Volume 6, Issue 4, April – 2021 International Journal of Innovative Science and Research Technology

ISSN No:-2456-2165

Genotype by Environment Interaction and Yield

Stability of Hybrid Maize Varieties Evaluated in
Three Locations of Mid Altitudes of Rwanda
Rwasimitana F.1*, Ukozehasi C1, Ngaboyisonga C.2, Eva J. 3
1
*University of Rwanda; College of Agriculture Animal sciences and Veterinary medicine(UR-CAVM)
2
Rwanda Agriculture and Animal Resources Development Board(RAB)
3
Swedish University of agriculture(SLU)

Abstract:- A multi-environment yield trial is important I. INTRODUCTION

to understand the genotype by environment interaction
and to select high performing and stable crop varieties. Agriculture is a major pillar of the Rwandan economy.
The aim of this study was to identify high yielding and It provides direct employment to more than 70% of
stable hybrid maize varieties for mid altitudes of Rwandan people, many of whom are small-scale farmers
Rwanda, to compare the performance of new hybrid living in rural areas and lacking adequate resources. The
varieties with commercial checks, and to determine the agricultural sector still contributes about 33 % of the
extent of genotype by environment interaction. Maize is national Growth Domestic Product (NISR, 2016). However,
a staple crop used to fight hunger and malnutrition in a main challenge is the current productivity gap, with a
developing countries. Different varieties have been difference in yield of 50% to 70% between farmers’ fields
released to increase yield including Open Pollinated and researchers’ plots for staple crops such as maize (2.2
Varieties (OPVs) and hybrids. Genotype by t/ha versus 7 t/ha on average) and soybean (0.8 t/ha versus
Environment interaction is an issue that all breeding 2.5 t/ha (FAO, 2016). The low agricultural productivity in
program need to overcome. In the future, improved Rwanda is a result of several constraints. Rwandan
varieties will be needed in order to increase income for agriculture is rain-fed, and hence more exposed to weather-
farmers and help in food security Field experiments related risks, especially to severe, frequent, and prolonged
were conducted to assess the performance and the drought periods. Declining in soil fertility, mainly due to the
stability of 27 maize varieties in the mid altitudes zone of lack of nutrient recovery causes important losses in
Rwanda in the Cyabayaga, Rubona and Bugarama sites. agricultural yields. The land holding is extremely small
The experimental design was alpha lattice (0,1) with a ranging from 0.5 ha to 1.1 ha leading to the high percentage
Randomized Complete Block Design (RCBD). Data were of small-scale subsistence farming. This affects the
collected for a number of characters i.e. silking, Antesis- economy at both household and national levels farmers. The
Silking Interval (ASI), plant height, plant aspect, ear per small land size per household is insufficient for supporting
plant, husk cover, ear aspect, ear rot and grain yield. household food needs throughout the year and for providing
Data were analyzed by GenS Stat statistical computer income-generating activities to its members. This, in turn,
package, Discovery Edition. ANOVA and AMMI leads to large imports of staple crops (World Bank and
analysis were applied to assess the performance and the CIAT, 2015) The use of agricultural inputs is very low. The
stability of varieties and the degree of genotype by use of improved seed ranges from 1% to 25% while mean
environment interaction (G×E). In addition, Principal fertilizer application is about 21% depending on the crop.
Component Analysis (PCA) and cluster analysis were The formal seed systems and value chains in Rwanda is still
conducted to assess relationships between varieties. The dominated by the public organizations for production,
results showed that RHM1706, RHMM1701, RHM1409, distribution and/or commercialization. The parallel informal
RHMM1707, WH509, RHMM1704, RHM407, WH101, seed systems supply 75% to 99% of the seed depending on
RHMM1710, RHMM1708, PAN53 and RHM104 were the crop commodity. The seed supplied through informal
stable across locations. Furthermore, the evaluated system is of poor quality and come predominantly from
varieties were found to cluster into five groups. Varieties farmer saved seed or is sourced from local markets (Promar,
found to be most stable are recommended for further 2012).
use.
Maize (Zea Mays) is a major cereal crop worldwide. It
Keywords:- Open Pollinated Varieties, Genotype By is used as food for human and feed for animals and
Environments, Hybrid Varieties. constitutes raw materials for food and non-food industries
(Khalil et al., 2011). It is a leading crop in Rwandan
agriculture and it is used by many farmers for food and feed
(Sallah et al., 2009). In 2007, the Government of Rwanda
(GoR) initiated the Crop Intensification Program (CIP) with
the aim of ensuring food security and improving livelihoods

