated varieties, which
are under production. Special emphasis was given to the comparison of the 88
LxT quality protein maize crosses with an already released three-way hybrid
(BHQ542). Those quality protein maize new crosses with better or comparable
yield as compared to the highest yielding hybrid or open pollinated varieties
check were identified. For the traits whose f-test was significant at either
or both locations, detailed LxT analysis of selected traits of pooled data was
done and presented in Table 2.
According to analysis of variance in Table 1, there was significant
differences among genotypes for days to emergence, days to tasseling, days to
silking, days to maturity, tassel branching, plant height, ear length, ear diameter,
No. of seed per plant, grain yield per plant, biomass, grain yield per hectare,
No. of kernels per row, No. of kernel rows per ear, No. of kernel per ear and
1000 seed weight at Awassa. Similarly, there was significant difference among
crosses for days to tasseling, days to silking, days to maturity, plant height,
ear length, tassel branching, biomass yield, grain yield and No. of kernels
per ear. At Melkassa, there was also significant difference among genotypes
for days to emergence, days to tasseling, days to silking, days to maturity,
tassel branching, plant height, ear height, ear length, ear diameter, disease
severity, number of seeds per plant, grain yield per plant, biomass yield, grain
yield, number of kernels per row, number of kernels row per ear, number of kernels
per ear and 1000 seed weight.
|| ANOVA of 20 traits at individual location and combined over
|*, **Significant at 5 and 1%, respectively, NS: Non significant,
df: Degree of freedom, Reps: Replication, DT: Days to tasseling, DS: Days
to silking, DM: Days to maturity, PH: Plant height, EH: Ear height, ED:
Ear diameter, YLD: Grain yield, KPE: Kernel per ear, TSW: Thousand seed
|| Linextester analysis pooled over locations for yield and
|*, **Significant at 5 and 1%, respectively, df: Degree of
freedom, GYLD: Grain yield, KPE: Kernel per ear, TSW: Thousand seed weight,
Env: Environment and NS: Non-significant
Similarly, there was significant variation among crosses for 14 traits out
of 20 at Melkassa site (Table 1). The mean squares of genotypes
at Awassa were non-significant for ear height, disease severity, lodging and
number of ears per plant. Similarly, at Melkassa non-significant difference
was observed for lodging and number of ears per plant. For all genotypes, traits
mean performance was recorded at individual location and across location.
In the analysis of variance pooled over location, all genotypes over location
was analyzed and all of the traits studied showed highly significant variation
(p<0.01) except disease severity, lodging and number of ears per plant. GenotypexEnvironment
interaction was also significant for 15 of the 20 traits including grain yield
(Table 2) and Genotypexenvironment interaction was highly
significant (p<0.01) for 15 traits except for ear height, disease severity,
lodging, No. of ears per plant and grain yield per plant. In the pooled analysis,
location also showed significant (p<0.05) and highly significant (p<0.01)
variation for some of the traits studied except for number of kernels row per
ear indicating that mean values of those traits were highly influenced by the
environment (Table 1). Hence, those genotypes, which showed
significant variation in the analysis of variance over location, were subjected
to LxT analysis and results are presented in Table 2.
The extent of variation among genotypes was ranged from 13.96 t ha-1
(L5xT2) to 2.6 t ha-1 (L38) and the extent of variation among crosses
was ranged from 13.96 t ha-1 (L5xT2) to 6.96 t ha-1 (L3xT1)
at Awassa for grain yield. The best check at Awassa was Melkassa 4 which gave
13.72 t ha-1. At Awassa those genotypes which gave comparable yield
with the best check are L5xT2, L36xT2, L37xT2 and L41xT2 and gave yield of 13.96,
13.67, 13.62 and 13.58 t ha-1, respectively. Only one cross (L5xT2)
surpassed the best check and gave yield of 13.96 t ha-1. The yield
of best check is unusual. This might be due to the 2007 growing condition, which
was more favorable than the previous seasons. The best check (BHQ542) gave yield
of 12.7 t ha-1 and 16 crosses were superior to this hybrid at Awassa.
The local check BH540 gave only yields of 10 t ha-1. Almost 72% of
crosses were superior to this hybrid which is widely produced around Awassa
with economic heterosis ranging from -29.92 (L3xT1) to 40.62 (L5xT2). At Awassa,
L5xT2, L41xT2, L36xT2 and L37xT2 manifested highest economic heterosis of 40.62,
36.78, 37.71 and 37.19%, respectively.
