International Journal of Plant Breeding and Crop Science

Vol. 4(3), pp. 291-299, October, 2017. www.premierpublishers.org. ISSN: XXXX-XXXX

IJPBCS

Research Article

Grain Yield Stability in Three-way Cross Hybrid Maize

Varieties using AMMI and GGE Biplot Analysis
1Sayo Sesay*, 2Alpha Bellah Jalloh, 3Vandi Alieu Sama
1,2,3
Crop Improvement Programme, Rokupr Agricultural Research Center, Rokupr - Sierra Leone Agricultural
Research Institute, PMB 1313, Tower Hill, Freetown, Sierra Leone.

A study to evaluate three-way cross hybrid maize varieties for wide adaptability and stability was
conducted in eight environments in Sierra Leone using AMMI and GGE biplot analysis. There were
significant genotype and environment main effects, and genotype x environment interactions
(GEI) effects. Differences due to environments accounted for 70.1% of the total treatments sum
of squares while genotypes and genotype x environment interaction accounted for 9.9% and
20.0%, respectively. The first four interaction principal component axes (IPCA) were also highly
significant and accounted for 38.7%, 25.2%, 14.3% and 8.6%, respectively of the total genotype x
environment interaction variation. The polygon view of the GGE biplot revealed that hybrid G14
produced the highest grain yield in environments E1, E5 and E7 whereas G24 was adaptive in
environments E6, E8, E3, E4 and E2. Hybrids G24, G9, G17 and G6 also produced high grain yields
and were relatively stable. Both AMMI and GGE biplot effectively partitioned treatments sum of
squares and were more appropriate in explaining genotype x environment interaction. The models
also identified G24 as the most desirable hybrid in terms of high grain yield and stability across
environments. Therefore, this hybrid is recommended for commercial release.

Keywords: Genotype x environment interaction, grain yield, three-way cross hybrid, biplot analysis.