IJISRT21APR259 www.ijisrt.com 184

Volume 6, Issue 4, April – 2021 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165
of small scale rural families. Within this program, maize (Ngaboyisonga et al., 2016) However, at present, some of
was together with seven other crops, i.e. rice, wheat, the available released varieties are limited to cultivation
cassava, potato, beans, soybean, cooking and dissert banana, within certain environments (Ngaboyisonga et al., 2014)
chosen as the ones mainly contributing to food security
(Minagri, 2010). Maize has an important role in the fight II. MATERIALS AND METHODS
against hunger in developing countries. Rwanda contributes
in this fight through a desire to increase the maize 2.1. Location of the study and Maize varieties used
production from 125,000MT up to 650,000MT. For success, This study was conducted in three sites of the mid
the area of cultivation needs to be increased. Also, improved altitudes of Rwanda that comprised Bugarama, Rubona and
varieties will be needed to boost yield through increased Cyabayaga sites. Rubona is located at the altitude of 1,692
resistance towards diseases, drought and lodging. The use of masl in the southern province. It has a bimodal rainfall with
hybrids may be preferential because of its high yielding 1,170 mm/year of precipitation. Foliar disease is not severe
capacity compared to open pollinated varieties (Kutka, and drought is occasionally occurring in this location. The
2011). Genotype by environment is of great concern in Cyabayaga site is located at the altitude of 1,370 masl. It has
order to maintain productivity of maize in different areas. a bimodal rainfall with 850 mm/year of precipitation. It is
The performance of varieties in terms of yield, disease vulnerable of Turcicum Leaf Blight (TLB), Grey Spot
resistance and drought tolerance need to be evaluated in (GLS) and there is frequent drought appearance compared to
different locations (Fatma, 2011). In the past years’ maize other site diseases. Bugarama is located at the altitude of
was grown in high altitudes in Rwanda, but nowadays it is 968 masl. It has a bimodal rainfall with 1,000 mm/year of
grown all around the country due to the development of new precipitation. It is a hot spot of Maize streak virus (MSV)
released varieties such as hybrids, open pollinated varieties and Drought happens occasionally (Table 2.1), Twenty-
and commercial (Ngaboyisonga et al., 2016). There is high seven maize varieties were used in the experimental trials.
demand of maize seed for the increase of production through They comprised 13 new Rwanda Agriculture Board (RAB)
the development and use of high yielding varieties. In hybrid varieties, six RAB pre-released hybrid varieties, six
Rwanda, the use of improved varieties will help in commercial hybrid varieties from regional seed companies
increasing yield. Major constraints are present that lead to a were used as checks together with two OPVs from RAB.
decrease in production of maize e.g.: drought, diseases and Among the six commercial checks, two were from Seed Co
less availability of improved varieties in the whole country Ltd, two from Western Seed Company Ltd (WSC) and two
(Mulinga, 2014), new policies like the crop intensification from PANNAR Ltd. All RAB hybrid varieties were Three
program(CIP) will contribute to the release of new varieties Way Cross Hybrid (TWCH) varieties (Table 2.2)
and will also make these varieties available to the farmers

2.2. Data Collection and Experimental design Two grains were planted per hill and after two weeks a
The experimental design for this study was an Alpha thinning was done to one plant/hill. Fertilizers were applied
lattice (4x6) with four replications, but it was analyzed as a both at planting with NPK and thereafter a top dressing with
Randomized Complete Block Design with four replications. urea was applied. The experiment was dependent on rain as
Every plot comprised of two rows of 5-m length, distances the main supplier of water. The first weeding was done after
within the rows were 0.75 m and between hills was 0.25 m.

