© The Authors, 2023, Published by the Universidad del Zulia*Corresponding author: hgungor78@hotmail.com
Keywords:
Grain yield
Yield components
Multi-environments
Stability
GGE biplot analysis of genotype by environment interaction of barley cultivars
Análisis de biplot GGE de interacción genotipo por ambiente de cultivares de cebada
Análise de biplot GGE da interação genótipo por ambiente de cultivares de cevada
Hüseyin Güngör
1*
Mehmet Fatih Çakır
2
Ziya Dumlupınar
3
Rev. Fac. Agron. (LUZ). 2023, 40(2): e234021
ISSN 2477-9407
DOI: https://doi.org/10.47280/RevFacAgron(LUZ).v40.n2.11
Crop production
Associate editor: Dr. Rosa Razz
University of Zulia, Faculty of Agronomy
Bolivarian Republic of Venezuela
1
Department of Field Crops, Faculty of Agriculture, Duzce
University, 81620, Duzce, Turkey.
2
Environment and Health Coordination Technical
Specialization, Duzce University, Duzce, Turkey.
3
Department of Agricultural Biotechnology, Faculty of
Agriculture, Kahramanmaras Sutcu Imam University,
Kahramanmaras, Turkey.
Received: 30-03-2023
Accepted: 16-05-2023
Published: 02-06-2023
Abstract
This study was conducted out to determine grain yield, yield components,
and some quality charecteristics of 17 barley (Hordeum vulgare L.) genotypes
at six environments in Thrace region of Turkey, using principal component
analysis (PCA) and genotype (G) + genotype × environment interaction
(GGE) biplot analysis to dene the genotypes with higher yield and desirable
quality traits during the 2016-2017 and 2017-2018 cropping seasons. Mean
values of the genotypes varied from 5106-6753 kg.ha
-1
for grain yield,
103.4-117.1 days for heading date, 94.6-110.3 cm for plant height, 6.26-
10.07 cm for spike length, 25.0-75.5 number of grains per spike, 1.20-2.99
g grain weight per spike, 35.0-50.5 g for thousand kernel weight and 56.4-
64.1 kg.hl
-1
for test weight. The relationships among the examined traits
and genotypes was 53.9 % as dened by PC biplot analyses. GGE biplot
analysis represented 94.77 % of the relationship of G + GE for grain yield.
Two mega circles were formed according to grain yield, Zeus genotype for
E1, E2 and E5 locations and Arcanda genotype for E3, E4 and E6 locations
were determined as prominent genotypes. Zeus and Arcanda cultivars have
been identied as the most ideal and stable genotypes.
This scientic publication in digital format is a continuation of the Printed Review: Legal Deposit pp 196802ZU42, ISSN 0378-7818.
Rev. Fac. Agron. (LUZ). 2023, 40(2): e234021. April-June. ISSN 2477-9407.
2-7 |
Resumen
Este estudio se realizó para determinar el rendimiento de grano,
los componentes de rendimiento y algunas características de calidad
de 17 genotipos de cebada (Hordeum vulgare L.) en seis ambientes
en la región de Tracia en Turquía, utilizando análisis de componentes
principales (PCA) y análisis de biplot de genotipo (G) + interacción
genotipo × ambiente (GGE) para denir los genotipos con mayor
rendimiento y características de calidad deseables durante las
temporadas de cultivo 2016-2017 y 2017-2018. Los valores medios
de los genotipos variaron de 5106-6753 kg.ha
-1
para el rendimiento
de grano, 103,4-117,1 días para la fecha de espigado, 94,6-110,3 cm
para la altura de la planta, 6,26-10,07 cm para la longitud de la espiga,
25,0-75,5 número de granos por espiga, 1,20-2,99 g de peso de grano
por espiga, 35,0-50,5 g para el peso de mil granos y 56,4-64,1 kg.hl
-1
para el peso de prueba. Las relaciones entre las características y los
genotipos examinados fueron del 53,9 % según los análisis de biplot
PC. El análisis de biplot GGE representó el 94,77 % de la relación
de G + GE para el rendimiento de grano. Se formaron dos mega
círculos según el rendimiento de grano, el genotipo Zeus para las
ubicaciones E1, E2 y E5 y el genotipo Arcanda para las ubicaciones
E3, E4 y E6 fueron determinados como genotipos prominentes. Los
cultivares Zeus y Arcanda han sido identicados como los genotipos
más ideales y estables.