INTRODUCTION

The need to provide farmers and seed growers with new
widely adapted maize varieties that will help boost their
production and livelihoods has been the concern and effort
of the maize breeding unit in Sierra Leone. In Sierra Leone,
maize is the second most important cereal crop grown
after rice, which is the countrys staple. The crop has a
wide scope of adaptability and can grow in almost all agro-
ecologies in the country. It is cherished by Sierra Leoneans
and is grown in homesteads in the uplands during the
raining season and inland valley swamps in the dry season
(MAFFS, 2005). It has a very high commercial value when
compared with other food crops grown in the country and
is both a food and economic crop, contributing to poverty
alleviation. Locally, maize is prepared into corn meal for
human consumption, or eaten boiled or roasted (MAFFS,
2005). It is an excellent feed for poultry and livestock.
Maize is also an excellent raw material for confectioneries,
Grain Yield Stability in Three-way Cross Hybrid Maize Varieties using AMMI and GGE Biplot Analysis
oil, starch and other useful products (Olawuyi et al., 2010;
Randjelovic et al., 2011; Bello et al., 2013; Orhun, 2013).
Despite all these benefits derived from maize, its potentials
have not been fully realized in Sierra Leone. Yields are still
very low, estimated to be 2tha-1 in 2015 (USDA, 2015) and
no new varieties with superior adaptability have been
released for several years.
*Corresponding Author: Sayo Sesay, Crop Improvement
Programme, Rokupr Agricultural Research Center, Rokupr
- Sierra Leone Agricultural Research Institute, PMB 1313,
Tower Hill, Freetown, Sierra Leone. Email:
osaiosesay@yahoo.com. Tel: +23278084923
Sesay et al. 292
Table 1: Descriptions of test locations
Region/Area District Location Latitude
South Bo Bandajuma 757'0"N 1140'0"W
Western Area Western Area Rural Newton 820'6.6"N 130'29.9"W
North Kambia Samu 95'5.9"N 1244'28.1"W 65
East Kenema Serabu 751'17.04"N 11 17'13.56"W 134
Environment code: Bandajuma 2012 = E1, Bandajuma 2013 = E2, Newton 2012 = E3, Newton 2013 = E4, Samu 2012 = E5, Samu
2013 = E6, Serabu 2012 = E7, and Serabu 2013 = E8.
Production is erratic and there are numerous constraints
to maize production and productivity- these include the
deterioration of well-known maize varieties (Western
Yellow, DMR-Yellow and TZR-ESR-yellow) under
cultivation due to varietal mixtures, pests and diseases
and other abiotic factors. There is the need to provide
farmers and seed growers with new widely adapted maize
varieties that will help boost their production and
livelihoods. This implies the development and/or
introduction of new maize genotypes with superior
performance and wide adaptability.
Although different methodologies and procedures have
been proposed for the interpretation of genotype x
environment interaction and the evaluation of the
adaptability and stability of genotypes, the AMMI (Additive
Main Effects and Multiplicative Interaction) and GGE
(Genotype main effects and Genotype x Environment
interaction) biplot analyses stand out for the larger number
of technical interpretations they provide (Duarte and
Vencovsky, 1999; Yan et al., 2000). Both AMMI and GGE
Biplot analyses combine rather than separate G and GxE
in mega-environment analysis and genotype evaluation
(Yan et al., 2007). Both models interpret the effects of
genotypes and environments as additive and the GxE
interaction as multiplicative, by principal component
analysis. However, several comparison studies between
the AMMI and GGE biplot analyses have been conducted
to evaluate genotypes, environments and genotype x
environment interactions. Whereas some studies showed
comparative advantages of the GGE biplot over the AMMI
(Yan et al., 2007), others found equal advantages or little
differences between the two models and conclude that
they should complement each other in genotypes and
environments evaluation. For instance, Mitrovic et al.
(2012) found no large difference between the AMMI and
GGE biplot analyses in the evaluation of experimental
maize hybrids in different climatic conditions and that both
methods can be used equally successfully. Miranda et al.
(2012) concluded that utilizing both AMMI and GGE biplot
simultaneously provides an innovative approach to the
interpretation of genotype x environment interactions.
Nassir (2013) further suggests that the combination of the
two models would be necessary to take advantage of the
respective areas of strength of each analysis to better
serve the purpose of recommending genotypes for a
location.
Grain Yield Stability in Three-way Cross Hybrid Maize Varieties using AMMI and GGE Biplot Analysis
Longitude
Altitude (m) Average annual Average annual
temperature (C) rainfall (mm)
98 26.2 2706
36 26 3392
26.8 2988
26.6 2853
To take advantage of the respective strengths of AMMI
and GGE biplot analyses for better decision makings in
recommending genotypes in specific locations, the
objective of this study was to evaluate three-way cross
hybrid maize varieties for wide adaptability and stability of
performance in eight environments in Sierra Leone.
MATERIALS AND METHODS
Experimental Sites, Treatments and Design
Trials were conducted in four different locations in the third
and fourth week of June in the wet cropping seasons of
2012 and 2013, comprising of eight environments. Each of
the test locations represented one District and were
selected based on their large maize production and
markets (Table 1).
The test genotypes were 23 three-way cross drought
tolerant hybrid maize varieties and one check, Western
Yellow. The hybrids were sourced from the International
Institute of Tropical Agriculture, IITA, Ibadan, Nigeria
(Table 2). At each environment, the land was well
ploughed and harrowed and trial plots were laid out in a
randomized complete block design with three replications.
Each plot consisted of two rows (1.5m wide x 5m long),
with 75cm between and 50cm within rows and two plants
per hill were maintained. NPK 15:15:15 and urea fertilizers
were applied at the rate of 120kg N/ha, 60kg P2O5/ha and
60kg K2O/ha. The nitrogen fertilizer was applied in two
splits - two and six weeks after planting. Weeding was
done three times with a hoe and other standard crop
management practices were applied. Grain yield (15%
moisture content) was recorded in kilograms at harvest
and later converted into tonnes per hectare.
Statistical Analysis
Grain yield data combined across the eight environments
were subjected to AMMI analysis of variance to determine
genotype and environment main effects, and genotype x
environment interactions effects. The mean yields of
genotypes in each environment and across environments
and their first interaction principal component axis (IPCA)
values were also analysed. The AMMI and GGE biplot
models were used to identify genotype stability, which
 Int. J. Plant Breed. Crop Sci. 293