IJISRT21APR259 www.ijisrt.com 185

Volume 6, Issue 4, April – 2021 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165
one week and was thereafter continued throughout the RHMM113, RHMM1706, RHM1407 and RHMM1705,
growing season when it was needed. reached silking inferior to 69 days on average, fourteen
varieties reached silking between 70 days and 75 days
This study was targeting high yielding varieties and whereas only two varieties, RHT132 and RHM150, reached
grain yield was the important trait to record. Yield was silking superior to 75 days. Moreover, all the varieties had
recorded in all plots by measuring the total number of ears ASI inferior to 2.5 days on average (Table 4.5).
harvested in a plot and obtaining the fresh weight in kg
(FW) after that a sample of ears was taken in the middle of The results indicated significant differences between
ten selected ears from each plot and this sample was used to varieties for plant height at P<0.001. The plant height varied
check the grain moisture content (GM) in % using a portable from 1.78 m (RHMM1709) to 2.60 m (WH509) with an
moisture-meter. Grain yield (GY) in t/ha at 15% of grain average of 2.27 m. The varieties RHMM170, RHMM1706,
moisture was calculated using the formula below, where A RHMM1707, RHMM1708, RHMM1710, RHMM1711,
is the distance (in m) between rows and B the distance (in RHMM1712, RHMM150, RHM104, RHM1402,
m) between hills at planting, C the length (in m) of RHM1407, RHM1409, RHMM113, RHT132, SC529,
harvested rows, and D the number of rows harvested: WH101, ISARM101, PAN-4m-21, SC403, Kigega and
FW 100- GM PAN53 had plant heights between 2.0 m and 2.5 m
GY = 10× × × 0.8 .
A × (B + C)× D 100- 15 respectively (Table 4.5). The analysis of variance showed
that differences between varieties for plant aspect were
significant at P=0.006. The plant aspect varied from 2.5
Other agronomic parameters collected included silking
(PAN53 and RHMM1704) to 3.2 (RHMM1709 and Kigega)
(d), the Anthesis to Silking Interval (ASI) (d), plant height
with an average of 2.86. The varieties RHMM1702,
(m), plant and ear aspects (scale of 1 to 5). Silking was
RHMM1705, RHMM1706, RHMM1707, RHMM1708,
recorded as observed in days after planting, where the
RHMM1710, RHMM1712, RHM104, RHM1402,
numbers of days were noted when up to a half of the plants
RHM1407, RHM1409, RHT132, PAN-4m-21, WH509,
in the plot showed silks. Thereafter ASI was calculated by
WH101, SC529 had plant aspect between 2.5 and 3.0, while
the difference between silking and Antesis. Hence, the
six varieties had plant aspect above 3.0 (Table 4.5).The
Antesis was also recorded as the number of days from
results of analysis of variance indicated significant
planting to when 50 % of the plants in the plot shed pollen.
differences between varieties for ear per plant at P<0.001.
Ten plant randomly selected to calculate average of plant
The EPP varied from 0.9 (RHMM1711) to 1.2
height in the plot from the plant base up to the point where
(RHMM1708) with an average of 1. A total of 17 varieties,
the tassel started to branch. The plant aspect was recorded as
RHMM1701, RHMM1702, RHMM1704, RHMM1705,
the brown husk stage when plants were still green and ears
RHMM1706, RHMM1709, RHMM1712, RHM1407,
fully attained development. For each plot, characteristics
RHMM113, RHT132, SC529, PAN53, ISARM101, SC403,
such as plant and ear height, uniformity of plants, diseases
Kigega and PAN-4m-21 had an average of EPP equal to 1.0.
and insect’s damages and lodging were noted. Thereafter, a
Furthermore, 7 varieties, RHMM1703, RHMM1707,
scale of “1” to “5” where “1= not a single damage, 2=some
RHMM1710, RHMM150, RHM104, RHM1402, RHM1409
damage, 3=initial damage, 4= damage, 5=severe damage
and WH101 had an average of EPP equal to 1.1 (Table
was used. The ear aspect was taken after harvest, but before
3.1).The analysis of variance showed that differences
taking a sample for moisture determination, by spreading the
between varieties for ear aspect were significant at P<0.001.
pile of ears in front of the plot and considering
The ear aspect varied from 2.4 (PAN53) to 3.4 (SC529) with
characteristics such as disease and insect damage, ear size,
an average of 3.0 Twelve varieties, RHMM1702,
grain filling, and uniformity of ears on a scale of “1” to “5”
RHMM1704, RHMM1707, RHMM1708, RHMM1710,
where “1= not a single damage, 2= some show damage,
RHMM1712, RHM1407, RHM1409, WH509, WH101,
3=initial damage, 4= damage, 5=severe damage.
PAN-4m-21and ISARM101 had ear aspect between 2.5 and
3.0. The varieties RHMM1701, RHMM1703, RHMM1705,
Data were analyzed by GenS stat computer package
RHMM1706, RHMM1709, RHMM1711, RHMM150,
and AMMI was used to analyze the G×E interaction for
RHM104, RHM1402, RHMM113, RHT132, SC529, Kigega
silking, plant height and grain yield. The means and IPCA 1
and SC403 had ear aspect equal or superior to 3.0 (Table
(Interaction Principal Component Axis) scores were used to
3.1).The analysis of variance showed significant differences
form the AMMI 1 biplot. The AMMI analysis of variance
between varieties for ear rot (P<0.001). Ear rot varied from
was performed using statistical computer package GenS stat,
1.9 (RHMM150) to 3.2 (SC529 and Kigega) with an
Discovery, 4th Edition using (Buysse et al., 2004) and
average of 2.5. Twelve varieties RHMM1701, RHMM1703,
Microsoft excel helped for biplot construction.
RHMM1708, RHMM1709, RHMM1710, RHM1407,
RHT132, WH509, ISARM101, PAN53, PAN-4m-21 and
III. RESULTS
WH101 had ear rot inferior to 2.5. The varieties,
RHMM1702, RHMM1704, RHMM1706, RHMM1707,
3.1 Combined performance across locations
RHMM1711, RHMM1712, RHM104, RHM1402,
The silking varied from 65.7 days (RHMM1712) to
RHM1409, RHMM113 and SC403 had ear rot equal or
77.8 days (RHMM150) with an average of 70.5 days.
superior to 2.5 and inferior to 3.0. Furthermore, three
Eleven varieties, RHMM1712, SC403, RHMM1708,
varieties Kigega, SC529 and RHMM1705 had ear rot
RHMM1710, RHMM1709, RHMM1711, RHMM1703,
superior or equal to 3.0 (Table 3.1). The results of analysis

IJISRT21APR259 www.ijisrt.com 186

Volume 6, Issue 4, April – 2021 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165
of variance revealed significant differences between whereas the varieties RHMM1701, RHMM1704,
varieties for grain yield (P<0.001). Grain yields varied on RHMM1706. RHMM1707, RHMM1709, RHM1409 and
average from 5.07 t/ha (Kigega) to 9.40 t/ha (RHM1402) WH509 had grain yields between 7.11 t/ha and 8.0 t/ha. Five
with an average of 7.11 t/ha. The varieties RHMM1702, varieties PAN53, RHMM1710, RHM1407, WH101 and
RHMM1703, RHMM1705, RHMM1711, RHMM1712, RHMM1708 had grain yields superior or equal 8.0t/ha and
RHMM150, RHM1402, RHMM113, RHT132, SC529, inferior 9.0t/ha (Table 3.1)
SC403 and PAN-4m-21 had grain yield inferior to 7.11 t/ha