The genotypes which manifested comparable grain yield at Awassa condition,
still gave yield of 11.7, 9.05, 8.85 and 10.35 t ha-1 at Melkassa
condition. At Melkassa, the best check BHQ542 gave grain yield of 11.8 t ha-1,
but surpassed by only four crosses (L18xT2, L21xT2, L30xT2 and L35xT2). When
the crosses compared with the local check Melkassa 2 (which gave 5.7 t ha-1),
all the crosses surpassed it with economic heterosis ranging from 2.44% (L17xT2)
to 108.94% (L35xT2). Average location yield of crosses was 10.94 and 8.86 t
ha-1 at Awassa and Melkassa, respectively. The 2007 growing season
was unique for Melkassa where the rain usually begins in the second week of
June and ends in the first week of September with effective growing season of
about 90 days.
During the 2007 growing season, the rain extended to the end of October and
the distribution was unusual. As a result, even late maturing varieties such
as Gibe composite and BH540 could reach maturity. This led to the unusually
high yield at Melkassa, which was comparable to that of Awassa. Therefore, not
all the top yielding crosses at Awassa repeated their performance at Melkassa.
Among the 20 top yielding crosses, only ten were common for both locations.
The common ones were L5xT2, L15xT2, L16xT2, L18xT2, L23xT2, L28xT2, L35xT2,
L38xT2, L39xT2 and L41xT2. These crosses was superior not only to the already
released QPM hybrids but also to the normal maize varieties under production
at two locations and they seem to be promising future varieties and should be
|| Discriminating power of the two testers
|*,**Significant at 5 and 1%, respectively, DE: Days to emergence,
DT: Days to tasseling, DS: Days to silking, DM: Days to maturity, TB: Tassel
branching, Ph: Plant height, EH: Era height, EL: Ear length, ED: Ear diameter,
DIS: Disease, LOG: Lodging, NEPP: No. of ears per plant, NSPP: No. of seed
per plant, GYPP: Grain yield per plant, BYLD: Biomass yield, GYLD: Grain
yield, KPR: Kernel per row, RPE: Row per ear, KPE: Kernel per ear and TSW:
1000 seed weight
In the combined data, LxT analysis revealed highly significant difference (p<0.01)
for all analyzed traits of genotypes (crosses and parents) (Table
2). Similarly crosses and parents also showed highly significant difference
(p<0.01) for all traits except for days to emergence, ear height, ear diameter,
No. of seed per plant and 1000 seed weight. Furthermore, crosses versus parents
interaction was highly significant for entire traits studied in LxT analysis
indicating the probability of potential crosses would come from parental genotype.
In the correlation analysis, the rank correlations (Table 3)
between the crosses of lines with the two testers were also low for most of
the traits and were statistically non-significant, indicating that the two testers
ranked the lines in a different manner. Although, the two testers differed in
only few traits, they ranked the lines in a completely different order. Hence,
in combined data over the two locations, the f-test of nine traits were statistically
significant for the crosses of both testers while only three traits were significant
for only one tester. It is surprising to note that Tester Obatampa per se performance
was superior to CML 144/ CML 159 in most of yield components such as number
of seeds per plant, grain yield per plant, No. of kernel per row, No. of kernel
rows per ear, No. of kernels per ear and 1000 seed weight; but the crosses gave
low yield per hectare. The high biomass production coupled with higher harvest
index of 0.37 and 0.35 for lines crossed with CML144/CML enabled them to produce
higher grain yields than Obatampa crosses. Consequently, all elite crosses at
each location are those of lines crossed with T2. Maximum yield 13.67 and 12.85
t ha-1 was from lines crossed with T2 while 12.8 and 9.65 t ha-1
was from lines crossed with T1 at Awassa and Melkassa, respectively. It is only
by their average grain yields over the two testers that the best yielding crosses
of lines were determined.
At a specific location (Awassa or Melkassa), the two testers were ranked the
44 lines differently. Hence, there were big rank changes in yield between the
two crosses of a single line. The crosses of the same line with T2 was the second
highest yielding in the combined data but was 41 1st when crossed with T1. The
rank correlation between the crosses of the two testers by the 20 traits was
therefore low and statistically non-significant for the majority of the traits
at Awasssa, Melkassa and in the combined data (Table 3).