Palabras clave: rendimiento de grano, componentes de rendimiento,
multi-ambientes, estabilidad.
Resumo
Este estudo foi realizado para determinar o rendimento de grãos,
componentes de rendimento e algumas características de qualidade
de 17 genótipos de cevada (Hordeum vulgare L.) em seis ambientes
na região de Trácia da Turquia, utilizando análise de componentes
principais (ACP) e análise de biplot GGE (genótipo G + interação
genótipo x ambiente GE) para denir os genótipos com maior
rendimento e características de qualidade desejáveis durante as safras
2016-2017 e 2017-2018. Os valores médios dos genótipos variaram
de 5106-6753 kg.ha
-1
para rendimento de grãos, de 103,4-117,1 dias
para data de perlhamento, de 94,6-110,3 cm para altura de planta,
de 6,26-10,07 cm para comprimento de espiga, de 25,0-75,5 número
de grãos por espiga, de 1,20-2,99 g de peso de grão por espiga, de
35,0-50,5 g para peso e número de sementes por mil, e de 56,4-64,1
kg.hl
-1
para peso de teste. As relações entre as características e os
genótipos examinados foram de 53,9 % como denido pelas análises
de biplot PC. A análise de biplot GGE representou 94,77 % da relação
de G + GE para o rendimento de grãos. Dois mega círculos foram
formados de acordo com o rendimento de grãos, o genótipo Zeus para
as localidades E1, E2 e E5 e o genótipo Arcanda para as localidades
E3, E4 e E6 foram determinados como genótipos proeminentes. As
cultivares Zeus e Arcanda foram identicadas como os genótipos
mais ideais e estáveis.
Palavras-chave: rendimento de grãos, componentes de rendimento,
multi-ambientes, estabilidade.
Introduction
The primary uses of barley (Hordeum vulgare ssp. vulgare L. 2n
= 2x = 14), one of the world’s oldest cultivars, are in the food, animal
feed, and malt industries (Meints et al., 2021).
Barley produces 157 million tons of grain annually, which puts
it fourth in the world behind corn, wheat, and rice according to data
from 2020 (Food and Agriculture Organization, 2023). Turkey’s grain
production area is made up of 28 % barley-growing land. In terms
of production amount and production area in 2021, barley comes in
third place behind wheat and corn (3.1 million ha) (5.75 million tons)
(TUIK, 2023).
Barley is less aected by drought and is tolerant to salinity and the
use of straw in animal nutrition is the factor that makes barley come
to the fore (Khalili et al., 2013; Vaezi et al., 2018).
Breeders are trying to develop superior varieties with desirable
characteristics tolerant of various ecological settings. In breeding
studies, it is crucial to identify the genotypes that function well and
exhibit a wide range of trait stability. The GGE combined analysis
method, also known as the GGE Biplot, combines the (G) genotype
and G×E (interaction) eects, or the two main components, on the
same graph. This allows plant breeders to assess the data in two
directions, visually. Plant breeders often use the biplot analysis
method because it enables the graphic display of multiple genotype
features and the visual comparison of the relationships between
genotypes and traits. The biplot analysis method has been accepted
as an eective evaluation method applied in plant breeding because it
provides the opportunity to evaluate many features at the same time
visually and aects success in selection (Yan et al., 2007; Hagos and
Abay, 2013; Gungor et al., 2022a).
Based on statistical data and comparative trial results, barley
breeding studies have been conducted recently in various countries
and have shown a signicant increase in yield in barley plants (Laidig
et al., 2017).
It is assumed that barley studies, particularly on adaptation and
improvement, will help to increase yield in the production areas.
Barley agriculture is signicant in Turkey in terms of production and
amount, and its importance has increased in the livestock sector.
This study aimed to identify the genotypes with high yield and
adaptability characteristics in various locations by evaluating some
barley genotypes’ yield components and quality characteristics in the
Thrace region using PCA and GGE biplot analysis methods.