Table 3. 1: General performance across locations

No Names SIL ASI HT PA EPP HC EA ER GYD
1 RHMM1701 74.8 1.1 2.14 3.0 1.0 3.3 3.0 2.2 7.27
2 RHMM1702 71.3 1.9 2.23 2.8 1.0 3.1 2.8 2.6 6.90
3 RHMM1703 68.6 2.3 2.34 3.0 1.1 3.0 3.0 2.1 6.63
4 RHMM1704 70.8 1.1 2.27 2.5 1.0 2.8 2.8 2.5 7.76
5 RHMM1705 68.9 2.0 2.18 2.7 1.0 3.0 3.1 3.0 6.19
6 RHMM1706 68.8 0.4 2.22 2.9 1.0 3.2 3.1 2.6 7.35
7 RHMM1707 69.6 1.8 2.30 2.8 1.1 3.1 2.9 2.5 7.79
8 RHMM1708 68.3 1.5 2.30 2.6 1.2 3.0 2.9 2.3 8.81
9 RHMM1709 68.5 1.3 1.98 3.2 1.0 3.3 3.1 2.4 7.10
10 RHMM1710 68.3 0.5 2.19 2.6 1.1 3.0 2.7 2.2 8.56
11 RHMM1711 68.5 1.9 2.30 3.1 0.9 3.2 3.3 2.6 5.33
12 RHMM1712 65.7 0.8 2.15 2.9 1.0 3.2 2.9 2.5 6.02
13 RHMM150 77.8 1.4 2.27 3.1 1.1 3.2 3.0 1.9 5.66
14 RHM104 70.0 0.9 2.31 2.8 1.1 2.8 3.0 2.5 9.40
15 RHM1402 71.0 1.2 2.29 2.8 1.1 2.8 3.1 2.8 6.98
16 RHM1407 68.8 0.9 2.22 2.8 1.0 3.0 2.8 2.3 8.08
17 RHM1409 70.0 0.6 2.26 2.7 1.1 2.9 2.9 2.7 7.73
18 RHMM113 68.6 2.1 2.21 3.1 1.0 3.1 3.3 2.5 6.31
19 RHT132 76.3 2.3 2.31 2.9 1.0 3.1 3.1 2.3 6.78
20 SC529 73.5 2.1 2.33 2.9 1.0 3.2 3.4 3.2 5.26
21 WH509 73.9 1.6 2.60 2.7 1.1 2.9 2.7 2.0 7.97
22 WH101 70.1 1.3 2.28 2.9 1.1 3.0 2.6 2.3 8.16
23 PAN53 73.8 1.8 2.39 2.5 1.0 3.0 2.4 2.2 8.99
24 PAN-4m-21 72.3 2.1 2.30 2.9 1.0 2.9 2.8 2.3 6.86
25 SC403 66.6 2.1 2.29 3.0 1.0 3.0 3.1 2.7 6.29
26 Kigega 69.3 2.3 2.30 3.2 1.0 2.9 3.3 3.2 5.07
27 ISARM101 69.5 1.8 2.39 2.8 1.0 3.1 2.8 2.3 6.79
Mean 70.5 1.5 2.27 2.86 1.0 3.0 3.0 2.5 7.11
C.V. (%) 3.5 95.1 8.3 17.3 12.2 14.5 15.3 30.3 18.7
F 16.16 1.94 3.83 1.91 2.06 1.14 3.03 2.24 9.07
P <0.001 0.006 <0.001 0.006 <0.001 0.297 <0.001 <0.001 <0.001

SIL: Silking (d) ASI: Anthesis-Silking Inerval (d) HT; Plant height (m)
PA: Plant Aspect (1-5) EPP: Ears per plant (Number) HC: Husk Cover (1-5) EA: Ear Aspect (1-5) ER:
Ear Rot (1-5) GY: Grain yield (t/ha@15%H2O)

3.2. Analysis of genotype by environment interaction The AMMI analysis of variance for plant height (Table
using AMMI model 3) of the 27 varieties (19 new varieties under evaluation)
The AMMI analysis of variance for silking (Table 4.5) across three sites showed that the variation due to
of the 27 varieties (19 new varieties under evaluation) across environments and G X E were highly significant (p<0.001).
three locations indicated that the variation due to varieties The varieties effects were not significant, locations effects
and locations were highly significant (p<0.001). The indicated 95% of the treatment Sums Squares (SS), while
varieties effects show only 1.7 % of the treatment Sums interaction of site and varieties effects explained 3.6% of the
Squares (SS), environments 79.3% while interaction effects treatment SS. It further showed that the first IPCAs
were not significant and explained only 2.9% of the (Interaction Principal Component Axis) were highly
treatment SS. It further reveal that the first IPCAs significant (P<0.001) and explained 28.3% of the interaction
(Interaction Principal Component Axis) was highly effects (Table 4.5).
significant (P<0.05) and explained 80.41% of the interaction
effects (Table 4.5).