Mean comparisons: The interaction of genotypes with the environments
and similarly the interaction of its components (crosses, parents and crosses
versus parents) was significant for all traits but for ear height (Table
2). The crossesxenvironment interaction was significant for most of the
traits but for days to emergence, ear height, ear diameter, No. of seeds per
plant, No. of kernel row per ear and 100 seed weight. The crossesxparentsxenvironment
component was non-significant for days to emergence, days to tasseling, ear
height, No. of kernel per row and No. of kernels per ear. The linesxenvironment
component were significant for all the 20 traits studied indicating the inconsistence
of performance of the lines over the environments. The linesx testersxcomponent
was non-significant for almost all traits except for days to tasseling, days
to silking, days to maturity, tassel branching, biomass and grain yield. The
difference between lines and testers was little affected by the change of the
The trend of genotypic variability of days to tasseling and days to silking
followed that of days to maturity, while the trend of other quantitative traits
followed that of the grain yield and 1000 seed weight (Fig. 1).
Wide variability among genotypes (lines, testers and crosses) was observed in
the mean comparisons. When the mean performance of genotypes (Crosses, lines
and testers) compared, at Awassa (Fig. 1), the mean values
of testers exceeded mean values of lines for different traits, which revealed
significant difference in the preliminary analysis at individual location and
across locations. The variation between mean values of yield of lines and testers
was detected and such variation is supporting crossing of lines with testers,
which were used as genotypes in this experiment. Moreover, the analysis of variance
of LxT analysis showed similar trend of variation (Fig. 1).
On the other hand at Melkassa, the mean comparison revealed that testers out
yielded lines and showed superior performance in some traits studied (Fig.
2). In this location great variation was observed for the traits such as
plant height, ear height, No. of seeds per plant, biomass yield, grain yield,
No. of kernels per ear and 1000 seed weight (Fig. 2). The
outstanding performance of testers was similar across locations for the trait,
which showed significant variation. In similar way, LxT analysis confirmed the
above mean comparisons of lines and testers so that the top crossing was made
and it was promising for the presence of heterosis of grain yield.
In the mean comparison of traits pooled over location, similar results were
observed for lines versus testers and crosses versus lines except No. of kernel
per row and No. of kernel per ear (Fig. 3). In this case,
the two traits of tester gave mean value of 37.75 and 694.25 while the same
traits in crosses observed to give the mean value of 36.43 and 606.08. Though
the testers mean value was greater for the above traits, it does not mean that
it yields too. In the case of yield and other yield related traits, crosses
surpassed the testers (Fig. 3). Thus, crosses were earlier
for days to tasseling, days to silking and days to maturity.
|| Comparisons at Awassa
|| Comparisons at Melkassa
|| Comparisons of pooled data
On the other hand, lines gave lower mean values of pant height, ear height,
kernel per plant, biomass yield, grain yield, kernel per row, kernel per ear
and 1000 seed weight than testers and crosses. This shows that the diverse variability
of genotypes during crossing brought the out-standing crosses with higher mean
values of grain yield and yield related traits. The mean square due to lines
was statistically significant for all the studied characters (Table
2). For example in the combined data, days to maturity ranged from 143-156
days, tassel branching from 11-24 branches, ear length from 10.4-16.1 cm, No.
of seeds per plant from 214-566, grain yield t ha-1 from 2.3-7.11
t ha-1, 1000 seed weight from 205-366 g. Some lines had 1000 seed
weight comparable to that of the crosses. The crosses also compared with testers
in the combined data. The crosses were taller than testers and produced more
biomass and grain yield than testers. It can be generalized that the maturity
trait manifested negative over-dominance for some crosses. As result, there
could be a possibility to identity the promising crosses based on yield performance,
heterosis and combing ability (not presented).
Heterosis: The heterosis of the F1 (the crosses) over the mid parents
and the better parents was computed and the results were presented in the Table
5 and discussed here to some extent for some characters in combined analysis.
For maturity traits such as days to tasseling, silking and days to maturity,
the better parent was the one, which was earlier from the parents. For other
traits, the better parent was the one with the higher value of the trait. The
deviations of the F1 mean from mid parents either in the positive or in the
negative direction was an indication of the presence of dominant gene action.