Materials and methods
Plant materials, experimental details, and layout
In the study, seventeen (eight two-row, nine six-row) barley
genotypes were used as material. The experiments were carried out at
six dierent environments (Kırklareli, Tekirdağ, and Edirne locations)
corresponding to the entire Thrace region, during the 2016-2017 and
2017-2018 cropping seasons (table 1). The research was carried out in
a randomized complete block design with four replications.
In both growing seasons, seeds were sown between the late
October and early November in plots with a 20 cm row spacing and
5 m in 6 rows, and 500 seeds per m
2
. In both sowing and harvest, the
experiment’s parcel sizes were carried out to be 6 m
2
. In the test plots,
weeds were manually controlled, and no pesticides were applied.
This scientic publication in digital format is a continuation of the Printed Review: Legal Deposit pp 196802ZU42, ISSN 0378-7818.
Güngör et al. Rev. Fac. Agron. (LUZ). 2023 40(2): e234021
3-7 |
Table 1. Cultivar, spike type, growing season, environments for trial.
Cultivars Spike type Owner Code Growing seasons Environments
Precipitation
(mm)
Arcanda 2 PG E1 2016-2017 Lüleburgaz 366.3
Imeon 2 IAK E2 2016-2017 Hayrabolu 451.7
Orfej 2 IAK E3 2016-2017 Edirne 408.0
Alena 2 PG E4 2017-2018 Lüleburgaz 696.3
Sladoran 2 TARI E5 2017-2018 Keşan 799.6
Harman 2 TARI E6 2017-2018 Babaeski 696.3
Hasat 2 TARI
Bolayır 2 TARI
Zeus 6 PG
IZ Bori 6 IAK
Bozhin 6 IAK
Hemus 6 IAK
Barberousse 6 TARI
Kıral 97 6 BDIARI
Veslet 6 IAK
Lord 6 TAREKS
Martı 6 TARI
PG: ProGen Seed Company, Turkey; TAREKS: Tareks Seed Company, Turkey; BDIARI: Bahri Dağdaş International Agricultural Research Institute, Turkey; TARI: Trakya
Agricultural Research Institute, Turkey; IAK: Institute of Agriculture Karnobat, Bulgaria.
With sowing, 50 kg.ha
-1
nitrogen (N) and 50 kg.ha
-1
phosphorus
(P
2
O
5
) were purely applied, the top fertilizer was divided into
two, and 90 kg.ha
-1
N was applied during the tillering period, and
60 kg.ha
-1
nitrogen was applied during the tillering period. Harvest
was performed in the rst week of July in both growing seasons.
In the research, plant height (PH), heading date (the number
of days between January 1st and the day when the plants are 50 %
spike in each plot-HD), spike length (SL), number of grains per spike
(KNS), weight per spike (KWS), thousand kernel weight (TKW), test
weight (TW) and grain yield (GY) were evaluated.
Statistical Analyses
Analysis of variance was conducted on the two years of data, and
the least signicant dierence (LSD) test of the mean data was made
by Duncan grouping test. Over the average data, principal component
analysis was calculated and evaluated using the biplot method
(SAS Institute Inc. JMP 15.1, 2020). Using average data across six
environments, GGE Biplot analyses were calculated using Genstat
14
th
(VSN International Ltd., 2011) software.
Results and discussion
Genotype by Environment Interaction (GEI)
The eects of G, E, and GE interaction were found important at
the P<0.01 level in the study using 17 barley genotypes. In this study,
the genotype, environment, and GE interactions were found to be
statistically signicant when the combined analysis of variance for all
traits was examined (table 2).
Table 2. Mean square values of investigated traits.