IJISRT21APR259 www.ijisrt.com 187

 Volume 6, Issue 4, April – 2021 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165
The AMMI analysis of variance for grain yield (Table interaction effects explained only 17.7% of the treatment
3) of the 27 varieties (19 new varieties under evaluation) SS. It further showed that the first IPCAs (Interaction
across three sites showed that the variation due to varieties, Principal Component Axis) were highly significant
environments and G X E were highly significant (p<0.001). (P<0.001) and explained 61.67% of the interaction effect.
(Table 4.5).
The varieties effects indicated that 23.1% of the
treatment Sums Squares (SS), environments 59.1% while

Table 3. 2: AMMI analysis of variance for silking, plant height and grain across locations

Sources of variation DF Silking Plant height Grain yield

SS MS F SS MS F SS MS F
Total 323 16,439 50.9 - 245.1 0.76 - 2,239.2 6.9 -
Treatments 80 14,875 185.9 29.73*** 220.1 2.75 30.31*** 1,799.1 22.5 12.74***
Varieties 26 2,627 101.1 16.16*** 2.9 0.11 1.23NS 416.1 16.0 9.07***
Sites 2 11,808 5,904.2 529.47*** 209.1 104.53 250.26*** 1,064.4 532.2 176.24***
Sites/Replications 9 100 11.2 1.78NS 3.8 0.42 4.60*** 27.2 3.0 1.71NS
Sites × Varieties 52 439 8.5 1.35NS 8.1 0.16 1.72** 318.6 6.1 3.47***
IPCA 27 353 13.1 2.09** 2.3 0.086 2.41*** 196.5 7.3 4.12***
Residuals 25 87 3.5 0.56NS 1.0 0.04 1.12NS 122.1 4.9 2.77***
Error 234 1463 6.3 - 8.3 0.04 - 412.9 1.8 -

3.3 Biplot of silking obtained by plotting the means axis (IPCA score <0.4). The varieties RHMM1407and
The AMMI 1 biplot (Figure 4.2) showed that the RHMM1709 had very large IPCA1 scores (IPCA1 scores
varieties: RHM1704, RHMM1701, RHT132, WH509, between +1.2 and 1.5 The varieties RHMM1402,
RHMM1702, PAN53, PAN-4m-21, RHMM1710, SC529 RHMM1702, PAN53, PAN-4m-21and RHMM150 were
and RHMM150 had average silking means superior to placed at Rubona whereas the cluster made by varieties:
overall means (70.5days). The variety RHM1M150 was the RHM1704, RHMM1701, RHT132, RHMM1704, and
highest silking means >77days) and was close to IPCA1 SC529 groped together with Cyabayaga site.

Figure 3: 3: Biplot of silking obtained by plotting the means (d) against IPCA1 (d0.5) for 27 maize varieties

3.4. Regression of altitude (masl) onto silking (d) regression captured 62% of the variation. Furthermore, the
The regression of altitude onto silking showed that the correlation coefficient between altitude and silking was 0.79
regression coefficient was positive indicating that silking implying that while the altitude was increasing it increased
increased significantly with increasing altitude. The tremendously the silking (Figure 4.3.)
coefficient of determination was 0.62 indicating that the

IJISRT21APR259 www.ijisrt.com 188

Volume 6, Issue 4, April – 2021 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165

Figure3.4: Regression of altitude (masl) onto silking (d)

3.5 Biplot of plant height and grain yield obtained by RHMM1707, WH509, RHMM1704, RHM407, WH101,
plotting the means (m) RHMM1710, RHMM1708, PAN53 and RHM104 had
The AMMI 1 biplot (Figure 3.5) showed that the average grain yield means superior to the overall means
varieties Kigega, RHM1704, RHMM1707, RHT132, (7.11 t/ha). The variety RHM104 showed the highest yield
WH509, RHMM1703, PAN53, PAN-4m-21, RHMM1710, (mean >9t/ha) and was close to IPCA 1 axis (IPCA score
SC529, ISARM101, RHM104, RHM1409 and <0.6). The cluster made by the varieties RHM1706,
RHMM17110 had average plant height means superior to RHMM1701, RHM1409, RHMM1707, WH509, RHM407,
overall means (2.3 m). The variety WH509 was the highest WH101, Kigega, RHMM1705, RHMM1703, ISARM101,
in plant height means (2.7 m) and was close to IPCA 1 axis PAN-4m-21, RHMM113, RHM1402, RHMM1702 and
(IPCA score <0.1). The varieties Kigega and RHMM1709 RHMM1708 had IPCA 1 scores between -0.5 and +0.5. In
had very large IPCA 1 scores (IPCA 1 scores between +0.4 this cluster, varieties RHMM113, RHMM1702, RHM1402,
and -0.4 The varieties RHMM1402, Kigega, RHMM1708, RHM1409 and WH101 were very close each other with
RHMM150, SC529, RHM1704, RHMM1707, RHMM1708, IPCA scores between -0.3 and 0. The variety SC529 had
PAN-4m-21 and WH509 were placed at Bugarama whereas very large IPCA 1 scores (IPCA 1 scores approximately to
the cluster made by varieties RHMM1711, RHT132, +1.2). The varieties RHM1706, RHMM1701 and RHM1409
ISARM101, PAN53 RHM1409, and RHM104 groped were placed at Rubona whereby the cluster made by
together with Cyabayaga site. varieties WH509, RHMM1704, RHM407, WH101,
RHMM1710, RHMM1708, PAN53 and RHM104 grouped
The AMMI 1 biplot (Figure 3.6) clearly showed that together with Bugarama site.
the varieties RHM1706, RHMM1701, RHM1409,