All crosses tasseled and matured earlier than the earliest parent, which was
the tester. The lines were flowered later than crosses and testers. At Awassa,
traits like days to tasseling, days to maturity, plant height, ear length, biomass
yield, grain yield, number of kernel per ear and number of kernel row per ear
showed significant heterosis over mid and better parents (table not presented
in this manuscript). In the analysis of heterosis, 98.8 and 81.8% of crosses
manifested positive and significant heterosis over mid and better parents, respectively
for grain yield indicated that those crosses have great potential for future
Regarding heterosis of lines crossed with testers, maximum negative heterosis
obtained on average from inbred line crossed with T1 (Obatampa). The range of
maturity of the lines was from 140-160 days. T1 matured in 150 days while T2
matured in 143 days. The range of maturity for the crosses (130-141 days) was
narrower than that of lines. Although, lines crossed with T1 were in general
earlier than that of lines crossed with T2, crosses with the highest better
parent heterosis were obtained from Top crosses of lines crossed with T2 for
grain yield and other agronomic traits except maturity traits. Better Parent
Heterosis (BPH) was computed by comparing earliness of the crosses with earliness
of the testers except in the 14 lines, which were earlier than the earliest
tester. In most cases, therefore, BPH for lines crossed with T2 was higher than
that of lines crossed with T1. The grain-filling period was longer for lines
crossed T2 than T1 and it matured about seven days later than T1 (143 vs. 150
From the inbred lines those crossed with T2 (tester two) (CML144/159) showed
higher performance and heterosis over mid parent and better parent for most
of the traits studied. Hence, regarding the most important economic trait (grain
yield), L36xT2, L38xT2, L5xT2, L26xT2 and L41xT2 manifested maximum heterosis
over mid parent and better parent at Awassa.
For grain yield, L36xT2 and L9xT2 expressed maximum and minimum heterosis of
189.47 and 45.21%, respectively over mid parents at Awassa. Referring inbred
line crossed with T1, L38xT1 gave positive and significant heterosis of 131.9
and 51.47% over mid and better parents. From the crosses with T1, 98.86% manifested
positive and significant heterosis over mid parent. Lines crossed with T1, showed
low heterosis compared with crosses of T2. Thus, lines crossed with T1 and T2
expressed significant heterosis ranging from 131.9% (L38xT1)-32.63% (L14xT1)
and from 189.47% (L36xT2)-54.4% (L4xT2), respectively over mid parents for grain
yield. On the other hand, lines crossed with T1 manifested positive and significant
better parent heterosis ranging from 51.47% (L38xT1)-25.26% (L31xT1), while
lines crossed with T2 had heterosis that ranged from 114.15% (L5xT2)-17.09%
(L9xT2). The cross L5xT2 gave the highest BPH of 7.4 t ha-1 for grain
yield. Moreover, the top yielder at Awassa L36xT2 expressed maximum heterosis
of 189.47 and 109.71% over mid and better parent, respectively.
Referring heterosis of grain yield at Melkassa, most of the lines crossed with
T1 showed no heterosis (over mid and better parent) but lines crossed with T2
revealed maximum and positive heterosis over mid and better parent. At this
site in general, the better parent performance was low for yield and yield related
traits. Here, L35xT2 gave the maximum better parent heterosis of 62.65% and
it was one of the best lines crossed with T2, which gave high yield. At Melkassa,
of all crosses studied, L18xT2 and L17xT2 gave maximum and minimum mid-parent
heterosis of 172.22 and -25.88%, respectively indicated that maximum and minimum
heterosis for grain yield was obtained from inbred lines crossed with T2.
According to the analysis of better parent heterosis, twenty lines crossed
with T2 were identified and marked in the table that can give the highest yield
in the combination.
|| Heterotic combinations of QPM inbred lines crossed with T1
|Category A: T1 (Obatampa), Category B: T2 (CML144xCML159),
Category C: Common well combiner to both testers
These combinations consists of some groups which can be suggested for immediate
use after a verification trail and others which requires further recurrent selection
for identifying of high yielding varieties.
Yield related traits, such as No. of kernels per ear, No. of kernels per row
and 1000 seed weight exhibited significant heterosis of 42.04, 36.33 and 6.82%
over mid and better parents, respectively. This variation found to be small
but can be an indication of importance for selection of grain yield in addition
to other important criterion for selection. This is referred as indirect selection.