Table 2: List of genotypes and their pedigrees used in the study

Genotypes Code Pedigree Source
WesternYellow G1
(Local check) Western Yellow (Local check)
M0826-1 G2 9071/1368xHIx4269-1x1368-7-2-B*4/P43SRC9FS100-1-1-8-#1-B1-13-B1-B*4 14A20600B
M0826-2 G3 9071/1368xHIx4269-1x1368-7-2-B*4/Obantapa-33-5-1-B*4 14A20602B
M0826-3 G4 9071/1368xHIx4269-1x1368-7-2-B*4/(TZMI501xKU1414x501)-1-4-3-1-B-BB 14A20604B
M0826-4 G5 9071/1368xHIx4269-1368-7-2-B*4/161 14A20606B
G6 (1368/Mi82-23-2-1-2-B*7/P43SRC9FS100-1-1-8-#1-B-13-B*7)-58-1/DTPL-W-C7-S2-7-1-1-1
M1124-15 -1-B-2-B*4/BabangoyoxMO17LPAxBabangoyo-23-1-3-1-B*6-9-1/IWD-SYN-STR-C3-70-2-BB 12A17508B
G7 (1368/Mi82-23-2-1-B*7/P43SRC9FS100-1-1-8-#1-B-13-B*7)-65-1/DTPL-W-C7-S2-7-1-1-1-1-B-2-
M1124-16 B*4/BabangoyoxMO17LPAxBabangoyo-23-1-3-1-B*6-23-1/IWD-SYN-STR-C3-70-2-BB 13A116888B
G8 (1368/Mi82-28-1-1-2-B*7/P43SRC9FS100-1-1-8-#1-B1-13-B1-B*6)-40-1/DTPL-W-C7-S2-1-2-1-1-
M1124-17 5-B-1-B*4/BabangoyoxMO17LPAxBabangoyo-23-4-3-3-B*6-46-1/IWD-SYN-STR-C3-70-2-BB 14A19475B
G9 (1368/Mi82-23-2-1-2-B*7/P43SRC9FS100-1-1-8-#1-B-13-B*7)-1-1/((ACR-86-8-1-2-1-1-1-
M1124-18 B-1-B*4/Babangoyo/MO17LPA/Babangoyo-28-1-2-1-B*6))-38-1/IWD-SYN-STR-C3-70-2-BB 12A17506B
G10 (1368/Mi82-23-2-1-2-B*7/P43SRC9FS100-1-1-8-#1-B-13-B*7)-58-1/DTPL-W-C7-S2-7-1-1-
M1124-23 1--B-2-B*4/BabangoyoxMO17LPAxBabangoyo-23-1-3-1-B*6-9-1/IWD-SYN-STR-C3-35-3-BB 12A17516B
G11 (1368/Mi82-23-2-1-2-B*7/P43SRC9FS100-1-1-8-#1-B-13-B*7)-65-1/DTPL-W-C7-S2-7-1-1-1--B-2-
M1124-24 B*4/BabangoyoxMO17LPAxBabangoyo-23-1-3-1-B*6-23-1/IWD-SYN-STR-C3-35-3-BB 12A17518B
G12 (1368/Mi82-23-2-1-2-B*7/P43SRC9FS100-1-1-8-#1-B-13-B*7)-1-1/((ACR-86-8-1-2-1-1-1-B-1-
M1124-26 B*4/(Babangoyo/MO17LPA/Babangoyo-28-1-2-1-B*6))-38-1/IWD-SYN-STR-C3-51-1-BB 12A17530B
G13 (1368/Mi82-23-2-1-2-B*7/P43SRC9FS100-1-1-8-#1-B-13-B*7)-58-1/DTPL-W-C7-S2-7-1-1-1-1-B-2-
M1124-27 B*4/BabangoyoxMO17LPAxBabangoyo-23-1-3-1-B*6-9-1/IWD-SYN-STR-C3-51-1-BB 12A17524B
G14 (1368/Mi82-23-2-1-2-B*7/P43SRC9FS100-1-1-8-#1-B-13-B*7)-65-1/DTPL-W-C7-S2-7-1-1-1-1-B-2-
M1124-28 B*4/BabangoyoxMO17LPAxBabangoyo-23-1-3-1-B*6-23-1/IWD-SYN-STR-C3-51-1-BB 12A17504B
(1368/Mi82-28-1-1-2-B*7/P43SRC9FS100-1-1-8-#1-B1-13-B1-B*6)-40-1/DTPL-W-C7-S2-1-2-1-1-
M1124-29 G15 5-B-1-B*4/BabangoyoxMO17LPAxBabangoyo-23-4-3-3-B*6-46-1/IWD-SYN-STR-C3-51-1-BB 12A17528B
G16 (KU1409/KU1414-SR/A619)-S2-2/9450xKI21-7-2-2-1-1-BB/(POP66SR/ACR91SUWAN1-
M1124-31 SRC1/ACR91SUWAN1-SRC1-6X(MP420x4001xMP420)-3-1-3-1-B)S2-5-B*5 12A17504B
Oba 98 G17 Oba 98 Premier
Oba Super 1 G18 Oba Super 1 Premier
Oba Super 7 G19 H05-01STR Premier
Oba Super 9 G20 H05-02STR Premier
SC621 G21 SC621 SEEDCO
SC627 G22 SC627 SEEDCO
SC637 G23 SC637 SEEDCO
SC719 G24 SC719 SEEDCO