Sources of
variations
Genotype
(G)
Environment
(E)
G x E C.V. (%)
Degrees of Freedom 16 5 80
Grain Yield 46000.1** 176862.7** 6322.2** 7.85
Variability (%) 34.62 41.59 23.79
Heading Date 429.3790** 200.8824** 23.377** 0.94
Variability (%) 70.50 10.31 19.19
Plant Height 491.537** 2020.104** 116.053** 2.88
Variability (%) 37.07 28.86 34.07
Spike Length / 27.78565** 13.37041** 2.73987** 9.52
Variability (%) 60.85 9.15 30.00
No. of Grain /Spike 6009.619** 503.355** 159.953** 15.34
Variability (%) 86.26 2.26 11.48
Grain Weight/Spike 7.045635** 2.969367** 0.658592** 19.34
Variability (%) 62.54 8.24 29.23
Test Weight 98.6434** 261.0733** 14.1094** 0.82
Variability (%) 39.34 32.53 28.13
Thousand Kernel
Weight
468.5870** 60.1145** 34.5890** 1.72
Variability (%) 70.96 2.84 26.19
** Signicant at the %1 level of probability.
This scientic publication in digital format is a continuation of the Printed Review: Legal Deposit pp 196802ZU42, ISSN 0378-7818.
Rev. Fac. Agron. (LUZ). 2023, 40(2): e234021. April-June. ISSN 2477-9407.
4-7 |
There was a 34.62 % genotype eect on grain yield, a 41.59 %
environment eect, and a 23.79 % genotype x environment interaction
eect (table 2). It has been found that environmental variation aects
grain yield more than other sources of variation. While the average
grain yield of the genotypes was 5901 kg.ha
-1
, the Lord (5106 kg.ha
-1
)
genotype had the lowest grain yield, and the Arcanda genotype had
the highest (6753 kg.ha
-1
) (table 3).
In similar studies, the environmental eect on grain yield; Vaezi
et al. (2017), 73.94 %, Ghazvini et al. (2022), 53.60 %, and Fana et
al. (2018) reported 47.29 %. Grain yield is a complex trait governed
by genotype (G), environment (E), and genotype x environment
interaction (GEI) eects. GEI interaction is essential for breeders
and reects yield variation that cannot be explained by genotype and
environmental eects (Yan et al., 2001; Vaezi et al., 2017).
The heading date was 70.50 % aected by genotype, 10.31 % by
the environment, and 19.19 % by genotype × environment interaction
(table 2). The earliest heading date was determined in the Harman
genotype with 103.4 days, and the latest heading date was determined
in the Kıral 97 genotype with 117.1 days (table 4). According to the
environmental averages, the mean heading date was determined as
108.9 days (table 5). The latest heading date was found at the E2
(111.9 days) location and the earliest at the E5 (106.8 days) location
(table 5).
The results showed that genotype had a 37.07 % eect on
plant height, the environment had a 28.86 % eect, and genotype
and environment interaction had a 34.07 % eect (table 2). The
genotypes’ average plant height ranged from 94.6 to 110.3 cm. The
Kıral 97 genotype produced the shortest plant height (94.6 cm), while
the Harman genotype produced the longest plant height (110.3 cm)
(table 4). The most extended plant height was E1 (111.6 cm), and the
shortest plant height was E4 (97.4 cm), according to environmental
averages (table 5).
Table 3. Mean grain yield (kg.ha
-1
) of cultivars across test
environments.
Cultivars E1 E2 E3 E4 E5 E6 Mean
Arcanda 7154 6712 6694 6522 6674 6763 6753 a
Imeon 7117 6515 5660 5633 5952 6276 6191 b
Orfey 7071 5377 5927 5773 6331 6657 6189 b
Alena 6969 5788 6458 5842 5581 6170 6134 b
Sladoran 6684 5796 5919 6184 5669 6421 6111 bc
Harman 6621 5891 5444 6345 5998 6025 6053 bc
Hasat 6288 5640 5548 5729 6204 5726 5855 cde
Bolayır 5679 5191 4546 5141 4979 6112 5274 gh
Zeus 7861 6636 5674 5953 6782 6527 6572 a
Izbori 7136 6488 4450 5794 6668 5566 6016 bcd
Bojin 7086 6301 5259 5765 5371 6358 6023 bcd
Hemus 6906 6142 5144 5142 5704 5703 5790 de
Barberousse 6767 6198 4923 5151 5909 5722 5778 de
Kıral 97 6575 6189 4415 5076 5895 5534 5613 ef
Veslet 6529 5592 4552 4624 5532 5805 5438 fg
Lord 6464 4262 4250 4822 5273
5567 5106 h
Martı 6021 5454 4701 4587 5622 6196 5430 fg
Mean 6760 a 5892 c 5268 e 5534 d 5890 c 6066 b 5901
The spike length was 60.85 % aected by genotype, 9.15 %
by the environment, and 30.00 % by genotype × environment
interaction (table 2). The spike length of the genotypes varied between
6.26-10.07 cm. The longest spike length was measured in the Imeon
(10.07 cm) genotype, and the lowest spike length was measured in the
Bojin (6.26 cm) genotype (table 4). When the environmental averages
were examined, E6 (8.35 cm) locations had the highest spike length,
and E1 and E2 (7.33 cm) locations had the lowest spike length (table 5).