For the physiological traits such as plant height and ear length 53.41 and
50.0% of crosses expressed positive and significant heterosis over mid parent,
respectively while none of crosses showed heterosis over better parent at Awassa
and for ear diameter, 69.32% of crosses over mid parent and 10.23% of crosses
over better parent exhibited positive and significant heterosis. At Melkassa
for plant height, ear length and ear diameter, 82.95, 39.77 and 12.55% of crosses
exhibited heterosis over mid parents revealing the dominance effects in their
inheritance, while no better parent heterosis was observed.
Heterotic patterns: Based on gca and sca effects (data not presented
here) and heterosis of mid and better parents across locations the best combination
patterns between testers and selected quality protein inbred lines was identified
and the results of some traits are presented in the Table 4.
In the analysis, almost all inbred lines crossed with T2 gave more or less higher
mean yield (Table 2), showed maximum heterosis over mid parent
and better parent and similar results were obserbed in the combined heterosis
analysis (Table 5). Thus, based on combined heterosis over
locations, 95.45 and 84.09% of inbred lines with T2 showed the presence of exploitable
heterosis over mid and better parents for grain yield, respectively. Those traits,
which showed significant heterosis across location, are presented in Table
4. Hence, those combinations, which showed exploitable positive heterosis
at individual location and across locations, performed consistently for yield
and other yield related traits in 2007 growing season. In Table
4 the best heterotic combinations are presented and could be used for further
The aim of this study was to see the performance of test crosses of QPM lines
with testers and to determine the heterosis and heterotic groups. In both locations
most of the crosses out yielded the testers and lines. Sixteen crosses were
superior to BHQ542 and 72% of the crosses was superior to BH540 at Hawassa where
as at Melkassa only four crosses surpassed the best check BHQ542 and all crosses
surpassed the local check Melkassa 2. Worku et al.
(1996) and Tulu et al. (2001) reported that
QPM hybrids produced higher or comparable grain yield as the normal maize check
BH540 cited by Worku et al. (2001a). In the comparisons
of QPM maize with the commercialized normal maize at Pakistan, four QPM hybrids
out yielded the top commercial maize varieties (Hussain et
|| Mid and better parent heterosis of yield related traits combined
|*,**Significant at 5 and 1%, respectively, BYLD: Biomass yield,
GYLD: Grain yield, KPE: Kernel per ear, KPR: Kernel per row and TSW: 1000
In our case, the normal maize lines crossed with the tester (which is called
the lysine and tryptophan donors) showed exiting performance in the two location
as compared to that in the Pakistan. This indicates the potential of gaining
many QPM maize commercial varieties to Ethiopian Farmers through crossing the
parent lines of top performing commercial maize varieties with the two testers
that has been used in this experiment. Taking into account the importance of
QPM maize for food security and the number of malnourished people in the developing
country, maize scientists has been convinced to go for quality maize research
since ninth (Hussain et al., 2006). In this study
most of the crosses in both locations showed significant performance and heterosis
as well. Strongly in line with this result Rodrigues and
Chaves (2002) reported that the significance of varietal performance of
crosses or QPM maize coupled with its significant average heterosis can be used
as the best indicator of heterotic pattern. The dominant gene action was reported
by different studies for the presence of heterosis but the genetic base of heterosis
was not clear so far. However, the report of Yao et
al. (2013) suggest that a major component of heterosis is a mechanism
that is modulated by dosage-sensitive factors that involves allelic diversity.
The interaction of genotypes with the environments and similarly the interaction
of its components (crosses, parents and crosses vs. parents) was significant
for all traits but for ear height (Table 3). Similar result
was reported by Rodrigues et al. (2006) for
crosses versus parents interaction across location. The crossesxlocation
interaction was significant for most of the traits but for days to emergence,
ear height, ear diameter, number of seeds per plant, number of kernel row per
ear and 100 seed weight. This is in agreement with the findings of Hussain
et al. (2006) who reported the interaction of crossesxlocation which
was statistically highly significant reflecting the importance of GxE.
Heterosis: According to the analysis of heterosis , 68.2 and 43.18%
of crosses over mid parents and better parents, respectively showed significant
heterosis for days to tasseling at Awassa in the opposite direction. Similarly,
for days to maturity all crosses manifested highly significant heterosis over
mid and better parents in the negative direction. In general, the crosses were
earlier than the earliest parent, which was in most case the tester and more
vigorous than both parents possessing high values for the quantitative traits
measured. Almost all of the crosses had negative mid parent and better parent
heterosis for the three maturity traits such as days to tasseling, days to silking
and days to maturity. For maturity trait, L17xT2 and L7xT1 exhibited the highest
and lowest heterosis of 3.7 and 14.4% over mid and better parents, respectively.