genotype (s) performed best in which environment (s), and
rank the genotypes based on high grain yield and stability.
GenStat release 12.1 (GenStat 2009) and Plant Breeding
Tools version 1.4 (PBTools, 2014) Softwares were used to
carry out the analyses.
RESULTS AND DISCUSSION
Analysis of Variance
The AMMI analysis of variance for grain yields of the 23
three-way cross hybrids grown in eight environments are
presented in Table 3. The variance component due to
treatments was highly significant and accounted for 96.3%
of the total variation. This large percentage indicated that
the AMMI model effectively partitioned treatments. The
effects of genotypes, environments and genotype x
environment interaction were significant. Differences due
to environments accounted for 70.1% of the total
treatments sum of squares while genotypes and genotype
x environment interaction accounted for 9.9% and 20.0%,
respectively. The largest portion accounted for by
environments indicated the significant influence of
Grain Yield Stability in Three-way Cross Hybrid Maize Varieties using AMMI and GGE Biplot Analysis
environments on grain yield in the study areas. Similar
results where the environment had the highest effect than
other factors had been reported by several researchers in
various crops (Negash et al., 2013; Kilic, 2014; Kendal and
Dogan., 2015; Edugbo et al., 2015; Oyekunle et al., 2017).
Assessment of genotypes in many locations and years
could increase the reliability of plant breeding programs,
but many location trials were more important than many
years trials in the same location in determining
performance of tested genotypes in terms of investigated
traits (Ezzat et al., 2010; Sayar et al., 2013). This
enhances the importance of our study conducted in many
locations in two years, which were different from one
another in terms of environmental conditions. Mosa et al.
(2011) and Khalil et al. (2011) found significant GxE
interaction for grain yield in maize genotypes. The first,
second, third and fourth interaction principal component
axes (IPCA) were also highly significant and accounted for
38.7%, 25.2%, 14.3% and 8.6%, respectively of the total
genotype x environment interaction variation. The first
IPCA mean square was almost six times and the second
IPCA mean square was only four times larger than the
residual mean square. The magnitude of the genotype x
environment interaction sum of squares was about two
Sesay et al. 294

Table 3: AMMI analysis of variance for grain yield (t/ha) of 24 three-way cross maize hybrids grown in eight environments in Sierra
Leone
Source Df SS MS % Total % Treatment % Interaction
Total 575 626.7 1.09
Treatments 191 603.6 3.16** 96.3
Genotypes 23 59.6 2.59** 9.9
Environments 7 423.5 60.49** 70.1
Block 16 2 0.13** 0.3
Interactions 161 120.5 0.75** 20.0
IPCA1 29 46.6 1.61** 38.7
IPCA2 27 30.4 1.13** 25.2
IPCA3 25 17.2 0.69** 14.3
IPCA4 23 10.4 0.45** 8.6
Residuals 57 15.9 0.28** 13.2
Error 368 21.1 0.06 3.4

Table 4: Mean grain yields (tha-1) and first IPCA values of 24 three-way cross maize hybrids grown in eight environments in Sierra
Leone
Genotype Code Bandajuma Bandajuma Newton Newton Samu Samu Serabu Serabu Mean First IPCA
2012 2013 2012 2013 2012 2013 2012 2013 Score
Check G1 2.0 1.4 4.4 3.4 1.3 1.8 2.1 2.3 2.3 -0.35
M0826-1 G2 2.5 1.8 4.4 3.7 1.9 3.3 2.0 3.2 2.8 -0.09
M0826-2 G3 2.1 1.7 4.6 3.6 1.3 2.3 1.6 3.0 2.5 -0.55
M0826-3 G4 2.6 2.1 4.9 4.0 1.9 2.9 2.2 3.3 3.0 -0.32
M0826-4 G5 2.6 1.8 4.5 3.8 2.1 3.2 2.6 3.0 3.0 0
M1124-15 G6 2.7 1.8 4.3 3.8 2.3 4.1 2.1 3.5 3.1 0.19
M1124-16 G7 2.8 1.5 3.8 3.6 2.5 4.3 2.9 3.0 3.1 0.69
M1124-17 G8 2.7 2.4 5.1 4.3 2.0 3.4 1.9 3.9 3.2 -0.48
M1124-18 G9 2.8 2.0 4.4 4.0 2.4 4.2 2.3 3.6 3.2 0.16
M1124-23 G10 2.2 1.5 4.1 3.5 1.7 3.3 1.7 3.1 2.6 0
M1124-24 G11 3.1 2.1 4.7 4.1 2.6 3.6 3.4 3.1 3.3 0.26
M1124-26 G12 2.7 1.6 4.1 3.7 2.3 4.0 2.6 3.1 3.0 0.44
M1124-27 G13 2.6 1.8 4.6 3.8 1.9 2.5 2.9 2.6 2.8 -0.09
M1124-28 G14 2.9 1.4 3.7 3.6 2.7 4.1 3.6 2.5 3.1 0.92
M1124-29 G15 2.5 1.6 4.3 3.7 2.0 3.1 2.6 2.8 2.8 0.1
M1124-31 G16 2.7 1.4 3.8 3.5 2.4 3.9 3.0 2.7 2.9 0.66
Oba 98 G17 2.8 1.9 4.5 3.9 2.3 3.5 2.8 3.1 3.1 0.18
Oba Super 1 G18 2.0 1.2 3.6 3.2 1.6 3.7 1.2 3.1 2.5 0.16
Oba Super 7 G19 2.8 2.0 4.8 4.0 2.2 2.9 3.1 2.9 3.1 -0.001
Oba Super 9 G20 2.3 1.8 4.3 3.7 1.8 3.7 1.3 3.7 2.8 -0.18
SC621 G21 1.7 1.3 4.1 3.2 1.0 2.1 1.2 2.6 2.1 -0.41
SC627 G22 2.0 1.8 4.8 3.6 1.2 2.0 1.5 3.0 2.5 -0.71
SC637 G23 2.0 1.7 4.5 3.5 1.2 2.4 1.3 3.1 2.5 -0.56
SC719 G24 2.9 2.3 4.8 4.2 2.4 4.4 2.1 4.1 3.4 -0.02
Mean 2.5 1.7 4.4 3.7 2.0 3.3 2.2 3.1 2.9
First IPCA score 0.19 -0.62 -1.01 -0.39 0.55 1.03 0.81 -0.56