IJISRT21APR259 www.ijisrt.com 189

Volume 6, Issue 4, April – 2021 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165
3.6. Principal Component Analysis (PCA) and cluster the cluster formed by RHM104, RHMM1704, RHM1407,
analysis RHMM1708 and RHM1710, the cluster formed by
The PCA biplot indicated that the PCA1 explained RHMM150 and RHT132, and the cluster formed by
24.8% of the variation and the PCA2 19.6%. PCA1 and RHM1409, WH101, ISARM101, RHM1402, RHMM1703
PCA2 together captured 44.4 % of the variation. and RHMM1702. None of these clusters comprised
Furthermore, it showed that that many varieties were commercial checks except the last cluster which contained
scattered whereas few of them formed distinct clusters such the check WH101 (Figure 4.6.).

RHMM1710
RHMM1712 RHMM1709
RHMM1708
2
RHM1407 RHMM1707
RHMM1704
RHM104
1 RHMM1706
RHM1409 SC403 RHMM1705
RHMM1702
PCA2 (19.6 %)

0 RHMM1703
WH101
RHM1402 RHMM1711
ISARM101 RHMM113

-1 PAN53 Kigega
PAN-4m-21
RHMM1701
-2

WH509
-3 RHMM150
SC59
RHT132

-4
-5 -4 -3 -2 -1 0 1 2 3 4
PCA1 (24.8 %)

Figure 3: 7: PCA biplot involving 27 maize varieties and 15 agronomic traits

3.7 Cluster analysis showing the similarities of 27 maize checks PAN53, WH509 were similar to RHM1708,
varieties using 15 agronomic traits RHM1707, RHM1407, RHM1710, RHM104 and RHM1704
The cluster analysis showed that the pre-released at 86%. Furthermore, it revealed that the pre-released
hybrid RHM113 was similar to SC403 at 98%, the variety RHM1407 was similar to RHM1710 at 96%,
commercial checks PAN-4m-21 and WH101 similar to the RHM104 similar to RHM1704 at 95%, both RHM1407 and
hybrid RHMM1703 at 93%, the check SC403 similar to RHM104 similar to RHMM1710 and RHMM1704 at 93%.
RHMM1711 at 93%, the check SC529 similar to Besides, they are similar RHM1707, RHM1708 at 90%
RHMM1702, RHM1705 and RHM1711 at 90% and the (Figure 4.7).

R HMM1 7 0 1
R HMM1 5 0
R HT1 3 2
R HMM1 7 0 9
R HMM1 7 0 2
R HMM1 7 0 5
SC 5 2 9
R HMM1 7 1 1
Kig eg a
R HMM1 1 3
SC 4 0 3
R HMM1 7 0 3
ISAR M1 0 1
W H1 0 1
PAN-4 m-2 1
R HMM1 7 0 6
R HMM1 7 1 2
R HM1 4 0 2
R HM1 4 0 9
R HMM1 7 0 4
R HM1 0 4
R HMM1 7 1 0
R HM1 4 0 7
R HMM1 7 0 7
R HMM1 7 0 8
W H5 0 9
PAN5 3

1. 00 0. 95 0. 90 0. 85 0. 80 0. 75 0. 70

Similarity
Figure4: 1: Dendogram showing the similarities of 27 maize varieties using 15 agronomic traits