At Awassa, maximum heterosis of 189.47% (L36xT2) was obtained and followed
by 189.09% (L15xT2) and 180.25% (L38xT2) over mid parent. These crosses also
showed better parent heterosis of 109.71, 100.72 and 94.26%, respectively. Similar
to the present findings, Hussain et al. (2006)
obtained four quality protein maize hybrids which are with better and comparable
performance to best commercial check. In addition Saleh
et al. (2002) reported superior and comparable performance of quality
protein maize varieties for grain yield and for some yield related characters.
For maturity, yield and quality characters Netravati et
al. (2013) reported the presence of heterosis over better parent and
standard check which is in line with the findings in this research.
Different studies reports the role of gene frequency for the expression of
heterosis of grain yield. Gardner and Eberhart (1966)
suggested that mid parent heterosis depends on the existence of dominance effects
and of gene frequency differences among population considered. Similarly, better
parent heterosis depends on the existence of differences among gene frequencies
of a variety in relation to the mean gene frequency of the parents in the same
group. Thus, a significant mid-parent heterosis effect indicates divergence
among the gene frequencies of the population for the loci displaying dominance.
Thus, a recurrent selection program lead to an increase in the heterosis among
these populations if there would be an increase in gene frequency divergence.
However, if such an increase is the same, both in direction and magnitude, in
the two populations, the heterosis will remain unchanged, as stated by Hallauer
and FIilho (1981).
Based on the heterosis and other parameter identifying the heterotic pattern,
heterotic combinations were reported for maximum exploitation of yielding potential
of a given crop. Similar to the heterotic pattern report in this study, several
authors have reported heterotic patterns used in the major maize production
regions of the world (Wellhausen, 1978; Parra
and Hallauer, 1997). In this research, line per se performance was compared
with its respective testcross and similar result was obtained to that of Kebede
et al. (2012). Acceptable level of protein quality in the QPM is
very important since the protein content in QPM is affected by heterotic pattern
(Ikramullah et al., 2011); indicating identification
of heterotic pattern as one step forward in the improvement of QPM varieties.
I would like to acknowledge the NORAD project in the University of Hawassa
for providing me financial support for the research study at Hawassa site and
Melkassa research center for financial support to the trails at Melkassa site.
I am also grateful to staff of Plant Science Department of Hawassa College of
Agriculture for their co-operation and material support during field works in
their farm. I wish to extend my heartfelt thanks to my supervisors Hussien Mohammed
(PhD) and Mandefro Nigussie(PhD), for their valuable and constructive comments
during the research and in the study work. I wish to add thanks to Dr. Mandefro
for providing me his own research materials (QPM inbred Lines and Testers) for
CIMMYT annual report 1999-2000: Science and sustenance. International Maize and Wheat Improvement Center (CIMMYT), Mexico City, Mexico, pp: 1-72. http://repository.cimmyt.org/xmlui/handle/10883/630.
Dowswell, C.R., R.L. Paliwal and R.P. Cantrell, 1996.
Maize in the Third World. West View Press, Colorado, USA., ISBN-13: 9780813389639, Pages: 268
Falconer, D.S., 1994.
Introduction to Quantitative Genetics. 3rd Edn., Longman, London
Gardner, C.O. and S.A. Eberhart, 1966.
Analysis and interpretation of the variety cross diallel and related populations. Biometrics, 22: 439-452.PubMed | Direct Link |
Huffnagel, H.P., 1961.
Agriculture in Ethiopia. Food and Agriculture Organization of the United Nation Organization, Rome, Pages: 465
Hallauer, A.R. and J.B.M. Filho, 1981.
Quantitative Genetics in Maize Breeding. 1st Edn., Iowa State University Press, Ames, USA., Pages: 468
Hussain, M., S.R. Chughtai, H.I. Javed, H.N. Malik and M.H. Munawwar, 2006.
Performance of locally constituted quality protein maize hybrids-a fortune for malnourished people and feed industry in Pakistan. Asian J. Plant Sci., 5: 385-389.CrossRef | Direct Link |
Ikramullah, I.H. Khalil, M. Noor and M.K.N. Shah, 2011.
Heterotic effects for yield and protein content in white quality protein maize. Sarhad J. Agric., 27: 403-409.Direct Link |
Kebede, A.Z., A.E. Melchinger, J.E. Cairns, J.L. Araus, D. Makumbi and G.N. Atlin, 2012.