times larger than that for genotype and 12.5 times larger
than the error mean square. This indicates that partitioning
of the interaction sum of squares by AMMI was very
effective (Makinde et al., 2013) and there were sizeable
differences in responses of the genotypes across
environments (Karimizadeh et al., 2013). According to Yan
and Kang (2003), this suggests the possible presence of
from 2.1 tha-1 for the hybrid SC621 to 3.4 tha-1 for SC719,
50% of the hybrids yielded above the grand mean. The
highest grain yields (4.4 tha-1 in 2012 and 3.7 tha-1 in 2013)
were obtained at Newton followed by 3.3 tha-1 at Samu in
2013. Conversely, the lowest grain yield were recorded at
Bandajuma in 2013 (1.7 tha-1) and Samu in 2012 (2.0 tha-
1); 50% of the environments produced grain yields above

different mega-environments with different winner
genotypes. This observation is in line with that of Adomou
et al. (1997), Makinde and Ariyo (2011) and Edugbo et al.
(2015).
Mean Performance of Genotypes
Table 4 presents genotype and environment means as
well as their first IPCA values. Average grain yield ranged
Grain Yield Stability in Three-way Cross Hybrid Maize Varieties using AMMI and GGE Biplot Analysis
the average. When a genotype and an environment have
the same sign, their interaction is positive; if different, their
interaction is negative (Tariku et al., 2013). Therefore,
hybrids M0826-4, M1124-15, M1124-16, M1124-18,
M1124-23, M1124-24, M1124-26, M1126-28, M1124-29,
M1124-31, Oba 98 and Oba Super1 had a positive
interaction with the environments Samu in 2012 and 2013,
Bandajuma, and Serabu in 2012 while the remaining
hybrids had a negative interaction with the environments