Table 4. Mean of yield component and quality traits of 17 barley
cultivars.
Cultivars HD PH SL KNS KWS TKW TW
Arcanda 113.9 b 101.8 f 9.59 b 26.8 g 1.57 e 50.5 a 63.9 ab
Imeon 111.8 d 99.5 g 10.07 a 31.3 f 1.38 efg 42.2 g 63.5 c
Orfey 106.4 gh 99.3 gh 8.85 c 27.5 g 1.26 fg 44.3 e 62.9 d
Alena 114.0 b 98.6 gh 8.69 cd 25.5 g 1.42 ef 49.8 b 64.1 a
Sladoran 105.0 ı 99.7 g 7.30 f 25.3 g 1.20 g 45.0 d 60.3 h
Harman 103.4 k 110.3 a 7.38 f 26.9 g 1.38 efg 45.3 d 63.7 bc
Hasat 104.5 ıj 108.1 b 8.41 de 25.8 g 1.30 fg 45.9 c 62.9 d
Bolayır 104.9 ı 97.7 h 8.28 de 25.0 g 1.20 g 43.4 f 63.8 ab
Zeus 112.6 c 107.1 bc 8.26 e 60.3 b 2.81 a 42.5 g 62.0 e
Iz Bori 106.9 g 103.2 f 6.32 h 60.4 c 2.04 c 39.2 ı 61.4 f
Bojin 112.8 c 103.1 f 6.26 h 56.8 bc 2.30 b 39.6 ı 60.5 gh
Hemus 108.1 f 105.9 cd 7.50 f 50.4 d 1.81 d 37.2 l 61.3 f
Barberousse 105.9 h 103.5 ef 6.82 g 45.6 e 1.78 d 37.1 l 62.0 e
Kıral 97 117.1 a 94.6 ı 6.76 g 75.5 a 2.99 a 35.0 m 56.4 j
Veslet 109.4 e 105.0 de 7.56 f 50.7 d 2.09 c 38.2 k 60.8 g
Lord 111.6 d 109.9 a 7.37 f 48.7 de 1.97 cd 38.7 j 64.0 a
Martı 104.2 j 107.1 bc 7.44 f 47.6 de 1.97 cd 40.6 h 59.7 ı
Mean 108.9 103.2 7.81 41.7 1.79 42.0 61.9
HD: Heading date, PH: Plant height, SL: Spike length, KNS: Number of grains
per spike, KWS: Grain weight per spike, TW: Test weight, TKW: Thousand kernel
weight.
The number of grains per spike was 86.26 % aected by genotype,
2.26 % by the environment, and 11.48 % by genotype-environment
interaction (table 2). According to genotype averages, the number
of grains per spike ranged from 25.0 to 75.5, with the genotype
Kıral 97 having the highest number of grains per spike (75.5) and
Bolayır having the lowest (25.0) (table 4). The E5 (45.1) location had
the highest value for the number of grains per spike, while the E2
(38.6) location had the lowest value, according to the environmental
averages (table 5).
The results showed that genotype had a 62.54 % eect on the grain
weight per spike, the environment had an 8.24 % eect, and genotype
and environment interaction had a 29.23 % eect (table 2). The Kıral
97 genotype had the most signicant grain weight per spike (2.99 g),
and the Sladoron and Bolayır genotypes had the lowest grain weight
per spike (1.2 g) (table 4). The grain weight per spike, as calculated
by the environmental averages, ranged from 1.52 to 2.01 g, with the
E5 location recording the most excellent value (2.01 g) and the E2
location recording the lowest value (1.52 g) (table 5).