IJISRT21APR259 www.ijisrt.com 190

Volume 6, Issue 4, April – 2021 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165
IV. DISCUSSION The AMMI analysis for silking, plant height and grain
yield of the 27 varieties across three sites reveal that the
The combined ANOVA for all three sites and showed variation due to varieties effects accounted less than of the
a significant difference for silking (P<0.001) among the sites and interaction effects. The variation due to sites was
evaluated varieties. The varieties were found affected by the higher than varieties together with interaction; which
environment, and especially in Cyabayaga site it was shown suggests that narrow adaptation was more important than
a delay was observed in silk emergence due to stressful broad adaptation. This means that varieties showed a ability
conditions such as occurrence of drought. Also significant to adapt to specific environments, which is in agreement
differences(P=0.006) was found for ASI, and it is known with previous studies conducted on different crops including
that a high ASI for more than three days’ results in plants maize indicated that environmental variation contributed
suffering of pollination problem thereby affecting yield. more than varieties and interaction effects (Ngaboyisonga et
This study also showed a low yield in the maize grown in al., 2016), (Mukherjee et al., 2013; Beyene et al., 2012).
Cyabayaga which is a similar result was found by Beyene et
al. (2012). The AMMI 1 biplot for yield showed an ideal
interpretation as varieties not far from the abscissa being
Plant height was also found significant different responsible for the main effects, but the pattern along the
(P<0.001) among the cultivars over the three cultivation ordinate indicate changes in interaction effects. Varieties
sites. Generally, maize grown in Bugarama resulted in the with a high mean superior to generalized mean and an IPCA
highest plant height as compared to other sites. It is a 1 score close to zero (near the abscissa) shows the variety
general perception that when plant perform better in being stable across environments (Yan et al., 2007). In this
vegetative stage i.e. grow more and higher, this will also study, the varieties RHM1706, RHMM1701, RHM1409,
lead to good performance in the reproductive stage, resulting RHMM1707, WH509, RHMM1704, RHM407, WH101,
in higher yield. Furthermore, the plant aspect differed RHMM1710, RHMM1708, PAN53 and RHM104 showed
significantly (P=0.006) among the varieties over the three average grain yield means superior to overall means and a
locations, and a good plant aspect is also known to be IPCA 1 close to the axis, therefore being considered stable
correlated with high yield (Yan and Tinker , 2006). across environment, following interpretations from previous
studies (Gauch, 2006; Yan et al., 2007). The varieties
The ear aspect was significantly (P<0.001) different PAN53 and RHM104 have also been found stable over
among varieties over the three locations. This character is a environments in previous studies (Ngaboyisonga et al.,
good indicator of the quality of the cultivation for farmers, 2016). PCA is known as a method helping breeders to
as a high value on the ear aspect normally result in high understand the pattern of a range of traits in a group of
yield and this character is normally easily understood by the varieties identifying those with most suitable traits
farmer., The ear rot was significantly (P<0.001) different combination (Ali et al,.2014). In this study, the varieties
among the varieties over the three locations. The ear rot is were grouped in five clusters, and only one commercial
known to affect grain yield by decreasing the grain output of variety in these, implying a relationship among the newly
the plant. Furthermore, the grain yield differed significantly released varieties. This study was conducted in three sites
(P<0.001) among varieties over the three locations. A for one season due to the limited time of a MSc education.
superior yield is always a major target goal within all During the season of evaluation, the Cyabayaga site was
breeding programs, and high yielding varieties are always vulnerable to prolonged droughts that lead to low
selected and recommended to farmers. In cases when yield performance in that site.
is not consistent across sites this is a problematic situation
for the breeder and indicates that the varieties are not stable V. CONCLUSION
over locations. Here the yield of the investigated varieties
was generally higher in the Bugarama site compared to the This study was intended to identify hybrid high yield
Rubona and Cyabayaga sites, respectively. The ANOVA and stable hybrid maize varieties for mid altitudes of
also showed significant differences for ASI (P=0.011), plant Rwanda. Therefore, trials were carried out in three different
height (P=0.002), ears per plant (P<0.001), ear rot which locations namely Cyabayaga, Rubona and Bugarama and a
could explain the differences in yield, similarly as has also total of 27 varieties were used. Performance of new hybrid
been the case in previous investigations (Dehghani et al., varieties was checked towards commercial check varieties to
2006; Adu et al, 2013). certify the identification of high yielding and stable varieties
adapted to the mid altitude of Rwanda. Furthermore, the
In the AMMI analysis, the treatment variation was extent of genotype by environment interaction was
categorized into three types of variations (variation due to evaluated in the study. The Results indicated that
varieties main effects, variation due to sites main effects and environments were diverse and a broad adaptation of the
variation due to combined varieties x sites effects). These maize varieties was more important than a narrow
three types of variations imply different suggestions: the adaptation. The varieties RHM1706, RHMM1701,
varieties variation refer to general adaptations, sites effects RHM1409, RHMM1707, WH509, RHMM1704, RHM407,
variation is related to narrow adaptations while varieties WH101, RHMM1710, RHMM1708, PAN53 and RHM104
variations combined together with varieties x sites represent were found stable across locations and therefore they could
large environments (Gauch, 2006). be used in mid altitude to increase maize production. This
study also suggested the present commercial checks, PAN53