Figure 1: AMMI biplot for mean grain yield of genotypes and environments (main effects) versus stability
(PC1) of 24 three-way cross maize hybrids grown in eight environments in Sierra Leone.
Newton in 2012 and 2013, Bandajuma and Serabu in 2013
- 50% of the hybrids had a positive interaction with 50% of
the environments. Likewise, 50% of the hybrids interacted
negatively with half of the environments. The hybrids
M0826-4 and M1124-23 (with PC1 scores of zero), and
Oba Super 7 and SC719 (with PC1 scores nearly zero)
can be fitted well by an addictive model. Hybrid SC719
gave the highest grain yields at Newton and Samu in 2012
and 2013, respectively and showed consistent
performance in the other environments, except location
Serabu in 2012.
AMMI Analysis
Figure 1 presents the AMMI biplot with the genotype and
environment main effects for grain yield on the X-axis and
PC1 scores on the Y-axis. The vertical line at the middle
represents the grand mean of the grain yield, while the
horizontal line at the middle is the PC1 value of zero
(Tolessa, 2015). The biplot revealed that genotypes G4,
G5, G6, G7, G8, G9, G11, G12, G14, G17, G19 and G24
are high yielding since they were placed on the right hand
side of the biplot. Similarly, the highest grain yield was
produced in environments E3, E4, E6 and E8. On the other
hand, the rest of the genotypes and environments were
considered to be low yielding and were placed on the left
hand side of the biplot. Genotypes or environments which
appear almost on the perpendicular line to the grand mean
Grain Yield Stability in Three-way Cross Hybrid Maize Varieties using AMMI and GGE Biplot Analysis
Int. J. Plant Breed. Crop Sci. 295
line have similar means, while those that occur on a
horizontal line to the zero PC1 line have similar interaction
patterns (Badu-Apraku et al., 2003). Thus, genotypes G2,
G13, G15, G20; G3, G18, G22, G23; G4, G5, G12; G6,
G7, G14, G17 and G19 had similar mean effects while G5,
G10, G19 and G24 showed the same interaction effects.
None of the environments had similar means and
interaction patterns, indicating that the environments were
extremely different. The AMMI biplot further reveals that
environment E1 and hybrids G18, G10, G15, G2, G13,
G20, G5, G19, G24, G17, G9 and G6 were the least
interactive while the highest genotype x environment
interaction was shown by hybrids G12, G7, G16 and G14
in environment E6, and by hybrid G8 in environments E3,
E4 and E8. However, stability should be accompanied with
high yielding performance, thus genotypes G5, G19, and
G24 were the most stable genotypes with high grain yields.
However, G24 was considered to be the most desirable
genotype because of its highest grain yield and absolute
stability. Genotypes G9, G17 and G6 also produced better
grain yields and were relatively stable.
GGE-biplot Analysis
Figure 2 shows the polygon view of which three-way cross
hybrids were the best in which environment(s). The two
principal component axes (PC1 and PC2) accounted for
63.8% of the total variation in grain yield, of which PC1
Sesay et al. 296
Figure 2: Polygon view of the GGE biplot showing which three-way cross maize hybrids had the best
grain yield in which environment.
explained 45.7% and PC2 explained 18.1%. This result is
in line with the finding of Oyekunle et al. (2017) who
observed that the two PC axes accounted for 59.8% of the
total variation, with PC1 and PC2 explained 44.7% and
15.1%, respectively of total variation for grain yield in early
maturing hybrids.
The biplot is divided into four sectors, which placed
genotypes into all the four sectors and test environments
into two sectors. Yan et al. (2000) pointed out that the
vertex hybrids in each sector represented the highest-
yielding hybrid in the location that fell within that particular
sector. Therefore, genotype G14 which is a vertex
genotype in the first sector had the highest grain yield in
environments E1, E5 and E7, while G24 which is the vertex
genotype in the second sector produced the highest grain
yield in environments E6 and E8. Genotype G24 also
produced consistently higher grain yields in environments
E3, E4 and E2. Genotypes G21 and G1 which are the
vertex genotypes in the third and fourth sectors,
respectively had the least grain yields in one or all of the
environments.
The ranking of the 24 genotypes based on their mean grain
yield and stability of performance and relative to an ideal
Grain Yield Stability in Three-way Cross Hybrid Maize Varieties using AMMI and GGE Biplot Analysis
genotype across eight environments is presented in Figure
3. The single arrowed line on the graph is the Average
Environment Axis (AEA) abscissa which points to a higher
genotype mean performance whereas the double-arrowed
line that passes through the biplot origin and is orthogonal
to the AEA abscissa is the AEA ordinate. Regardless of
direction, it shows greater GEI effect and greater instability
(Yan and Kang, 2003). The stability and instability of
genotypes are associated with the vector line of the AEA
abscissa on the biplot graph (Yan, 2001). A genotype with
shorter vector from the AEC abscissa is considered more
stable while genotypes with longer vector are unstable
(Kaya et al., 2006; Yan, 2001). Thus, hybrids G14, G16,
G7, G11, G20, G23, G22, G3, G8 and G19 with longer
vector lines away from the AEA line are considered to be
unstable, while G1, G17, G18, G10, G5, G9, G4, G6, G2
and G12 with shorter vector lines are regarded as being
stable. Hybrid G14 was the most unstable hybrid since it
had the longest vector line away from the AEA line
whereas G5 with the shortest vector line was considered
to be the most stable hybrid. The stability of hybrids is
meaningful only when considered simultaneously with high
grain yield performance (Yan and Tinker, 2006).
Therefore, hybrids G5, G6, G9, G12 and G17 had high
grain yield performance and relative stability. Hybrids G1,
 Int. J. Plant Breed. Crop Sci. 297

Figure 3: GGE biplot showing the ranking of 24 three-way cross maize hybrids for both mean yield and stability of
performance relative to an ideal genotype across eight environments in Sierra Leone.

G10 and G18 were very stable but were among the low
yielding hybrids and cannot, therefore, be recommended
for commercial release. Hybrid G24 was considered to be
the most desirable hybrid because it gave the highest grain
yield and had absolute stability and is, therefore,
recommended for commercial release. This hybrid is found
near the ideal genotype represented by the small circle
with an arrow pointing to it (Yan and Kang 2003).
partitioned treatments sum of squares and were more
appropriate in explaining genotype x environment
interaction. The models also identified hybrid G24 as the
most desirable hybrid. This hybrid is therefore
recommended for release for commercial production.
CONFLICT OF INTERESTS

CONCLUSIONS There is no conflict of interest for this paper

There were significant differences between genotypes and

environments and significant genotype x environment
interactions. The polygon view of the GGE biplot identified
the hybrid G14 as the most adaptive in environments E1,
E5, and E7 whereas hybrid G24 was adaptive in
environments E6, E8, E3, E4 and E2. Genotypes G24, G9,
G17 and G6 produced high grain yields and were relatively
stable. Both the AMMI and GGE biplot effectively
ACKNOWLEDGEMENTS
The authors are grateful to the Alliance for a Green
Revolution in Africa (AGRA) for financial support and the
International Institute of Tropical Agriculture (IITA), Ibadan,
Nigeria for providing planting materials and technical
support.