The eect of genotype on test weight was found to be 39.34 %,
the eect of the environment as 32.53 %, and the eect of genotype
× environment interaction was 28.13 % (table 2). Alena (64.1 kg.hl
-1
)
This scientic publication in digital format is a continuation of the Printed Review: Legal Deposit pp 196802ZU42, ISSN 0378-7818.
Güngör et al. Rev. Fac. Agron. (LUZ). 2023 40(2): e234021
5-7 |
genotype had the highest test weight and the lowest Kıral 97 (56.4
kg.hl
-1
) genotype (table 4). According to the environmental averages,
the lowest test weight was determined in the E6 location (59.5 kg.hl
-1
)
and the highest in the E2 (64.5 kg.hl
-1
) location (table 5).
The environment had a 2.84 % impact, the genotype had a 70.96
% inuence, and the genotype x environment combination had a 26.19
% impact on thousand kernel weight (table 2). The genotype Kıral 97
(35.0 g) had the lowest value, and the genotype Arcanda (50.5 g) had
the most incredible thousand kernel weight (table 4). According to
environmental averages the thousand kernel weight varied between
40.6-43.2 g, with the lowest measurement coming from location E2
(40.6 g) and the highest value coming from location E6 (43.2 g) (table 5).
Table 5. Average of yield component and quality traits in six
environments.
Environments HD PH SL KNS KWS TKW TW
E1 108.4 d 111.6 a 7.33 d 39.7 b 1.65 b 42.6 b 63.6 b
E2 111.9 a 101.5 d 7.33 d 38.6 b 1.52 c 40.6 e 64.5 a
E3 109.5 b 107.1 b 7.62 c 39.1 b 1.66 b 42.5 b 62.8 c
E4 108.1 d 97.4 e 8.09 b 43.0 a 2.00 a 42.0 c 60.5 e
E5 106.8 e 103.2 c 8.17 ab 45.1 a 2.01 a 41.3 d 60.9 d
E6 109.0 c 98.2 e 8.35 a 43.8 a 1.92 a 43.2 a 59.5 f
Mean 108.9 101.4 7.81 41.5 1.79 42.0 61.9
HD: Heading date, PH: Plant height, SL: Spike length, KNS: Number of grains
per spike, KWS: Grain weight per spike, TW: Test weight, TKW: Thousand kernel
weight.
Principal Component and GGE-Biplot Analysis
Principal Component two-dimensional score accounted for 53.9
% of the total variation, PC1 had 36.2 %, and PC2 had 17.7 %. The
locations were divided into two groups, E1, E2, and E3 circles formed
a group, E4, E5, and E6 circles formed a dierent group (gure 1).
Grain yield (GY) correlated positively with the traits SL, TKW, TW,
and PH but negatively with the traits KNS, KWS, and HD. In terms
of GY, SL, TKW, TW, and PH characteristics, two-row and six-row
barley genotypes, stood out, respectively (gure 1).
According to Yan and Tinker (2006), there is a positive
relationship between the features in these slices when the angle
between the vectors is between 0°-90°, a negative relationship when
the angle is between 90°-180°, and no relationship when the angle is
90° evaluated as.
Figure. 1. Relationship among genotypes and distance origin and
angles among the vectors.
Figure. 2. Position of environments on GGE biplot graph
investigated traits for grain yield.
The GGE-biplot graph of the positions of the locations explained
94.77 %. PC1 81.07 %, PC2 13.70 % were recorded. E1, E2, and E5
locations had positive PC2 values, while E3, E4 and E6 had negative
PC2 values. All circles had a positive correlation. While the E1 and
E2 locations vectors had the lowest angles, the vectors of E2 and E3
locations had the widest angles (gure 2). While the variation between
genotypes increases as the vector moves away from the origin, the
variation between genotypes decreases as the vector approaches the
origin (Abate et al., 2015).
According to the GGE biplot scatters graph analysis, there are
two mega environments and seven sectors. The rst mega-circle
comprised E1, E2, and E5, and the second comprised E3, E4, and E6.