IJISRT21APR259 www.ijisrt.com 191

Volume 6, Issue 4, April – 2021 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165
and WH509, to be replaced by RHM1708, RHM1707, [11]. Minagri. (2010). Evaluation report on crop
RHM1407, RHM1710, RHM104 and RHM1704. The intensification program. Kigali: International Center
locations used in this study were found not enough to for Soil Fertility and Agricultural.
represent all zones of the mid altitude area of Rwanda, so [12]. Mukherjee A.K Mohapatra N.K., B. L. (2013).
more trials are needed to be conducted to understand a full Additive main effects and multiplicative interaction
picture. The results also showed that not all varieties are (AMMI) analysis of GxE interactions in rice-blast
stable in mid altitudes of Rwanda and these varieties should pathosystem to identify stable resistant genotypes.
therefore not be recommended to be used in the mid altitude. Africa Journal of Agriculture Res , Vol :8, pp 5492-
Furthermore, genotype by environment interaction was 5507.
found influencing the performance of the varieties. [13]. Mulinga, N. N. (2014). Performance of hybrid maize
Bugarama was the best site to grow maize and the varieties introductions in the mid- altitudes zone of Rwanda.
performed better at this site as compared to the two other Scholarly agriculture journal of sciences
sites used, were Rubona was found the second best site and Vol.4(11),November , 553-557.
Cyabayaga was the last in performance. [14]. Ngabitsinze, J. C. (2014). Analysis of Economic
Efficiency of Maize Production in Huye District in
REFERENCES Rwanda. International Journal of Agriculture
Innovations and Research , 2319-1473.
[1]. Adu, G.D., R. Akromah, M.S. Abdulai, K. Obeng- [15]. Ngaboyisonga, Nyombayire A, Gafishi M.K,
Antwi, A.W. Kena, K.M.L. Tengan m and H. Alidu. Nizeyimana F, Uwera A, Ndayishimiye T, Karemera
(2013). Assessment of Genotype by Environment F. X, Mutanyagwa P, Gumisiriza G. and Gahakwa.
interactions and Grain yield Performance of Extra- (2016). Adaptability and genotype by envirnoment
Early Maize (Zea Mays L.) Hybrids. Journal of Interaction of maize commercial hybrid Varieties from
Biology, Agriculture and Healthcare , Vol.3, No.12,pp East African seed companies in Rwanda
7-15. Envirnoments. Global Journal of Agricultural
[2]. Ali Kahraman, Mustafa Onder and Ercan Ceyhan. Research Vol 4,No 2 , 32-40.
(2014). Cluster Analysis in Common Bean Genotypes [16]. Ngaboyisonga. C, F. Nizeyimana, A. Nyombayire,
(Phaseolus vulgaris L.). Turkish journal of M.K. Gafishi,. (2014). Identification of Elite, High
Agricultural and natural science , 1:1030-1035. Yielding and Stable Maize Cultivars for Rwandan
[3]. Beyene, Y., Mugo, S. Mutinda, C., Tefera, T., Karaya, Mid-altitude Environments. Springer International
H., Ajanga, S., Shuma, J., Tenda, R, and Publishing Switzerland , 165-178.
Kanga,.(2011). Genotype by environment interaction [17]. NISR, N. i. (2016). Statistical yearbook. KIgali-
and yield stability of stem borer resistant maize hybrid Rwanda: SYB(2016).
in Kenya. .African journal of biotechnology vol 10 , [18]. Sallah, P.Y.K., Mukakalisa S., Nyombayire A. and
10: 4752-4758. Mutanyagwa P. (2009). Response of two maize
[4]. Beyene, Y., Mugo, S., Tefera, T., Gethi, J., Gakunga, varieties to density and nitrogen fertilizer in the
J., Ajanga, S., Karaya, H, Musila, R.,Muasya. W., highland zone of Rwanda. Journal of applied
Tende, R. and Njoka, S. . (2012). Yield stability of biosciences 20 , 1194-1202.
stem borer resistantmaize hybrids evaluated in regional [19]. Tyagi, S. D and Khan, M. H. . ( 2009). Genotypeby
trials in East Africa. African Journal of Biotechnology environment interaction and stability analysis for yield
, Vol. 6(2), pp. 77-83. and its components in soybean (Glycine max L.
[5]. Buysse W, Stern R, Coe R. (2004). GenStat discovery Department of plant breeding and genetics K.P. ,
edition for everyday use. Nairob: ICRAF. kmudasirhafiz@yahoo.com.
[6]. Dehghani, H, Ebadi, A. and Yousefi, A..(2006). Biplot [20]. USDA, U. S. (2018). World Agricultural Supply and
analysis of genotype by environment interaction for Demand Estimates. Washington DC: USDA Foreign
barley yield in Iran. Agronomy journal , 98:388-393. Agricultural Service Office of Global Analysis IPA
[7]. Fatma, E. I(2011). Evaluation by envirnoment division.
interactions in maize hybrids using GGE byplots [21]. Yan, W. and Tinker, N. A. . (2006). Biplot analysis of
analysis. crop breeding and aplied biotechnology , multi-environment trial data: Principles and
11:1-9. application. Canadian Journal of Plant Science , . 86:
[8]. Gauch, H. (2006). Statistical analysis of yield trials by 623-645.
AMMI and GGE. crop sciences , 42:619-633.
[9]. Khalil, A.I, Ur- Rahman, H., Ur-Rehman, N., Arif, M,
Khalil, I. H, Iqbal, M., Hidayatullah, A.K., Sajjad, M.
and Ishaq, M. (2011). Evaluation of maize hybrids for
grain yield stability in North-West of Pakistan. Sarhad
Journal of Agriculture , Vol.27, No.2.
[10]. Kutka, F. (2011). Open-Pollinated vs. Hybrid Maize
Cultivars. Sustainability. 3: , 1531-1554.

IJISRT21APR259 www.ijisrt.com 192