Grain Yield Stability in Three-way Cross Hybrid Maize Varieties using AMMI and GGE Biplot Analysis
Sesay et al. 298
REFERENCES
Adomou M, Ntare BR, Williams JH (1997). Stability of pod
yields and parameters of a simple physiological model
for yield among peanut lines in Northern Benin. Peanut
Sci. 24(2): 107-112.
Badu-Apraku B, Abamu FJ, Menkir A, Fakorede MAB,
Obeng-Antwi K, Th C. (2003). Genotype x
Environment Interactions in the Regional Early Maize
Variety Trials in West and Central Africa. Maydica, 48:
93-104.
Bello OB, Ige SA, Abdulmaliq SY, Afolabi MS. (2013).
Interrelationship and Path Coefficient Analysis of
Morpho-physiological Traits among Maize (Zea mays
L.) Diallelic Crosses in the Southern Guinea Savanna
of Nigeria. Contemp. Agric. 62(3-4): 255-265.
Duarte JB, Vencovsky R (1999). Genotype environment
interaction: an introduction to AMMI analysis =
Interao gentipos ambientes: uma introduo
anlise AMMI. Sociedade Brasileira de Gentica,
Ribeiro Preto, SP, Brazil, 60p (in Portuguese).
Edugbo RE, Nwofia GE, Fayeun LS. (2015). An
Assessment of Soybean (Glycine max, L. Merrill) Grain
Yield in different Environments using AMMI and GGE
Biplot Models in Humidorest Fringes of Southeast
Nigeria. Agricultura Tropica Et Subtropica, 48(3-4), 82-
90. DOI: 10.1515/ats-2015-0012.
Ezzat EM, Ali MA, Mahmoud AM. (2010). Agronomic
performance, genotype x environment interactions and
stability analysis of grain sorghum (S. bicolor L.
Moench) Asian J. Crop Sci. 2 (4): 250-260.
GenStat release 12.1 Statistical Software for PC/Windows
Vista. GenStat VSN International Ltd. (2009).
Karimizadeh R, Mohammadi M, Sabaghni N, Mahmoodi
AA, Roustami B, Seyyedi F, Akbari F. (2013). GGE
Biplot Analysis of Yield Stability in Multi-environment
Trials of Lentil Genotypes under Rainfed Condition.
Notulae Scientia Biologicae 5(2): 256-262.
Kaya Y, Akcura M, Taner S. (2006). GGE-Biplot Analysis
of Multi-environment Yield Trials in Bread Wheat.
Turkey J. Agric. 30:325-337.
Kendal E, Doan Y (2015). Stability of a Candidate and
Cultivars (Hordeum Vulgare L) by GGE Biplot Analysis
of Multi-Environment Yield Trial in Spring Barley.
Agriculture & Forestry, 61(4): 307-318.
Khalil IA, Rahman H, Rehman NU, Arif M, Khalil IH, Iqbal
M, Ullah H, Afridi K, Sajjad M, Ishaq M. (2011).
Evaluation of Maize Hybrids for Grain Yield Stability in
North-west of Pakistan. Sarhad J. Agric. 27(2): 213-
218.
Kili H (2014). Additive Main Effects and Multiplicative
Interactions (AMMI) Analysis of Grain Yield in Barley
Genotypes across Environments. Tarm Bilimleri
Dergisi J. Agric. Sci. 20:337-344.
MAFFS (Ministry of Agriculture Forestry and Food
Security), Crop Production Guidelines for Sierra
Grain Yield Stability in Three-way Cross Hybrid Maize Varieties using AMMI and GGE Biplot Analysis
Leone, National Agricultural Research Coordinating
Council (NARCC), Institute of Agricultural Research
(IAR), Rice Research Station (RRS) and Njala
University (NU), 10 July, 2005, pp. 17- 18.
Makinde SCO, Ariyo OJ (2011). Analysis of Genotype
Environment interaction of groundnut (Arachis
hypogaea L.). Malaysian J. Applied Biology 40 (2): 19-
26.
Makinde SCO, Ariyo OJ, Akinbowale RI. (2013).
Assessment of groundnut performance in different
environments using Additive Main effects and
Multiplicative Interaction (AMMI) model. Canadian J.
Plant Breeding 1(2): 60-66.
Miranda GV, de Souza LV, Guimares LJM, Namorato H,
Oliveira LR, Soares MO. (2012). Multivariate analyses
of genotype x environment interaction of popcorn Pesq.
agropec. bras., Braslia, 44(1): 45-50.
Mitrovi B, Stanisavljevi D, Treski S, Stojakovi M,
Ivanovi M, Bekavac G, Rajkovi M. (2012). Evaluation
of experimental maize hybrids tested in multi-location
trials using ammi and gge biplot analyses. Turkish J.
Field Crops, 17(1): 35-40.
Mosa HE, Motawei AA, Abd El-Aal AMM, Abd El-Azeem
MEM. (2011). Yield Stability of some Promising Maize
(Zea mays L.) Hybrids under Varying Locations. J.
Agric. Res. Kafer El-Sheikh University,37(1): 99-109.
Nassir AL (2013). Genotype x environment analysis of
some yield components of upland rice (Oryza sativa L.)
under two ecologies in Nigeria. Intl. J. Plant Breeding
and Genetics. 7(2): 105-114.
Negash AW, Mwambi H, Zewotir T, Taye G. (2013).
Additive main effects and multiplicative interactions
model (AMMI) and genotype main effect and genotype
by environment interaction (GGE) biplot analysis of
multi-environmental wheat variety trials. Afr. J. Agric.
Res. 8(12): 1033-1040.
Olawuyi OJ, Odebode AC, Alfar-Abdullahi M, Olakojo SA,
Adesoye AI. (2010). Performance of maize genotypes
and arbuscular mycorrhizal fungi in Samara district of
south west region of Doha-Qatar. Nig. J. Mycol.
3(1):86-100.
Orhun GE (2013). Maize for Life. Intl. J. Food Sci. Nutrition
Engineering, 3(2): 13-16. DOI:
10.5923/j.food.20130302.01.
Oyekunle M, Haruna A, Badu-Apraku B, Usman IS, Mani
H, Ado SG, Olaoye G, Obeng-Antwi K, Abdulmalik RO,
Ahmed HO. (2017). Assessment of Early-Maturing
Maize Hybrids and Testing Sites Using GGE Biplot
Analysis. Crop Sci. 57: 1-9.
PBTools, version 1.4. 2014. Biometrics and Breeding
Informatics, PBGB Division, International Rice
Research Institute, Los Baos, Laguna.
Randjelovic V, Prodanovic S, Tomic Z, Simic A. (2011).
Genotype x year effect on grain yield and nutritive
values of maize (Zea mays L.). J. Ani. Vet. Sci. Adv.
10(7): 835-840.