The two-row barley genotypes were found in the second mega-circle,
whereas the six-row barley genotypes were found in the rst mega
environment. For the rst and second mega environments, Zeus and
Arcanda took center stage respectively. The farthest genotypes from
mega circles were found to be the Martı and Lord genotypes (gure 3). In
their mega environment, high performing genotypes are those located
at the top of the polygon (Yan, 2014).
The stability of the genotypes regarding grain yield is shown in
the mean circumference axis (AEC) Figure 4 was also evaluated.
Zeus genotype had the highest PC1 score. Izbon, Arcanda, Imeon,
Bojin, Hemus, Barberousse, Orfey, Kıral 97 and Alena genotypes are
in front of the AEC ordinate, while Sladoran, Veslet, Harman, Lord,
Hasat, Bolayır, and Martı genotypes are behind the AEC ordinate.
The Martı genotype had the lowest PC1 value. E1 and E2 locations
had the highest PC1 value (gure 4).
The Zeus and Imeon genotypes were placed next to the Arcanda,
which was closest to the centeral circle. The farthest genotypes from the
centeral circle were found to be Martı, Bolayır, and Lord (gure 5). In
GGE Biplot analyses, circles closer to the centeral circle are considered
ideal circles. It has been determined that the E4 and E5 locations are
This scientic publication in digital format is a continuation of the Printed Review: Legal Deposit pp 196802ZU42, ISSN 0378-7818.
Rev. Fac. Agron. (LUZ). 2023, 40(2): e234021. April-June. ISSN 2477-9407.
6-7 |
located closer to the central circle. E6 location was determined as the
furthest perimeter from the central circle (gure 6).
Figure. 3. Scatter plot graph of GGE biplot analysis.
Figure. 4. Average environment coordinate, grain yield
performance of the cultivars, vectors to AEC.
The mean coordination of environments is used to interpret
the yield stability of genotypes (AEC). The AEC ordinate’s
line of stability, which intersects the origin and passes the ideal
circumference, is the line with the arrow on it. The genotypes with the
highest values are those that are on the right side of the general mean
line, and the genotypes with the lowest values are those that are on the
left. A genotype is less stable the longer its projection is in absolute
terms. Genotypes and environments close to the center circle in the
comparison plot are considered ideal genotypes and environments
(Yan and Tinker, 2006).
Figure. 5. Comparison biplot “ideal genotype’’using GGE biplot
with scaling focused on genotypes.
Figure. 6. Comparison biplot of ‘’ideal environment’’ for grain
yield using GGE biplot.
In similar studies conducted, they stated that genotypes could
be evaluated in terms of many traits with GGE Biplot methods, and
it provides convenience in visually examining genotypes that show
high performance in more than one trait and provides in selection
(Vaezi et al., 2017; Gungor et al., 2022b).
Conclusions
In this study, seventeen barley genotypes Arcanda and Zeus
genotypes as the ideal genotypes. In contrast, Martı and Bolayır
genotypes were determined as non-ideal genotypes in the study
carried out in six dierent environments. While the E1 location
had the highest grain yield, E4 and E5 locations were determined
as the most representative circles. As the study was conducted over
two years in dierent environments in the Thrace region, Zeus and
Arcanda genotypes were found outstanding genotypes in terms of
both high yielding and stable.
Literature cited
Abate, F., Mekbib, F. & Dessalegn, F. (2015). GGE biplot analysis of multi-
environment yield trials of durum wheat (Triticum turgidum Desf.)
genotypes in North Western Ethiopia. Journal of Expanded Agriculture
International, 8, 120-129.https://doi.org/10.9734/AJEA/2015/9994
Fana, G., Tadese, D., Sebsibe, H. & Verma, R.P. (2018). Multi-environment
trial analysis of food barley in Ethiopio using AMMI and GGE biplot
methods. Journal of Plant Breeding and Genetics, 6, 75-85. https://doi.
org/10.33687/pbg.006.03.2579
FAOSTAT. (2023). Food and Agriculture Organization of the United Nations.