Relationship of line per se and testcross performance for grain yield of tropical maize in drought and well-watered trials. Crop Sci., 53: 1228-1236.CrossRef | Direct Link |
Krivanek, A.F., H. De Groote, N.S. Gunaratna, A.O. Diallo and D. Friesen, 2007.
Breeding and disseminating Quality Protein Maize (QPM) for Africa. Afr. J. Biotechnol., 6: 312-324.Direct Link |
Worku, M., J. Abdurehman, L. Tulu, H. Tuna and L. Wolde et al
Development of improved maize germplasm for the mid and low altitude sub-humid agro-ecologies of Ethiopia. Proceedings of the 2nd National Maize Workshop of Ethiopia, November 12-16, 2001, Addis Ababa, Ethiopia, pp: 27-30
Mulatu, K., G. Bogale, B. Tolessa, M. Worku, Y. Dessalegne and A. Afeta, 1993.
Maize production trends and research in Ethiopia. Proceedings of the 1st National Maize Workshop of Ethiopia, May 5-7, 1992, IAR/CIMMYT, Addis Ababa, Ethiopia, pp: 4-12
Netravati, H., G. Shantakumar, S. Adiger, L. Malkannavar and P. Gangashetty, 2013.
Heterosis breeding for maturity, yield and quality characters in maize (Zea mays
L.). Mol. Plant Breed., 4: 44-49.CrossRef |
Parra, J.R. and A.R. Hallauer, 1997.
Utilization of exotic maize germplasm. Plant Breed. Rev., 14: 165-187.
Pixley, K.V., 2003.
The development and promotion of quality protein maize in sub-Saharan Africa: Progress report 2003. International Maize and Wheat Improvement Center (CIMMYT), Harare (Zimbabwe), pp: 56. http://ibsa.mx:8080/xmlui/handle/10883/787.
Prasanna, B.M., S.K. Vasal, B. Kassahun and N.N. Singh, 2001.
Quality protein maize. Curr. Sci. Assoc., 81: 1308-1319.Direct Link |
Rodrigues, M.C., L.J. Chaves and C.A.P. Pacheco, 2006.
Heterosis in crosses among white grain maize populations with high quality protein. Pesquisa Agropecuaria Brasileira, 41: 59-66.CrossRef | Direct Link |
Saleh, G.B., D. Abdullah and A.R. Anuar, 2002.
Performance, heterosis and heritability in selected tropical maize single, double and three-way cross hybrids. J. Agric. Sci., 138: 21-28.CrossRef | Direct Link |
Singh, R.K. and B.D. Chaudhary, 1977.
Biometrical Methods in Quantitative Genetic Analysis. 3rd Edn., Kalyani Publishers, New Delhi, India, ISBN: 9788176633079, Pages: 288
Tulu, L., M. Worku, G. Seboksa, J. Abdurehman, H. Tuna, S. Tariku and S. Twumasi-Afriyie, 2001.
Quality protein maize research in Ethiopia. Proceedings of the 2nd National Maize Workshop of Ethiopia, November 12-16, 2001, Addis Ababa, Ethiopia, pp: 39-44
Wellhausen, E.J., 1978.
Recent Developments in Maize Breeding in the Tropics. In: Maize Breeding in the Tropics, Walden, D.B. (Ed.). John Wiley & Sons, New York, pp: 59-91
Worku, M., H. Tuna, M.Nigussie, A. Deressa, D. Tanner and S. Twumasi-Afriyie, 2001.
Maize production trends and research in Ethiopia. Proceedings of the 2nd National Maize Workshop of Ethiopia, November 12-16, 2001, Addis Ababa, Ethiopia, pp: 10-14
Worku, M., L. Wolde, B. Tolessa, K. Mulatu and L. Tula, 1996.
Heterotic patterns of some intermediate maturing maize germplasm. Afr. Crop Sci. J., 4: 497-501.Direct Link |
Yao, H., A.D. Gray, D.L. Auger and J.A. Birchler, 2013.
Genomic dosage effects on heterosis in triploid maize. Proc. Natl. Acad. Sci., 110: 2665-2669.Direct Link |
Rodrigues, M.C. and L.J. Chaves, 2002.
Heterosis and its components in crosses among high quality protein maize populations. Crop Breed. Applied Biotechnol., 2: 281-290.Direct Link |