Sayar MS, Anlarsal AE, Basbag M. (2013). Genotype
environment interactions and stability analysis for dry-
matter yield and seed yield in Hungarian vetch
(Viciapannonica CRANTZ.). Turk. J. Field Crops, 18(2):
238-246.
Tariku S, Lakew T, Bitew M, Asfaw M. (2013). Genotype
by environment interaction and grain yield stability
analysis of rice (Oryza sativa L.) genotypes evaluated
in north western Ethiopia. Net J. Agric. Sci. 1, 1016.
United States Department of Agriculture (2015). Sierra
Leone corn yield by year.
https://www.indexmundi.com/agriculture/?country=sl&
commodity=corn&graph=yield.
Yan W (2001). GGE biplot - A Windows Application for
Graphical Analysis of Multi-environment Trial Data and
Other Types of Two-way Data. Agron. J., 93: 1111-
1118.
Yan W, Hunt LA, Sheng Q, Szlavnics Z. (2000). Cultivar
evaluation and mega-environment investigation based
on the GGE biplot. Crop Sci. 40:597605.
doi:10.2135/cropsci2000.403597x.
Yan W, Kang MS (2003). GGE Biplot Analysis: A Graphical
Tool for Breeders, Geneticists, and Agronomists. 1st
Edition, CRC Press LLC., Boca Raton, Florida, USA.
pp: 271.
Grain Yield Stability in Three-way Cross Hybrid Maize Varieties using AMMI and GGE Biplot Analysis
Int. J. Plant Breed. Crop Sci. 299
Yan W, Kang MS, Ma B, Woods S, and Cornelius PL.
(2007). GGE Biplot vs. AMMI Analysis of Genotype-by-
Environment Data. Crop Sci 47:643655.
Yan W, Tinker NA (2006). Biplot Analysis of Multi-
environment Trial Data: Principles and Applications.
Canadian J. Plant Sci., 86: 623-645.
Yan WK, Hunt LA, Sheng QL, Szlavnics Z. (2000). Cultivar
evaluation and mega-environment investigation based
on the GGE Biplot. Crop Sci. 40: 597605.
Accepted 29 September 2017
Citation: Sesay S, Jalloh AB, Sama VA (2017). Grain
Yield Stability in Three-way Cross Hybrid Maize Varieties
using AMMI and GGE Biplot Analysis. International
Journal of Plant Breeding and Crop Science 4(2): 268-275.
Copyright: 2017 Sesay et al. This is an open-access
article distributed under the terms of the Creative
Commons Attribution License, which permits unrestricted
use, distribution, and reproduction in any medium,
provided the original author and source are cited.