Rome-Italy. (Online Database). http://www.faostat/en/
Ghazvini, H., Bagherikia, S., Pour-Aboughdareh, A., Sharifalhossaini, M.,
Razarvi, S. A., Mohammadi, S., Ghasemi-Kalkhoren, M., Fathihafshejani,
A. & Khakizade, G. (2022). GGE biplot analysis of promising barley lines
in the cold regions of Iran. Journal of Crop Improvement, 36(4), 461-472.
https://doi.org/10.1080/15427528.2021.1977448
Gungor, H., Cakir, M.F., & Dumlupinar, Z. (2022a). Evaluation of triticale:
genotype by environment interaction and GGE biplot analysis. Journal
of Animal and Plant Science, 32(6), 1637-1647. https://doi.org/10.36889/
JAPS.2022.6.0573
Gungor, H., Cakir, M.F., & Dumlupinar, Z. (2022b). Evaluation of wheat
genotypes: genotype × environment interaction and GGE biplot analysis.
Turkish Journal of Field Crops, 27(1), 149-157. https://doi.org/10.17557/
tjfc.1081513
Hagos, G. & Abay, F. (2013). AMMI and GGE biplot analysıs of bread wheat
genotypes in the Northen part of Ethiopia. Journal of Plant Breeding
Genetics, 1:12-18. https://esciencepress.net/journals/index.php/JPBG/
article/view/50/76
Khalili, M., Pour-Aboughadareh, A. & Naghavi, M.R. (2013). Screening of
drought tolerant cultivars in barley using morphophysiological traits
This scientic publication in digital format is a continuation of the Printed Review: Legal Deposit pp 196802ZU42, ISSN 0378-7818.
Güngör et al. Rev. Fac. Agron. (LUZ). 2023 40(2): e234021
7-7 |
and integrated selection index under water decit stress condition.
Advanced Crop Science., 3, 462–471. https://www.researchgate.net/
publication/257208136
Laidig, F., Piepho, H. P., Rentel, D, Drobek, T. & Meyer, U. (2017). Breeding
progress, genotypic and environmental variation and correlation of
quality traits in malting barley in German ocial variety trials between
1983 and 2015. Theorethical and Applied Genetics, 130, 2411–2429.
https://doi.org/10.1007/s00122-017-2967-4
Meints, B., Vallejos, C. & Hayes, P. (2021). Multi-use naked barley: A new
frontier. Journal of Cereal Science, 102, 103370. https://doi.org/10.1016/j.
jcs.2021.103370
SAS. (2020). JMP® 15 Design of Experiments Guide. Cary, NC 27513. USA.
TUIK. (2023). Turkish Statistical Institute. Retrieved from 30 January 2023 from
https://www.tuik.gov.tr
Vaezi, B., Pour-Aboughadareh, A., Mohammadi, R., Armion, M., Mehraban, A.,
Hossein-Pour, T. & Dorii, M. (2017). GGE biplot and AMMI analysis
of barley yield performance in Iran. Cereal Research Communications,
45(3), 500-511. https://doi.org/10.1556/0806.45.2017.019
Vaezi, B., Pour-Aboughadareh, A., Mehraban, A., Hossein-Pour, T., Mohammadi,
R., Armion, M. & Dorii, M. (2018). The use of parametric and non-
parametric measures for selecting stable and adapted barley lines.
Archives of Agronomy and Soil Science, 64(5), 597–611. https://doi.org/
10.1080/03650340.2017.1369529
VSN-INTERNATIONAL. (2011). Genstat for windows 14. ed. VSN International,
Hemel Hempstead, UK. https://vsni.co.uk/software/genstat
Yan, W., Cornelius, P. L., Crossa, J. & Hunt, L. A. (2001). Two types of GGE
biplots for analyzing multi-environment trial data. Crop Science, 41,
656–663. https://doi.org/10.2135/cropsci2001.413656x
Yan, W. & Tinker, N.A. (2006). Biplot analysis of multi-environment trial data:
principles and applications. Canadian Journal of Plant Science, 86, 623-
645. http://doi.org/10.4141/P05-169
Yan, W, Kang, M. S., Ma, B., Woods, S. & Cornelius, P.L. (2007). GGE Biplot
vs. AMMI analysis of genotype by environment data. Crop Science, 47,
643-655. https://doi.org/10.2135/cropsci2006.06.0374
Yan, W. (2014). Crop Variety Trials: Data Management and Analysis. John Wiley
& Sons. pp 360. https://doi.org/10.2135/cropsci2014.08.0002br